Fundamentals Of Fluid Mechanics 8th Edition | Ashish Sv

Transcript

Free ebooks ==> www.Ebook777.com M U N S O N , YO U N G A N D O K I I S H I ’ S Eighth Edition Philip M. Gerhart Andrew L. Gerhart www.Ebook777.com John I. Hochstein Free ebooks ==> www.Ebook777.com www.Ebook777.com WileyPLUS is a research-based online environment for effective teaching and learning. WileyPLUS builds students’ confidence because it takes the guesswork out of studying by providing students with a clear roadmap: • • • what to do how to do it if they did it right It offers interactive resources along with a complete digital textbook that help students learn more. With WileyPLUS, students take more initiative so you’ll have greater impact on their achievement in the classroom and beyond. For more information, visit www.wileyplus.com ALL THE HELP, RESOURCES, AND PERSONAL SUPPORT YOU AND YOUR STUDENTS NEED! www.wileyplus.com/resources Student Partner Program 2-Minute Tutorials and all of the resources you and your students need to get started Student support from an experienced student user Collaborate with your colleagues, find a mentor, attend virtual and live events, and view resources www.WhereFacultyConnect.com Quick Start © Courtney Keating/iStockphoto Pre-loaded, ready-to-use assignments and presentations created by subject matter experts Technical Support 24/7 FAQs, online chat, and phone support www.wileyplus.com/support Your WileyPLUS Account Manager, providing personal training and support Free ebooks ==> www.Ebook777.com Munson, Young, and Okiishi’s 8e Fundamentals of Fluid Mechanics Philip M. Gerhart College of Engineering and Computer Science University of Evansville Evansville, Indiana Andrew L. Gerhart A. Leon Linton Department of Mechanical Engineering Lawrence Technological University Southfield, Michigan John I. Hochstein Department of Mechanical Engineering University of Memphis Memphis, Tennessee www.Ebook777.com Executive Editor: Linda Ratts Executive Publisher: Don Fowley Senior Product Designer: Jennifer Welter Senior Content Manager: Valerie Zaborski Associate Product Designer: Wendy Ashenberg Design Director: Harry Nolan Senior Designer: Madelyn Lesure Executive Marketing Manager: Dan Sayre Photo Researcher: Billy Ray Editorial Assistant: Courtney Jordan Senior Production Editor: John Curley Production Management Services: Aptara Inc./Peggy Kellar Cover Design: Wendy Lai Cover Photo: Graham Jeffery/Sensitive Light This book was set in 10/12 STIX Regular by Aptara®, Inc., and printed and bound by Courier Kendallville. The cover was printed by Courier Kendallville. This book is printed on acid free paper. ∞ Founded in 1807, John Wiley & Sons, Inc. has been a valued source of knowledge and understanding for more than 200 years, helping people around the world meet their needs and fulfill their aspirations. Our company is built on a foundation of principles that include responsibility to the communities we serve and where we live and work. In 2008, we launched a Corporate Citizenship Initiative, a global effort to address the environmental, social, economic, and ethical challenges we face in our business. Among the issues we are addressing are carbon impact, paper specifications and procurement, ethical conduct within our business and among our vendors, and community and charitable support. For more information, please visit our website: www.wiley.com/go/citizenship. Copyright © 2016, 2013, 2009, 2006, 2002, 1999, 1994, 1990 by John Wiley & Sons, Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Sections 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc. 222 Rosewood Drive, Danvers, MA 01923, website www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030-5774, (201)748-6011, fax (201)748-6008, website http://www. wiley.com/go/permissions. Evaluation copies are provided to qualified academics and professionals for review purposes only, for use in their courses during the next academic year. These copies are licensed and may not be sold or transferred to a third party. Upon completion of the review period, please return the evaluation copy to Wiley. Return instructions and a free of charge return shipping label are available at www.wiley.com/go/returnlabel. Outside of the United States, please contact your local representative. ISBN 978-1-119-08070-1 (Binder-Ready Version) The inside back cover will contain printing identification and country of origin if omitted from this page. In addition, if the ISBN on the back cover differs from the ISBN on this page, the one on the back cover is correct Printed in the United States of America 10 9 8 7 6 5 4 3 2 1 About the Authors Philip M. Gerhart, Dean of Engineering and Computer Science and Professor of Mechanical and Civil Engineering at the University of Evansville, received his BSME degree from Rose-Hulman Institute of Technology in 1968 and his M.S. and Ph.D. degrees in Mechanical Engineering from the University of Illinois at Urbana-Champaign in 1969 and 1971. Before becoming Chair of Mechanical and Civil Engineering at the University of Evansville, he was on the Mechanical Engineering faculty at the University of Akron from 1971 to 1984. Dr. Gerhart has taught a variety of courses in fluid mechanics and other thermo-fluid sciences. He has consulted widely in the power generation and process industries and has authored or coauthored two previous books on fluid mechanics and fluid machinery. Since 1975, he has been deeply involved in the development of the American Society of Mechanical Engineers Performance Test Codes. He served as ASME Vice President for Performance Test Codes from 1998 to 2001, and is currently a member and vice-chair of the Committee on Fans, chair of the Committee on Fired Steam Generators, and a member of the Standing Committee on Performance Test Codes. Dr. Gerhart is a member of the American Society for Engineering Education and is a Life Fellow of the American Society of Mechanical Engineers. His honors and awards include the Outstanding Teacher Award from the Faculty Senate of the United Methodist Church, and the Performance Test Codes Medal from ASME. Andrew L. Gerhart, Associate Professor of Mechanical Engineering at Lawrence Technological University, received his BSME degree from the University of Evansville in 1996, his MSME from the University of Wyoming, and his Ph.D. in Mechanical Engineering from the University of New Mexico. At Lawrence Tech, Dr. Gerhart has developed both undergraduate and graduate courses in viscous flow, turbulence, creative problem solving, and first-year introductory engineering. He has codeveloped college-wide curriculum in engineering design and university-wide curriculum in leadership. He is the supervisor of the Thermal Science and Aerodynamics Laboratories, Coordinator of the Aeronautical Engineering Minor/Certificate, chair of the First Year Engineering curriculum committee, and faculty advisor for the student branch of the American Institute of Aeronautics and Astronautics and the SAE Aero Design team. Dr. Gerhart facilitates workshops worldwide, having trained hundreds of faculty members in active, collaborative, and problem-based learning, as well as training professional engineers and students in creative problem solving and innovation. He is a member of the American Society for Engineering Education and has received four best paper awards from their Annual Conferences. Dr. Gerhart was awarded the 2010 Michigan Professor of the Year by the Carnegie Foundation for the Advancement of Teaching and the Council for Advancement and Support of Education, Lawrence Tech’s Henry and Barbara Horldt Excellence in Teaching Award, the Engineering Society of Detroit’s (ESD) Outstanding Young Engineer, and ESD’s Council Leadership Award. He was elected to ESD’s College of Fellows, and is actively involved with The American Society of Mechanical Engineers, serving on the Performance Test Code Committee for Air-Cooled Condensers. John I. Hochstein, Professor of Mechanical Engineering at the University of Memphis, received a BE from Stevens Institute of Technology in 1973, an M.S. in Mechanical Engineering from the Pennsylvania State University in 1979, and his Ph.D. in Mechanical Engineering from the University of Akron in 1984. He has been on the faculty of the mechanical engineering department at the University of Memphis since 1991 and served as department chair from 1996 to 2014. Working as an engineer in nonacademic positions, Dr. Hochstein contributed to the design of the Ohio-Class submarines at the Electric Boat Division of General Dynamics and to the design of the Clinch River Breeder Reactor while an engineer at the Babcock & Wilcox Company. The focus of his doctoral studies was computational modeling of spacecraft cryogenic propellant management systems, and he has remained involved with NASA research on this topic since that time. ix x About the Authors Dr. Hochstein has twice been a NASA Summer Faculty Fellow for two consecutive summers: once at the NASA Lewis (now Glenn) Research Center, and once at the NASA Marshall Space Flight Center. Dr. Hochstein’s current primary research focus is on the capture of hydrokinetic energy to produce electricity. Dr. Hochstein is an Associate Fellow of AIAA and has served on the Microgravity Space Processes Technical Committee since 1986. He joined ASME as an undergraduate student and served for 4 years on the K20 Computational Heat Transfer Committee. He is a member of ASEE and has served the profession as an ABET Program Evaluator since 2002. About the New Authors xi A Quarter-Century of Excellence Bruce R. Munson, Professor Emeritus of Engineering Mechanics at Iowa State University, received his B.S. and M.S. degrees from Purdue University and his Ph.D. degree from the Aerospace Engineering and Mechanics Department of the University of Minnesota in 1970. Prior to joining the Iowa State University faculty in 1974, Dr. Munson was on the mechanical engineering faculty of Duke University from 1970 to 1974. From 1964 to 1966, he worked as an engineer in the jet engine fuel control department of Bendix Aerospace Corporation, South Bend, Indiana. Dr. Munson’s main professional activity has been in the area of fluid mechanics education and research. He has been responsible for the development of many fluid mechanics courses for studies in civil engineering, mechanical engineering, engineering science, and agricultural engineering and is the recipient of an Iowa State University Superior Engineering Teacher Award and the Iowa State University Alumni Association Faculty Citation. He has authored and coauthored many theoretical and experimental technical papers on hydrodynamic stability, low Reynolds number flow, secondary flow, and the applications of viscous incompressible flow. He is a member of The American Society of Mechanical Engineers. Donald F. Young, Anson Marston Distinguished Professor Emeritus in Engineering, received his B.S. degree in mechanical engineering, his M.S. and Ph.D. degrees in theoretical and applied mechanics from Iowa State University, and has taught both undergraduate and graduate courses in fluid mechanics at Iowa State for many years. In addition to being named a Distinguished Professor in the College of Engineering, Dr. Young has also received the Standard Oil Foundation Outstanding Teacher Award and the Iowa State University Alumni Association Faculty Citation. He has been engaged in fluid mechanics research for more than 35 years, with special interests in similitude and modeling and the interdisciplinary field of biomedical fluid mechanics. Dr. Young has contributed to many technical publications and is the author or coauthor of two textbooks on applied mechanics. He is a Fellow of The American Society of Mechanical Engineers. Ted H. Okiishi, Professor Emeritus of Mechanical Engineering at Iowa State University, joined the faculty there in 1967 after receiving his undergraduate and graduate degrees from that institution. From 1965 to 1967, Dr. Okiishi served as a U.S. Army officer with duty assignments at the National Aeronautics and Space Administration Lewis Research Center, Cleveland, Ohio, where he participated in rocket nozzle heat transfer research, and at the Combined Intelligence Center, Saigon, Republic of South Vietnam, where he studied seasonal river flooding problems. Professor Okiishi and his students have been active in research on turbomachinery fluid dynamics. Some of these projects have involved significant collaboration with government and industrial laboratory researchers, with two of their papers winning the ASME Melville Medal (in 1989 and 1998). Dr. Okiishi has received several awards for teaching. He has developed undergraduate and graduate courses in classical fluid dynamics as well as the fluid dynamics of turbomachines. He is a licensed professional engineer. His professional society activities include having been a vice president of The American Society of Mechanical Engineers (ASME) and of the American Society for Engineering Education. He is a Life Fellow of The American Society of Mechanical Engineers and past editor of its Journal of Turbomachinery. He was recently honored with the ASME R. Tom Sawyer Award. Wade W. Huebsch, Associate Professor in the Department of Mechanical and Aerospace Engineering at West Virginia University, received his B.S. degree in aerospace engineering from San Jose State University where he played college baseball. He received his M.S. degree in mechanical engineering and his Ph.D. in aerospace engineering from Iowa State University in 2000. Dr. Huebsch specializes in computational fluid dynamics research and has authored multiple journal articles in the areas of aircraft icing, roughness-induced flow phenomena, and boundary xi Free ebooks ==> www.Ebook777.com xii A Quarter-Century of Excellence layer flow control. He has taught both undergraduate and graduate courses in fluid mechanics and has developed a new undergraduate course in computational fluid dynamics. He has received multiple teaching awards such as Outstanding Teacher and Teacher of the Year from the College of Engineering and Mineral Resources at WVU as well as the Ralph R. Teetor Educational Award from SAE. He was also named as the Young Researcher of the Year from WVU. He is a member of the American Institute of Aeronautics and Astronautics, the Sigma Xi research society, the Society of Automotive Engineers, and the American Society of Engineering Education. Alric P. Rothmayer, Professor of Aerospace Engineering at Iowa State University, received his undergraduate and graduate degrees from the Aerospace Engineering Department at the University of Cincinnati, during which time he also worked at NASA Langley Research Center and was a visiting graduate research student at the Imperial College of Science and Technology in London. He joined the faculty at Iowa State University (ISU) in 1985 after a research fellowship sponsored by the Office of Naval Research at University College in London. Dr. Rothmayer has taught a wide variety of undergraduate fluid mechanics and propulsion courses for over 25 years, ranging from classical low and high speed flows to propulsion cycle analysis. Dr. Rothmayer was awarded an ISU Engineering Student Council Leadership Award, an ISU Foundation Award for Early Achievement in Research, an ISU Young Engineering Faculty Research Award, and a National Science Foundation Presidential Young Investigator Award. He is an Associate Fellow of the American Institute of Aeronautics and Astronautics (AIAA), and was chair of the 3rd AIAA Theoretical Fluid Mechanics Conference. Dr. Rothmayer specializes in the integration of Computational Fluid Dynamics with asymptotic methods and low order modeling for viscous flows. His research has been applied to diverse areas ranging from internal flows through compliant tubes to flow control and aircraft icing. In 2001, Dr. Rothmayer won a NASA Turning Goals into Reality (TGIR) Award as a member of the Aircraft Icing Project Team, and also won a NASA Group Achievement Award in 2009 as a member of the LEWICE Ice Accretion Software Development Team. He was also a member of the SAE AC-9C Aircraft Icing Technology Subcommittee of the Aircraft Environmental Systems Committee of SAE and the Fluid Dynamics Technical Committee of AIAA. www.Ebook777.com Preface This book is intended to help undergraduate engineering students learn the fundamentals of fluid mechanics. It was developed for use in a first course on fluid mechanics, either one or two semesters/ terms. While the principles of this course have been well-established for many years, fluid mechanics education has evolved and improved. With this eighth edition, a new team of authors is working to continue the distinguished tradition of this text. As it has throughout the past seven editions, the original core prepared by Munson, Young, and Okiishi remains. We have sought to augment this fine text, drawing on our many years of teaching experience. Based on our experience and suggestions from colleagues and students, we have made a number of changes to this edition. The changes (listed below, and indicated by the word New in descriptions in this preface) are made to clarify, update, and expand certain ideas and concepts. New to This Edition In addition to the continual effort of updating the scope of the material presented and improving the presentation of all of the material, the following items are new to this edition. Self-Contained: Material that had been removed from the text and provided only on-line has been brought back into the text. Most notable are Section 5.4 on the second law of thermodynamics and useful energy loss and Appendix E containing units conversion factors. Compressible Flow: Chapter 11 on compressible flow has been extensively reorganized and a limited amount of new material added. There are ten new example problems; some of them replace previous examples. All have special emphasis on engineering applications of the material. Example solutions employ tabulated compressible flow functions as well as graphs. Appendices: Appendix A has been expanded. Compressible flow function tables have been added to Appendix D. A new extensive set of units conversion factors in a useful and compact format appears in Appendix E. Computational Fluid Dynamics (CFD): A still unsettled issue in introductory fluid mechanics texts is what to do about computational fluid dynamics. A complete development of the subject is well beyond the scope of an introductory text; nevertheless, highly complex, highly capable CFD codes are being employed for engineering design and analysis in a continually expanding number of industries. We have chosen to provide a description of many of the challenges and practices that characterize widely used CFD codes. Our aim is twofold: to show how reasonably complex flows can be computed and to foster a healthy skepticism in the nonspecialist. This material is presented in an expanded Appendix A. Problems and Examples: Many new examples and problems emphasize engineering applications. Approximately 30% new homework problems have been added for this edition, and there are additional problems in WileyPLUS. Value: Nearly everyone is concerned about the upward spiral of textbook prices (yes, even authors and publishers!). We have taken a few modest steps to keep the price of this book reasonable. Most of these steps involve the removal of “bells and whistles.” For example, the thumbnail photos that accompanied the video icons in the 7th edition have been dropped. Wiley has also developed a number of different products to meet differing student needs and budgets. xiii xiv Preface y Key Features Fr < 1 Fr = 1 Fr > 1 E (© Photograph courtesy of Pend Oreille Public Utility District.) Illustrations, Photographs, and Videos Fluid mechanics has always been a “visual” subject—much can be learned by viewing various characteristics of fluid flow. Fortunately this visual component is becoming easier to incorporate into the learning environment, for both access and delivery, and is an important help in learning fluid mechanics. Thus, many photographs and illustrations have been included in the book. Some of these are within the text material; some are used to enhance the example problems; and some are included as margin figures of the type shown in the left margin to more clearly illustrate various points discussed in the text. Numerous video segments illustrate many interesting and practical applications of real-world fluid phenomena. Each video segment is identified at the appropriate location in the text material by a video icon of the type shown in the left margin. Each video segment has a separate associated text description of what is shown in the video. There are many homework problems that are directly related to the topics in the videos. Examples VIDEO V1.9 Floating razor blade One of our aims is to represent fluid mechanics as it really is—an exciting and useful discipline. To this end, we include analyses of numerous everyday examples of fluid-flow phenomena to which students and faculty can easily relate. In this edition there are numerous examples that provide detailed solutions to a variety of problems. Many of the examples illustrate engineering applications of fluid mechanics, as is appropriate in an engineering textbook. Several illustrate what happens if one or more of the parameters is changed. This gives the student a better feel for some of the basic principles involved. In addition, many of the examples contain photographs of the actual device or item involved in the example. Also, all of the examples are outlined and carried out with the problem solving methodology of “Given, Find, Solution, and Comment” as discussed in the “Note to User” before Example 1.1. The Wide World of Fluids The set of approximately 60 short “The Wide World of Fluids” stories reflect some important, and novel, ways that fluid mechanics affects our lives. Many of these stories have homework problems associated with them. The title of this feature has been changed from the 7th edition’s “Fluids in the News” because the stories cover more than just the latest developments in fluid mechanics. Homework Problems A wide variety of homework problems (approximately 30% new to this edition) stresses the practical application of principles. The problems are grouped and identified according to topic. The following types of problems are included: 1) 2) 3) 4) 5) 6) 7) 8) “standard” problems, computer problems, discussion problems, supply-your-own-data problems, problems based on “The Wide World of Fluids” topics, problems based on the videos, “Lifelong learning” problems, problems that require the user to obtain a photograph/image of a given flow situation and write a brief paragraph to describe it, Preface xv Computer Problems—Several problems are designated as computer problems. Depending on the preference of the instructor or student, any of the problems with numerical data may be solved with the aid of a personal computer, a programmable calculator, or even a smartphone. Lifelong Learning Problems—Each chapter has lifelong learning problems that involve obtaining additional information about various fluid mechanics topics and writing a brief report about this material. Well-Paced Concept and Problem-Solving Development Since this is an introductory text, we have designed the presentation of material to allow for the gradual development of student confidence in fluid mechanics problem solving. Each important concept or notion is considered in terms of simple and easy-to-understand circumstances before more complicated features are introduced. Many pages contain a brief summary (a highlight) sentence in the margin that serves to prepare or remind the reader about an important concept discussed on that page. Several brief elements have been included in each chapter to help the student see the “big picture” and recognize the central points developed in the chapter. A brief Learning Objectives section is provided at the beginning of each chapter. It is helpful to read through this list prior to reading the chapter to gain a preview of the main concepts presented. Upon completion of the chapter, it is beneficial to look back at the original learning objectives. Additional reinforcement of these learning objectives is provided in the form of a Chapter Summary and Study Guide at the end of each chapter. In this section a brief summary of the key concepts and principles introduced in the chapter is included along with a listing of important terms with which the student should be familiar. These terms are highlighted in the text. All items in the Learning Objectives and the Study Guide are “action items” stating something that the student should be able to do. A list of the main equations in the chapter is included in the chapter summary. System of Units Three systems of units are used throughout the text: the International System of Units (newtons, kilograms, meters, and seconds), the British Gravitational System (pounds, slugs, feet, and seconds), and the English Engineering System, sometimes called the U.S. Customary System (pounds (or pounds force), pounds mass, feet, and seconds). Distribution of the examples and homework problems between the three sets of units is about 50%, 40%, 10%. Prerequisites and Topical Organization A first course in Fluid Mechanics typically appears in the junior year of a traditional engineering curriculum. Students should have studied statics and dynamics, and mechanics of materials should be at least a co-requisite. Prior mathematics should include calculus, with at least the rudiments of vector calculus, and differential equations. In the first four chapters of this text the student is made aware of some fundamental aspects of fluid mechanics, including important fluid properties, flow regimes, pressure variation in fluids at rest and in motion, fluid kinematics, and methods of flow description and analysis. The Bernoulli equation is introduced in Chapter 3 to draw attention, early on, to some of the interesting effects and applications of the relationship between fluid motion and pressure in a flow field. We believe that this early consideration of elementary fluid dynamics increases student enthusiasm for the more complicated material that follows. In Chapter 4 we convey the essential elements of flow kinematics, including Eulerian and Lagrangian descriptions of flow fields, and indicate the vital relationship between the two views. For instructors who wish to consider kinematics in detail before the material on elementary fluid dynamics, Chapters 3 and 4 can be interchanged without loss of continuity. Chapters 5, 6, and 7 expand on the basic methods generally used to solve or to begin solving fluid mechanics problems. Emphasis is placed on understanding how flow phenomena are described mathematically and on when and how to use infinitesimal or finite control volumes. The effects of fluid friction on pressure and velocity are also considered in some detail. Although Chapter 5 xvi Preface considers fluid energy and energy dissipation, a formal course in thermodynamics is not a necessary prerequisite. Chapter 7 features the advantages of using dimensional analysis and similitude for organizing data and for planning experiments and the basic techniques involved. Owing to the growing importance of computational fluid dynamics (CFD) in engineering design and analysis, material on this subject is included in Appendix A. This material may be omitted without any loss of continuity to the rest of the text. Chapters 8 through 12 offer students opportunities for the further application of the principles learned earlier in the text. Also, where appropriate, additional important notions such as boundary layers, transition from laminar to turbulent flow, turbulence modeling, and flow separation are introduced. Practical concerns such as pipe flow, open-channel flow, flow measurement, drag and lift, the effects of compressibility, and the fundamental fluid mechanics of turbomachinery are included. Students who study this text and solve a representative set of the problems will have acquired a useful knowledge of the fundamentals of fluid mechanics. Faculty who use this text are provided with numerous topics to select from in order to meet the objectives of their own courses. More material is included than can be reasonably covered in one term. There is sufficient material for a second course, most likely titled “Applied Fluid Mechanics.” All are reminded of the fine collection of supplementary material. We have cited throughout the text various articles and books that are available for enrichment. Instructor Resources WileyPLUS provides instructor resources, such as the Instructor Solutions Manual, containing complete, detailed solutions to all of the problems in the text, and figures from the text appropriate for use in lecture slides. Sign up for access at www.wileyplus.com. Student Resources: Through a registration process, WileyPLUS also provides access to students for appropriate resources such as fluids videos and additional problems, among others. Harnessing the full power of WileyPLUS: If an educator chooses to require WileyPLUS for their course, the educator will set up the WileyPLUS course in advance and request students to register and use it. Students obtain access via a registration code that may be added to a print edition or purchased for online-only access. WileyPLUS builds students’ confidence because it takes the guesswork out of studying by providing students with a clear roadmap: what to do, how to do it, if they did it right. This interactive approach focuses on: CONFIDENCE: Research shows that students experience a great deal of anxiety over studying. That’s why we provide a structured learning environment that helps students focus on what to do, along with the support of immediate resources. MOTIVATION: To increase and sustain motivation throughout the semester, WileyPLUS helps students learn how to do it at a pace that’s right for them. Our integrated resources—available 24/7—function like a personal tutor, directly addressing each student’s demonstrated needs with specific problem-solving techniques. SUCCESS: WileyPLUS helps to assure that each study session has a positive outcome by putting students in control. Through instant feedback and study objective reports, students know if they did it right, and where to focus next, so they achieve the strongest results. With WileyPLUS, our efficacy research shows that students improve their outcomes by as much as one letter grade. WileyPLUS helps students take more initiative, so you’ll have greater impact on their achievement in the classroom and beyond. Preface xvii What do students receive with WileyPLUS? ■ ■ ■ ■ The complete digital textbook, saving students up to 60% off the cost of a printed text. Question assistance, including links to relevant sections in the online digital textbook. Immediate feedback and proof of progress, 24/7. Integrated, multimedia resources—including fluids phenomena videos, problem-solving videos, What An Engineer Sees animations, practice reading questions, and much more—that provide multiple study paths and encourage more active learning. What do instructors receive with WileyPLUS? ■ ■ ■ Reliable resources that reinforce course goals inside and outside of the classroom. The ability to easily identify those students who are falling behind. Media-rich course materials and assessment content including Instructor Solutions Manual, figures from the text appropriate for use in lecture slides, Fluids Phenomena Videos, autogradable Reading Questions that can be used in Pre-Lecture Quizzes, autogradable concept questions, autogradable Homework Problems, and much more. Sign up for access at www.wileyplus.com. Acknowledgments First, we wish to express our gratitude to Bruce Munson, Donald Young, Ted Okiishi, Wade Huebsch, and Alric Rothmayer for their part in producing seven editions of this excellent book. Also we thank the people at Wiley, especially Don Fowley, Linda Ratts, and Jenny Welter, for trusting us to assume responsibility for this text. Finally, we thank our families for their continued encouragement during the writing of this edition. Working with students and colleagues over the years has taught us much about fluid mechanics education. We have drawn from this experience for the benefit of users of this book. Obviously we are still learning, and we welcome any suggestions and comments from you. Philip M. Gerhart Andrew L. Gerhart John I. Hochstein Contents 1 INTRODUCTION Learning Objectives 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1 1 2.13 Chapter Summary and Study Guide 79 References 80, Problems 81 Some Characteristics of Fluids 3 Dimensions, Dimensional Homogeneity, and Units 4 1.2.1 Systems of Units 6 Analysis of Fluid Behavior 11 Measures of Fluid Mass and Weight 11 1.4.1 Density 11 1.4.2 Specific Weight 12 1.4.3 Specific Gravity 12 Ideal Gas Law 12 Viscosity 14 Compressibility of Fluids 20 1.7.1 Bulk Modulus 20 1.7.2 Compression and Expansion of Gases 21 1.7.3 Speed of Sound 22 Vapor Pressure 23 Surface Tension 24 A Brief Look Back in History 27 Chapter Summary and Study Guide 29 3 ELEMENTARY FLUID DYNAMICS— THE BERNOULLI EQUATION 101 Learning Objectives 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 References 30, Problems 31 2 FLUID STATICS Learning Objectives 2.1 2.2 2.3 40 40 Pressure at a Point 40 Basic Equation for Pressure Field 41 Pressure Variation in a Fluid at Rest 43 2.3.1 Incompressible Fluid 44 2.3.2 Compressible Fluid 46 2.4 Standard Atmosphere 48 2.5 Measurement of Pressure 50 2.6 Manometry 52 2.6.1 Piezometer Tube 52 2.6.2 U-Tube Manometer 53 2.6.3 Inclined-Tube Manometer 55 2.7 Mechanical and Electronic Pressure-Measuring Devices 56 2.8 Hydrostatic Force on a Plane Surface 59 2.9 Pressure Prism 65 2.10 Hydrostatic Force on a Curved Surface 68 2.11 Buoyancy, Flotation, and Stability 70 2.11.1 Archimedes’ Principle 70 2.11.2 Stability 73 2.12 Pressure Variation in a Fluid with Rigid-Body Motion 74 2.12.1 Linear Motion 75 2.12.2 Rigid-Body Rotation 77 3.9 101 Newton’s Second Law 101 F = ma along a Streamline 104 F = ma Normal to a Streamline 108 Physical Interpretations and Alternate Forms of the Bernoulli Equation 110 Static, Stagnation, Dynamic, and Total Pressure 113 Examples of Use of the Bernoulli Equation 117 3.6.1 Free Jets 118 3.6.2 Confined Flows 120 3.6.3 Flowrate Measurement 126 The Energy Line and the Hydraulic Grade Line 131 Restrictions on Use of the Bernoulli Equation 134 3.8.1 Compressibility Effects 134 3.8.2 Unsteady Effects 135 3.8.3 Rotational Effects 137 3.8.4 Other Restrictions 138 Chapter Summary and Study Guide 138 References 139, Problems 140 4 FLUID KINEMATICS 156 Learning Objectives 156 4.1 4.2 4.3 4.4 The Velocity Field 156 4.1.1 Eulerian and Lagrangian Flow Descriptions 159 4.1.2 One-, Two-, and ThreeDimensional Flows 160 4.1.3 Steady and Unsteady Flows 161 4.1.4 Streamlines, Streaklines, and Pathlines 161 The Acceleration Field 165 4.2.1 Acceleration and the Material Derivative 165 4.2.2 Unsteady Effects 168 4.2.3 Convective Effects 168 4.2.4 Streamline Coordinates 171 Control Volume and System Representations 173 The Reynolds Transport Theorem 175 4.4.1 Derivation of the Reynolds Transport Theorem 177 xix xx Contents 4.5 4.4.2 Physical Interpretation 182 4.4.3 Relationship to Material Derivative 182 4.4.4 Steady Effects 183 4.4.5 Unsteady Effects 184 4.4.6 Moving Control Volumes 185 4.4.7 Selection of a Control Volume 186 Chapter Summary and Study Guide 187 References 188, Problems 189 5 FINITE CONTROL VOLUME ANALYSIS 197 Learning Objectives 197 5.1 5.2 5.3 5.4 5.5 Conservation of Mass—The Continuity Equation 198 5.1.1 Derivation of the Continuity Equation 198 5.1.2 Fixed, Nondeforming Control Volume 200 5.1.3 Moving, Nondeforming Control Volume 206 5.1.4 Deforming Control Volume 208 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 211 5.2.1 Derivation of the Linear Momentum Equation 211 5.2.2 Application of the Linear Momentum Equation 212 5.2.3 Derivation of the Moment-ofMomentum Equation 226 5.2.4 Application of the Moment-ofMomentum Equation 227 First Law of Thermodynamics— The Energy Equation 234 5.3.1 Derivation of the Energy Equation 234 5.3.2 Application of the Energy Equation 237 5.3.3 The Mechanical Energy Equation and the Bernoulli Equation 241 5.3.4 Application of the Energy Equation to Nonuniform Flows 247 5.3.5 Combination of the Energy Equation and the Moment-ofMomentum Equation 250 Second Law of Thermodynamics— Irreversible Flow 251 5.4.1 Semi-infinitesimal Control Volume Statement of the Energy Equation 251 5.4.2 Semi-infinitesimal Control Volume Statement of the Second Law of Thermodynamics 251 5.4.3 Combination of the Equations of the First and Second Laws of Thermodynamics 252 Chapter Summary and Study Guide 253 References 255, Problems 256 6 DIFFERENTIAL ANALYSIS OF FLUID FLOW 277 Learning Objectives 277 6.1 Fluid Element Kinematics 278 6.1.1 Velocity and Acceleration Fields Revisited 279 6.1.2 Linear Motion and Deformation 279 6.1.3 Angular Motion and Deformation 280 6.2 Conservation of Mass 283 6.2.1 Differential Form of Continuity Equation 283 6.2.2 Cylindrical Polar Coordinates 286 6.2.3 The Stream Function 286 6.3 The Linear Momentum Equation 289 6.3.1 Description of Forces Acting on the Differential Element 290 6.3.2 Equations of Motion 292 6.4 Inviscid Flow 293 6.4.1 Euler’s Equations of Motion 293 6.4.2 The Bernoulli Equation 293 6.4.3 Irrotational Flow 295 6.4.4 The Bernoulli Equation for Irrotational Flow 297 6.4.5 The Velocity Potential 297 6.5 Some Basic, Plane Potential Flows 300 6.5.1 Uniform Flow 301 6.5.2 Source and Sink 302 6.5.3 Vortex 304 6.5.4 Doublet 307 6.6 Superposition of Basic, Plane Potential Flows 309 6.6.1 Source in a Uniform Stream— Half-Body 309 6.6.2 Rankine Ovals 312 6.6.3 Flow around a Circular Cylinder 314 6.7 Other Aspects of Potential Flow Analysis 319 6.8 Viscous Flow 320 6.8.1 Stress–Deformation Relationships 320 6.8.2 The Navier–Stokes Equations 321 6.9 Some Simple Solutions for Laminar, Viscous, Incompressible Flows 322 6.9.1 Steady, Laminar Flow between Fixed Parallel Plates 323 6.9.2 Couette Flow 325 6.9.3 Steady, Laminar Flow in Circular Tubes 327 6.9.4 Steady, Axial, Laminar Flow in an Annulus 330 6.10 Other Aspects of Differential Analysis 332 6.10.1 Numerical Methods 332 6.11 Chapter Summary and Study Guide 333 References 334, Problems 335 Contents 7 DIMENSIONAL ANALYSIS, SIMILITUDE, AND MODELING 8.4 346 Learning Objectives 346 7.1 7.2 7.3 7.4 The Need for Dimensional Analysis 347 Buckingham Pi Theorem 349 Determination of Pi Terms 350 Some Additional Comments about Dimensional Analysis 355 7.4.1 Selection of Variables 355 7.4.2 Determination of Reference Dimensions 357 7.4.3 Uniqueness of Pi Terms 358 7.5 Determination of Pi Terms by Inspection 359 7.6 Common Dimensionless Groups in Fluid Mechanics 361 7.7 Correlation of Experimental Data 366 7.7.1 Problems with One Pi Term 366 7.7.2 Problems with Two or More Pi Terms 367 7.8 Modeling and Similitude 370 7.8.1 Theory of Models 370 7.8.2 Model Scales 373 7.8.3 Practical Aspects of Using Models 374 7.9 Some Typical Model Studies 376 7.9.1 Flow through Closed Conduits 376 7.9.2 Flow around Immersed Bodies 378 7.9.3 Flow with a Free Surface 382 7.10 Similitude Based on Governing Differential Equations 385 7.11 Chapter Summary and Study Guide 388 8.5 8.6 8.7 9 FLOW OVER IMMERSED BODIES 482 Learning Objectives 482 9.1 9.2 401 Learning Objectives 401 8.1 8.2 8.3 General Characteristics of Pipe Flow 402 8.1.1 Laminar or Turbulent Flow 403 8.1.2 Entrance Region and Fully Developed Flow 405 8.1.3 Pressure and Shear Stress 406 Fully Developed Laminar Flow 407 8.2.1 From F = ma Applied Directly to a Fluid Element 408 8.2.2 From the Navier–Stokes Equations 412 8.2.3 From Dimensional Analysis 413 8.2.4 Energy Considerations 415 Fully Developed Turbulent Flow 417 8.3.1 Transition from Laminar to Turbulent Flow 417 8.3.2 Turbulent Shear Stress 419 8.3.3 Turbulent Velocity Profile 423 8.3.4 Turbulence Modeling 427 8.3.5 Chaos and Turbulence 427 Dimensional Analysis of Pipe Flow 427 8.4.1 Major Losses 428 8.4.2 Minor Losses 433 8.4.3 Noncircular Conduits 443 Pipe Flow Examples 446 8.5.1 Single Pipes 446 8.5.2 Multiple Pipe Systems 456 Pipe Flowrate Measurement 460 8.6.1 Pipe Flowrate Meters 460 8.6.2 Volume Flowmeters 465 Chapter Summary and Study Guide 466 References 467, Problems 468 References 389, Problems 390 8 VISCOUS FLOW IN PIPES xxi 9.3 9.4 9.5 General External Flow Characteristics 483 9.1.1 Lift and Drag Concepts 484 9.1.2 Characteristics of Flow Past an Object 487 Boundary Layer Characteristics 491 9.2.1 Boundary Layer Structure and Thickness on a Flat Plate 491 9.2.2 Prandtl/Blasius Boundary Layer Solution 495 9.2.3 Momentum Integral Boundary Layer Equation for a Flat Plate 499 9.2.4 Transition from Laminar to Turbulent Flow 504 9.2.5 Turbulent Boundary Layer Flow 506 9.2.6 Effects of Pressure Gradient 510 9.2.7 Momentum Integral Boundary Layer Equation with Nonzero Pressure Gradient 514 Drag 515 9.3.1 Friction Drag 516 9.3.2 Pressure Drag 517 9.3.3 Drag Coefficient Data and Examples 519 Lift 533 9.4.1 Surface Pressure Distribution 535 9.4.2 Circulation 541 Chapter Summary and Study Guide 545 References 546, Problems 547 10 OPEN-CHANNEL FLOW 559 Learning Objectives 559 10.1 General Characteristics of Open-Channel Flow 559 10.2 Surface Waves 561 10.2.1 Wave Speed 561 10.2.2 Froude Number Effects 564 Free ebooks ==> www.Ebook777.com xxii Contents 10.3 Energy Considerations 566 10.3.1 Energy Balance 566 10.3.2 Specific Energy 567 10.4 Uniform Flow 570 10.4.1 Uniform Flow Approximations 570 10.4.2 The Chezy and Manning Equations 571 10.4.3 Uniform Flow Examples 574 10.5 Gradually Varied Flow 579 10.6 Rapidly Varied Flow 581 10.6.1 The Hydraulic Jump 582 10.6.2 Sharp-Crested Weirs 587 10.6.3 Broad-Crested Weirs 590 10.6.4 Underflow (Sluice) Gates 593 10.7 Chapter Summary and Study Guide 594 11.10 Analogy between Compressible and OpenChannel Flows 667 11.11 Two-Dimensional Supersonic Flow 668 11.12 Chapter Summary and Study Guide 670 References 673, Problems 673 12 TURBOMACHINES 12.1 12.2 12.3 12.4 References 596, Problems 596 11 COMPRESSIBLE FLOW 604 Learning Objectives 604 11.1 11.2 11.3 11.4 11.5 11.6 11.7 11.8 11.9 Ideal Gas Thermodynamics 605 Stagnation Properties 610 Mach Number and Speed of Sound 612 Compressible Flow Regimes 617 Shock Waves 621 11.5.1 Normal Shock 622 Isentropic Flow 626 11.6.1 Steady Isentropic Flow of an Ideal Gas 627 11.6.2 Incompressible Flow and Bernoulli’s Equation 630 11.6.3 The Critical State 631 One-Dimensional Flow in a Variable Area Duct 632 11.7.1 General Considerations 632 11.7.2 Isentropic Flow of an Ideal Gas With Area Change 635 11.7.3 Operation of a Converging Nozzle 641 11.7.4 Operation of a Converging– Diverging Nozzle 643 Constant-Area Duct Flow With Friction 647 11.8.1 Preliminary Consideration: Comparison with Incompressible Duct Flow 647 11.8.2 The Fanno Line 648 11.8.3 Adiabatic Frictional Flow (Fanno Flow) of an Ideal Gas 652 Frictionless Flow in a Constant-Area Duct with Heating or Cooling 660 11.9.1 The Rayleigh Line 660 11.9.2 Frictionless Flow of an Ideal Gas with Heating or Cooling (Rayleigh Flow) 662 11.9.3 Rayleigh Lines, Fanno Lines, and Normal Shocks 666 679 Learning Objectives 679 12.5 12.6 12.7 12.8 12.9 12.10 Introduction 680 Basic Energy Considerations 681 Angular Momentum Considerations 685 The Centrifugal Pump 687 12.4.1 Theoretical Considerations 688 12.4.2 Pump Performance Characteristics 692 12.4.3 Net Positive Suction Head (NPSH) 694 12.4.4 System Characteristics, Pump-System Matching, and Pump Selection 696 Dimensionless Parameters and Similarity Laws 700 12.5.1 Special Pump Scaling Laws 702 12.5.2 Specific Speed 703 12.5.3 Suction Specific Speed 704 Axial-Flow and Mixed-Flow Pumps 705 Fans 707 Turbines 708 12.8.1 Impulse Turbines 709 12.8.2 Reaction Turbines 716 Compressible Flow Turbomachines 719 12.9.1 Compressors 719 12.9.2 Compressible Flow Turbines 723 Chapter Summary and Study Guide 725 References 726, Problems 727 A Computational Fluid Dynamics B Physical Properties of Fluids 737 756 C Properties of the U.S. Standard Atmosphere 761 D Compressible Flow Functions for an Ideal Gas 763 E Comprehensive Table of Conversion Factors 771 Answers See WileyPLUS for this material Index I-1 www.Ebook777.com 1 Introduction Learning Objectives After completing this chapter, you should be able to: ■ list the dimensions and units of physical quantities. ■ identify the key fluid properties used in the analysis of fluid behavior. ■ calculate values for common fluid properties given appropriate information. ■ explain effects of fluid compressibility. ■ use the concepts of viscosity, vapor pressure, and surface tension. 𝓵 p V (Photograph courtesy of CIRRUS Design Corporation.) Fluid mechanics is the discipline within the broad field of applied mechanics that is concerned with the behavior of liquids and gases at rest or in motion. It covers a vast array of phenomena that occur in nature (with or without human intervention), in biology, and in numerous engineered, invented, or manufactured situations. There are few aspects of our lives that do not involve fluids, either directly or indirectly. The immense range of different flow conditions is mind-boggling and strongly dependent on the value of the numerous parameters that describe fluid flow. Among the long list of parameters involved are (1) the physical size of the flow, ℓ; (2) the speed of the flow, V; and (3) the pressure, p, as indicated in the figure in the margin for a light aircraft parachute recovery system. These are just three of the important parameters that, along with many others, are discussed in detail in various sections of this book. To get an inkling of the range of some of the parameter values involved and the flow situations generated, consider the following. ■ ■ Size, ℓ Every flow has a characteristic (or typical) length associated with it. For example, for flow of fluid within pipes, the pipe diameter is a characteristic length. Pipe flows include the flow of water in the pipes in our homes, the blood flow in our arteries and veins, and the airflow in our bronchial tree. They also involve pipe sizes that are not within our everyday experiences. Such examples include the flow of oil across Alaska through a 4-foot-diameter, 799-mile-long pipe and, at the other end of the size scale, the new area of interest involving flow in nano scale pipes whose diameters are on the order of 10−8 m. Each of these pipe flows has important characteristics that are not found in the others. Characteristic lengths of some other flows are shown in Fig. 1.1a. Speed, V As we note from The Weather Channel, on a given day the wind speed may cover what we think of as a wide range, from a gentle 5-mph breeze to a 100-mph hurricane or a 250-mph 1 2 Chapter 1 ■ Introduction VIDEO V1.1 Mt. St. Helens eruption ■ tornado. However, this speed range is small compared to that of the almost imperceptible flow of the fluid-like magma below the Earth’s surface that drives the continental drift motion of the tectonic plates at a speed of about 2 × 10−8 m/s or the hypersonic airflow around a meteor as it streaks through the atmosphere at 3 × 104 m/s. Characteristic speeds of some other flows are shown in Fig. 1.1b. Pressure, p The pressure within fluids covers an extremely wide range of values. We are accustomed to the 35 psi (lb/in.2) pressure within our car’s tires, the “120 over 70” typical blood pressure reading, or the standard 14.7 psi atmospheric pressure. However, the large 10,000 psi pressure in the hydraulic ram of an earth mover or the tiny 2 × 10−6 psi pressure of a sound wave generated at ordinary talking levels are not easy to comprehend. Characteristic pressures of some other flows are shown in Fig. 1.1c. The list of fluid mechanics applications goes on and on. But you get the point. Fluid mechanics is a very important, practical subject that encompasses a wide variety of situations. It is very likely that during your career as an engineer you will be involved in the analysis and design of systems that require a good understanding of fluid mechanics. Although it is not possible to adequately cover all of the important areas of fluid mechanics within one book, it is hoped that this introductory text will provide a sound foundation of the fundamental aspects of fluid mechanics. V1.2 E. coli swimming 108 Jupiter red spot diameter 106 Ocean current diameter 106 Diameter of hurricane Meteor entering atmosphere Mt. St. Helens plume 104 Average width of middle Mississippi River 102 102 Boeing 787 NACA Ames wind tunnel 10-2 Water pipe diameter 100 10-2 10-6 Water jet cutter width Amoeba Thickness of lubricating oil layer in journal bearing Diameter of smallest blood vessel 102 Mississippi River Syrup on pancake Raindrop 10-4 104 Water from fire hose nozzle Flow past bike rider Diameter of Space Shuttle main engine exhaust jet Outboard motor prop Rocket nozzle exhaust Speed of sound in air Tornado V, m/s 100 106 Space Shuttle reentry 10-4 Microscopic swimming animal 10-6 Glacier flow p, lb/in.2 104 𝓵, m VIDEO 100 Nano scale devices (a) Fire hydrant Auto tire Standard atmosphere “Excess pressure” on hand held out of car traveling 60 mph Atmospheric pressure on Mars 10-2 Pressure change causing ears to “pop” in elevator Pressure at 40-mile altitude 10-4 Artificial kidney filter pore size 10-8 Water jet cutting Mariana Trench in Pacific Ocean Hydraulic ram Scuba tank Car engine combustion Vacuum pump 10-8 Continental drift (b) 10-6 Sound pressure at normal talking (c) ■ Figure 1.1 Characteristic values of some fluid flow parameters for a variety of flows: (a) object size, (b) fluid speed, (c) fluid pressure. 1.1 1.1 Some Characteristics of Fluids 3 Some Characteristics of Fluids Both liquids and gases are fluids. One of the first questions we need to explore is––what is a fluid? Or we might ask–what is the difference between a solid and a fluid? We have a general, vague idea of the difference. A solid is “hard” and not easily deformed, whereas a fluid is “soft” and is easily deformed (we can readily move through air). Although quite descriptive, these casual observations of the differences between solids and fluids are not very satisfactory from a scientific or engineering point of view. A closer look at the molecular structure of materials reveals that matter that we commonly think of as a solid (steel, concrete, etc.) has densely spaced molecules with large intermolecular cohesive forces that allow the solid to maintain its shape, and to not be easily deformed. However, for matter that we normally think of as a liquid (water, oil, etc.), the molecules are spaced farther apart, the intermolecular forces are smaller than for solids, and the molecules have more freedom of movement. Thus, liquids can be easily deformed (but not easily compressed) and can be poured into containers or forced through a tube. Gases (air, oxygen, etc.) have even greater molecular spacing and freedom of motion with negligible cohesive intermolecular forces, and as a consequence are easily deformed (and compressed) and will completely fill the volume of any container in which they are placed. Both liquids and gases are fluids. THE WIDE WORLD OF FLUIDS Will what works in air work in water? For the past few years submarine must be designed to withstand external pressures of a San Francisco company has been working on small, maneuver- nearly 700 pounds per square inch greater than that inside the able submarines designed to travel through water using wings, vehicle. On the other hand, at high altitude where commercial controls, and thrusters that are similar to those on jet airplanes. jets fly, the exterior pressure is 3.5 psi rather than standard sea- After all, water (for submarines) and air (for airplanes) are both level pressure of 14.7 psi, so the vehicle must be pressurized fluids, so it is expected that many of the principles governing the internally for passenger comfort. In both cases, however, the flight of airplanes should carry over to the “flight” of winged design of the craft for minimal drag, maximum lift, and efficient submarines. Of course, there are differences. For example, the thrust is governed by the same fluid dynamic concepts. F Surface Although the differences between solids and fluids can be explained qualitatively on the basis of molecular structure, a more specific distinction is based on how they deform under the action of an external load. Specifically, a fluid is defined as a substance that deforms continuously when acted on by a shearing stress of any magnitude. A shearing stress (force per unit area) is created whenever a tangential force acts on a surface as shown by the figure in the margin. When common solids such as steel or other metals are acted on by a shearing stress, they will initially deform (usually a very small deformation), but they will not continuously deform (flow). However, common fluids such as water, oil, and air satisfy the definition of a fluid—that is, they will flow when acted on by a shearing stress. Some materials, such as slurries, tar, putty, toothpaste, and so on, are not easily classified since they will behave as a solid if the applied shearing stress is small, but if the stress exceeds some critical value, the substance will flow. The study of such materials is called rheology and does not fall within the province of classical fluid mechanics. Thus, all the fluids we will be concerned with in this text will conform to the definition of a fluid. Although the molecular structure of fluids is important in distinguishing one fluid from another, it is not yet practical to study the behavior of individual molecules when trying to describe the behavior of fluids at rest or in motion. Rather, we characterize the behavior by considering the average, or macroscopic, value of the quantity of interest, where the average is evaluated over a small volume containing a large number of molecules. Thus, when we say that the velocity at a certain point in a fluid is so much, we are really indicating the average velocity of the molecules in a small volume surrounding the point. The volume is small compared with the physical dimensions of the system of interest, but large compared with the average distance between molecules. Is this a reasonable way to describe the behavior of a fluid? The answer is generally yes, since the spacing between molecules is typically very small. For gases at normal pressures and temperatures, the spacing is on the order of 10−6 mm, and for liquids it is on the order of 10−7 mm. The number of 4 Chapter 1 ■ Introduction molecules per cubic millimeter is on the order of 10 18 for gases and 10 21 for liquids. It is thus clear that the number of molecules in a very tiny volume is huge and the idea of using average values taken over this volume is certainly reasonable. We thus assume that all the fluid characteristics we are interested in (pressure, velocity, etc.) vary continuously throughout the fluid—that is, we treat the fluid as a continuum and we refer to the very small volume as a point in the flow. This concept will certainly be valid for all the circumstances considered in this text. One area of fluid mechanics for which the continuum concept breaks down is in the study of rarefied gases such as would be encountered at very high altitudes. In this case the spacing between air molecules can become large and the continuum concept is no longer acceptable. 1.2 Fluid characteristics can be described qualitatively in terms of certain basic quantities such as length, time, and mass. Dimensions, Dimensional Homogeneity, and Units Since in our study of fluid mechanics we will be dealing with a variety of fluid characteristics, it is necessary to develop a system for describing these characteristics both qualitatively and quantitatively. The qualitative aspect serves to identify the nature, or type, of the characteristics (such as length, time, stress, and velocity), whereas the quantitative aspect provides a numerical measure of the characteristics. The quantitative description requires both a number and a standard by which various quantities can be compared. A standard for length might be a meter or foot, for time an hour or second, and for mass a slug or kilogram. Such standards are called units, and several systems of units are in common use as described in the following section. The qualitative description is conveniently given in terms of certain primary quantities, such as length, L, time, T, mass, M, and temperature, ϴ. These primary quantities can then be used to provide a qualitative description of any other secondary quantity: for example, area ≐ L 2, velocity ≐ LT −1, density ≐ ML−3, and so on, where the symbol ≐ is used to indicate the dimensions of the secondary quantity in terms of the primary quantities. Thus, to describe qualitatively a velocity, V, we would write V ≐ LT −1 and say that “the dimensions of a velocity equal length divided by time.” The primary quantities are also referred to as basic dimensions. For a wide variety of problems involving fluid mechanics, only the three basic dimensions, L, T, and M are required. Alternatively, L, T, and F could be used, where F is the basic dimensions of force. Since Newton’s law states that force is equal to mass times acceleration, it follows that F ≐ MLT −2 or M ≐ FL−1 T 2. Thus, secondary quantities expressed in terms of M can be expressed in terms of F through the relationship above. For example, stress, σ, is a force per unit area, so that σ ≐ FL−2, but an equivalent dimensional equation is σ ≐ ML−1T −2. Table 1.1 provides a list of dimensions for a number of common physical quantities. All theoretically derived equations are dimensionally homogeneous—that is, the dimensions of the left side of the equation must be the same as those on the right side, and all additive separate terms must have the same dimensions. We accept as a fundamental premise that all equations describing physical phenomena must be dimensionally homogeneous. If this were not true, we would be attempting to equate or add unlike physical quantities, which would not make sense. For example, the equation for the velocity, V, of a uniformly accelerated body is V = V0 + at (1.1) where V0 is the initial velocity, a the acceleration, and t the time interval. In terms of dimensions the equation is LT −1 ≐ LT −1 + LT −2T and thus Eq. 1.1 is dimensionally homogeneous. Some equations that are known to be valid contain constants having dimensions. The equation for the distance, d, traveled by a freely falling body can be written as d = 16.1t 2 (1.2) 1.2 5 Dimensions, Dimensional Homogeneity, and Units ■ Table 1.1 Dimensions Associated with Common Physical Quantities FLT System MLT System Acceleration Angle Angular acceleration Angular velocity Area LT −2 F 0L0T 0 T −2 T −1 L2 LT −2 M 0L0T 0 T −2 T −1 L2 Density Energy Force Frequency Heat FL−4T 2 FL F T −1 FL ML−3 ML2T −2 MLT −2 T −1 ML2T −2 Length Mass Modulus of elasticity Moment of a force Moment of inertia (area) L FL−1T 2 FL−2 FL L4 L M ML−1T −2 ML2T −2 L4 Moment of inertia (mass) Momentum FLT 2 FT ML2 MLT −1 FLT System MLT System Power Pressure Specific heat FLT −1 FL−2 L2T −2 ϴ −1 ML2T −3 ML−1T −2 L2T −2 ϴ −1 Specific weight Strain Stress Surface tension Temperature FL−3 F 0L0 T 0 FL−2 FL−1 ϴ ML−2 T −2 M 0L0 T 0 ML−1T −2 MT −2 ϴ Time Torque Velocity Viscosity (dynamic) Viscosity (kinematic) T FL LT −1 FL−2 T L2 T −1 T ML2T −2 LT −1 ML−1T −1 L2T −1 Volume Work L3 FL L3 ML2T −2 and a check of the dimensions reveals that the constant must have the dimensions of LT −2 if the equation is to be dimensionally homogeneous. Actually, Eq. 1.2 is a special form of the well-known equation from physics for freely falling bodies, d= General homogeneous equations are valid in any system of units. gt 2 2 (1.3) in which g is the acceleration of gravity. Equation 1.3 is dimensionally homogeneous and valid in any system of units. For g = 32.2 fts2 the equation reduces to Eq. 1.2 and thus Eq. 1.2 is valid only for the system of units using feet and seconds. Equations that are restricted to a particular system of units can be denoted as restricted homogeneous equations, as opposed to equations valid in any system of units, which are general homogeneous equations. The preceding discussion indicates one rather elementary, but important, use of the concept of dimensions: the determination of one aspect of the generality of a given equation simply based on a consideration of the dimensions of the various terms in the equation. The concept of dimensions also forms the basis for the powerful tool of dimensional analysis, which is considered in detail in Chapter 7. Note to the users of this text. All of the examples in the text use a consistent problemsolving methodology, which is similar to that in other engineering courses such as statics. Each example highlights the key elements of analysis: Given, Find, Solution, and Comment. The Given and Find are steps that ensure the user understands what is being asked in the problem and explicitly list the items provided to help solve the problem. The Solution step is where the equations needed to solve the problem are formulated and the problem is actually solved. In this step, there are typically several other tasks that help to set up the solution and are required to solve the problem. The first is a drawing of the problem; where appropriate, it is always helpful to draw a sketch of the problem. Here the relevant geometry and coordinate system to be used as well as features such as control volumes, forces and pressures, velocities, and mass flow rates are included. This helps in gaining an understanding of the problem. Making appropriate assumptions to solve the problem is the second task. In a realistic engineering problemsolving environment, the necessary assumptions are developed as an integral part of the solution process. Assumptions can provide appropriate simplifications or offer useful constraints, both of 6 Chapter 1 ■ Introduction which can help in solving the problem. Throughout the examples in this text, the necessary assumptions are embedded within the Solution step, as they are in solving a real-world problem. This provides a realistic problem-solving experience. The final element in the methodology is the Comment. For the examples in the text, this section is used to provide further insight into the problem or the solution. It can also be a point in the analysis at which certain questions are posed. For example: Is the answer reasonable, and does it make physical sense? Are the final units correct? If a certain parameter were changed, how would the answer change? Adopting this type of methodology will aid in the development of problemsolving skills for fluid mechanics, as well as other engineering disciplines. EXAMPLE 1.1 Restricted and General Homogeneous Equations GIVEN A liquid flows through an orifice located in the side of a tank as shown in Fig. E1.1. A commonly used equation for determining the volume rate of flow, Q, through the orifice is h Q = 0.61 A √2gh Q where A is the area of the orifice, g is the acceleration of gravity, and h is the height of the liquid above the orifice. A FIND Investigate the dimensional homogeneity of this formula. (b) (a) ■ Figure E1.1 SOLUTION The dimensions of the various terms in the equation are Q = volume/time ≐ L3T −1, A = area ≐ L2, g = acceleration of gravity ≐ LT −2, and h = height ≐ L. and, therefore, the equation expressed as Eq. 1 can only be dimensionally correct if the number 4.90 has the dimensions of L1 2T −1. Whenever a number appearing in an equation or formula has dimensions, it means that the specific value of the number will depend on the system of units used. Thus, for the case being considered with feet and seconds used as units, the number 4.90 has units of ft1 2s. Equation 1 will only give the correct value for Q (in ft3s) when A is expressed in square feet and h in feet. Thus, Eq. 1 is a restricted homogeneous equation, whereas the original equation is a general homogeneous equation that would be valid for any consistent system of units.   These terms, when substituted into the equation, yield the dimensional form: (L3T −1 ) ≐ (0.61) (L2 ) ( √2 ) (LT −2 ) 1 2 (L) 1 2       or (L3T −1 ) ≐ [0.61√2](L3T −1 ) It is clear from this result that the equation is dimensionally homogeneous (both sides of the formula have the same dimensions of L3T −1 ), and the number 0.61 √2 is dimensionless. If we were going to use this relationship repeatedly, we might be tempted to simplify it by replacing g with its standard value of 32.2 ft s2 and rewriting the formula as Q = 4.90 A √h (1) A quick check of the dimensions reveals that L3T −1 ≐ (4.90) (L5 2 )   COMMENT A quick check of the dimensions of the various terms in an equation is a useful practice and will often be helpful in eliminating errors—that is, as noted previously, all physically meaningful equations must be dimensionally homogeneous. We have briefly alluded to units in this example, and this important topic will be considered in more detail in the next section. 1.2.1 Systems of Units In addition to the qualitative description of the various quantities of interest, it is generally necessary to have a quantitative measure of any given quantity. For example, if we measure the width of this page in the book and say that it is 10 units wide, the statement has no meaning until the unit of length is defined. If we indicate that the unit of length is a meter, and define the meter as some standard length, a unit system for length has been established (and a numerical value can be given to the page width). In addition to length, a unit must be established for each of the remaining basic 1.2 7 Dimensions, Dimensional Homogeneity, and Units quantities (force, mass, time, and temperature). There are several systems of units in use, and we shall consider three systems that are commonly used in engineering. International System (SI). In 1960 the Eleventh General Conference on Weights and Measures, the international organization responsible for maintaining precise uniform standards of measurements, formally adopted the International System of Units as the international standard. This system, commonly termed SI, has been widely adopted worldwide and is widely used (although certainly not exclusively) in the United States. It is expected that the long-term trend will be for all countries to accept SI as the accepted standard and it is imperative that engineering students become familiar with this system. In SI the unit of length is the meter (m), the time unit is the second (s), the mass unit is the kilogram (kg), and the temperature unit is the kelvin (K). Note that there is no degree symbol used when expressing a temperature in kelvin units. The kelvin temperature scale is an absolute scale and is related to the Celsius (centigrade) scale (°C) through the relationship K = °C + 273.15 Although the Celsius scale is not in itself part of SI, it is common practice to specify temperatures in degrees Celsius when using SI units. The force unit, called the newton (N), is defined from Newton’s second law as 1 N = (1 kg)(1 m s2 ) In mechanics it is very important to distinguish between weight and mass. Thus, a 1-N force acting on a 1-kg mass will give the mass an acceleration of 1 ms2. Standard gravity in SI is 9.807 ms2 (commonly approximated as 9.81 ms2) so that a 1-kg mass weighs 9.81 N under standard gravity. Note that weight and mass are different, both qualitatively and quantitatively! The unit of work in SI is the joule (J), which is the work done when the point of application of a 1-N force is displaced through a 1-m distance in the direction of a force. Thus, 1J=1N∙m The unit of power is the watt (W) defined as a joule per second. Thus, 1 W = 1 Js = 1 N ∙ ms Prefixes for forming multiples and fractions of SI units are given in Table 1.2. For example, the notation kN would be read as “kilonewtons” and stands for 103 N. Similarly, mm would be read as “millimeters” and stands for 10−3 m. The centimeter is not an accepted unit of length in the SI system, so for most problems in fluid mechanics in which SI units are used, lengths will be expressed in millimeters or meters. British Gravitational (BG) System. In the BG system the unit of length is the foot (ft), the time unit is the second (s), the force unit is the pound (lb), and the temperature unit is the degree Fahrenheit (°F) or the absolute temperature unit is the degree Rankine (°R), where °R = °F + 459.67 ■ Table 1.2 Prefixes for SI Units Factor by Which Unit Is Multiplied Prefix Symbol 1015 1012 109 106 103 102 10 10−1 peta tera giga mega kilo hecto deka deci P T G M k h da d Factor by Which Unit Is Multiplied Prefix Symbol 10−2 10−3 10−6 10−9 10−12 10−15 10−18 centi milli micro nano pico femto atto c m μ n p f a 8 Chapter 1 ■ Introduction The mass unit, called the slug, is defined from Newton’s second law (force = mass × acceleration) as 1 lb = (1 slug)(1 ft s2 ) Two systems of units that are widely used in engineering are the British Gravitational (BG) System and the International System (SI). This relationship indicates that a 1-lb force acting on a mass of 1 slug will give the mass an acceleration of 1 fts2. The weight, 𝒲 (which is the force due to gravity, g), of a mass, m, is given by the equation 𝒲 = mg and in BG units 𝒲(lb) = m (slugs) g(fts2 ) Since Earth’s standard gravity is taken as g = 32.174 fts2 (commonly approximated as 32.2 fts2 ), it follows that a mass of 1 slug weighs 32.2 lb under standard gravity. THE WIDE WORLD OF FLUIDS How long is a foot? Today, in the United States, the common French commission proposed that a new universal length unit length unit is the foot, but throughout antiquity the unit used called a meter (metre) be defined as the distance of one- to measure length has quite a history. The first length units quarter of the Earth’s meridian (north pole to the equator) were based on the lengths of various body parts. One of the divided by 10 million. Although controversial, the meter was earliest units was the Egyptian cubit, first used around 3000 B.C. accepted in 1799 as the standard. With the development of and defined as the length of the arm from elbow to extended advanced technology, the length of a meter was redefined in fingertips. Other measures followed, with the foot simply taken 1983 as the distance traveled by light in a vacuum during the as the length of a man’s foot. Since this length obviously varies time interval of 1/299,792,458 s. The foot is now defined as from person to person it was often “standardized” by using the 0.3048 meter. Our simple rulers and yardsticks indeed have an length of the current reigning royalty’s foot. In 1791 a special intriguing history. English Engineering (EE) System. In the EE system, units for force and mass are defined independently; thus special care must be exercised when using this system in conjunction with Newton’s second law. The basic unit of mass is the pound mass (lbm), and the unit of force is the pound (lb).1 The unit of length is the foot (ft), the unit of time is the second (s), and the absolute temperature scale is the degree Rankine (°R). To make the equation expressing Newton’s second law dimensionally homogeneous we write it as F= ma gc (1.4) where gc is a constant of proportionality, which allows us to define units for both force and mass. For the BG system, only the force unit was prescribed and the mass unit defined in a consistent manner such that gc = 1. Similarly, for SI the mass unit was prescribed and the force unit defined in a consistent manner such that gc = 1. For the EE system, a 1-lb force is defined as that force which gives a 1 lbm a standard acceleration of gravity, which is taken as 32.174 fts2. Thus, for Eq. 1.4 to be both numerically and dimensionally correct 1 lb = (1 lbm) (32.174 fts2 ) gc so that gc = 1 (1 lbm) (32.174 fts2 ) (1 lb) It is also common practice to use the notation, lbf, to indicate pound force. 1.2 1.0 m 3 ft 1.0 0.06 kg 2 0.5 0.5 0 Length lbm 1.0 N 9 0.2 lb 0.04 1 0 slug 2 Dimensions, Dimensional Homogeneity, and Units 1 0.5 0.1 0.02 0 0 Mass 0 0 Force ■ Figure 1.2 Comparison of SI, BG, and EE units. With the EE system, weight and mass are related through the equation 𝒲= mg gc where g is the local acceleration of gravity. Under conditions of standard gravity (g = gc ) the weight in pounds and the mass in pound mass are numerically equal. Also, since a 1-lb force gives a mass of 1 lbm an acceleration of 32.174 ft s2 and a mass of 1 slug an acceleration of 1 ft s2, it follows that 1 slug = 32.174 lbm When solving problems it is important to use a consistent system of units, e.g., don’t mix BG and SI units. We cannot overemphasize the importance of paying close attention to units when solving problems. It is very easy to introduce huge errors into problem solutions through the use of incorrect units. Get in the habit of using a consistent system of units throughout a given solution. It really makes no difference which system you use as long as you are consistent; for example, don’t mix slugs and newtons. If problem data are specified in SI units, then use SI units throughout the solution. If the data are specified in BG units, then use BG units throughout the solution. The relative sizes of the SI, BG, and EE units of length, mass, and force are shown in Fig. 1.2. Extensive tables of conversion factors between unit systems, and within unit systems, are provided in Appendix E. For your convenience, abbreviated tables of conversion factors for some quantities commonly encountered in fluid mechanics are presented in Tables 1.3 and 1.4 on the inside back cover (using a slightly different format than Appendix E). Note that numbers in these tables are presented in computer exponential notation. For example, the number 5.154 E+2 is the number 5.154×102 in scientific notation. You should note that each conversion factor can be thought of as a fraction in which the numerator and denominator are equivalent. For example, an entry for “Length” from Table 1.4 instructs the user “To convert from … m … to … ft … Multiply by 3.281.” Therefore 1 m is the same length as 3.281 ft. Therefore a fraction formed with a numerator of 1 m and a denominator of 3.281 ft is the very definition of a fraction with a value of one, as is its reciprocal. This may seem obvious when the units of the denominator and numerator are of the same dimension. It is equally true for the more complicated conversion factors that include multiple dimensions and therefore a greater number of units. You already know that you can multiply any quantity by one without changing its value. Likewise, you can multiply (or divide) any quantity by any conversion factor in the tables, provided you use both the number and the units. The result will not be incorrect, even if it does not yield the result you hoped for. 10 Chapter 1 ■ Introduction EXAMPLE 1.2 BG and SI Units GIVEN A tank of liquid having a total mass of 36 kg rests on a support in the equipment bay of the Space Shuttle. FIND Determine the force (in newtons) that the tank exerts on the support shortly after lift off when the shuttle is accelerating upward as shown in Fig. E1.2a at 15 ft s2. SOLUTION A free-body diagram of the tank is shown in Fig. E1.2b, where 𝒲 is the weight of the tank and liquid, and Ff is the reaction of the floor on the tank. Application of Newton’s second law of motion to this body gives ∑ F = ma or Ff − 𝒲 = ma (1) where we have taken upward as the positive direction. Since 𝒲 = mg, Eq. 1 can be written as Ff = m (g + a) (2) Before substituting any number into Eq. 2, we must decide on a system of units, and then be sure all of the data are expressed in these units. Since we want Ff in newtons, we will use SI units so that Ff = 36 kg [9.81 m s2 + (15 ft s2 ) (0.3048 m ft)] = 518 kg ∙ m s2 ■ Figure E1.2a (Photograph courtesy of NASA.) Since 1 N = 1 kg ∙ m s2 , it follows that Ff = 518 N  (downward on floor) (Ans) The direction is downward since the force shown on the freebody diagram is the force of the support on the tank so that the force the tank exerts on the support is equal in magnitude but opposite in direction. 𝒲 a COMMENT As you work through a large variety of problems in this text, you will find that units play an essential role in arriving at a numerical answer. Be careful! It is easy to mix units and cause large errors. If in the above example the acceleration had been left as 15 fts2 with m and g expressed in SI units, we would have calculated the force as 893 N and the answer would have been 72% too large! Ff ■ Figure E1.2b THE WIDE WORLD OF FLUIDS Units and space travel A NASA spacecraft, the Mars Climate maneuvering commands sent from earth caused the Orbiter to Orbiter, was launched in December 1998 to study the Martian sweep within 37 miles of the surface rather than the intended geography and weather patterns. The spacecraft was slated to 93 miles. The subsequent investigation revealed that the errors begin orbiting Mars on September 23, 1999. However, NASA offi- were due to a simple mix-up in units. One team controlling the cials lost communication with the spacecraft early that day and Orbiter used SI units, whereas another team used BG units. This it is believed that the spacecraft broke apart or overheated costly experience illustrates the importance of using a consis- because it came too close to the surface of Mars. Errors in the tent system of units. 1.4 1.3 Measures of Fluid Mass and Weight 11 Analysis of Fluid Behavior The study of fluid mechanics involves the same fundamental laws you have encountered in physics and other mechanics courses. These laws include Newton’s laws of motion, conservation of mass, and the first and second laws of thermodynamics. Thus, there are strong similarities between the general approach to fluid mechanics and to rigid-body and deformable-body solid mechanics. This is indeed helpful since many of the concepts and techniques of analysis used in fluid mechanics will be ones you have encountered before in other courses. The broad subject of fluid mechanics can be generally subdivided into fluid statics, in which the fluid is at rest, and fluid dynamics, in which the fluid is moving. In the following chapters we will consider both of these areas in detail. Before we can proceed, however, it will be necessary to define and discuss certain fluid properties that are intimately related to fluid behavior. It is obvious that different fluids can have grossly different characteristics. For example, gases are light and compressible, whereas liquids are heavy (by comparison) and relatively incompressible. A syrup flows slowly from a container, but water flows rapidly when poured from the same container. To quantify these differences, certain fluid properties are used. In the following several sections, properties that play an important role in the analysis of fluid behavior are considered. 1.4 Measures of Fluid Mass and Weight 1.4.1 Density The density of a fluid, designated by the Greek symbol ρ (rho), is defined as its mass per unit volume. Density is typically used to characterize the mass of a fluid system. In the BG system, ρ has units of slugsft3 and in SI the units are kgm3. The value of density can vary widely between different fluids, but for liquids, variations in pressure and temperature generally have only a small effect on the value of ρ. The small change in the density of water with large variations in temperature is illustrated in Fig. 1.3. Tables 1.5 and 1.6 list values of density for several common liquids. The density of water at 60 °F is 1.94 slugsft3 or 999 kgm3. The large numerical difference between those two values illustrates the importance of paying attention to units! Unlike liquids, the density of a gas is strongly influenced by both pressure and temperature, and this difference will be discussed in the next section. The specific volume, υ, is the volume per unit mass and is therefore the reciprocal of the density—that is, υ= 1 ρ (1.5) This property is not commonly used in fluid mechanics but is used in thermodynamics. 1000 990 Density, ρ kg/m3 The density of a fluid is defined as its mass per unit volume. @ 4°C ρ = 1000 kg/m3 980 970 960 950 0 20 40 60 Temperature, °C ■ Figure 1.3 Density of water as a function of temperature. 80 100 12 Chapter 1 ■ Introduction 1.4.2 Specific Weight Specific weight is weight per unit volume; specific gravity is the ratio of fluid density to the density of water at a certain temperature. The specific weight of a fluid, designated by the Greek symbol γ (gamma), is defined as its weight per unit volume. Thus, specific weight is related to density through the equation γ = ρg (1.6) where g is the local acceleration of gravity. Just as density is used to characterize the mass of a fluid system, the specific weight is used to characterize the weight of the system. In the BG system, γ has units of lbft3 and in SI the units are Nm3. Under conditions of standard gravity (g = 32.174 fts2 = 9.807 ms2 ), water at 60 °F has a specific weight of 62.4 lbft3 and 9.80 kNm3. Tables 1.5 and 1.6 list values of specific weight for several common liquids (based on standard gravity). More complete tables for water can be found in Appendix B (Tables B.1 and B.2). 1.4.3 Specific Gravity The specific gravity of a fluid, designated as SG, is defined as the ratio of the density of the fluid to the density of water at some specified temperature. Usually the specified temperature is taken as 4 °C (39.2 °F), and at this temperature the density of water is 1.94 slugsft3 or 1000 kgm3. In equation form, specific gravity is expressed as SG = 13.55 Water ρ ρH2O@4 °C (1.7) and since it is the ratio of densities, the value of SG does not depend on the system of units used. For example, the specific gravity of mercury at 20 °C is 13.55. This is illustrated by the figure in the margin. Thus, the density of mercury can be readily calculated in either BG or SI units through the use of Eq. 1.7 as ρHg = (13.55)(1.94 slugsft3 ) = 26.3 slugsft3 Mercury 1 1.5 or ρHg = (13.55) (1000 kgm3 ) = 13.6 × 103 kgm3 It is clear that density, specific weight, and specific gravity are all interrelated, and from a knowledge of any one of the three the others can be calculated. Ideal Gas Law Gases are highly compressible in comparison to liquids, with changes in gas density directly related to changes in pressure and temperature through the equation ρ= p RT or, in the more standard form, p = ρRT, In the ideal gas law, absolute pressures and temperatures must be used. (1.8) where p is the absolute pressure, ρ the density, T the absolute temperature,2 and R is a gas constant. Equation 1.8 is commonly termed the ideal or perfect gas law, or the equation of state for an ideal gas. It is known to closely approximate the behavior of real gases under typical conditions when the gases are not approaching liquefaction. Pressure in a fluid at rest is defined as the normal force per unit area exerted on a plane surface (real or imaginary) immersed in a fluid and is created by the bombardment of the surface with the fluid molecules. From the definition, pressure has the dimension of FL−2 and in BG units 2 We will use T to represent temperature in thermodynamic relationships although T is also used to denote the basic dimension of time. 1.5 44.7 30 14.7 0 Ideal Gas Law 13 is expressed as lbft2 (psf) or lbin.2 (psi) and in SI units as Nm2. In SI, 1 Nm2 defined as a pascal, abbreviated as Pa, and pressures are commonly specified in pascals. The pressure in the ideal gas law must be expressed as an absolute pressure, denoted (abs), which means that it is measured relative to absolute zero pressure (a pressure that would only occur in a perfect vacuum). Standard sea-level atmospheric pressure (by international agreement) is 14.696 psi (abs) or 101.33 kPa (abs). For most calculations these pressures can be rounded to 14.7 psi and 101 kPa, respectively. In engineering it is common practice to measure pressure relative to the local atmospheric pressure, and when measured in this fashion it is called gage pressure. Thus, the absolute pressure can be obtained from the gage pressure by adding the value of the atmospheric pressure. For example, as shown by the figure in the margin, a pressure of 30 psi (gage) in a tire is equal to 44.7 psi (abs) at standard atmospheric pressure. Pressure is a particularly important fluid characteristic and it will be discussed more fully in the next chapter. 0 –14.7 (gage) (abs) p, psi E X AMPLE 1.3 Ideal Gas Law GIVEN The compressed air tank shown in Fig. E1.3a has a volume of 0.84 ft3. The temperature is 70 °F and the atmospheric pressure is 14.7 psi (abs). FIND When the tank is filled with air at a gage pressure of 50 psi, determine the density of the air and the weight of air in the tank. SOLUTION The air density can be obtained from the ideal gas law (Eq. 1.8) ρ= p RT so that (50 lbin.2 + 14.7 lbin.2 ) (144 in.2ft2 ) (1716 ft ∙ lbslug ∙ °R)[(70 + 460)°R] = 0.0102 slugs/ft3 ρ= (Ans) Note that both the pressure and temperature were changed to absolute values. The weight, 𝒲, of the air is equal to 𝒲 = ρg × (volume) = (0.0102 slugft3 ) (32.2 fts2 ) (0.84 ft3 ) = 0.276 slug ∙ ft/s2 ■ Figure E1.3a (Photograph courtesy of Jenny Products, Inc.) 0.5 so that since 1 lb = 1 slug ∙ ft s2 𝒲 = 0.276 lb W, lb 0.4 0.3 COMMENT By repeating the calculations for various values of the pressure, p, the results shown in Fig. E1.3b are obtained. Note that doubling the gage pressure does not double the amount of air in the tank, but doubling the absolute pressure does. Thus, a scuba diving tank at a gage pressure of 100 psi does not contain twice the amount of air as when the gage reads 50 psi. (50 psi, 0.276 lb) 0.2 0.1 –20 0 (Ans) 0 20 40 p, psi ■ Figure E1.3b 60 80 100 14 Chapter 1 ■ Introduction The gas constant, R, which appears in Eq. 1.8, depends on the particular gas and is related to the molecular weight of the gas. Values of the gas constant for several common gases are listed in Tables 1.7 and 1.8. Also in these tables the gas density and specific weight are given for standard atmospheric pressure and gravity and for the temperature listed. More complete tables for air at standard atmospheric pressure can be found in Appendix B (Tables B.3 and B.4). 1.6 Viscosity VIDEO V1.3 Viscous fluids VIDEO V1.4 No-slip condition Real fluids, even though they may be moving, always “stick” to the solid boundaries that contain them. The properties of density and specific weight are measures of the “heaviness” of a fluid. It is clear, however, that these properties are not sufficient to uniquely characterize how fluids behave since two fluids (such as water and oil) can have approximately the same value of density but behave quite differently when flowing. Apparently, some additional property is needed to describe the “fluidity” of the fluid. To determine this additional property, consider a hypothetical experiment in which a material is placed between two very wide parallel plates as shown in Fig. 1.4a. The bottom plate is rigidly fixed, but the upper plate is free to move. If a solid, such as steel, were placed between the two plates and loaded with the force P as shown, the top plate would be displaced through some small distance, δa (assuming the solid was mechanically attached to the plates). The vertical line AB would be rotated through the small angle, δβ, to the new position AB′. We note that to resist the applied force, P, a shearing stress, τ, would be developed at the plate–material interface, and for equilibrium to occur, P = τA where A is the effective upper plate area (Fig. 1.4b). It is well known that for elastic solids, such as steel, the small angular displacement, δβ (called the shearing strain), is proportional to the shearing stress, τ, that is developed in the material. What happens if the solid is replaced with a fluid such as water? We would immediately notice a major difference. When the force P is applied to the upper plate, it will move continuously with a velocity, U (after the initial transient motion has died out) as illustrated in Fig. 1.5. This behavior is consistent with the definition of a fluid—that is, if a shearing stress is applied to a fluid it will deform continuously. A closer inspection of the fluid motion between the two plates would reveal that the fluid in contact with the upper plate moves with the plate velocity, U, and the fluid in contact with the bottom fixed plate has a zero velocity. The fluid between the two plates moves with velocity u = u (y) that would be found to vary linearly, u = Uyb, as illustrated in Fig. 1.5. Thus, a velocity gradient, dudy, is developed in the fluid between the plates. In this particular case the velocity gradient is a constant since dudy = Ub, but in more complex flow situations, such as that shown by the photograph in the margin, this is not true. The experimental observation that the fluid “sticks” to the solid boundaries is a very important one in fluid mechanics and is usually referred to as the no-slip condition. All fluids, both liquids and gases, satisfy this condition for typical flows. δa P P B B' A δβ τA b Fixed plate (a) (b) ■ Figure 1.4 (a) Deformation of material placed between two parallel plates. (b) Forces acting on upper plate. δa U P u b B B' y A δβ ■ Figure 1.5 Behavior of a fluid placed between two Fixed plate parallel plates. Free ebooks ==> www.Ebook777.com 1.6 y u = u(y) 15 In a small time increment, δt, an imaginary vertical line AB in the fluid would rotate through an angle, δβ, so that u = 0 on surface Solid body Viscosity δβ ≈ tan δβ = δa b Since δa = U δt, it follows that δβ = U δt b We note that in this case, δβ is a function not only of the force P (which governs U) but also of time. Thus, it is not reasonable to attempt to relate the shearing stress, τ, to δβ as is done for solids. Rather, we consider the rate at which δβ is changing and define the rate of shearing strain, γ˙, as γ˙ = lim δt→0 δβ δt which in this instance is equal to γ˙ = U du = b dy A continuation of this experiment would reveal that as the shearing stress, τ, is increased by increasing P (recall that τ = P/A), the rate of shearing strain is increased in direct proportion— that is, τ ∝ γ˙ or τ∝ VIDEO V1.5 Capillary tube viscometer du dy This result indicates that for common fluids such as water, oil, gasoline, and air the shearing stress and rate of shearing strain (velocity gradient) can be related with a relationship of the form τ=μ Dynamic viscosity is the fluid property that relates shearing stress and fluid motion. du dy (1.9) where the constant of proportionality is designated by the Greek symbol μ (mu) and is called the absolute viscosity, dynamic viscosity, or simply the viscosity of the fluid. In accordance with Eq. 1.9, plots of τ versus dudy should be linear with the slope equal to the viscosity as illustrated in Fig. 1.6. The actual value of the viscosity depends on the particular fluid, and for a particular fluid the viscosity is also highly dependent on temperature as illustrated in Fig. 1.6 with the two curves for water. Fluids for which the shearing stress is linearly related to the rate of shearing strain (also referred to as the rate of angular deformation) are designated as Newtonian fluids after Isaac Newton (1642–1727). Fortunately, most common fluids, both liquids and gases, are Newtonian. A more general formulation of Eq. 1.9 which applies to more complex flows of Newtonian fluids is given in Section 6.8.1. THE WIDE WORLD OF FLUIDS An extremely viscous fluid Pitch is a derivative of tar once is quite small. In fact, to date only seven drops have fallen used for waterproofing boats. At elevated temperatures it from the end of the funnel, although the eighth drop is flows quite readily. At room temperature it feels like a solid—it poised ready to fall “soon.” While nobody has actually seen a can even be shattered with a blow from a hammer. However, drop fall from the end of the funnel, a beaker below the it is a liquid. In 1927 Professor Parnell heated some pitch and funnel holds the previous drops that fell over the years. It is poured it into a funnel. Since that time it has been allowed to estimated that the pitch is about 100 billion times more flow freely (or rather, drip slowly) from the funnel. The flowrate viscous than water. www.Ebook777.com 16 Chapter 1 ■ Introduction Crude oil (60 °F) Shearing stress, τ μ 1 Water (60 °F) Water (100 °F) Air (60 °F) du Rate of shearing strain, __ dy The various types of non-Newtonian fluids are distinguished by how their apparent viscosity changes with shear rate. of shearing stress with rate of shearing strain for common fluids. Fluids for which the shearing stress is not linearly related to the rate of shearing strain are designated as non-Newtonian fluids. Although there is a variety of types of non-Newtonian fluids, the simplest and most common are shown in Fig. 1.7. The slope of the shearing stress versus rate of shearing strain graph is denoted as the apparent viscosity, μap. For Newtonian fluids the apparent viscosity is the same as the viscosity and is independent of shear rate. For shear thinning fluids the apparent viscosity decreases with increasing shear rate—the harder the fluid is sheared, the less viscous it becomes. Many colloidal suspensions and polymer solutions are shear thinning. For example, latex paint does not drip from the brush because the shear rate is small and the apparent viscosity is large. However, it flows smoothly onto the wall because the thin layer of paint between the wall and the brush causes a large shear rate and a small apparent viscosity. For shear thickening fluids the apparent viscosity increases with increasing shear rate—the harder the fluid is sheared, the more viscous it becomes. Common examples of this type of fluid include water–corn starch mixture and water–sand mixture (“quicksand”). Thus, the difficulty in removing an object from quicksand increases dramatically as the speed of removal increases. The other type of behavior indicated in Fig. 1.7 is that of a Bingham plastic, which is neither a fluid nor a solid. Such material can withstand a finite, nonzero shear stress, τyield, the yield stress, without motion (therefore, it is not a fluid), but once the yield stress is exceeded it flows like a fluid (hence, it is not a solid). Toothpaste and mayonnaise are common examples of Bingham plastic Bingham plastic τyield Shearing stress, τ For non-Newtonian fluids, the apparent viscosity is a function of the shear rate. ■ Figure 1.6 Linear variation Shear thinning Newtonian μ ap 1 Shear thickening du Rate of shearing strain, dy ■ Figure 1.7 Variation of shearing stress with rate of shearing strain for several types of fluids, including common non-Newtonian fluids. 1.6 Viscosity 17 4.0 2.0 1.0 8 6 4 Gl yc Dynamic viscosity, μ, N • s/m2 2 SA E 1 × 10-1 8 6 4 10 W er in oi l 2 -2 1 × 10 8 6 4 2 1 × 10-3 8 6 4 Water 2 -4 1 × 10 8 6 4 Air 2 1 × 10-5 8 6 -20 Hydrogen 0 20 40 60 Temperature, °C τ < τ yield τ > τ yield VIDEO V1.6 NonNewtonian behavior 80 100 120 ■ Figure 1.8 Dynamic (absolute) viscosity of some common fluids as a function of temperature. materials. As indicated in the figure in the margin, mayonnaise can sit in a pile on a slice of bread (the shear stress less than the yield stress), but it flows smoothly into a thin layer when the knife increases the stress above the yield stress. From Eq. 1.9 it can be readily deduced that the dimensions of viscosity are FTL−2. Thus, in BG units viscosity is given as lb ∙ sft2 and in SI units as N ∙ sm2. Values of viscosity for several common liquids and gases are listed in Tables 1.5 through 1.8. A quick glance at these tables reveals the wide variation in viscosity among fluids. Viscosity is only mildly dependent on pressure and the effect of pressure is usually neglected. However, as previously mentioned, and as illustrated in Fig. 1.8, viscosity is very sensitive to temperature. For example, as the temperature of water changes from 60 to 100 °F the density decreases by less than 1%, but the viscosity decreases by about 40%. It is thus clear that particular attention must be given to temperature when determining viscosity. Figure 1.8 shows in more detail how the viscosity varies from fluid to fluid and how for a given fluid it varies with temperature. It is to be noted from this figure that the viscosity of liquids decreases with an increase in temperature, whereas for gases an increase in temperature causes an increase in viscosity. This difference in the effect of temperature on the viscosity of liquids and gases can again be traced back to the difference in molecular structure. The liquid molecules are closely spaced, with strong cohesive forces between molecules, and the resistance to relative motion between adjacent layers of fluid is related to these intermolecular forces. As the temperature increases, these cohesive forces are reduced with a corresponding reduction in resistance to motion. Since viscosity is an index of this resistance, it follows that the viscosity is reduced by an increase in temperature. In gases, however, the molecules are widely spaced and intermolecular forces negligible. In this case, resistance to relative motion arises due to the exchange of momentum of gas molecules between adjacent layers. As molecules are transported by random motion from a region of low bulk velocity to mix with molecules in a region of higher bulk velocity (and vice versa), there is an effective 18 Chapter 1 ■ Introduction momentum exchange that resists the relative motion between the layers. As the temperature of the gas increases, the random molecular activity increases with a corresponding increase in viscosity. The effect of temperature on viscosity can be closely approximated using two empirical formulas. For gases the Sutherland equation can be expressed as Viscosity is very sensitive to temperature. CT 3 2 T+S   μ= (1.10) where C and S are empirical constants, and T is absolute temperature. Thus, if the viscosity is known at two temperatures, C and S can be determined. Or, if more than two viscosities are known, the data can be correlated with Eq. 1.10 by using some type of curve-fitting scheme. For liquids an empirical equation that has been used is μ = DeBT (1.11) where D and B are constants and T is absolute temperature. This equation is often referred to as Andrade’s equation. As was the case for gases, the viscosity must be known at least for two temperatures so the two constants can be determined. A more detailed discussion of the effect of temperature on fluids can be found in Ref. 1. EXAMPLE 1.4 Viscosity and Dimensionless Quantities GIVEN A dimensionless combination of variables that is important in the study of viscous flow through pipes is called the Reynolds number, Re, defined as ρVDμ where, as indicated in Fig. E1.4, ρ is the fluid density, V the mean fluid velocity, D the pipe diameter, and μ the fluid viscosity. A Newtonian fluid having a viscosity of 0.38 N ∙ sm2 and a specific gravity of 0.91 flows through a 25-mm-diameter pipe with a velocity of 2.6 ms. V ρ, μ FIND Determine the value of the Reynolds number using D (a) SI units and (b) BG units. SOLUTION (a) The fluid density is calculated from the specific gravity as ρ = SG ρH2O@4 °C = 0.91 (1000 kgm3 ) = 910 kg/m3 ■ Figure E1.4 and from the definition of the Reynolds number (910 kgm3 ) (2.6 ms) (25 mm) (10−3 mmm) ρVD = μ 0.38 N ∙ sm2 = 156 (kg ∙ ms2 )N Re = However, since 1 N = 1 kg ∙ ms it follows that the Reynolds number is unitless—that is, 2 Re = 156 Re = (1.77 slugsft3 ) (8.53 fts) (8.20 × 10−2 ft) 7.94 × 10−3 lb ∙ sft2 = 156 (slug ∙ fts2 )lb = 156 (Ans) since 1 lb = 1 slug ∙ fts . 2 (Ans) The value of any dimensionless quantity does not depend on the system of units used if all variables that make up the quantity are expressed in a consistent set of units. To check this, we will calculate the Reynolds number using BG. units. (b) We first convert all the SI values of the variables appearing in the Reynolds number to BG values. Thus, ρ= V= D= μ= and the value of the Reynolds number is (910 kgm3 ) (1.940 × 10−3 ) = 1.77 slugsft3 (2.6 ms) (3.281) = 8.53 fts (0.025 m) (3.281) = 8.20 × 10−2 ft (0.38 N ∙ sm2 ) (2.089 × 10−2 ) = 7.94 × 10−3 lb ∙ s/ft2 COMMENTS The values from part (a) and part (b) are the same, as expected. Dimensionless quantities play an important role in fluid mechanics, and the significance of the Reynolds number as well as other important dimensionless combinations will be discussed in detail in Chapter 7. It should be noted that in the Reynolds number it is actually the ratio μρ that is important, and this is the property that is defined as the kinematic viscosity. 1.6 E X AMPLE 1.5 19 Newtonian Fluid Shear Stress GIVEN The velocity distribution for the flow of a Newtonian fluid between two fixed wide, parallel plates (see Fig. E1.5a) is given by the equation y 2 3V u= 1 − (h) ] 2 [ where V is the mean velocity. The fluid has a viscosity of 0.04 lb ∙ sft2 . Also, V = 2 fts and h = 0.2 in. FIND Determine: (a) the shearing stress acting on the bottom wall, and (b) the shearing stress acting on a plane parallel to the walls and passing through the centerline (midplane). SOLUTION du τ=μ dy u h ■ Figure E1.5a (1) Thus, if the velocity distribution u = u(y) is known, the shearing stress can be determined at all points by evaluating the velocity gradient, dudy. For the distribution given 3Vy du =− 2 dy h y h For this type of parallel flow the shearing stress is obtained from Eq. 1.9, (a) Viscosity you explain the change of sign? [Hint: Sketch a sheared fluid particle adjacent to the top plate.] (2) u Along the bottom wall y = −h so that (from Eq. 2) du 3V = dy h ■ Figure E1.5b and therefore the shearing stress is 15 (Ans) A positive shear stress seems intuitively satisfying. To check your intuition, sketch a sheared fluid particle adjacent to the bottom plate (Fig. E1.5b). The fluid motion seems to be trying to drag the plate in the flow direction. Reference to the section in which the concept of viscosity was introduced, and Fig. 1.4 in particular, makes clear that the shear stress on the fluid at the wall is indeed positive. A review of Eq. 2 and the answers to parts a and b of this example makes it clear that the shear stress varies linearly across the flow field. (b) Along the midplane where y = 0 it follows from Eq. 2 that 10 5 T (Ib/ft2) (0.04 lb ∙ sft2 ) (3) (2 fts) 3V τbottom = μ = (h) (0.2 in.) (1 ft12 in.) wall 2 = 14.4 lbft (in direction of flow) 0 –0.2 –0.1 0 du =0 dy –5 and thus the shearing stress is τmidplane = 0 (Ans) COMMENT Equation 2 has been evaluated at the bottom plate and mid-plane to provide the answers to parts a and b of this example. If you use it to compute the shear stress at the top plate, you will notice that it produces a shear stress of the same magnitude as at the bottom plate, but of opposite sign. A plot of τ versus y between the plates, Fig. E1.5b, confirms both the difference in sign and the earlier observation of the linear relationship. Can –10 –15 y (in.) ■ Figure E1.5c 0.1 0.2 Free ebooks ==> www.Ebook777.com 20 Chapter 1 ■ Introduction Quite often viscosity appears in fluid flow problems combined with the density in the form μ ν= ρ Kinematic viscosity is defined as the ratio of the absolute viscosity to the fluid density. 1.7 This ratio is called the kinematic viscosity and is denoted with the Greek symbol ν (nu). The dimensions of kinematic viscosity are L2 T, and the BG units are ft2 s and SI units are m2 s. Values of kinematic viscosity for some common liquids and gases are given in Tables 1.5 through 1.8. More extensive tables giving both the dynamic and kinematic viscosities for water and air can be found in Appendix B (Tables B.1 through B.4), and graphs showing the variation in both dynamic and kinematic viscosity with temperature for a variety of fluids are also provided in Appendix B (Figs. B.1 and B.2). Although in this text we are primarily using BG and SI units, dynamic viscosity is often expressed in the metric CGS (centimeter-gram-second) system with units of dyne ∙ scm2. This combination is called a poise, abbreviated P. In the CGS system, kinematic viscosity has units of cm2 s, and this combination is called a stoke, abbreviated St. Compressibility of Fluids 1.7.1 Bulk Modulus p V p + dp V – dV VIDEO V1.7 Water balloon An important question to answer when considering the behavior of a particular fluid is how easily can the volume (and thus the density) of a given mass of the fluid be changed when there is a change in pressure? That is, how compressible is the fluid? A property that is commonly used to characterize compressibility is the bulk modulus, Eυ, defined as dp Eυ = − (1.12) dVV where dp is the differential change in pressure needed to create a differential change in volume, dV, of a volume V. This is illustrated by the figure in the margin. The negative sign is included since an increase in pressure will cause a decrease in volume. Since a decrease in volume of a given mass, m = ρV, will result in an increase in density, Eq. 1.12 can also be expressed as dp Eυ = (1.13) dρρ The bulk modulus (also referred to as the bulk modulus of elasticity) has dimensions of pressure, FL−2. In BG units, values for Eυ are usually given as lbin.2 (psi) and in SI units as Nm2 (Pa). Large values for the bulk modulus indicate that the fluid is relatively incompressible—that is, it takes a large pressure change to create a small change in volume. As expected, values of Eυ for common liquids are large (see Tables 1.5 and 1.6). For example, at atmospheric pressure and a temperature of 60 °F it would require a pressure of 3120 psi to compress a unit volume of water 1%. This result is representative of the compressibility of liquids. Since such large pressures are required to effect a change in volume, we conclude that liquids can be considered as incompressible for most practical engineering applications. As liquids are compressed the bulk modulus increases, but the bulk modulus near atmospheric pressure is usually the one of interest. The use of bulk modulus as a property describing compressibility is most prevalent when dealing with liquids, although the bulk modulus can also be determined for gases. THE WIDE WORLD OF FLUIDS This water jet is a blast Usually liquids can be treated as or 3000 to 4000 atmospheres), the water is compressed (i.e., the incompressible fluids. However, in some applications the com- volume reduced) by about 10 to 15%. When a fast-opening valve pressibility of a liquid can play a key role in the operation of a within the pressure vessel is opened, the water expands and device. For example, a water pulse generator using compressed produces a jet of water that upon impact with the target material water has been developed for use in mining operations. It can produces an effect similar to the explosive force from conven- fracture rock by producing an effect comparable to a conven- tional explosives. Mining with the water jet can eliminate various tional explosive such as gunpowder. The device uses the energy hazards that arise with the use of conventional chemical explo- stored in a water-filled accumulator to generate an ultrahigh- sives, such as those associated with the storage and use of pressure water pulse ejected through a 10- to 25-mm-diameter explosives and the generation of toxic gas by-products that discharge valve. At the ultrahigh pressures used (300 to 400 MPa, require extensive ventilation. (See Problem 1.110.) www.Ebook777.com 1.7 Compressibility of Fluids 21 1.7.2 Compression and Expansion of Gases When gases are compressed (or expanded), the relationship between pressure and density depends on the nature of the process. If the compression or expansion takes place under constant temperature conditions (isothermal process), then from Eq. 1.8 p = constant ρ Isentropic (k = 1.4) (1.14) If the compression or expansion is frictionless and no heat is exchanged with the surroundings (isentropic process), then p Isothermal p ρk ρ The value of the bulk modulus depends on the type of process involved. = constant (1.15) where k is the ratio of the specific heat at constant pressure, cp, to the specific heat at constant volume, cυ (i.e., k = cp cυ ). The two specific heats are related to the gas constant, R, through the equation R = cp − cυ. As was the case for the ideal gas law, the pressure in both Eqs. 1.14 and 1.15 must be expressed as an absolute pressure. Values of k for some common gases are given in Tables 1.7 and 1.8 and for air over a range of temperatures, in Appendix B (Tables B.3 and B.4). The pressure–density variations for isothermal and isentropic conditions are illustrated in the margin figure. With explicit equations relating pressure and density, the bulk modulus for gases can be determined by obtaining the derivative dpdρ from Eq. 1.14 or 1.15 and substituting the results into Eq. 1.13. It follows that for an isothermal process Eυ = p (1.16) Eυ = kp (1.17) and for an isentropic process, Note that in both cases the bulk modulus varies directly with pressure. For air under standard atmospheric conditions with p = 14.7 psi (abs) and k = 1.40, the isentropic bulk modulus is 20.6 psi. A comparison of this figure with that for water under the same conditions (Eυ = 312,000 psi) shows that air is approximately 15,000 times as compressible as water. It is thus clear that in dealing with gases, greater attention will need to be given to the effect of compressibility on fluid behavior. However, as will be discussed further in later sections, gas flows can often be treated as incompressible flows if the changes in pressure are small. E X AMPLE 1.6 Isentropic Compression of a Gas GIVEN A cubic foot of air at an absolute pressure of 14.7 psi is compressed isentropically to Fig. E1.6a. FIND 1 2 ft3 by the tire pump shown in What is the final pressure? SOLUTION For an isentropic compression pi ρki = pf ρkf where the subscripts i and f refer to initial and final states, respectively. Since we are interested in the final pressure, pf, it follows that ρf k pf = ( ρi ) pi ■ Figure E1.6a 22 Chapter 1 ■ Introduction 400 As the volume, V. is reduced by one-half, the density must double, since the mass, m = ρ V. of the gas remains constant. Thus, with k = 1.40 for air pf = (2) 1.40 (14.7 psi) = 38.8 psi (abs) 350 300 (Ans) COMMENT By repeating the calculations for various values of the ratio of the final volume to the initial volume, Vf Vi , the results shown in Fig. E1.6b are obtained. Note that even though air is often considered to be easily compressed (at least compared to liquids), it takes considerable pressure to significantly reduce a given volume of air as is done in an automobile engine where the compression ratio is on the order of Vf Vi = 1/8 = 0.125. pf, psi 250 200 150 100 (0.5, 38.8 psi) 50 0 0 0.2 0.4 0.6 0.8 1 Vf /Vi ■ Figure E1.6b 1.7.3 Speed of Sound Another important consequence of the compressibility of fluids is that disturbances introduced at some point in the fluid propagate at a finite velocity. For example, if a fluid is flowing in a pipe and a valve at the outlet is suddenly closed (thereby creating a localized disturbance), the effect of the valve closure is not felt instantaneously upstream. It takes a finite time for the increased pressure created by the valve closure to propagate to an upstream location. Similarly, a loudspeaker diaphragm causes a localized disturbance as it vibrates, and the small change in pressure created by the motion of the diaphragm is propagated through the air with a finite velocity. The velocity at which these small disturbances propagate is called the acoustic velocity or the speed of sound, c. It will be shown in Chapter 11 that the speed of sound is related to changes in pressure and density of the fluid medium through the equation The velocity at which small disturbances propagate in a fluid is called the speed of sound. 6000 water c= 4000 (1.18) c, ft/s or in terms of the bulk modulus defined by Eq. 1.13 c= 2000 air 0 dp B dρ 0 100 T, °F 200 Eυ Bρ (1.19) Since the disturbance is small, there is negligible heat transfer and the process is assumed to be isentropic. Thus, the pressure–density relationship used in Eq. 1.18 is that for an isentropic process. For gases undergoing an isentropic process, Eυ = kp (Eq. 1.17) so that c= kp Bρ and making use of the ideal gas law, it follows that c = √kRT VIDEO V1.8 As fast as a speeding bullet (1.20) Thus, for ideal gases the speed of sound is proportional to the square root of the absolute temperature. For example, for air at 60 °F with k = 1.40 and R = 1716 ft ∙ lbslug ∙ °R, it follows that c = 1117 fts. The speed of sound in air at various temperatures can be found in Appendix B (Tables B.3 and B.4). Equation 1.19 is also valid for liquids, and values of Eυ can be used to determine the speed of sound in liquids. For water at 20 °C, Eυ = 2.19 GNm2 and ρ = 998.2 kgm3 so that c = 1481 ms or 4860 fts. As shown by the figure in the margin, the speed of sound is much higher in water than in air. If a fluid were truly incompressible (Eυ = ∞) the speed of sound 1.8 Vapor Pressure 23 would be infinite. The speed of sound in water for various temperatures can be found in Appendix B (Tables B.1 and B.2). E X AMPLE 1.7 Speed of Sound and Mach Number GIVEN A jet aircraft flies at a speed of 550 mph at an altitude FIND Determine the ratio of the speed of the aircraft, V, to that of 35,000 ft, where the temperature is −66 °F and the specific heat ratio is k = 1.4. of the speed of sound, c, at the specified altitude. SOLUTION 0.9 From Eq. 1.20 the speed of sound can be calculated as c = √kRT = √(1.40) (1716 ft lb/slug °R) (−66 + 460) °R = 973 ft /s 0.8 (–66 °F, 0.829) V= Ma = V/c Since the air speed is (550 mi/ hr) (5280 ft/mi) = 807 ft /s (3600 s / hr) 0.7 the ratio is V 807 ft /s = 0.829 = c 973 ft /s 0.6 (Ans) COMMENT This ratio is called the Mach number, Ma. If Ma < 1.0 the aircraft is flying at subsonic speeds, whereas for Ma > 1.0 it is flying at supersonic speeds. The Mach number is an important dimensionless parameter used in the study of the flow of gases at high speeds and will be further discussed in Chapters 7 and 11. By repeating the calculations for different temperatures, the results shown in Fig. E1.7 are obtained. Because the speed of 1.8 Liquid Vapor, pυ Liquid A liquid boils when the pressure is reduced to the vapor pressure. 0.5 –100 –50 0 T, °F 50 100 ■ Figure E1.7 sound increases with increasing temperature, for a constant airplane speed, the Mach number decreases as the temperature increases. Vapor Pressure It is a common observation that liquids such as water and gasoline will evaporate if they are simply placed in a container open to the atmosphere. Evaporation takes place because some liquid molecules at the surface have sufficient momentum to overcome the intermolecular cohesive forces and escape into the atmosphere. If the container is closed with a small air space left above the surface, and this space evacuated to form a vacuum, a pressure will develop in the space as a result of the vapor that is formed by the escaping molecules. When an equilibrium condition is reached so that the number of molecules leaving the surface is equal to the number entering, the vapor is said to be saturated and the pressure that the vapor exerts on the liquid surface is termed the vapor pressure, pv . Similarly, if the end of a completely liquid-filled container is moved as shown in the figure in the margin without letting any air into the container, the space between the liquid and the end becomes filled with vapor at a pressure equal to the vapor pressure. Since the development of a vapor pressure is closely associated with molecular activity, the value of vapor pressure for a particular liquid depends on temperature. Values of vapor pressure for water at various temperatures can be found in Appendix B (Tables B.1 and B.2), and the values of vapor pressure for several common liquids at room temperatures are given in Tables 1.5 and 1.6. Boiling, which is the formation of vapor bubbles within a fluid mass, is initiated when the absolute pressure in the fluid reaches the vapor pressure. As commonly observed in the kitchen, water at standard atmospheric pressure will boil when the temperature reaches 212 °F (100 °C)— Boiling temperature, °F 24 Chapter 1 ■ Introduction 250 150 50 0 0 20 40 60 Altitude, thousands of feet In flowing liquids it is possible for the pressure in localized regions to reach vapor pressure, thereby causing cavitation. 1.9 Surface Tension Surface tension, lb/ft VIDEO 6 V1.9 Floating razor blade × 10−3 4 Water 2 0 that is, the vapor pressure of water at 212 °F is 14.7 psi (abs). However, if we attempt to boil water at a higher elevation, say 30,000 ft above sea level (the approximate elevation of Mt. Everest), where the atmospheric pressure is 4.37 psi (abs), we find that boiling will start when the temperature is about 157 °F. At this temperature the vapor pressure of water is 4.37 psi (abs). For the U.S. Standard Atmosphere (see Section 2.4), the boiling temperature is a function of altitude as shown in the figure in the margin. Thus, boiling can be induced at a given pressure acting on the fluid by raising the temperature, or at a given fluid temperature by lowering the pressure. An important reason for our interest in vapor pressure and boiling lies in the common observation that in flowing fluids it is possible to develop very low pressure due to the fluid motion, and if the pressure is lowered to the vapor pressure, boiling will occur. For example, this phenomenon may occur in flow through the irregular, narrowed passages of a valve or pump. When vapor bubbles are formed in a flowing fluid, they are swept along into regions of higher pressure where they suddenly collapse with sufficient intensity to actually cause structural damage. The formation and subsequent collapse of vapor bubbles in a flowing fluid, called cavitation, is an important fluid flow phenomenon to be given further attention in Chapters 3 and 7. 0 50 100 150 200 Temperature, °F At the interface between a liquid and a gas, or between two immiscible liquids, forces develop in the liquid surface that cause the surface to behave as if it were a “skin” or “membrane” stretched over the fluid mass. Although such a skin is not actually present, this conceptual analogy allows us to explain several commonly observed phenomena. For example, a steel needle or a razor blade will float on water if placed gently on the surface because the tension developed in the hypothetical skin supports it. Small droplets of mercury will form into spheres when placed on a smooth surface because the cohesive forces in the surface of the mercury tend to hold all the molecules together in a compact shape. Similarly, discrete bubbles will form in a liquid. (See the photograph at the beginning of Chapter 1.) These various types of surface phenomena are due to the unbalanced cohesive forces acting on the liquid molecules at the fluid surface. Molecules in the interior of the fluid mass are surrounded by molecules that are attracted to each other equally. However, molecules along the surface are subjected to a net force toward the interior. The apparent physical consequence of this unbalanced force along the surface is to create the hypothetical skin or membrane. A tensile force may be considered to be acting in the plane of the surface along any line in the surface. The intensity of the molecular attraction per unit length along any line in the surface is called the surface tension and is designated by the Greek symbol σ (sigma). For a given liquid the surface tension depends on temperature as well as the other fluid it is in contact with at the interface. The dimensions of surface tension are FL−1 with BG units of lbft and SI units of Nm. Values of surface tension for some common liquids (in contact with air) are given in Appendix B (Tables B.1 and B.2) for water at various temperatures. As indicated by the figure in the margin, the value of the surface tension decreases as the temperature increases. THE WIDE WORLD OF FLUIDS Walking on water Water striders are insects commonly found speed video to examine in detail the movement of the water on ponds, rivers, and lakes that appear to “walk” on water. A striders. They found that each stroke of the insect’s legs creates typical length of a water strider is about 0.4 in., and they can dimples on the surface with underwater swirling vortices suffi- cover 100 body lengths in one second. It has long been recog- cient to propel it forward. It is the rearward motion of the vortices nized that it is surface tension that keeps the water strider from that propels the water strider forward. To further substantiate sinking below the surface. What has been puzzling is how they their explanation, the MIT team built a working model of a water propel themselves at such a high speed. They can’t pierce the strider, called Robostrider, which creates surface ripples and water surface or they would sink. A team of mathematicians and underwater vortices as it moves across a water surface. engineers from the Massachusetts Institute of Technology (MIT) Waterborne creatures, such as the water strider, provide an inter- applied conventional flow visualization techniques and high- esting world dominated by surface tension. (See Problem 1.131.) Free ebooks ==> www.Ebook777.com 1.9 25 Surface Tension R σ σ Δ p π R2 ■ Figure 1.9 Forces acting on one-half of a liquid drop. The pressure inside a drop of fluid can be calculated using the free-body diagram in Fig. 1.9. If the spherical drop is cut in half (as shown), the force developed around the edge due to surface tension is 2πRσ. This force must be balanced by the pressure difference, Δp, between the internal pressure, pi, and the external pressure, pe, acting over the circular area, πR2. Thus, VIDEO V1.10 Capillary rise 2πRσ = Δp πR2 or Δp = pi − pe = R VIDEO (1.21) It is apparent from this result that the pressure inside the drop is greater than the pressure surrounding the drop. (Would the pressure on the inside of a bubble of water be the same as that on the inside of a drop of water of the same diameter and at the same temperature?) Among common phenomena associated with surface tension is the rise (or fall) of a liquid in a capillary tube. If a small open tube is inserted into water, the water level in the tube will rise above the water level outside the tube, as is illustrated in Fig. 1.10a. In this situation we have a liquid–gas– solid interface. For the case illustrated there is an attraction (adhesion) between the wall of the tube and liquid molecules which is strong enough to overcome the mutual attraction (cohesion) of the molecules and pull them up the wall. Hence, the liquid is said to wet the solid surface. The height, h, is governed by the value of the surface tension, σ, the tube radius, R, the specific weight of the liquid, γ, and the angle of contact, θ, between the fluid and tube. From the freebody diagram of Fig. 1.10b we see that the vertical force due to the surface tension is equal to 2πRσ cos θ and the weight is γπR2h, and these two forces must balance for equilibrium. Thus, h ~ _1_ R h 2σ R V1.11 Contact angle γπR2h = 2πRσ cos θ so that the height is given by the relationship h= Capillary action in small tubes, which involves a liquid– gas–solid interface, is caused by surface tension. 2σ cos θ γR (1.22) The angle of contact is a function of both the liquid and the surface. For water in contact with clean glass θ ≈ 0°. It is clear from Eq. 1.22 that the height is inversely proportional to the tube radius, and therefore, as indicated by the figure in the margin, the rise of a liquid in a tube as a result of capillary action becomes increasingly pronounced as the tube radius is decreased. If adhesion of molecules to the solid surface is weak compared to the cohesion between molecules, the liquid will not wet the surface and the level in a tube placed in a nonwetting liquid will actually be depressed, as shown in Fig. 1.10c. Mercury is a good example of a nonwetting liquid when it is in contact with a glass tube. For nonwetting liquids the angle of contact is greater than 90°, and for mercury in contact with clean glass θ ≈ 130°. θ 2π Rσ θ 2 h γπ R h h 2R (a) (b) (c) ■ Figure 1.10 Effect of capillary action in small tubes. (a) Rise of column for a liquid that wets the tube. (b) Free-body diagram for calculating column height. (c) Depression of column for a nonwetting liquid. www.Ebook777.com 26 Chapter 1 ■ Introduction EXAMPLE 1.8 Capillary Rise in a Tube GIVEN Pressures are sometimes determined by measuring the FIND What diameter of clean glass tubing is required so that height of a column of liquid in a vertical tube. the rise of water at 20 °C in a tube due to capillary action (as opposed to pressure in the tube) is less than h = 1.0 mm? SOLUTION From Eq. 1.22 h= 2σ cos θ γR R= 2σ cos θ γh obtained. Note that as the allowable capillary rise is decreased, the diameter of the tube must be significantly increased. There is always some capillarity effect, but it can be minimized by using a large enough diameter tube. so that 100 For water at 20 °C (from Table B.2), σ = 0.0728 Nm and γ = 9.789 kNm3. Since θ ≈ 0° it follows that for h = 1.0 mm, (9.789 × 10 Nm ) (1.0 mm) (10 = 0.0149 m 3 3 −3 D, mm 2(0.0728 Nm) (1) R= mmm) 60 40 (1 mm, 29.8 mm) 20 and the minimum required tube diameter, D, is D = 2R = 0.0298 m = 29.8 mm 80 (Ans) 0 0 0.5 of the capillary rise, h, the results shown in Fig. E1.8 are 1 1.5 2 h, mm COMMENT By repeating the calculations for various values ■ Figure E1.8 Surface tension effects play a role in many fluid mechanics problems, including the movement of liquids through soil and other porous media, flow of thin films, formation of drops and bubbles, and the breakup of liquid jets. For example, surface tension is a main factor in the formation of drops from a leaking faucet, as shown in the photograph in the margin. Surface phenomena associated with liquid–gas, liquid–liquid, and liquid–gas–solid interfaces are exceedingly complex, and a more detailed and rigorous discussion of them is beyond the scope of this text. Fortunately, in many fluid mechanics problems, surface phenomena, as characterized by surface tension, are not important, since inertial, gravitational, and viscous forces are much more dominant. (Photograph copyright 2007 by Andrew Davidhazy, Rochester Institute of Technology.) THE WIDE WORLD OF FLUIDS Spreading of oil spills With the large traffic in oil tankers there There are many factors that influence the ability of an oil slick to is great interest in the prevention of and response to oil spills. spread, including the size of the spill, wind speed and direction, As evidenced by the famous Exxon Valdez oil spill in Prince and the physical properties of the oil. These properties include William Sound in 1989, oil spills can create disastrous environ- surface tension, specific gravity, and viscosity. The higher the sur- mental problems. A more recent example of this type of catas- face tension the more likely a spill will remain in place. Since the trophe is the oil spill that occurred in the Gulf of Mexico in 2010. specific gravity of oil is less than one, it floats on top of the It is not surprising that much attention is given to the rate at water, but the specific gravity of an oil can increase if the lighter which an oil spill spreads. When spilled, most oils tend to spread substances within the oil evaporate. The higher the viscosity of horizontally into a smooth and slippery surface, called a slick. the oil, the greater the tendency to stay in one place. 1.10 1.10 Some of the earliest writings that pertain to modern fluid mechanics can be traced back to the ancient Greek civilization and subsequent Roman Empire. A Brief Look Back in History 27 A Brief Look Back in History Before proceeding with our study of fluid mechanics, we should pause for a moment to consider the history of this important engineering science. As is true of all basic scientific and engineering disciplines, their actual beginnings are only faintly visible through the haze of early antiquity. But we know that interest in fluid behavior dates back to the ancient civilizations. Through necessity there was a practical concern about the manner in which spears and arrows could be propelled through the air, in the development of water supply and irrigation systems, and in the design of boats and ships. These developments were, of course, based on trial-and-error procedures. However, it was the accumulation of such empirical knowledge that formed the basis for further development during the emergence of the ancient Greek civilization and the subsequent rise of the Roman Empire. Some of the earliest writings that pertain to modern fluid mechanics are those of Archimedes (287–212 B.C.), a Greek mathematician and inventor who first expressed the principles of hydrostatics and flotation. Elaborate water supply systems were built by the Romans during the period from the fourth century B.C. through the early Christian period, and Sextus Julius Frontinus (A.D. 40–103), a Roman engineer, described these systems in detail. However, for the next 1000 years during the Middle Ages (also referred to as the Dark Ages), there appears to have been little added to further understanding of fluid behavior. As shown in Fig. 1.11, beginning with the Renaissance period (about the fifteenth century) a rather continuous series of contributions began that forms the basis of what we consider to be the science of fluid mechanics. Leonardo da Vinci (1452–1519) described through sketches and writings many different types of flow phenomena. The work of Galileo Galilei (1564–1642) marked the beginning of experimental mechanics. Following the early Renaissance period and during the seventeenth and eighteenth centuries, numerous significant contributions were made. These include theoretical and mathematical advances associated with the famous names of Newton, Bernoulli, Euler, and d’Alembert. Experimental aspects of fluid mechanics were also advanced during this period, but unfortunately the two different approaches, theoretical and experimental, developed along separate paths. Hydrodynamics was the term associated with the theoretical or mathematical study of idealized, frictionless fluid behavior, with the term hydraulics being used to describe the applied or experimental aspects of real fluid behavior, particularly the behavior of water. Further contributions and refinements were made to both theoretical hydrodynamics and experimental hydraulics during the nineteenth century, with the general differential equations describing fluid motions that are used in modern fluid mechanics being developed in this period. Experimental hydraulics became more of a science, and many of the results of experiments performed during the nineteenth century are still used today. At the beginning of the twentieth century, both the fields of theoretical hydrodynamics and experimental hydraulics were highly developed, and attempts were being made to unify the two. In 1904 a classic paper was presented by a German professor, Ludwig Prandtl (1875–1953), who introduced the concept of a “fluid boundary layer,” which laid the foundation for the unification of the theoretical and experimental aspects of fluid mechanics. Prandtl’s idea was that for flow next to a Geoffrey Taylor Theodor von Karman Ludwig Prandtl Osborne Reynolds Ernst Mach George Stokes Jean Poiseuille Louis Navier Leonhard Euler Daniel Bernoulli Isaac Newton Galileo Galilei Leonardo da Vinci 1200 1300 1400 1500 1600 1700 1800 1900 Year ■ Figure 1.11 Time line of some contributors to the science of fluid mechanics. 2000 28 Chapter 1 ■ Introduction The rich history of fluid mechanics is fascinating, and many of the contributions of the pioneers in the field are noted in the succeeding chapters. solid boundary a thin fluid layer (boundary layer) develops in which friction (viscous force) is important, but outside this layer the fluid behaves very much like a frictionless fluid. This relatively simple concept provided the necessary impetus for the resolution of the conflict between the hydrodynamicists and the hydraulicists. Prandtl is generally accepted as the founder of modern fluid mechanics. Also, during the first decade of the twentieth century, powered flight was first successfully demonstrated with the subsequent vastly increased interest in aerodynamics. Because the design of aircraft required an understanding of fluid flow and an ability to make accurate predictions of the effect of airflow on bodies, the field of aerodynamics provided a great stimulus for the many rapid developments in fluid mechanics that took place during the twentieth century. As we proceed with our study of the fundamentals of fluid mechanics, we will continue to note the contributions of many of the pioneers in the field. Table 1.9 provides a chronological listing ■ Table 1.9 Chronological Listing of Some Contributors to the Science of Fluid Mechanics Noted in the Texta ARCHIMEDES (287– 212 B.C.) Established elementary principles of buoyancy and flotation. SEXTUS JULIUS FRONTINUS (A.D. 40–103) Wrote treatise on Roman methods of water distribution. Leonardo da Vinci Isaac Newton LEONARDO da VINCI (1452–1519) Expressed elementary principle of continuity; observed and sketched many basic flow phenomena; suggested designs for hydraulic machinery. LOUIS MARIE HENRI NAVIER (1785–1836) Extended equations of motion to include “molecular” forces. AUGUSTIN LOUIS de CAUCHY (1789–1857) Contributed to the general field of theoretical hydrodynamics and to the study of wave motion. EVANGELISTA TORRICELLI (1608–1647) Related barometric height to weight of atmosphere, and form of liquid jet to trajectory of free fall. GOTTHILF HEINRICH LUDWIG HAGEN (1797–1884) Conducted original studies of resistance in and transition between laminar and turbulent flow. BLAISE PASCAL (1623–1662) Finally clarified principles of barometer, hydraulic press, and pressure transmissibility. JEAN LOUIS POISEUILLE (1799–1869) Performed meticulous tests on resistance of flow through capillary tubes. ISAAC NEWTON (1642–1727) Explored various aspects of fluid resistance— inertial, viscous, and wave; discovered jet contraction. HENRI PHILIBERT GASPARD DARCY (1803–1858) Performed extensive tests on filtration and pipe resistance; initiated open-channel studies carried out by Bazin. DANIEL BERNOULLI (1700–1782) Experimented and wrote on many phases of fluid motion, coining name “hydrodynamics”; devised manometry technique and adapted primitive energy principle to explain velocity-head indication; proposed jet propulsion. LEONHARD EULER (1707–1783) First explained role of pressure in fluid flow; formulated basic equations of motion and so-called Bernoulli theorem; introduced concept of cavitation and principle of centrifugal machinery. Ernst Mach GIOVANNI BATTISTA VENTURI (1746–1822) Performed tests on various forms of mouthpieces— in particular, conical contractions and expansions. GALILEO GALILEI (1564–1642) Indirectly stimulated experimental hydraulics; revised Aristotelian concept of vacuum. HENRI de PITOT (1695–1771) Constructed double-tube device to indicate water velocity through differential head. Daniel Bernoulli ANTOINE CHEZY (1718–1798) Formulated similarity parameter for predicting flow characteristics of one channel from measurements on another. JEAN le ROND d’ALEMBERT (1717–1783) Originated notion of velocity and acceleration components, differential expression of continuity, and paradox of zero resistance to steady nonuniform motion. JULIUS WEISBACH (1806–1871) Incorporated hydraulics in treatise on engineering mechanics, based on original experiments; noteworthy for flow patterns, nondimensional coefficients, weir, and resistance equations. WILLIAM FROUDE (1810–1879) Developed many towing-tank techniques, in particular the conversion of wave and boundary layer resistance from model to prototype scale. ROBERT MANNING (1816–1897) Proposed several formulas for open-channel resistance. GEORGE GABRIEL STOKES (1819–1903) Derived analytically various flow relationships ranging from wave mechanics to viscous resistance— particularly that for the settling of spheres. ERNST MACH (1838–1916) One of the pioneers in the field of supersonic aerodynamics. 1.11 Chapter Summary and Study Guide 29 ■ Table 1.9 (continued) OSBORNE REYNOLDS (1842–1912) Described original experiments in many fields— cavitation, river model similarity, pipe resistance— and devised two parameters for viscous flow; adapted equations of motion of a viscous fluid to mean conditions of turbulent flow. Osborne Reynolds JOHN WILLIAM STRUTT, LORD RAYLEIGH (1842–1919) Investigated hydrodynamics of bubble collapse, wave motion, jet instability, laminar flow analogies, and dynamic similarity. VINCENZ STROUHAL (1850–1922) Investigated the phenomenon of “singing wires.” EDGAR BUCKINGHAM (1867–1940) Stimulated interest in the United States in the use of dimensional analysis. Ludwig Prandtl MORITZ WEBER (1871–1951) Emphasized the use of the principles of similitude in fluid flow studies and formulated a capillarity similarity parameter. a LUDWIG PRANDTL (1875–1953) Introduced concept of the boundary layer and is generally considered to be the father of present-day fluid mechanics. LEWIS FERRY MOODY (1880–1953) Provided many innovations in the field of hydraulic machinery. Proposed a method of correlating pipe resistance data that is widely used. THEODOR VON KÁRMÁN (1881–1963) One of the recognized leaders of twentieth century fluid mechanics. Provided major contributions to our understanding of surface resistance, turbulence, and wake phenomena. PAUL RICHARD HEINRICH BLASIUS (1883–1970) One of Prandtl’s students who provided an analytical solution to the boundary layer equations. Also demonstrated that pipe resistance was related to the Reynolds number. Used by permission of IIHR—Hydroscience & Engineering, The University of Iowa. of some of these contributors and reveals the long journey that makes up the history of fluid mechanics. This list is certainly not comprehensive with regard to all past contributors but includes those who are mentioned in this text. As mention is made in succeeding chapters of the various individuals listed in Table 1.9, a quick glance at this table will reveal where they fit into the historical chain. It is, of course, impossible to summarize the rich history of fluid mechanics in a few paragraphs. Only a brief glimpse is provided, and we hope it will stir your interest. References 2 to 5 are good starting points for further study, and in particular Ref. 2 provides an excellent, broad, easily read history. Try it—you might even enjoy it! 1.11 Chapter Summary and Study Guide This introductory chapter discussed several fundamental aspects of fluid mechanics. Methods for describing fluid characteristics both quantitatively and qualitatively are considered. For a quantitative description, units are required. The concept of dimensions is introduced in which basic dimensions such as length, L, time, T, and mass, M, are used to provide a description of various quantities of interest. The use of dimensions is helpful in checking the generality of equations, as well as serving as the basis for the powerful tool of dimensional analysis discussed in detail in Chapter 7. Various important fluid properties are defined, including fluid density, specific weight, specific gravity, viscosity, bulk modulus, speed of sound, vapor pressure, and surface tension. The ideal gas law is introduced to relate pressure, temperature, and density in common gases, along with a brief discussion of the compression and expansion of gases. The distinction between absolute and gage pressure is introduced. This important idea is explored more fully in Chapter 2. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed you should be able to ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. determine the dimensions of common physical quantities. determine whether an equation is a general or restricted homogeneous equation. Free ebooks ==> www.Ebook777.com 30 Chapter 1 ■ Introduction fluid units basic dimensions dimensionally homogeneous density specific weight specific gravity ideal gas law absolute pressure gage pressure no-slip condition rate of shearing strain absolute viscosity Newtonian fluid non-Newtonian fluid kinematic viscosity bulk modulus speed of sound vapor pressure surface tension ■ ■ ■ ■ ■ ■ ■ ■ correctly use units and systems of units in your analyses and calculations. calculate the density, specific weight, or specific gravity of a fluid from a knowledge of any two of the three. calculate the density, pressure, or temperature of an ideal gas (with a given gas constant) from a knowledge of any two of the three. relate the pressure and density of a gas as it is compressed or expanded using Eqs. 1.14 and 1.15. use the concept of viscosity to calculate the shearing stress in simple fluid flows. calculate the speed of sound in fluids using Eq. 1.19 for liquids and Eq. 1.20 for gases. determine whether boiling or cavitation will occur in a liquid using the concept of vapor pressure. use the concept of surface tension to solve simple problems involving liquid–gas or liquid– solid–gas interfaces. Some of the important equations in this chapter are: γ = ρg Specific weight SG = Specific gravity ρ ρH2O@4 °C (1.6) (1.7) Ideal gas law p = ρRT (1.8) Newtonian fluid shear stress τ=μ du dy (1.9) Eυ = − Bulk modulus dp d VV Speed of sound in an ideal gas c = √kRT Capillary rise in a tube h= 2σ cos θ γR (1.12) (1.20) (1.22) References 1. Reid, R. C., Prausnitz, J. M., and Sherwood, T. K., The Properties of Gases and Liquids, 3rd Ed., McGraw-Hill, New York, 1977. 2. Rouse, H. and Ince, S., History of Hydraulics, Iowa Institute of Hydraulic Research, Iowa City, 1957, Dover, New York, 1963. 3. Tokaty, G. A., A History and Philosophy of Fluid Mechanics, G. T. Foulis and Co., Ltd., Oxfordshire, Great Britain, 1971. 4. Rouse, H., Hydraulics in the United States 1776–1976, Iowa Institute of Hydraulic Research, Iowa City, Iowa, 1976. 5. Garbrecht, G., ed., Hydraulics and Hydraulic Research—A Historical Review, A. A. Balkema, Rotterdam, Netherlands, 1987. 6. Brenner, M. P., Shi, X. D., Eggens, J., and Nagel, S. R., Physics of Fluids, Vol. 7, No. 9, 1995. 7. Shi, X. D., Brenner, M. P., and Nagel, S. R., Science, Vol. 265, 1994. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are no unique answers to these problems. www.Ebook777.com Problems 31 Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. Section 1.2 Dimensions, Dimensional Homogeneity, and Units 1.1 The force, F, of the wind blowing against a building is given by F = CD ρV 2 A/2, where V is the wind speed, ρ the density of the air, A the cross-sectional area of the building, and CD is a constant termed the drag coefficient. Determine the dimensions of the drag coefficient. 1.2 The Mach number is a dimensionless ratio of the velocity of an object in a fluid to the speed of sound in the fluid. For an airplane flying at velocity V in air at absolute temperature T, the Mach number Ma is Ma = V flow rate is given in units of mass per unit time, show that the momentum flux can be expressed in units of force. 1.12 An equation for the frictional pressure loss Δ p (inches H2O) in a circular duct of inside diameter d (in.) and length L (ft) for air flowing with velocity V (ft/min) is Δp = 0.027( V 1.82 , d 1.22 )( V0 ) L where V0 is a reference velocity equal to 1000 ft/min. Find the units of the “constant” 0.027. 1.13 The volume rate of flow, Q, through a pipe containing a slowly moving liquid is given by the equation Q= πR4 Δp 8μℓ where k is a dimensionless constant and R is the specific gas constant for air. Show that Ma is dimensionless. where R is the pipe radius, Δp the pressure drop along the pipe, μ a fluid property called viscosity (FL−2T), and ℓ the length of pipe. What are the dimensions of the constant π/8? Would you classify this equation as a general homogeneous equation? Explain. 1.3 Verify the dimensions, in both the FLT and MLT systems, of the following quantities which appear in Table 1.1: (a) volume, (b) acceleration, (c) mass, (d) moment of inertia (area), and (e) work. 1.14 Show that each term in the following equation has units of lb/ft3. Consider u a velocity, y a length, x a length, p a pressure, and μ an absolute viscosity. √kRT , 1.4 Verify the dimensions, in both the FLT and MLT systems, of the following quantities which appear in Table 1.1: (a) angular velocity, (b) energy, (c) moment of inertia (area), (d) power, and (e) pressure. 1.5 Verify the dimensions, in both the FLT system and the MLT system, of the following quantities which appear in Table 1.1: (a) frequency, (b) stress, (c) strain, (d) torque, and (e) work. 1.6 If u is a velocity, x a length, and t a time, what are the dimensions (in the MLT system) of (a) ∂u∂t, (b) ∂ 2u∂x∂t, and (c) ʃ ( ∂u/∂t) dx? 1.7 Verify the dimensions, in both the FLT system and the MLT system, of the following quantities which appear in Table 1.1: (a) acceleration, (b) stress, (c) moment of a force, (d) volume, and (e) work. 1.8 If p is a pressure,V a velocity, and ρ a fluid density, what are the dimensions (in the MLT system) of (a) p/ρ, (b) pVρ, and (c) p/ρV2 ? 1.9 If P is a force and x a length, what are the dimensions (in the FLT system) of (a) dP/dx, (b) d3P/dx3, and (c) ʃ P dx? 1.10 If V is a velocity, ℓ a length, and ν a fluid property (the kinematic viscosity) having dimensions of L2T −1, which of the following combinations are dimensionless: (a) Vℓν, (b) Vℓν, (c) V 2ν, (d) Vℓν?   1.11 The momentum flux (discussed in Chapter 5) is given by the ˙ V, where m˙ is mass flow rate and V is velocity. If mass product m 0=− ∂p ∂ 2u + μ 2. ∂x ∂y 1.15 The pressure difference, Δp, across a partial blockage in an artery (called a stenosis) is approximated by the equation Δp = Kυ 2 μV A0 + Ku( − 1) ρV 2 D A1 where V is the blood velocity, μ the blood viscosity (FL−2T ), ρ the blood density (ML−3 ), D the artery diameter, A0 the area of the unobstructed artery, and A1 the area of the stenosis. Determine the dimensions of the constants Kυ and Ku. Would this equation be valid in any system of units? 1.16 Assume that the speed of sound, c, in a fluid depends on an elastic modulus, Eυ , with dimensions FL−2, and the fluid density, ρ, in the form c = (Eυ ) a (ρ) b. If this is to be a dimensionally homogeneous equation, what are the values for a and b? Is your result consistent with the standard formula for the speed of sound? (See Eq. 1.19.) 1.17 A formula to estimate the volume rate of flow, Q, flowing over a dam of length, B, is given by the equation Q = 3.09 BH3/2 where H is the depth of the water above the top of the dam (called the head). This formula gives Q in ft3/s when B and H are in feet. Is the constant, 3.09, dimensionless? Would this equation be valid if units other than feet and seconds were used? 32 Chapter 1 ■ Introduction 1.18 A commercial advertisement shows a pearl falling in a bottle of shampoo. If the diameter D of the pearl is quite small and the shampoo sufficiently viscous, the drag 𝒟 on the pearl is given by Stokes’s law, 𝒟 = 3πμVD, where V is the speed of the pearl and μ is the fluid viscosity. Show that the term on the right side of Stokes’s law has units of force. †1.19 Cite an example of a restricted homogeneous equation contained in a technical article found in an engineering journal in your field of interest. Define all terms in the equation, explain why it is a restricted equation, and provide a complete journal citation (title, date, etc.). 1.20 Express the following quantities in SI units: (a) 10.2 in.min, (b) 4.81 slugs, (c) 3.02 lb, (d) 73.1 fts2, (e) 0.0234 lb ∙ sft2. 1.21 Express the following quantities in BG units: (a) 14.2 km, (b) 8.14 Nm3, (c) 1.61 kgm3, (d) 0.0320 N ∙ ms, (e) 5.67 mmhr. newtons) when located on a planet with an acceleration of gravity equal to 4.0 fts2. 1.33 The density of a certain type of jet fuel is 775 kg/m3. Determine its specific gravity and specific weight. 1.34 At 4 °C a mixture of automobile antifreeze (50% water and 50% ethylene glycol by volume) has a density of 1064 kg/m3. If the water density is 1000 kg/m3, find the density of the ethylene glycol. 1.35 A hydrometer is used to measure the specific gravity of liquids. (See Video V2.8.) For a certain liquid, a hydrometer reading indicates a specific gravity of 1.15.What is the liquid’s density and specific weight?Express your answer in SI units. 1.36 An open, rigid-walled, cylindrical tank contains 4 ft3 of water at 40 °F. Over a 24-hour period of time the water temperature varies from 40 to 90 °F. Make use of the data in Appendix B to determine how much the volume of water will change. For a tank diameter of 2 ft, would the corresponding change in water depth be very noticeable? Explain. 1.22 Express the following quantities in SI units: (a) 160 acres, †1.37 Estimate the number of pounds of mercury it would take to fill your bathtub. List all assumptions and show all calculations. (b) 15 gallons (U.S.), (c) 240 miles, (d) 79.1 hp, (e) 60.3 °F. 1.23 Water flows from a large drainage pipe at a rate of 1.38 A mountain climber’s oxygen tank contains 1 lb of oxygen trip at sea level where the acceleration of grav1200 galmin. What is this volume rate of flow in (a) m3s, when he begins his 2 ity is 32.174 ft/s . What is the weight of the oxygen in the tank 3 (b) litersmin, and (c) ft s? when he reaches the top of Mt. Everest where the acceleration of 1.24 The universal gas constant R0 is equal to 49,700 ft2(s2 ∙ °R), gravity is 32.082 ft/s2? Assume that no oxygen has been removed or 8310 m2/(s2 ∙ K). Show that these two magnitudes are equal. from the tank; it will be used on the descent portion of the climb. 1.25 Dimensionless combinations of quantities (commonly called dimensionless parameters) play an important role in fluid mechanics. Make up five possible dimensionless parameters by using combinations of some of the quantities listed in Table 1.1. 1.39 The information on a can of pop indicates that the can contains 355 mL. The mass of a full can of pop is 0.369 kg, while an empty can weighs 0.153 N. Determine the specific weight, density, and specific gravity of the pop and compare your results with the corresponding values for water at 20 °C. Express your results in SI units. 1.26 An important dimensionless parameter in certain types of fluid flow problems is the Froude number defined as V √gℓ, *1.40 The variation in the density of water, ρ, with temperature, T, where V is a velocity, g the acceleration of gravity, and ℓ a length. in the range 20 °C ≤ T ≤ 50 °C, is given in the following table. Determine the value of the Froude number for V = 10 fts, g = 32.2 fts2, and ℓ = 2 ft. Recalculate the Froude number using Density (kgm3) 998.2 997.1 995.7 994.1 992.2 990.2 988.1 SI units for V, g, and ℓ. Explain the significance of the results of Temperature (°C) 20 25 30 35 40 45 50 these calculations. Section 1.4 Measures of Fluid Mass and Weight 1.27 Obtain a photograph/image of a situation in which the density or specific weight of a fluid is important. Print this photo and write a brief paragraph that describes the situation involved. 1.28 A tank contains 500 kg of a liquid whose specific gravity is 2. Determine the volume of the liquid in the tank. 1.29 A stick of butter at 35 °F measures 1.25 in. × 1.25 in. × 4.65 in. and weighs 4 ounces. Find its specific weight. Use these data to determine an empirical equation of the form ρ = c1 + c2T + c3T 2 which can be used to predict the density over the range indicated. Compare the predicted values with the data given. What is the density of water at 42.1 °C? 1.41 If 1 cup of cream having a density of 1005 kg/m3 is turned into 3 cups of whipped cream, determine the specific gravity and specific weight of the whipped cream. 1.42 With the exception of the 410 bore, the gauge of a shotgun barrel indicates the number of round lead balls, each having the bore diameter of the barrel, that together weigh 1 lb. For example, a 1 shotgun is called a 12-gauge shotgun if a 12-lb lead ball fits the bore of the barrel. Find the diameter of a 12-gauge shotgun in inches and millimeters. Lead has a specific weight of 0.411 lb/in3. 1.30 Clouds can weigh thousands of pounds due to their liquid water content. Often this content is measured in grams per cubic meter (g/m3). Assume that a cumulus cloud occupies a volume of one cubic kilometer, and its liquid water content is 0.2 g/m3. †1.43 The presence of raindrops in the air during a heavy rainstorm (a) What is the volume of this cloud in cubic miles? (b) How much increases the average density of the air–water mixture. Estimate by does the water in the cloud weigh in pounds? what percent the average air–water density is greater than that of 1.31 A tank of oil has a mass of 25 slugs. (a) Determine its weight just still air. State all assumptions and show calculations. in pounds and in newtons at the Earth’s surface. (b) What would be its mass (in slugs) and its weight (in pounds) if located on the Section 1.5 Ideal Gas Law moon’s surface where the gravitational attraction is approximately 1.44 A regulation basketball is initially flat and is then inflated one-sixth that at the Earth’s surface? to a pressure of approximately 24 lb/in2 absolute. Consider the air 1.32 A certain object weighs 300 N at the Earth’s surface. temperature to be constant at 70 °F. Find the mass of air required Determine the mass of the object (in kilograms) and its weight (in to inflate the basketball. The basketball’s inside radius is 4.67 in. Problems 1.45 Nitrogen is compressed to a density of 4 kg/m3 under an absolute pressure of 400 kPa. Determine the temperature in degrees Celsius. 1.46 The temperature and pressure at the surface of Mars during a Martian spring day were determined to be −50 °C and 900 Pa, respectively. (a) Determine the density of the Martian atmosphere for these conditions if the gas constant for the Martian atmosphere is assumed to be equivalent to that of carbon dioxide. (b) Compare the answer from part (a) with the density of the Earth’s atmosphere during a spring day when the temperature is 18 °C and the pressure 101.6 kPa (abs). 1.47 A closed tank having a volume of 2 ft3 is filled with 0.30 lb of a gas. A pressure gage attached to the tank reads 12 psi when the gas temperature is 80 °F. There is some question as to whether the gas in the tank is oxygen or helium. Which do you think it is? Explain how you arrived at your answer. 1.48 Assume that the air volume in a small automobile tire is constant and equal to the volume between two concentric cylinders 13 cm high with diameters of 33 cm and 52 cm. The air in the tire is initially at 25 °C and 202 kPa. Immediately after air is pumped into the tire, the temperature is 30 °C and the pressure is 303 kPa. What mass of air was added to the tire? What would be the air pressure after the air has cooled to a temperature of 0 °C? 33 1.55 For flowing water, what is the magnitude of the velocity gradient needed to produce a shear stress of 1.0 N/m2? 1.56 Make use of the data in Appendix B to determine the dynamic viscosity of glycerin at 85 °F. Express your answer in both SI and BG units. One type of capillary-tube viscometer is shown in Video 1.57 V1.5 and in Fig. P1.57. For this device the liquid to be tested is drawn into the tube to a level above the top etched line. The time is then obtained for the liquid to drain to the bottom etched line.The kinematic viscosity, ν, in m2/s is then obtained from the equation ν = KR4t where K is a constant, R is the radius of the capillary tube in mm, and t is the drain time in seconds. When glycerin at 20 °C is used as a calibration fluid in a particular viscometer, the drain time is 1430 s. When a liquid having a density of 970 kg/m3 is tested in the same viscometer the drain time is 900 s. What is the dynamic viscosity of this liquid? Glass strengthening bridge 1.49 A compressed air tank contains 5 kg of air at a temperature of 80 °C. A gage on the tank reads 300 kPa. Determine the volume of the tank. Etched lines 1.50 A rigid tank contains air at a pressure of 90 psia and a temperature of 60 °F. By how much will the pressure increase as the temperature is increased to 110 °F? Capillary tube 1.51 The density of oxygen contained in a tank is 2.0 kg/m3 when the temperature is 25 °C. Determine the gage pressure of the gas if the atmospheric pressure is 97 kPa. 1.52 The helium-filled blimp shown in Fig. P1.52 is used at various athletic events. Determine the number of pounds of helium within it if its volume is 68,000 ft3 and the temperature and pressure are 80 °F and 14.2 psia, respectively. ■ Figure P1.52 *1.53 Develop a computer program for calculating the density of an ideal gas when the gas pressure in pascals (abs), the temperature in degrees Celsius, and the gas constant in Jkg ∙ K are specified. Plot the density of helium as a function of temperature from 0 °C to 200 °C and pressures of 50, 100, 150, and 200 kPa (abs). ■ Figure P1.57 1.58 The viscosity of a soft drink was determined by using a capillary tube viscometer similar to that shown in Fig. P1.58 and Video V1.5. For this device the kinematic viscosity, ν, is directly proportional to the time, t, that it takes for a given amount of liquid to flow through a small capillary tube. That is, ν = Kt. The following data were obtained from regular pop and diet pop. The corresponding measured specific gravities are also given. Based on these data, by what percent is the absolute viscosity, μ, of regular pop greater than that of diet pop? Regular pop Diet pop t(s) 377.8 300.3 SG 1.044 1.003 1.59 The viscosity of a certain fluid is 5 × 10−4 poise. Determine its viscosity in both SI and BG units. 1.60 The kinematic viscosity and specific gravity of a liquid are 3.5 × 10−4 m2/s and 0.79, respectively. What is the dynamic viscosity of the liquid in SI units? Section 1.6 Viscosity (also see Lab Problems 1.1LP and 1.2LP) 1.61 A liquid has a specific weight of 59 lb/ft3 and a dynamic viscosity of 2.75 lb ∙ s/ft2. Determine its kinematic viscosity. 1.54 Obtain a photograph/image of a situation in which the viscosity of a fluid is important. Print this photo and write a brief paragraph that describes the situation involved. 1.62 The kinematic viscosity of oxygen at 20 °C and a pressure of 150 kPa (abs) is 0.104 stokes. Determine the dynamic viscosity of oxygen at this temperature and pressure. 34 Chapter 1 ■ Introduction *1.63 Fluids for which the shearing stress, τ, is not linearly related to the rate of shearing strain, γ˙ , are designated as nonNewtonian fluids. Such fluids are commonplace and can exhibit unusual behavior, as shown in Video V1.6. Some experimental data obtained for a particular non-Newtonian fluid at 80 °F are shown below. τ (lb/ft2) γ˙ (s−1) 0 0 2.11 50 7.82 100 18.5 150 31.7 200 Calculate the value of the viscosity at 50 °C and compare with the value given in Table B.2. (Hint: Rewrite the equation in the form ln μ = (B) 1 + ln D T and plot ln μ versus 1T. From the slope and intercept of this curve, B and D can be obtained. If a nonlinear curve-fitting program is available, the constants can be obtained directly from Eq. 1.11 without rewriting the equation.) Plot these data and fit a second-order polynomial to the data using a suitable graphing program. What is the apparent viscosity of this fluid when the rate of shearing strain is 70 s −1 ? Is this apparent viscosity larger or smaller than that for water at the same temperature? 1.71 For a certain liquid μ = 7.1 × 10−5 lb ∙ s/ft2 at 40 °F and μ = 1.9 × 10−5 lb ∙ s/ft2 at 150 °F. Make use of these data to determine the constants D and B which appear in Andrade’s equation (Eq. 1.11). What would be the viscosity at 80 °F? 1.64 Water flows near a flat surface and some measurements of the water velocity, u, parallel to the surface, at different heights, y, above the surface are obtained. At the surface y = 0. After an analysis of the data, the lab technician reports that the velocity distribution in the range 0 < y < 0.1 ft is given by the equation 1.72 For a parallel plate arrangement of the type shown in Fig. 1.5 it is found that when the distance between plates is 2 mm, a shearing stress of 150 Pa develops at the upper plate when it is pulled at a velocity of 1 m/s. Determine the viscosity of the fluid between the plates. Express your answer in SI units. u = 0.81 + 9.2y + 4.1 × 103y3 1.73 Two flat plates are oriented parallel above a fixed lower plate as shown in Fig. P1.73. The top plate, located a distance b above the fixed plate, is pulled along with speed V. The other thin plate is located a distance cb, where 0 < c < 1, above the fixed plate. This plate moves with speed V1, which is determined by the viscous shear forces imposed on it by the fluids on its top and bottom. The fluid on the top is twice as viscous as that on the bottom. Plot the ratio V1/V as a function of c for 0 < c < 1. with u in ft/s when y is in ft. (a) Do you think that this equation would be valid in any system of units? Explain. (b) Do you think this equation is correct? Explain. You may want to look at Video 1.4 to help you arrive at your answer. 1.65 Calculate the Reynolds numbers for the flow of water and for air through a 4-mm-diameter tube, if the mean velocity is 3 ms and the temperature is 30 °C in both cases (see Example 1.4). Assume the air is at standard atmospheric pressure. 1.66 SAE 30 oil at 60 °F flows through a 2-in.-diameter pipe with a mean velocity of 5 ft/s. Determine the value of the Reynolds number (see Example 1.4). 1.67 For air at standard atmospheric pressure the values of the constants that appear in the Sutherland equation (Eq. 1.10) are C = 1.458 × 10−6 kg(m ∙ s ∙ K12 ) and S = 110.4 K. Use these values to predict the viscosity of air at 10 °C and 90 °C and compare with values given in Table B.4 in Appendix B. *1.68 Use the values of viscosity of air given in Table B.4 at temperatures of 0, 20, 40, 60, 80, and 100 °C to determine the constants C and S which appear in the Sutherland equation (Eq. 1.10). Compare your results with the values given in Problem 1.67. (Hint: Rewrite the equation in the form T 3 2 S 1 = ( )T + μ C C   and plot T 3 2μ versus T. From the slope and intercept of this curve, C and S can be obtained.) V 2μ b V1 cb ■ Figure P1.73 1.74 Three large plates are separated by thin layers of ethylene glycol and water, as shown in Fig. P1.74. The top plate moves to the right at 2 m/s. At what speed and in what direction must the bottom plate be moved to hold the center plate stationary? Copper plate   1.69 The viscosity of a fluid plays a very important role in determining how a fluid flows. (See Video V1.3.) The value of the viscosity depends not only on the specific fluid but also on the fluid temperature. Some experiments show that when a liquid, under the action of a constant driving pressure, is forced with a low velocity, V, through a small horizontal tube, the velocity is given by the equation V = K/μ. In this equation K is a constant for a given tube and pressure, and μ is the dynamic viscosity. For a particular liquid of interest, the viscosity is given by Andrade’s equation (Eq. 1.11) with D = 5 × 10 −7 lb ∙ s/ft2 and B = 4000 °R. By what percentage will the velocity increase as the liquid temperature is increased from 40 °F to 100 °F? Assume all other factors remain constant. *1.70 Use the value of the viscosity of water given in Table B.2 at temperatures of 0, 20, 40, 60, 80, and 100 °C to determine the constants D and B which appear in Andrade’s equation (Eq. 1.11). μ 0.1 cm Water (20°C) 2m/s Plastic plate 0.2 cm Steel plate Ethylene glycol (20°C) ■ Figure P1.74 1.75 There are many fluids that exhibit non-Newtonian behavior (see, for example, Video V1.6). For a given fluid the distinction between Newtonian and non-Newtonian behavior is usually based on measurements of shear stress and rate of shearing strain. Assume that the viscosity of blood is to be determined by measurements of shear stress, τ, and rate of shearing strain, du/dy, obtained from a small blood sample tested in a suitable viscometer. Based on the data given below, determine if the blood Problems is a Newtonian or non-Newtonian fluid. Explain how you arrived at your answer. τ(N/m2) du/dy (s 0.04 0.06 −1 0.12 0.18 0.30 0.52 1.12 2.10 ) 2.25 4.50 11.25 22.5 45.0 90.0 450 225 1.76 The sled shown in Fig. P1.76 slides along on a thin horizontal layer of water between the ice and the runners. The horizontal force that the water puts on the runners is equal to 1.2 lb when the sled’s speed is 50 ft/s. The total area of both runners in contact with the water is 0.08 ft2, and the viscosity of the water is 3.5 × 10−5 lb ∙ s/ft2. Determine the thickness of the water layer under the runners. Assume a linear velocity distribution in the water layer. 35 0.1 mm gap V 20° ■ Figure P1.80 1.81 A layer of water flows down an inclined fixed surface with the velocity profile shown in Fig. P1.81. Determine the magnitude and direction of the shearing stress that the water exerts on the fixed surface for U = 2 m/s and h = 0.1 m. h U u y u y y2 __ = 2 __ – __2 U ■ Figure P1.76 h h ■ Figure P1.81 1.77 A 25-mm-diameter shaft is pulled through a cylindrical bearing as shown in Fig. P1.77. The lubricant that fills the 0.3-mm gap between the shaft and bearing is an oil having a kinematic viscosity of 8.0 × 10−4 m2/s and a specific gravity of 0.91. Determine the force P required to pull the shaft at a velocity of 3 m/s. Assume the velocity distribution in the gap is linear. 1.82 Oil (absolute viscosity = 0.0003 lb ∙ s /ft2, density = 50 lbm/ft3) flows in the boundary layer, as shown in Fig. P1.82. The plate is 1 ft wide perpendicular to the paper. Calculate the shear stress at the plate surface. Edge of boundary layer Bearing Shaft Lubricant y u u = 1500 y – 5 × 106 × y3 ft/sec 0.01 ft P Plate ■ Figure P1.82 0.5 m ■ Figure P1.77 1.78 A hydraulic lift in a service station has a 32.50-cm-diameter ram that slides in a 32.52-cm-diameter cylinder. The annular space is filled with SAE 10 oil at 20 °C. The ram is traveling upward at the rate of 0.10 m/s. Find the frictional force when 3.0 m of the ram is engaged in the cylinder. 1.79 A piston having a diameter of 5.48 in. and a length of 9.50 in. slides downward with a velocity V through a vertical pipe. The downward motion is resisted by an oil film between the piston and the pipe wall. The film thickness is 0.002 in., and the cylinder weighs 0.5 lb. Estimate V if the oil viscosity is 0.016 lb ∙ s/ft2. Assume the velocity distribution in the gap is linear. 1.80 A 10-kg block slides down a smooth inclined surface as shown in Fig. P1.80. Determine the terminal velocity of the block if the 0.1-mm gap between the block and the surface contains SAE 30 oil at 60 °F. Assume the velocity distribution in the gap is linear, and the area of the block in contact with the oil is 0.1 m2. *1.83 Standard air flows past a flat surface, and velocity measurements near the surface indicate the following distribution: y (ft) 0.005 0.01 0.02 0.04 0.06 0.08 u (fts) 0.74 1.51 3.03 6.37 10.21 14.43 The coordinate y is measured normal to the surface and u is the velocity parallel to the surface. (a) Assume the velocity distribution is of the form u = C1y + C2y3 and use a standard curve-fitting technique to determine the constants C1 and C2. (b) Make use of the results of part (a) to determine the magnitude of the shearing stress at the wall (y = 0) and at y = 0.05 ft. 1.84 A new computer drive is proposed to have a disc, as shown in Fig. P1.84. The disc is to rotate at 10,000 rpm, and the reader head is to be positioned 0.0005 in. above the surface of the disc. 36 Chapter 1 ■ Introduction Estimate the shearing force on the reader head as a result of the air between the disc and the head. Stationary reader head Torque (ft ∙ lb) Angular velocity (rads) 13.1 26.0 39.5 52.7 64.9 78.6 1.0 2.0 3.0 4.0 5.0 6.0 For this viscometer Ro = 2.50 in., Ri = 2.45 in., and ℓ = 5.00 in. Make use of these data and a standard curve-fitting program to determine the viscosity of the liquid contained in the viscometer. 0.2-in.dia. 10,000 rpm 0.0005 in. 2 in. Rotating disc 𝒯 Liquid Fixed outer cylinder ω Ri ■ Figure P1.84 Ro Rotating inner cylinder 1.85 The space between two 6-in.-long concentric cylinders is filled with glycerin (viscosity = 8.5 × 10−3 lb ∙ sft2 ). The inner cylinder has a radius of 3 in. and the gap width between cylinders is 0.1 in. Determine the torque and the power required to rotate the inner cylinder at 180 revmin. The outer cylinder is fixed. Assume the velocity distribution in the gap to be linear. 1.86 A pivot bearing used on the shaft of an electrical instrument is shown in Fig. P1.86. An oil with a viscosity of μ = 0.010 lb ∙ s/ft2 fills the 0.001-in. gap between the rotating shaft and the stationary base. Determine the frictional torque on the shaft when it rotates at 5000 rpm. 5000 rpm ℓ ■ Figure P1.87 1.88 The concentric cylinder viscometer shown in Fig. P1.88 has a cylinder height of 10.0 cm, a cylinder radius of 3.0 cm, and a uniform gap between the cylinder and the container (bottom and sides) of 0.10 cm. The pulley has a radius of 3.0 cm. Determine the weight required to produce a constant rotational speed of 30 rpm if the gap is filled with: (a) water, (b) gasoline, (c) glycerin. 3.0 cm 0.2 in. 𝒲 H = 10.0 cm 30° 0.001 in. 0.10 cm 1.87 The viscosity of liquids can be measured through the use of a rotating cylinder viscometer of the type illustrated in Fig. P1.87. In this device the outer cylinder is fixed and the inner cylinder is rotated with an angular velocity, ω. The torque 𝒯 required to develop ω is measured and the viscosity is calculated from these two measurements. (a) Develop an equation relating μ, ω, 𝒯, ℓ, Ro, and Ri. Neglect end effects and assume the velocity distribution in the gap is linear. (b) The following torque-angular velocity data were obtained with a rotating cylinder viscometer of the type discussed in part (a). 0.10 cm 3.0 cm μ = 0.010 lb • s/ft2 ■ Figure P1.86 R= ■ Figure P1.88 1.89 A 12-in.-diameter circular plate is placed over a fixed bottom plate with a 0.1-in. gap between the two plates filled with glycerin as shown in Fig. P1.89. Determine the torque required to rotate the circular plate slowly at 2 rpm. Assume that the velocity distribution in the gap is linear and that the shear stress on the edge of the rotating plate is negligible. Torque Rotating plate 0.1-in. gap ■ Figure P1.89 Problems †1.90 Vehicle shock absorbers damp out oscillations caused by road roughness. Describe how a temperature change may affect the operation of a shock absorber. 1.91 Some measurements on a blood sample at 37 °C (98.6 °F) indicate a shearing stress of 0.52 Nm2 for a corresponding rate of shearing strain of 200 s−1 . Determine the apparent viscosity of the blood and compare it with the viscosity of water at the same temperature. Section 1.7 Compressibility of Fluids 1.92 Obtain a photograph/image of a situation in which the compressibility of a fluid is important. Print this photo and write a brief paragraph that describes the situation involved. 1.93 A sound wave is observed to travel through a liquid with a speed of 1500 m/s. The specific gravity of the liquid is 1.5. Determine the bulk modulus for this fluid. 1.94 A rigid-walled cubical container is completely filled with water at 40 °F and sealed. The water is then heated to 100 °F. Determine the pressure that develops in the container when the water reaches this higher temperature. Assume that the volume of the container remains constant and the value of the bulk modulus of the water remains constant and equal to 300,000 psi. 1.95 Estimate the increase in pressure (in psi) required to decrease a unit volume of mercury by 0.1%. 1.96 A 1-m3 volume of water is contained in a rigid container. Estimate the change in the volume of the water when a piston applies a pressure of 35 MPa. 1.97 Determine the speed of sound at 20 °C in (a) air, (b) helium, and (c) natural gas (methane). Express your answer in m/s. 1.98 Air is enclosed by a rigid cylinder containing a piston. A pressure gage attached to the cylinder indicates an initial reading of 25 psi. Determine the reading on the gage when the piston has compressed the air to one-third its original volume. Assume the compression process to be isothermal and the local atmospheric pressure to be 14.7 psi. 1.99 Repeat Problem 1.98 if the compression process takes place without friction and without heat transfer (isentropic process). 1.100 Carbon dioxide at 30 °C and 300 kPa absolute pressure expands isothermally to an absolute pressure of 165 kPa. Determine the final density of the gas. 1.101 Natural gas at 70 °F and standard atmospheric pressure of 14.7 psi (abs) is compressed isentropically to a new absolute pressure of 70 psi. Determine the final density and temperature of the gas. 1.102 A compressed air tank in a service station has a volume of 10 ft3. It contains air at 70 °F and 150 psia. How many tubeless tires can it fill to 44.7 psia at 70 °F if each tire has a volume of 1.5 ft3 and the compressed air tank is not refilled? The tank air temperature remains constant at 70 °F because of heat transfer through the tank’s large surface area. 1.103 Oxygen at 30 °C and 300 kPa absolute pressure expands isothermally to an absolute pressure of 120 kPa. Determine the final density of the gas. 1.104 Compare the isentropic bulk modulus of air at 101 kPa (abs) with that of water at the same pressure. 37 1.105 Develop a computer program for calculating the final gage pressure of gas when the initial gage pressure, initial and final volumes, atmospheric pressure, and the type of process (isothermal or isentropic) are specified. Use BG units. Check your program against the results obtained for Problem 1.98. 1.106 Often the assumption is made that the flow of a certain fluid can be considered as incompressible flow if the density of the fluid changes by less than 2%. If air is flowing through a tube such that the air pressure at one section is 9.0 psi and at a downstream section it is 8.6 psi at the same temperature, do you think that this flow could be considered an incompressible flow? Support your answer with the necessary calculations. Assume standard atmospheric pressure. 1.107 An important dimensionless parameter concerned with very high-speed flow is the Mach number, defined as V/c, where V is the speed of the object such as an airplane or projectile, and c is the speed of sound in the fluid surrounding the object. For a projectile traveling at 800 mph through air at 50 °F and standard atmospheric pressure, what is the value of the Mach number? 1.108 The “power available in the wind” of velocity V through an area A is ˙ = 2ρAV3, W 1 where ρ is the air density (0.075 lbm/ft3). For an 18-mph wind, find the wind area A that will supply a power of 4 hp. 1.109 Air enters the converging nozzle shown in Fig. P1.72 at T1 = 70 °F and V1 = 50 ft/s. At the exit of the nozzle, V2 is given by V2 = √V 21 + 2cp (T1 − T2 ), where cp = 187 ft ∙ lb/lbm ∙ °F and T2 is the air temperature at the exit of the nozzle. Find the temperature T2 for which V2 = 1000 ft/s. V1 V2 2 1 ■ Figure P1.109 1.110 (See The Wide World of Fluids article titled “This water jet is a blast,” Section 1.7.1.) By what percent is the volume of water decreased if its pressure is increased to an equivalent to 3000 atmospheres (44,100 psi)? Section 1.8 Vapor Pressure 1.111 During a mountain climbing trip it is observed that the water used to cook a meal boils at 90 °C rather than the standard 100 °C at sea level. At what altitude are the climbers preparing their meal? (See Tables B.2 and C.2 for data needed to solve this problem.) 1.112 When a fluid flows through a sharp bend, low pressures may develop in localized regions of the bend. Estimate the minimum absolute pressure (in psi) that can develop without causing cavitation if the fluid is water at 160 °F. 38 Chapter 1 ■ Introduction 1.113 A partially filled closed tank contains ethyl alcohol at 68 °F. If the air above the alcohol is evacuated, what is the minimum absolute pressure that develops in the evacuated space? 1.114 Estimate the minimum absolute pressure (in pascals) that can be developed at the inlet of a pump to avoid cavitation if the fluid is carbon tetrachloride at 20 °C. 1.115 When water at 70 °C flows through a converging section of pipe, the pressure decreases in the direction of flow. Estimate the minimum absolute pressure that can develop without causing cavitation. Express your answer in both BG and SI units. 1.116 At what atmospheric pressure will water boil at 35 °C? Express your answer in both SI and BG units. Section 1.9 Surface Tension 1.117 Obtain a photograph/image of a situation in which the surface tension of a fluid is important. Print this photo and write a brief paragraph that describes the situation involved. When a 2-mm-diameter tube is inserted into a liquid in 1.118 an open tank, the liquid is observed to rise 10 mm above the free surface of the liquid (see Video V1.10). The contact angle between the liquid and the tube is zero, and the specific weight of the liquid is 1.2 × 104 N/m3. Determine the value of the surface tension for this liquid. 1.119 A soda straw with an inside diameter of 0.125 in. is inserted into a pan of water at 60 °F. The water in the straw rises to a height of 0.150 in. above the water surface in the pan. Determine the value of θ, the contact angle of the water with the straw (see Fig. 1.10). 1.120 Small droplets of carbon tetrachloride at 68 °F are formed with a spray nozzle. If the average diameter of the droplets is 200 μm, what is the difference in pressure between the inside and outside of the droplets? 1.121 A 12-mm-diameter jet of water discharges vertically into the atmosphere. Due to surface tension the pressure inside the jet will be slightly higher than the surrounding atmospheric pressure. Determine this difference in pressure. 1.122 A method used to determine the surface tension of a liquid is to determine the force necessary to raise a wire ring through the air–liquid interface. What is the value of the surface tension if a force of 0.015 N is required to raise a 4-cm-diameter ring? Consider the ring weightless, as a tensiometer (used to measure the surface tension) “zeroes” out the ring weight. 1.123 Calculate the pressure difference between the inside and 1 outside of a spherical water droplet having a diameter of 32 in. and a temperature of 50 °F. 1.124 As shown in Video V1.9, surface tension forces can be strong enough to allow a double-edge steel razor blade to “float” on water, but a single-edge blade will sink. Assume that the surface tension forces act at an angle θ relative to the water surface as shown in Fig. P1.124. (a) The mass of the double-edge blade is 0.64 × 10 −3 kg, and the total length of its sides is 206 mm. Determine the value of θ required to maintain equilibrium between the blade weight and the resultant surface tension force. (b) The mass of the single-edge blade is 2.61 × 10 −3 kg, and the total length of its sides is 154 mm. Explain why this blade sinks. Support your answer with the necessary calculations. Surface tension force Blade θ ■ Figure P1.24 1.125 Explain how sweat soldering of copper pipe works from a fluid mechanics viewpoint. 1.126 Under the right conditions, it is possible, due to surface tension, to have metal objects float on water. (See Video V1.9.) Consider placing a short length of a small diameter steel (γ = 490 lb/ft3) rod on a surface of water. What is the maximum diameter that the rod can have before it will sink? Assume that the surface tension forces act vertically upward. Note: A standard paper clip has a diameter of 0.036 in. Partially unfold a paper clip and see if you can get it to float on water. Do the results of this experiment support your analysis? 1.127 An open, clean glass tube, having a diameter of 3 mm, is inserted vertically into a dish of mercury at 20 °C (see Video V1.10). How far will the column of mercury in the tube be depressed? 1.128 An open, clean glass tube (θ = 0°) is inserted vertically into a pan of water (see Video V1.10). What tube diameter is needed if the water level in the tube is to rise one tube diameter (due to surface tension)? 1.129 Determine the height that water at 60 °F will rise due to 1 capillary action in a clean, 4 -in.-diameter tube (see Video V1.10). What will be the height if the diameter is reduced to 0.01 in.? 1.130 Two vertical, parallel, clean glass plates are spaced a distance of 2 mm apart. If the plates are placed in water, how high will the water rise between the plates due to capillary action? 1.131 (See The Wide World of Fluids article titled “Walking on water,” Section 1.9.) (a) The water strider bug shown in Fig. P1.131 is supported on the surface of a pond by surface tension acting along the interface between the water and the bug’s legs. Determine the minimum length of this interface needed to support the bug. Assume the bug weighs 10−4 N and the surface tension force acts vertically upwards. (b) Repeat part (a) if surface tension were to support a person weighing 750 N. ■ Figure P1.131 Problems ■ Lifelong Learning Problems 1.1 LL Although there are numerous non-Newtonian fluids that occur naturally (quicksand and blood among them), with the advent of modern chemistry and chemical processing, many new manufactured non-Newtonian fluids are now available for a variety of novel applications. Obtain information about the discovery and use of newly developed non-Newtonian fluids. Summarize your findings in a brief report. 1.2 LL For years, lubricating oils and greases obtained by refining crude oil have been used to lubricate moving parts in a wide variety of machines, motors, and engines. With the increasing cost of crude oil and the potential for the reduced availability of it, the need for non-petroleum-based lubricants has increased considerably. Obtain information about non-petroleum-based lubricants. Summarize your findings in a brief report. 39 1.3 LL It is predicted that nano technology and the use of nano sized objects will allow many processes, procedures, and products that, as of now, are difficult for us to comprehend. Among new nano technology areas is that of nano scale fluid mechanics. Fluid behavior at the nano scale can be entirely different than that for the usual everyday flows with which we are familiar. Obtain information about various aspects of nano fluid mechanics. Summarize your findings in a brief report. Free ebooks ==> www.Ebook777.com 2 Fluid Statics Learning Objectives After completing this chapter, you should be able to: ■ determine the pressure at various locations in a fluid at rest. ■ explain the concept of manometers and apply appropriate equations to determine pressures. ■ calculate the hydrostatic pressure force on a plane or curved submerged surface. ■ calculate the buoyant force and determine the stability of floating or submerged objects. In this chapter we will consider an important class of problems in which the fluid is either at rest or moving in such a manner that there is no relative motion between adjacent particles. In both instances there will be no shearing stresses in the fluid, and the only forces that develop on the surfaces of the particles will be due to the pressure. Thus, our principal concerns are to investigate pressure and its variation throughout a fluid and the force on submerged surfaces due to that pressure variation. The absence of shearing stresses greatly simplifies the analysis and, as we will see, allows us to obtain relatively simple solutions to many important practical problems. 2.1 Pressure at a Point As we briefly discussed in Chapter 1, the term pressure is used to indicate the normal force per unit area at a given point acting on a given plane within the fluid mass of interest. A question that immediately arises is how the pressure at a point varies with the orientation of the plane passing through the point. To answer this question, consider the free-body diagram, illustrated in Fig. 2.1, that represents a small triangular wedge of fluid from some arbitrary location within a fluid mass. Since we are considering the situation in which there are no shearing stresses, the only external forces acting on the wedge are due to the pressure and gravity. For simplicity the forces in the x direction are not shown, and the z axis is taken as the vertical axis so the weight acts in the negative z direction. Although we are primarily interested in fluids at rest, to make the analysis as general as possible, we will allow the fluid element to have accelerated motion. The assumption of zero shearing stresses will still be valid as long as the fluid mass moves as a rigid body; that is, there is no relative motion between adjacent elements. 40 www.Ebook777.com 2.2 Basic Equation for Pressure Field 41 z θ ps δ x δ s δs py δ x δ z y δz θ δx δy x δx δyδz γ ________ 2 pz δ x δ y ■ Figure 2.1 Forces on an arbitrary wedge-shaped element of fluid. The equations of motion (Newton’s second law, F = ma) in the y and z directions are, respectively, δx δy δz ay 2 δx δy δz δx δy δz ∑ Fz = pz δx δy − ps δx δs cos θ − γ =ρ az 2 2 ∑ Fy = py δx δz − ps δx δs sin θ = ρ where ps, py, and pz are the average pressures on the faces, γ and ρ are the fluid specific weight and density, respectively, and ay, az the accelerations. Note that a pressure must be multiplied by an appropriate area to obtain the force due to the pressure. It follows from the geometry that δy = δs cos θ  δz = δs sin θ so that the equations of motion can be rewritten as The pressure at a point in a fluid at rest is independent of direction. py − ps = ρay δy 2 pz − ps = (ρaz + γ) δz 2 Since we are really interested in what is happening at a point, we take the limit as δx, δy, and δz approach zero (while maintaining the angle θ), and it follows that py = ps  pz = ps py = pz py 2.2 pz or ps = py = pz. The angle θ was arbitrarily chosen so we can conclude that the pressure at a point in a fluid at rest, or in motion, is independent of direction as long as there are no shearing stresses present. This important result is known as Pascal’s law, named in honor of Blaise Pascal (1623– 1662), a French mathematician who made important contributions in the field of hydrostatics. Thus, as shown by the photograph in the margin, at the junction of the side and bottom of the beaker, the pressure is the same on the side as it is on the bottom. In Chapter 6 it will be shown that for moving fluids in which there is relative motion between particles (so that shearing stresses develop), the normal stress at a point, which corresponds to pressure in fluids at rest, is not necessarily the same in all directions. In such cases the pressure is defined as the average of any three mutually perpendicular normal stresses at the point. Basic Equation for Pressure Field Although we have answered the question of how the pressure at a point varies with direction, we are now faced with an equally important question—how does the pressure in a fluid in which there are no shearing stresses vary from point to point? To answer this question, consider a small rectangular Free ebooks ==> www.Ebook777.com 42 Chapter 2 ■ Fluid Statics ( ) ∂p δ z p + ––– ––– δ x δ y ∂z 2 z δz ( ( ) ∂p δy p – ––– ––– δ x δ z ∂y 2 δx ) ∂p δy p + ––– ––– δ x δ z ∂y 2 γδ x δ y δ z δy ( ^ k y ^ ^ i ) ∂p δ z p – ––– ––– δ x δ y ∂z 2 j x ■ Figure 2.2 Surface and body forces acting on small fluid element. The pressure may vary across a fluid particle. ∂p δy ––– ––– ∂y 2 element of fluid removed from some arbitrary position within the mass of fluid of interest as illustrated in Fig. 2.2. There are two types of forces acting on this element: surface forces due to the pressure and a body force equal to the weight of the element. Other possible types of body forces, such as those due to magnetic fields, will not be considered in this text. If we let the pressure at the center of the element be designated as p, then the average pressure on the various faces can be expressed in terms of p and its derivatives, as shown in Fig. 2.2. We are actually using a Taylor series expansion of the pressure at the element center to approximate the pressures a short distance away and neglecting higher order terms that will vanish as we let δx, δy, and δz approach zero. This is illustrated by the figure in the margin. For simplicity the surface forces in the x direction are not shown. The resultant surface force in the y direction is ∂p δy ∂p δy δFy = (p − δx δz − p + δx δz ( ∂y 2 ) ∂y 2 ) p or δy ––– 2 y δFy = − ∂p δx δy δz ∂y Similarly, for the x and z directions the resultant surface forces are δFx = − ∂p ∂p δx δy δz  δFz = − δx δy δz ∂x ∂z The resultant surface force acting on the element can be expressed in vector form as δFs = δFxˆi + δFy ˆj + δFzkˆ The resultant surface force acting on a small fluid element depends only on the pressure gradient if there are no shearing stresses present. or δFs = −( ∂p ∂p ∂p ˆi + jˆ + kˆ δx δy δz ∂x ∂y ∂z ) (2.1) where ˆi, ˆj, and kˆ are the unit vectors along the coordinate axes shown in Fig. 2.2. The group of terms in parentheses in Eq. 2.1 represents in vector form the pressure gradient and can be written as ∂p ∂p ∂p ˆi + ˆj + kˆ = ∇p ∂x ∂y ∂z www.Ebook777.com 2.3 Pressure Variation in a Fluid at Rest 43 where ∇ ( ) = ∂ ( ) ˆ ∂ ( ) ˆ ∂ ( ) ˆ i+ j+ k ∂x ∂y ∂z and the symbol ∇ is the gradient or “del” vector operator. Thus, the resultant surface force per unit volume can be expressed as δFs = −∇p δx δy δz Since the z axis is vertical, the weight of the element is −δ 𝒲kˆ = −γ δx δy δz kˆ where the negative sign indicates that the force due to the weight is downward (in the negative z direction). Newton’s second law, applied to the fluid element, can be expressed as ∑ δF = δm a where ∑ δF represents the resultant force acting on the element, a is the acceleration of the element, and δm is the element mass, which can be written as ρ δx δy δz. It follows that ∑ δF = δFs − δ 𝒲kˆ = δm a or −∇p δx δy δz − γ δx δy δz kˆ = ρ δx δy δz a and, therefore, −∇p − γkˆ = ρa (2.2) Equation 2.2 is the general equation of motion for a fluid in which there are no shearing stresses. We will use this equation in Section 2.12 when we consider the pressure distribution in a moving fluid. For the present, however, we will restrict our attention to the special case of a fluid at rest. 2.3 Pressure Variation in a Fluid at Rest For a fluid at rest a = 0 and Eq. 2.2 reduces to ∇p + γkˆ = 0 or in component form ∂p ∂p ∂p = 0   = 0   = −γ ∂x ∂y ∂z For liquids or gases at rest, the pressure gradient in the vertical direction at any point in a fluid depends only on the specific weight of the fluid at that point. (2.3) These equations show that for the coordinates defined, pressure is not a function of x or y. Thus, as we move from point to point in a horizontal plane (any plane parallel to the x–y plane), the pressure does not change. Since p depends only on z, the last of Eqs. 2.3 can be written as the ordinary differential equation dp = −γ dz (2.4) Equation 2.4 is the fundamental equation for fluids at rest and can be used to determine how pressure changes with elevation. This equation and the figure in the margin indicate that the pressure gradient in the vertical direction is negative; that is, the pressure decreases as we move upward in a fluid at rest. There is no requirement that γ be a constant. Thus, it is valid for fluids with constant specific weight, such as liquids, as well as fluids whose specific weight may vary with elevation, 44 z Chapter 2 ■ Fluid Statics dp Δ p= −γ ––– = ––– dz Δz p pressure increases with decreasing elevation such as air or other gases. However, to proceed with the integration of Eq. 2.4 it is necessary to stipulate how the specific weight varies with z. If the fluid is flowing (i.e., not at rest with a = 0), then the pressure variation is usually much more complex than that given by Eq. 2.4. For example, the pressure distribution on your car as it is driven along the road varies in a complex manner with x, y, and z. This idea is covered in detail in Chapters 3, 6, and 9. 2.3.1 Incompressible Fluid Since the specific weight is equal to the product of fluid density and acceleration due to gravity (γ = ρg), changes in γ are caused by a change in either ρ or g. For most engineering applications the variation in g is negligible, so our main concern is with the possible variation in the fluid density. In general, a fluid with constant density is called an incompressible fluid. For liquids the variation in density is usually negligible, even for large changes in pressure, so that the assumption of constant specific weight when dealing with liquids is usually a good one. For this instance, Eq. 2.4 can be directly integrated ∫ p2 dp = −γ p1 ∫ z2 dz z1 to yield p2 − p1 = −γ(z2 − z1 ) or p1 − p2 = γ(z2 − z1 ) VIDEO V2.1 Pressure on a car pA = 0 A = 1 in.2 23.1 ft 𝒲 = 10 lb where p1 and p2 are pressures at the vertical elevations z1 and z2, as is illustrated in Fig. 2.3. Equation 2.5 can be written in the compact form p1 − p2 = γh (2.6) p1 = γh + p2 (2.7) or where h is the distance, z2 − z1, which is the depth of fluid measured downward from the location of p2. This type of pressure distribution is commonly called a hydrostatic distribution, and Eq. 2.7 shows that in an incompressible fluid at rest the pressure varies linearly with depth. The pressure must increase with depth to “hold up” the fluid above it. It can also be observed from Eq. 2.6 that the pressure difference between two points can be specified by the distance h since Water h= p1 − p2 γ pA = 10 lb VIDEO (2.5) V2.2 Demonstration of atmospheric pressure. Free surface (pressure = p0) p2 z h = z2 – z1 z2 p1 z1 y x ■ Figure 2.3 Notation for pressure variation in a fluid at rest with a free surface. 2.3 Pressure Variation in a Fluid at Rest 45 In this case h is called the pressure head and is interpreted as the height of a column of fluid of specific weight γ required to give a pressure difference p1 − p2. For example, a pressure difference of 10 psi can be specified in terms of pressure head as 23.1 ft of water (γ = 62.4 lbft3 ), or 518 mm of Hg (γ = 133 kNm3 ). As illustrated by the figure in the margin, a 23.1-ft-tall column of water with a cross-sectional area of 1 in.2 weighs 10 lb. THE WIDE WORLD OF FLUIDS Giraffe’s blood pressure A giraffe’s long neck allows it to graze leg regions, giraffes have a tight sheath of thick skin over their up to 6 m above the ground. It can also lower its head to drink at lower limbs that acts like an elastic bandage in exactly the same ground level. Thus, in the circulatory system there is a significant way as do the g-suits of fighter pilots. In addition, valves in the hydrostatic pressure effect due to this elevation change. To upper neck prevent backflow into the head when the giraffe low- maintain blood flow to its head throughout this change in eleva- ers its head to ground level. It is also thought that blood vessels tion, the giraffe must maintain a relatively high blood pressure at in the giraffe’s kidney have a special mechanism to prevent large heart level—approximately two and a half times that in humans. changes in filtration rate when blood pressure increases or To prevent rupture of blood vessels in the high-pressure lower decreases with its head movement. (See Problem 2.17.) When one works with liquids there is often a liquid-vapor interface we call a free surface, as is illustrated in Fig. 2.3, and it is convenient to use this surface as a reference plane. The reference pressure p0 would correspond to the pressure acting on the free surface (which would frequently be atmospheric pressure). If we let p2 = p0 in Eq. 2.7 it follows that the pressure p at any depth h below the free surface is given by the equation: p = γh + p0 (2.8) As is demonstrated by Eq. 2.7 or 2.8, the pressure in a homogeneous, incompressible fluid at rest depends only on the depth of the fluid relative to some reference plane, and it is not influenced by the size or shape of container in which the fluid is held. Thus, in Fig. 2.4 the pressure is the same at all points along the line AB, even though the containers have very irregular shapes. The actual value of the pressure along AB depends only on the depth, h, the surface pressure, p0, and the specific weight, γ, of the liquid in the container. Liquid surface (p = p0) h B A Specific weight γ ■ Figure 2.4 Fluid pressure in containers of arbitrary shape. 46 Chapter 2 ■ Fluid Statics EXAMPLE 2.1 Pressure–Depth Relationship Open GIVEN Because of a leak in a buried gasoline storage tank, water has seeped in to the depth shown in Fig. E2.1. The specific gravity of the gasoline is SG = 0.68. FIND Determine the pressure at the gasoline–water interface and at the bottom of the tank. Express the pressure in units of lbft2, lbin.2, and as a pressure head in feet of water. 17 ft Gasoline SOLUTION (1) Since we are dealing with liquids at rest, the pressure distribution will be hydrostatic, and therefore the pressure variation can be found from the equation: p = γh + p0 With p0 corresponding to the pressure at the free surface of the gasoline, then the pressure at the interface is p1 = (SG) (γH2O ) h + p0 = (0.68) (62.4 lbft3 ) (17 ft) + p0 = 721 + p0 (lbft2 ) If we measure the pressure relative to atmospheric pressure (gage pressure), it follows that p0 = 0, and therefore p1 = 721 lbft2 (Ans) 2 p1 = 721 lbft 144 in.2ft2 = 5.01 lbin.2 p1 721 lbft2 = = 11.6 ft γH2O 62.4 lbft3 The transmission of pressure throughout a stationary fluid is the principle upon which many hydraulic devices are based. (Ans) (Ans) Water (2) 3 ft ■ Figure E2.1 It is noted that a rectangular column of water 11.6 ft tall and 1 ft2 in cross section weighs 721 lb. A similar column with a 1-in.2 cross section weighs 5.01 lb. We can now apply the same relationship to determine the pressure at the tank bottom; that is, p2 = γH2O hH2O + p1 = (62.4 lbft3 ) (3 ft) + 721 lbft2 = 908 lbft2 908 lbft2 p2 = = 6.31 lbin.2 144 in.2ft2 p2 908 lbft2 = = 14.6 ft γH2O 62.4 lbft3 (Ans) (Ans) (Ans) COMMENT Observe that if we wish to express these pressures in terms of absolute pressure, we would have to add the local atmospheric pressure (in appropriate units) to the previous results. A further discussion of gage and absolute pressure is given in Section 2.5. The required equality of pressures at equal elevations throughout a system is important for the operation of hydraulic jacks (see Fig. 2.5a), lifts, and presses, as well as hydraulic controls on aircraft and other types of heavy machinery. The fundamental idea behind such devices and systems is demonstrated in Fig. 2.5b. A piston located at one end of a closed system filled with a liquid, such as oil, can be used to change the pressure throughout the system, and thus transmit an applied force F1 to a second piston where the resulting force is F2. Since the pressure p acting on the faces of both pistons is the same (the effect of elevation changes is usually negligible for this type of hydraulic device), it follows that F2 = (A2 A1 )F1. The piston area A2 can be made much larger than A1 and therefore a large mechanical advantage can be developed; that is, a small force applied at the smaller piston can be used to develop a large force at the larger piston. The applied force could be created manually through some type of mechanical device, such as a hydraulic jack, or through compressed air acting directly on the surface of the liquid, as is done in hydraulic lifts commonly found in service stations. 2.3.2 Compressible Fluid We normally think of gases such as air, oxygen, and nitrogen as being compressible fluids because the density of the gas can change significantly with modest changes in pressure and temperature. Thus, although Eq. 2.4 applies at a point in a gas, it is necessary to consider the possible variation in γ before the equation can be integrated. However, as was discussed in Chapter 1, the specific 2.3 Pressure Variation in a Fluid at Rest 47 A2 F2 = pA2 F1 = pA1 A1 A2 A1 (b) (a) ■ Figure 2.5 (a) Hydraulic jack, (b) Transmission of fluid pressure. weights of common gases are small when compared with those of liquids. For example, the specific weight of air at sea level and 60 °F is 0.0763 lbft3, whereas the specific weight of water under the same conditions is 62.4 lbft3. Since the specific weights of gases are comparatively small, it follows from Eq. 2.4 that the pressure gradient in the vertical direction is correspondingly small, and even over distances of several hundred feet the pressure will remain essentially constant for a gas. This means we can neglect the effect of elevation changes on the pressure for stationary gases in tanks, pipes, and so forth in which the distances involved are small. For those situations in which the variations in heights are large, on the order of thousands of feet, attention must be given to the variation in the specific weight. As is described in Chapter 1, the equation of state for an ideal (or perfect) gas is p = ρRT If the specific weight of a fluid varies significantly as we move from point to point, the pressure will no longer vary linearly with depth. where p is the absolute pressure, R is the gas constant, and T is the absolute temperature. This relationship can be combined with Eq. 2.4 to give dp gp =− dz RT and by separating variables ∫ p2 p1 1 p2 / p1 Isothermal 0.8 0.6 p2 dp g = ln = − p p1 R 5000 10,000 z2 – z1, ft z1 dz T (2.9) where g and R are assumed to be constant over the elevation change from z1 to z2. Although the acceleration of gravity, g, does vary with elevation, the variation is very small (see Tables C.1 and C.2 in Appendix C), and g is usually assumed constant at some average value for the range of elevation involved. Before completing the integration, one must specify the nature of the variation of temperature with elevation. For example, if we assume that the temperature has a constant value T0 over the range z1 to z2 (isothermal conditions), it then follows from Eq. 2.9 that Incompressible 0 ∫ z2 p2 = p1 exp − [ g(z2 − z1 ) RT0 ] (2.10) 48 Chapter 2 ■ Fluid Statics This equation provides the desired pressure–elevation relationship for an isothermal layer. As shown in the margin figure, even for a 10,000-ft altitude change, the difference between the constant temperature (isothermal) and the constant density (incompressible) results is relatively minor. For nonisothermal conditions a similar procedure can be followed if the temperature–elevation relationship is known, as is discussed in the following section. EXAMPLE 2.2 Incompressible and Isothermal Pressure–Depth Variations GIVEN In 2010, the world’s tallest building, the Burj Khalifa skyscraper, was completed and opened in the United Arab Emirates. The final height of the building, which had remained a secret until completion, is 2717 ft (828 m). FIND (a) Estimate the ratio of the pressure at the 2717-ft top of the building to the pressure at its base, assuming the air to be at a common temperature of 59 °F. (b) Compare the pressure calculated in part (a) with that obtained by assuming the air to be incompressible with γ = 0.0765 lbft3 at 14.7 psi (abs) (values for air at standard sea level conditions). SOLUTION (a) For the assumed isothermal conditions, and treating air as a compressible fluid, Eq. 2.10 can be applied to yield p2 g(z2 − z1 ) = exp [ − p1 RT0 ] (32.2 fts2 ) (2717 ft) = exp {− (1716 ft ˙ lbslug ˙ °R)[(59 + 460)°R] } = 0.906 (Ans) (b) If the air is treated as an incompressible fluid we can apply Eq. 2.5. In this case p2 = p1 − γ(z2 − z1 ) ■ Figure E2.2 (Figure courtesy of Emaar Properties, Dubai, UAE.) or γ(z2 − z1 ) p2 =1− p1 p1 (0.0765 lbft3 ) (2717 ft) =1− = 0.902 (14.7 lbin.2 ) (144 in.2ft2 ) (Ans) COMMENTS Note that there is little difference between the two results. Since the pressure difference between the bottom and top of the building is small, it follows that the variation in fluid density is small and, therefore, the compressible fluid and incompressible fluid analyses yield essentially the same result. 2.4 We see that for both calculations the pressure decreases by approximately 10% as we go from ground level to the top of this tallest building. It does not require a very large pressure difference to support a 2717-ft-tall column of fluid as light as air. This result supports the earlier statement that the changes in pressures in air and other gases due to elevation changes are very small, even for distances of hundreds of feet. Thus, the pressure differences between the top and bottom of a horizontal pipe carrying a gas, or in a gas storage tank, are negligible since the distances involved are very small. Standard Atmosphere An important application of Eq. 2.9 relates to the variation in pressure in the earth’s atmosphere. Ideally, we would like to have measurements of pressure versus altitude over the specific range for the specific conditions (temperature, reference pressure) for which the pressure is to be determined. 2.4 The standard atmosphere is an idealized representation of mean conditions in the earth’s atmosphere. 300 Space shuttle 150 100 49 However, this type of information is usually not available. Thus, a “standard atmosphere” has been determined that can be used in the design of aircraft, missiles, and spacecraft and in comparing their performance under standard conditions. The concept of a standard atmosphere was first developed in the 1920s, and since that time many national and international committees and organizations have pursued the development of such a standard. The currently accepted standard atmosphere is based on a report published in 1962 and updated in 1976 (see Refs. 1 and 2), defining the so-called U.S. standard atmosphere, which is an idealized representation of middle-latitude, year-round mean conditions of the earth’s atmosphere. Several important properties for standard atmospheric conditions at sea level are listed in Table 2.1, and Fig. 2.6 shows the temperature profile for the U.S. standard atmosphere. As is shown in this figure, the temperature decreases with altitude in the region nearest the earth’s surface (troposphere), then becomes essentially constant in the next layer (stratosphere), and subsequently starts to increase in the next layer. Typical events that occur in the atmosphere are shown in the figure in the margin. Since the temperature variation is represented by a series of linear segments, it is possible to integrate Eq. 2.9 to obtain the corresponding pressure variation. For example, in the troposphere, which extends to an altitude of about 11 km ( ∼36,000 ft), the temperature variation is of the form T = Ta − βz (2.11) where Ta is the temperature at sea level (z = 0) and β is the lapse rate (the rate of change of temperature with elevation). For the standard atmosphere in the troposphere, β = 0.00650 K/m or 0.00357 °Rft. Meteor ■ Table 2.1 Properties of U.S. Standard Atmosphere at Sea Level* Mt. Everest SI Units BG Units Temperature, T Pressure, p 288.15 K (15 °C) 101.33 kPa (abs) Density, ρ Specific weight, γ Viscosity, μ 1.225 kgm3 12.014 Nm3 1.789 × 10−5 N ∙ sm2 518.67 °R (59.00 °F) 2116.2 lbft2 (abs) [14.696 lbin.2 (abs)] 0.002377 slugsft3 0.07647 lbft3 3.737 × 10−7 lb ∙ sft2 *Acceleration of gravity at sea level = 9.807 ms2 = 32.174 fts2. -2.5 °C 0 Property 50 47.3 km (p = 0.1 kPa) -44.5 °C Ozone layer Thunder storm Commercial jet 40 32.2 km (p = 0.9 kPa) 30 20 Stratosphere -56.5 °C 50 Altitude z, km Altitude z, km Aurora Standard Atmosphere 10 20.1 km (p = 5.5 kPa) 11.0 km (p = 22.6 kPa) p = 101.3 kPa (abs) 15 °C Troposphere 0 -100 -80 -60 -40 -20 Temperature T, °C 0 +20 ■ Figure 2.6 Variation of temperature with altitude in the U.S. standard atmosphere. Free ebooks ==> www.Ebook777.com 50 Chapter 2 ■ Fluid Statics Equation 2.11 used with Eq. 2.9 yields p = pa(1 − βz gRβ Ta ) (2.12) where pa is the absolute pressure at z = 0. With pa, Ta, and g obtained from Table 2.1, and with the gas constant R = 286.9 Jkg ∙ K or 1716 ft ∙ lbslug ∙ °R, the pressure variation throughout the troposphere can be determined from Eq. 2.12. This calculation shows that at the outer edge of the troposphere, where the temperature is −56.5 °C (−69.7 °F), the absolute pressure is about 23 kPa (3.3 psia). It is to be noted that modern jetliners cruise at approximately this altitude. Pressures at other altitudes are shown in Fig. 2.6, and tabulated values for temperature, acceleration of gravity, pressure, density, and viscosity for the U.S. standard atmosphere are given in Tables C.1 and C.2 in Appendix C. Pressure is designated as either absolute pressure or gage pressure. Measurement of Pressure Many different devices and techniques have been developed to measure pressure because it is so central to understanding fluid mechanics and to solving problems of practical interest. As is noted briefly in Chapter 1, the pressure at a point within a fluid mass is reported as either an absolute pressure or a gage pressure. Absolute pressure is measured relative to a perfect vacuum (absolute zero pressure), whereas gage pressure is measured relative to the local atmospheric pressure. Thus, a gage pressure of zero corresponds to a pressure that is equal to the local atmospheric pressure. Absolute pressures are always positive, but gage pressures can be either positive or negative depending on whether the pressure is above atmospheric pressure (a positive value) or below atmospheric pressure (a negative value). A negative gage pressure is also referred to as a suction or vacuum pressure. For example, 10 psi (abs) could be expressed as −4.7 psi (gage), if the local atmospheric pressure is 14.7 psi, or alternatively 4.7 psi suction or 4.7 psi vacuum. The concept of gage and absolute pressure is illustrated graphically in Fig. 2.7 for two typical pressures located at points 1 and 2. In addition to the reference used for the pressure measurement, the units used to express the value are obviously of importance. As is described in Section 1.5, pressure is a force per unit area, and the units in the BG system are lbft2 or lbin.2, commonly abbreviated psf or psi, respectively. In the SI system the units are Nm2; this combination is called the pascal and written as Pa (1 Nm2 ≡ 1 Pa). As noted earlier, pressure can also be expressed as the height of a column of liquid. Then the units will refer to the height of the column (in., ft, mm, m, etc.), and in addition, the liquid in the column must be specified (H2O, Hg, etc.). For example, standard atmospheric pressure can be expressed as 760 mm Hg (abs). In this text, pressures will be assumed to be gage pressures unless specifically designated absolute. For example, 10 psi or 100 kPa would be gage pressures, whereas 10 psia or 100 kPa (abs) would refer to absolute pressures. It is to be noted that pressure differences are independent of the reference, so that no special notation is required in this case. 1 Gage pressure @ 1 Pressure 2.5 Local atmospheric pressure reference 2 Absolute pressure @1 Gage pressure @ 2 (suction or vacuum) Absolute pressure @2 Absolute zero reference ■ Figure 2.7 Graphical representation of gage and absolute pressure. www.Ebook777.com 2.5 51 Measurement of Pressure pvapor A h patm B Mercury Water ■ Figure 2.8 Mercury barometer. The measurement of atmospheric pressure is usually accomplished with a mercury barometer, which in its simplest form consists of a glass tube closed at one end with the open end immersed in a container of mercury as shown in Fig. 2.8. The tube is initially filled with mercury (with its open end up) and then turned upside down (open end down), with the open end in the container of mercury. The column of mercury will come to an equilibrium position where its weight plus the force due to the vapor pressure (which develops in the space above the column) balances the force due to the atmospheric pressure. Thus, patm = γh + pvapor (2.13) where γ is the specific weight of mercury. For most practical purposes the contribution of the vapor pressure can be neglected since it is very small [for mercury, the fluid most commonly used in barometers, pvapor = 0.000023 lbin.2 (abs) at a temperature of 68 °F], so that patm ≈ γh. It is conventional to specify atmospheric pressure in terms of the height, h, in millimeters or inches of mercury. Note that if water were used instead of mercury, the height of the column would have to be approximately 34 ft rather than 29.9 in. of mercury for an atmospheric pressure of 14.7 psia! This is shown to scale in the figure in the margin. The concept of the mercury barometer is an old one, with the invention of this device attributed to Evangelista Torricelli in about 1644. Mercury E X AMPLE 2.3 Barometric Pressure GIVEN A mountain lake has an average temperature of 10 °C and a maximum depth of 40 m. The barometric pressure is 598 mm Hg. FIND Determine the absolute pressure (in pascals) at the deepest part of the lake. SOLUTION The pressure in the lake at any depth, h, is given by the equation p = γh + p0 where p0 is the pressure at the surface. Since we want the absolute pressure, p0 will be the local barometric pressure expressed in a consistent system of units; that is pbarometric = 598 mm = 0.598 m γHg and for γHg = 133 kNm3 p0 = (0.598 m) (133 kNm3 ) = 79.5 kNm2 From Table B.2, γH2 O = 9.804 kNm3 at 10 °C and therefore p = (9.804 kNm3 ) (40 m) + 79.5 kNm2 = 392 kNm2 + 79.5 kNm2 = 472 kPa (abs) (Ans) COMMENT This simple example illustrates the need for close attention to the units used in the calculation of pressure; that is, be sure to use a consistent unit system, and be careful not to add a pressure head (m) to a pressure (Pa). 52 Chapter 2 ■ Fluid Statics THE WIDE WORLD OF FLUIDS Weather, barometers, and bars One of the most important 1914 and defined as 105 N/m2 . The definition of a bar is probably indicators of weather conditions is atmospheric pressure. In gen- related to the fact that standard sea-level pressure is eral, a falling or low pressure indicates bad weather; rising or 1.0133 × 105 N/m2 , that is, only slightly larger than one bar. For high pressure, good weather. During the evening TV weather typical weather patterns, “sea-level equivalent” atmospheric report in the United States, atmospheric pressure is given as so pressure remains close to one bar. However, for extreme many inches (commonly around 30 in.). This value is actually the weather conditions associated with tornadoes, hurricanes, or height of the mercury column in a mercury barometer adjusted typhoons, dramatic changes can occur. The lowest atmospheric to sea level. To determine the true atmospheric pressure at a sea-level pressure ever recorded was associated with a typhoon, particular location, the elevation relative to sea level must be Typhoon Tip, in the Pacific Ocean on October 12, 1979. The value known. Another unit used by meteorologists to indicate atmo- was 0.870 bar (25.8 in. Hg). (See Problem 2.24.) spheric pressure is the bar, first used in weather reporting in 2.6 Manometry Manometers use vertical or inclined liquid columns to measure pressure. A standard technique for measuring pressure involves the use of liquid columns in vertical or inclined tubes. Pressure-measuring devices based on this technique are called manometers. The mercury barometer is an example of one type of manometer, but there are many other configurations possible depending on the particular application. Three common types of manometers include the piezometer tube, the U-tube manometer, and the inclined-tube manometer. 2.6.1 Piezometer Tube Column of mercury The simplest type of manometer consists of a vertical tube, open at the top, and attached to the container in which knowledge of the pressure is desired, as illustrated in Fig. 2.9. The figure in the margin shows an important device whose operation is based on this principle. It is a sphygmomanometer, the traditional instrument used to measure blood pressure. Since manometers involve columns of fluids at rest, the fundamental equation describing their function is Eq. 2.8 Container of mercury p = γh + p0 Tube open at top Arm cuff which gives the pressure at any elevation within a homogeneous fluid in terms of a reference pressure p0 and the vertical distance h between p and p0. Remember that in a fluid at rest pressure will increase with increasing depth. Application of this equation to the piezometer tube of Fig. 2.9 indicates that the pressure pA can be determined by a measurement of h1 through the relationship pA = γ1h1 where γ1 is the specific weight of the liquid in the container. Note that since the tube is open at the top, the pressure p0 can be set equal to zero (we are now using gage pressure), with the height h1 measured from the upper surface to point (1). Since point (1) and point A within the container are at the same elevation, pA = p1. Open γ1 A h1 (1) ■ Figure 2.9 Piezometer tube. Free ebooks ==> www.Ebook777.com 2.6 Manometry 53 Open γ1 A h2 (1) h1 (2) (3) γ2 (gage fluid) ■ Figure 2.10 Simple U-tube manometer. Although the piezometer tube is a very simple pressure-measuring device, it has several disadvantages. It is suitable only if the pressure in the container is greater than atmospheric pressure (otherwise air would be sucked into the system), and the pressure to be measured must be relatively small so the required height of the column is reasonable. Also the fluid in the container in which the pressure is to be measured must be a liquid rather than a gas. 2.6.2 U-Tube Manometer Forming the manometer tube in a U-shape as shown in Fig. 2.10 provides the option of introducing a gauge fluid with a higher specific weight than the fluid in the container in which the pressure is to be measured. The first, most obvious, capability added by the introduction of the U-shape with a gauge fluid is the ability to measure the pressure in a gas. Again, the pressure to be measured, pA, is related to the column heights. A numerical value can be computed by repeated application of Eq. 2.8. The following sequence of equations is generated by taking an imaginary “walk” though the manometer from pA to the open end, and repeatedly applying Eq. 2.8. step 1:  p1 = pA (same elevation) step 2:  p2 = p1 + γ1h1 = pA + γ1h1 step 3:  p3 = p2 step 4: patm = p3 + γ2 (−h2 ) = pA + γ1h1 − γ2h2 or pA = patm − γ1h1 + γ2h2 This process can be codified into a simple manometer rule by recalling the implications of Eq. 2.8: 1. Write the pressure at either end of the manometer; 2. Proceed through the manometer, adding γh if moving to a greater depth or subtracting γh if moving to a lesser depth; 3. Stop at the far end, or any point in between, and set the expression equal to the local pressure. The contribution of gas columns in manometers is usually negligible since the weight of the gas is so small. VIDEO V2.3 Blood pressure measurement If you start to apply the manometer rule at the “far” end of the manometer, the equation you seek is produced in a single line of work. For the manometer of Fig. 2.10: patm + γ2h2 − γ1h1 = pA (2.14) Thinking about this result quickly reveals two useful results. First, if the pressure to be measured is in a gas, γ2 >> γ1 , so to a very good approximation: pA = patm + γ2h2 , the same result as was obtained for the piezometer. Second, we can adjust the sensitivity, and therefore the resolution and range, of the pressure measurement device. Increasing the density of the gauge fluid allows practical measurement of much larger pressures but sacrifices resolution. Decreasing the density of the gauge fluid limits the practical range of pressure measurement but increases the resolution. The most common gauge fluids are water, mercury (SG ≈ 13.6), and oil (SG ≈ 0.8). www.Ebook777.com 54 Chapter 2 ■ Fluid Statics EXAMPLE 2.4 Simple U-Tube Manometer Pressure gage GIVEN A closed tank contains compressed air and oil (SGoil = 0.90) as is shown in Fig. E2.4. A U-tube manometer using mercury (SGHg = 13.6) is connected to the tank as shown. The column heights are h1 = 36 in., h2 = 6 in., and h3 = 9 in. Air FIND Determine the pressure reading (in psi) of the gage. Open h1 SOLUTION Oil Applying the manometer rule, starting at the open end of the manometer: h3 h2 patm + γHgh3 − γoilh2 − γoilh1 = pAir (1) The hydrostatic variation of pressure within the air has been neglected because its density is much smaller than that of either liquid. Using our convention that, unless otherwise stated, pressures will be gauge pressures: Hg ■ Figure E2.4 pAir = 0 + (SGHg ) (γH2O )h3 + (SGoil ) (γH2O ) (−h2 −h1 ) lb 9 6 + 36 pAir = 62.4 3 (13.6) ft) − (0.9)( ft)] ( )[ ( 12 12 ft lb 1 ft2 pAir = 440 2 × = 3.06 psi (Ans) ft 144 in2 (both in the tank and in the tube). It is not just the mercury in the manometer that is important. Assume that the gage pressure remains at 3.06 psi, but the manometer is altered so that it contains only oil. That is, the mercury is replaced by oil. A simple calculation shows that in this case the vertical oil-filled tube would need to be h3 = 11.3 ft tall, rather than the original h3 = 9 in. There is an obvious advantage of using a heavy fluid such as mercury in manometers. COMMENTS Note that the air pressure is a function of the height of the mercury in the manometer and the depth of the oil Manometers are often used to measure the difference in pressure between two points. (2) The U-tube manometer is also widely used to measure the difference in pressure between two containers or two points in a given system. Consider a manometer connected between containers A and B as is shown in Fig. 2.11. Application of the manometer rule produces an equation for the pressure difference between the containers. Starting at pB: pB + γ3h3 + γ2h2 − γ1h1 = pA or pA − pB = γ3h3 + γ2h2 − γ1h1. Carefully include units when replacing symbols with information provided in a problem statement to help ensure a correct solution. B (5) h3 γ3 (4) γ1 γ2 A (1) h2 h1 (2) (3) ■ Figure 2.11 Differential U-tube manometer. 2.6 55 Manometry Capillarity due to surface tension at the two-fluid interfaces in the manometer is usually negligible for common gauge fluids, common tube materials, and typical tube bore diameters (see Section 1.9). For highly accurate measurements it may be necessary to account for the temperature dependence of the fluids’ density. EX AMPLE 2.5 U-Tube Manometer GIVEN As will be discussed in Chapter 3, the volume rate of flow, Q, through a pipe can be determined by means of a flow nozzle located in the pipe as illustrated in Fig. E2.5a. The nozzle creates a pressure drop, pA − pB, along the pipe that is related to the flow through the equation Q = K √pA − pB, where K is a constant depending on the pipe and nozzle size. The pressure drop is frequently measured with a differential U-tube manometer of the type illustrated. FIND (a) Determine an equation for pA − pB in terms of the specific weight of the flowing fluid, γ1, the specific weight of the gage fluid, γ2, and the various heights indicated. (b) For γ1 = 9.80 kNm3, γ2 = 15.6 kNm3, h1 = 1.0 m, and h2 = 0.5 m, what is the value of the pressure drop, pA − pB? γ1 (4) (5) (3) γ1 h2 (1) SOLUTION (2) h1 (a) Although the fluid in the pipe is moving, the fluids in the columns of the manometer are at rest so that the pressure variation in the manometer tubes is hydrostatic. Applying the manometer rule, starting at A: pA − γ1h1 − γ2h2 + γ1 (h1 + h2 ) = pB or pA − pB = (γ2 − γ1 )h2 γ2 A Flow B Flow nozzle ■ Figure E2.5a (Ans) COMMENT It is to be noted that the only column height of importance is the differential reading, h2. The differential manometer could be placed 0.5 or 5.0 m above the pipe (h1 = 0.5 m or h1 = 5.0 m), and the value of h2 would remain the same. h2 can be large enough to provide an accurate reading provided the value of γ2 − γ1 is also small. 3 (15.6 kN/m3, 2.90 kPa) pA − pB = (0.5 m) (15.6 kNm3 − 9.80 kNm3 ) = 2.90 kPa (Ans) COMMENT By repeating the calculations for manometer fluids with different specific weights, γ2, the results shown in Fig. E2.5b are obtained. Note that relatively small pressure differences can be measured if the manometer fluid has nearly the same specific weight as the flowing fluid. It is the difference in the specific weights, γ2 − γ1, that is important. Hence, by rewriting the answer as h2 = (pA − pB )(γ2 − γ1 ) it is seen that even if the value of pA − pB is small, the value of pA – pB, kPa (b) The specific value of the pressure drop for the data given is 2 1 γ2 = γ1 0 8 10 12 γ2, kN/m3 14 16 ■ Figure E2.5b 2.6.3 Inclined-Tube Manometer To measure small pressure changes, a manometer of the type shown in Fig. 2.12 is frequently used. One leg of the manometer is inclined at an angle θ, and the differential reading ℓ2 is measured along the inclined tube. The difference in pressure pA − pB can be expressed as pA + γ1h1 − γ2ℓ2 sin θ − γ3 h3 = pB or pA − pB = γ2ℓ2 sin θ + γ3 h3 − γ1h1 (2.15) 56 Chapter 2 ■ Fluid Statics γ3 B γ1 h3 A (2) γ2 h1 ℓ2 (1) θ ■ Figure 2.12 Inclined-tube manometer. Inclined-tube manometers can be used to measure small pressure differences accurately. where it is to be noted the pressure difference between points (1) and (2) is due to the vertical distance between the points, which can be expressed as ℓ2 sin θ. Thus, for relatively small angles the differential reading along the inclined tube can be made large even for small pressure differences. The inclined-tube manometer is often used to measure small differences in gas pressures so that if pipes A and B contain a gas, then pA − pB = γ2ℓ2 sin θ or ℓ2 ~ 1 sinθ ℓ2 0 30 60 ℓ2 = 90 θ, deg 2.7 (2.16) where the contributions of the gas columns h1 and h3 have been neglected. Equation 2.16 and the figure in the margin show that the differential reading ℓ2 (for a given pressure difference) of the inclined-tube manometer can be increased over that obtained with a conventional U-tube manometer by the factor 1sin θ. Recall that sin θ → 0 as θ → 0. Mechanical and Electronic Pressure-Measuring Devices A Bourdon tube pressure gage uses a hollow, elastic, and curved tube to measure pressure. VIDEO pA − pB γ2 sin θ V2.4 Bourdon gage Manometers are widely used because they are simple, inexpensive, and reliable. However, they are limited in range, respond relatively slowly, are unsuitable for environments that might result in loss of gauge fluid, and are not easily interfaced with automated data acquisition systems. To overcome some of these problems numerous other types of pressure-measuring instruments have been developed. Most of these make use of the idea that when a pressure acts on an elastic structure the structure will deform, and this deformation can be related to the magnitude of the pressure. Probably the most familiar device of this kind is the Bourdon pressure gage, which is shown in Fig. 2.13a. The essential mechanical element in this gage is the hollow, elastic curved tube (Bourdon tube), which is connected to the pressure source as shown in Fig. 2.13b. As the pressure within the tube increases the tube tends to straighten, and although the deformation is small, it can be translated into the motion of a pointer on a dial as illustrated. Since it is the difference in pressure between the outside of the tube (atmospheric pressure) and the inside of the tube that causes the movement of the tube, the indicated pressure is gage pressure. The Bourdon gage must be calibrated so that the dial reading can directly indicate the pressure in suitable units such as psi, psf, or pascals. A zero reading on the gage indicates that the measured pressure is equal to the local atmospheric pressure. This type of gage can be used to measure a negative gage pressure (vacuum) as well as positive pressures. The aneroid barometer is another type of mechanical gage that is used for measuring atmospheric pressure. Since atmospheric pressure is specified as an absolute pressure, the conventional Bourdon gage is not suitable for this measurement. The common aneroid barometer contains a hollow, closed, elastic element that is evacuated so that the pressure inside the element is near absolute zero. As the external atmospheric pressure changes, the element deflects, and this motion can be translated into the movement of an attached dial. As with the Bourdon gage, the dial can be calibrated to give atmospheric pressure directly, with the usual units being millimeters or inches of mercury. 2.7 Mechanical and Electronic Pressure-Measuring Devices (a) 57 (b) ■ Figure 2.13 (a) Liquid-filled Bourdon pressure gages for various pressure ranges. (b) Internal elements of Bourdon gages. The “C-shaped” Bourdon tube is shown on the left, and the “coiled spring” Bourdon tube for high pressures of 1000 psi and above is shown on the right. (Photographs courtesy of Weiss Instruments, Inc.) For many applications in which pressure measurements are required, the pressure must be measured with a device that converts the pressure into an electrical output. For example, it may be desirable to continuously monitor a pressure that is changing with time. This type of pressuremeasuring device is called a pressure transducer, and many different designs are used. One design simply connects a Bourdon tube to a linear variable differential transformer (LVDT), as illustrated in Fig. 2.14, to produce a pressure transducer. The core of the LVDT is connected to the free end of the Bourdon tube. As pressure is applied, motion of the tube end moves the core though the coil to produce an output voltage. This voltage is a linear function of the pressure and could be recorded on an oscillograph or digitized for storage or processing on a computer. THE WIDE WORLD OF FLUIDS Tire pressure warning Proper tire inflation on vehicles is system fits within the tire and contains a pressure transducer important for more than ensuring long tread life. It is critical in (usually either a piezo-resistive or a capacitive-type transduc- preventing accidents such as rollover accidents caused by er) and a transmitter that sends the information to an elec- underinflation of tires. The National Highway Traffic Safety tronic control unit within the vehicle. Information about tire Administration is developing a regulation regarding four-tire tire- pressure and a warning when the tire is underinflated is dis- pressure monitoring systems that can warn a driver when a tire played on the instrument panel. The environment (hot, cold, is more than 25% underinflated. Some of these devices are cur- vibration) in which these devices must operate, their small size, rently in operation on select vehicles; it is expected that they will and required low cost provide challenging constraints for the soon be required on all vehicles. A typical tire-pressure monitoring design engineer. Bourdon C-tube Core LVDT Output Mounting block Pressure line ■ Figure 2.14 Pressure transducer Input Spring that combines a linear variable differential transformer (LVDT) with a Bourdon gage. (From Ref. 4, used by permission.) 58 Chapter 2 ■ Fluid Statics It is relatively complicated to make pressure transducers for the accurate measurement of pressures that vary rapidly with time. One disadvantage of a pressure transducer using a Bourdon tube as the elastic sensing element is that it is limited to the measurement of pressures that are static or only changing slowly (quasistatic). Because of the relatively large mass of the Bourdon tube, it cannot respond to rapid changes in pressure. To overcome this difficulty, a different type of transducer is used in which the sensing element is a thin, elastic diaphragm that is in contact with the fluid. As the pressure changes, the diaphragm deflects, and this deflection can be sensed and converted into an electrical voltage. One way to accomplish this is to locate strain gages either on the surface of the diaphragm not in contact with the fluid, or on an element attached to the diaphragm. These gages can accurately sense the small strains induced in the diaphragm and provide an output voltage proportional to pressure. This type of transducer is capable of measuring accurately both small and large pressures, as well as both static and dynamic pressures. For example, strain-gage pressure transducers of the type shown in Fig. 2.15 are used to measure arterial blood pressure, which is a relatively small pressure that varies periodically with a fundamental frequency of about 1 Hz. The transducer is usually connected to the blood vessel by means of a liquid-filled, small-diameter tube called a pressure catheter. Although the strain-gage type of transducer can be designed to have very good frequency response (a) Case Diaphragm stop Electrical connections Armature Diaphragm Link pin Beam (strain gages deposited on beam) (b) ■ Figure 2.15 (a) Photograph of a typical pressure transducer with a male thread fitting in front of the diaphragm for system connection and an electrical connector in the rear of the device. (b) Schematic diagram of a typical pressure transducer device (male thread connector not shown). Deflection of the diaphragm due to pressure is measured with a silicon beam on which strain gages and an associated bridge circuit have been deposited. 2.8 Hydrostatic Force on a Plane Surface 59 (up to approximately 10 kHz), they become less sensitive at the higher frequencies since the diaphragm must be made stiffer to achieve the higher frequency response. As an alternative, the diaphragm can be constructed of a piezoelectric crystal to be used as both the elastic element and the sensor. When a pressure is applied to the crystal, a voltage develops because of the deformation of the crystal. This voltage is directly related to the applied pressure. Depending on the design, this type of transducer can be used to measure both very low and high pressures (up to approximately 100,000 psi) at high frequencies. Additional information on pressure transducers can be found in Refs. 3, 4, and 5. 2.8 VIDEO Hydrostatic Force on a Plane Surface V2.5 Hoover dam The resultant force of a static fluid on a plane surface is due to the hydrostatic pressure distribution on the surface. When a surface is submerged in a fluid, forces develop on the surface due to the fluid. The determination of these forces is important in the design of storage tanks, ships, dams, and other hydraulic structures. For fluids at rest we know that the force must be perpendicular to the surface since there are no shearing stresses present. We also know that the pressure will vary linearly with depth as shown in Fig. 2.16 if the fluid is incompressible. For a horizontal surface, such as the bottom of a liquid-filled tank (Fig. 2.16a), the magnitude of the resultant force is simply FR = pA, where p is the uniform pressure on the bottom and A is the area of the bottom. For the open tank shown, p = γh. Note that if atmospheric pressure acts on both sides of the bottom, as is illustrated, the resultant force on the bottom is simply due to the liquid in the tank. Since the pressure is constant and uniformly distributed over the bottom, the resultant force acts through the centroid of the area as shown in Fig. 2.16a. As shown in Fig. 2.16b, the pressure on the ends of the tank is not uniformly distributed. Determination of the force due to hydrostatic pressure on the tank ends is more challenging because the pressure is not constant, the ends may not be rectangular plates, and in general they may not be vertical. Figure 2.17 depicts a submerged inclined plane surface and provides labels for locations and distances that will be required to determine the resultant force and its line of action. For the present we will assume that the fluid surface is open to the atmosphere. Let the plane in which the surface lies intersect the free surface at 0 and make an angle θ with this surface as in Fig. 2.17. The x–y coordinate system is defined so that 0 is the origin and y = 0 (i.e., the x axis) is directed along the surface as shown. The surface may be of arbitrary shape provided it remains in the x–y plane. We wish to determine the direction, location, and magnitude of the resultant force acting on one side of this area due to the liquid in contact with the area. At any given depth, h, the force acting on dA (the differential area of Fig. 2.17) is dF = γh dA and is perpendicular to the surface. Thus, the magnitude of the resultant force can be found by summing these differential forces over the entire surface. In equation form FR = ∫ γh dA = A Free surface p=0 FR γy sin θ dA A Free surface p=0 Specific weight = γ h ∫ Specific weight = γ p = γh p = γh p=0 (a) Pressure on tank bottom p=0 (b) Pressure on tank ends ■ Figure 2.16 (a) Pressure distribution and resultant hydrostatic force on the bottom of an open tank. (b) Pressure distribution on the ends of an open tank. Free ebooks ==> www.Ebook777.com 60 Chapter 2 ■ Fluid Statics Free surface 0 θ h hc y yc yR dF FR x x A c y xc dA CP xR Centroid, c Location of resultant force (center of pressure, CP) ■ Figure 2.17 Notation for hydrostatic force on an inclined plane surface of arbitrary shape. where h = y sin θ. For constant γ and θ The magnitude of the resultant fluid force depends on only the depth and is equal to the pressure acting at the centroid of the area multiplied by the total area. FR = γ sin θ ∫ y dA (2.17) A The integral appearing in Eq. 2.17 is the first moment of the area with respect to the x axis, so we can write ∫ y dA = yc A A where yc is the y coordinate of the centroid of area A measured from the x axis, which passes through 0. Equation 2.17 can thus be written as FR = γAyc sin θ or more simply as γ hc FR = γhc A c A FR = γhc A (2.18) where, as shown by the figure in the margin, hc is the vertical distance from the fluid surface to the centroid of the area. As indicated by the figure in the margin, it depends only on the specific weight of the fluid, the total area, and the depth of the centroid of the area below the surface. In effect, Eq. 2.18 indicates that the magnitude of the resultant force is equal to the pressure at the centroid of the area multiplied by the total area. Since all the differential forces that were summed to obtain FR are perpendicular to the surface, the resultant FR must also be perpendicular to the surface. www.Ebook777.com 2.8 Hydrostatic Force on a Plane Surface 61 Although our intuition might suggest that the resultant force should pass through the centroid of the area, this is not the case. The y coordinate, yR, of the resultant force can be determined by summation of moments around the x axis. That is, the moment of the resultant force must equal the moment of the distributed pressure force, or FR yR =   ∫ ∫ y dF = A γ sin θ y2 dA A and, noting that FR = γAyc sin θ, ∫ y2 dA A yR = yc A The integral in the numerator is the second moment of the area (moment of inertia), Ix, with respect to an axis formed by the intersection of the plane containing the surface and the free surface (x axis). Thus, we can write yR = Ix yc A Use can now be made of the parallel axis theorem to express Ix as Ix = Ixc + Ay2c yc FR where Ixc is the second moment of the area with respect to an axis passing through its centroid and parallel to the x axis. Thus, c Ixc + yc yc A yR = Ixc yc A (2.19) As shown by Eq. 2.19 and the figure in the margin, the resultant force does not pass through the centroid but for nonhorizontal surfaces is always below it, since Ixc yc A > 0. The x coordinate, xR, for the resultant force can be determined in a similar manner by summing moments about the y axis. Thus, The resultant fluid force does not pass through the centroid of the area. ∫ FR xR = γ sin θ xy dA A and, therefore, xR = ∫ xy dA A yc A = Ixy yc A where Ixy is the product of inertia with respect to the x and y axes. Again, using the parallel axis theorem,1 we can write xR = Ixyc yc A + xc (2.20) where Ixyc is the product of inertia with respect to an orthogonal coordinate system passing through the centroid of the area and formed by a translation of the x–y coordinate system. If the submerged 1 Recall that the parallel axis theorem for the product of inertia of an area states that the product of inertia with respect to an orthogonal set of axes (x–y coordinate system) is equal to the product of inertia with respect to an orthogonal set of axes parallel to the original set and passing through the centroid of the area, plus the product of the area and the x and y coordinates of the centroid of the area. Thus, Ixy = Ixyc + Axcyc. Free ebooks ==> www.Ebook777.com 62 Chapter 2 ■ Fluid Statics c a –– 2 x a –– 2 y b –– 2 b –– 2 A = ba A = π R2 1 ba3 Ixc = ––– R 12 x c 1 ab3 Iyc = ––– 12 Ixyc = 0 (b) Circle π R2 A = ––––– ab A = ––– d 2 2 R x 4R ––– 3π R Iyc = 0.3927R4 ba3 Ixc = –––-– 36 2 Ixyc = –ba –––– (b – 2d) Ixc = 0.1098R4 c y 4 Ixyc = 0 y (a) Rectangle π R4 Ixc = Iyc = ––––– 72 a c x y Ixyc = 0 a –– 3 b+d ––––––– 3 b (c) Semicircle (d) Triangle π R2 A = ––––– 4R ––– 3π 4R ––– 3π c 4 Ixc = Iyc = 0.05488R4 x R Ixyc = –0.01647R4 y (e) Quarter circle ■ Figure 2.18 Geometric properties of some common shapes. Gate c FR FR right left area is symmetrical with respect to an axis passing through the centroid and parallel to either the x or y axis, the resultant force must lie along the line x = xc, since Ixyc is identically zero in this case. The point through which the resultant force acts is called the center of pressure. It is to be noted from Eqs. 2.19 and 2.20 that as yc increases the center of pressure moves closer to the centroid of the area. Since yc = hc sin θ, the distance yc will increase if the depth of submergence, hc, increases, or, for a given depth, the area is rotated so that the angle, θ, decreases. Thus, the hydrostatic force on the right-hand side of the gate shown in the margin figure acts closer to the centroid of the gate than the force on the left-hand side. Centroidal coordinates and moments of inertia for some common areas are given in Fig. 2.18. THE WIDE WORLD OF FLUIDS The Three Gorges Dam Spanning China’s Yangtze River, the dam, is 483 m long with 23 bottom outlets and 22 surface sluice Three Gorges Dam became the world’s largest power generating gates. The maximum discharge capacity is 102,500 m3/s. After facility when the last of its main turbines began operation in July more than 10 years of construction, the dam gates were finally 2012. The dam is of the concrete gravity type, having a length of closed, and on June 10, 2003, the reservoir had been filled to its 2309 m with a height of 185 m. The main elements of the project interim level of 135 m. Due to the large depth of water at the dam include the dam, two power plants, and navigation facilities con- and the huge extent of the storage pool, hydrostatic pressure sisting of a ship lock and lift. The power plants will contain forces have been a major factor considered by engineers. When 26 Francis-type turbines, each with a capacity of 700 mega- filled to its normal pool level of 175 m, the total reservoir storage watts. The spillway section, which is the center section of the capacity is 39.3 billion m3. (See Problem 2.111.) www.Ebook777.com 2.8 E X AMPLE 2.6 Hydrostatic Force on a Plane Surface 63 Hydrostatic Force on a Plane Circular Surface GIVEN The 4-m-diameter circular gate of Fig. E2.6a is located in the inclined wall of a large reservoir containing water (γ = 9.80 kNm3 ). The gate is mounted on a shaft along its horizontal diameter, and the water depth is 10 m at the shaft. 60° si –––m yR n 60–°–– x 10 m –––10 y Stop = Determine (a) the magnitude and location of the resultant force exerted on the gate by the water and 0 yc FIND 0 Shaft (b) the moment that would have to be applied to the shaft to open the gate. FR A 4m c c SOLUTION (b) A (a) To find the magnitude of the force of the water we can apply Eq. 2.18, Oy Center of pressure FR (a) FR = γhc A c and since the vertical distance from the fluid surface to the centroid of the area is 10 m, it follows that FR = (9.80 × 10 Nm ) (10 m) (4π m ) = 1230 × 103 N = 1.23 MN 3 3 Ox M 𝒲 (c) 2 (Ans) ■ Figure E2.6a–c To locate the point (center of pressure) through which FR acts, we use Eqs. 2.19 and 2.20, xR = Ixyc yc A + xc  yR = Ixc + yc yc A For the coordinate system shown, xR = 0 since the area is symmetric about the y-axis, and the center of pressure must lie along the diameter A-A. To obtain yR, we have from Fig. 2.18 Ixc = πR4 4 (b) The moment required to open the gate can be obtained with the aid of the free-body diagram of Fig. E2.6c. In this diagram 𝒲 is the weight of the gate and Ox and Oy are the horizontal and vertical reactions of the shaft on the gate. We can now sum moments about the shaft ∑ Mc = 0 and, therefore, M = FR (yR − yc ) = (1230 × 103 N) (0.0866 m) = 1.07 × 105 N ˙ m and yc is shown in Fig. E2.6b. Thus, 4 (π4) (2 m) 10 m + sin 60° (10 msin 60°) (4π m2 ) = 0.0866 m + 11.55 m = 11.6 m yR = 0.5 and the distance (along the gate) below the shaft to the center of pressure is 0.4 (Ans) We can conclude from this analysis that the force on the gate due to the water has a magnitude of 1.23 MN and acts through a point along its diameter A-A at a distance of 0.0866 m (along the gate) below the shaft. The force is perpendicular to the gate surface as shown in Fig. E2.6b. COMMENT By repeating the calculations for various values of the depth to the centroid, hc, the results shown in Fig. E2.6d are obtained. Note that as the depth increases, the distance between the center of pressure and the centroid decreases. yR – yc, m yR − yc = 0.0866 m (Ans) 0.3 0.2 (10m, 0.0886 m) 0.1 0 0 5 10 15 hc, m ■ Figure E2.6d 20 25 30 64 Chapter 2 ■ Fluid Statics EXAMPLE 2.7 Hydrostatic Pressure Force on a Plane Triangular Surface GIVEN An aquarium contains seawater (γ = 64.0 lbft3 ) to a FIND Determine depth of 1 ft as shown in Fig. E2.7a. To repair some damage to one corner of the tank, a triangular section is replaced with a new section as illustrated in Fig. E2.7b. (a) the magnitude of the force of the seawater on this triangular area, and (b) the location of this force. SOLUTION (a) The various distances needed to solve this problem are shown in Fig. E2.7c. Since the surface of interest lies in a vertical plane, yc = hc = 0.9 ft, and from Eq. 2.18 the magnitude of the force is FR = γhc A = (64.0 lb/ft3 ) (0.9 ft)[(0.3 ft) 22] = 2.59 lb (Ans) COMMENT Note that this force is independent of the tank length. The result is the same if the tank is 0.25 ft, 25 ft, or 25 miles long. 0.3 ft 1 ft 0.3 ft 0.9 ft 2.5 ft (b) The y coordinate of the center of pressure (CP) is found from Eq. 2.19, yR = (b) Ixc + yc yc A x y and from Fig. 2.18 Ixc = (0.3 ft) (0.3 ft) 3 0.0081 4 = ft 36 36 yc 1 ft so that yR = 0.008136 ft4 yR c (0.9 ft) (0.092 ft2 ) = 0.00556 ft + 0.9 ft = 0.906 ft (Ans) c CP xR 0.1 ft (c) Similarly, from Eq. 2.20 xR = Ixyc yc A + xc δA 0.2 ft 0.1 ft + 0.9 ft Median line CP 0.15 ft 0.15 ft (d ) ■ Figure E2.7b–d and from Fig. 2.18 Ixyc = (0.3 ft) (0.3 ft) 2 0.0081 4 ft (0.3 ft) = 72 72 so that xR = 0.008172 ft4 (0.9 ft) (0.092 ft2 ) + 0 = 0.00278 ft (Ans) COMMENT Thus, we conclude that the center of pressure is ■ Figure E2.7a (Photograph courtesy of Tenecor Tanks, Inc.) 0.00278 ft to the right of and 0.00556 ft below the centroid of the area. If this point is plotted, we find that it lies on the median line for the area as illustrated in Fig. E2.7d. Since we can think of the total area as consisting of a number of small rectangular strips of area δA (and the fluid force on each of these small areas acts through its center), it follows that the resultant of all these parallel forces must lie along the median. 2.9 2.9 Pressure Prism 65 Pressure Prism An informative and useful graphical interpretation can be made for the force developed by a fluid acting on a plane rectangular area. Consider the pressure distribution along a vertical wall of a tank of constant width b, which contains a liquid having a specific weight γ. Since the pressure must vary linearly with depth, we can represent the variation as is shown in Fig. 2.19a, where the pressure is equal to zero at the upper surface and equal to γh at the bottom. It is apparent from this diagram that the average pressure occurs at the depth h2 and, therefore, the resultant force acting on the rectangular area A = bh is h FR = pav A = γ ( ) A 2 which is the same result as obtained from Eq. 2.18. The pressure distribution shown in Fig. 2.19a applies across the vertical surface, so we can draw the three-dimensional representation of the pressure distribution as shown in Fig. 2.19b. The base of this “volume” in pressure-area space is the plane surface of interest, and its altitude at each point is the pressure. This volume is called the pressure prism, and it is clear that the magnitude of the resultant force acting on the rectangular surface is equal to the volume of the pressure prism. Thus, for the prism of Fig. 2.19b the fluid force is The magnitude of the resultant fluid force is equal to the volume of the pressure prism and passes through its centroid. FR = volume = 1 h (γh)(bh) = γ ( ) A 2 2 where bh is the area of the rectangular surface, A. The resultant force must pass through the centroid of the pressure prism. For the volume under consideration the centroid is located along the vertical axis of symmetry of the surface and at a distance of h3 above the base (since the centroid of a triangle is located at h3 above its base). This result can readily be shown to be consistent with that obtained from Eqs. 2.19 and 2.20. This same graphical approach can be used for plane rectangular surfaces that do not extend up to the fluid surface, as illustrated in Fig. 2.20a. In this instance, the cross section of the pressure prism is trapezoidal. However, the resultant force is still equal in magnitude to the volume of the pressure prism, and it passes through the centroid of the volume. Specific values can be obtained by decomposing the pressure prism into two parts, ABDE and BCD, as shown in Fig. 2.20b. Thus, FR = F1 + F2 where the components can readily be determined by inspection for rectangular surfaces. The location of FR can be determined by summing moments about some convenient axis, such as one passing through A. In this instance FRyA = F1y1 + F2 y2 and y1 and y2 can be determined by inspection. h p h CP FR h– 3 h– 3 FR b γh ( a) γh (b) ■ Figure 2.19 Pressure prism for vertical rectangular area. Free ebooks ==> www.Ebook777.com 66 Chapter 2 ■ Fluid Statics h1 γ h1 B A y1 h2 F1 FR p yA y2 F2 C D E γ (h2 - h1) (a) The use of the pressure prism concept to determine the force on a submerged area is best suited for plane rectangular surfaces. The resultant fluid force acting on a submerged area is affected by the pressure at the free surface. (b) ■ Figure 2.20 Graphical representation of hydrostatic forces on a vertical rectangular surface. For inclined plane rectangular surfaces the pressure prism can still be developed, and the cross section of the prism will generally be trapezoidal, as is shown in Fig. 2.21. Although it is usually convenient to measure distances along the inclined surface, the pressures developed depend on the vertical distances as illustrated. The use of pressure prisms for determining the force on submerged plane areas is convenient if the area is rectangular so the volume and centroid can be easily determined. However, for other nonrectangular shapes, integration would generally be needed to determine the volume and centroid. In these circumstances it is more convenient to use the equations developed in the previous section, in which the necessary integrations have been made and the results presented in a convenient and compact form that is applicable to submerged plane areas of any shape. The effect of atmospheric pressure on a submerged area has not yet been considered, and we may ask how this pressure will influence the resultant force. If we again consider the pressure distribution on a plane vertical wall, as is shown in Fig. 2.22a, the pressure varies from zero at the surface to γh at the bottom. Since we are setting the surface pressure equal to zero, we are using atmospheric pressure as our datum, and thus the pressure used in the determination of the fluid force is gage pressure. If we wish to include atmospheric pressure, the pressure distribution will be as is shown in Fig. 2.22b. We note that in this case the force on one side of the wall now consists of FR as a result of the hydrostatic pressure distribution, plus the contribution of the atmospheric pressure, patm A, where A is the area of the surface. However, if we are going to include the effect of atmospheric pressure on one side of the wall, we must realize that this same pressure acts on the outside surface (assuming it is exposed to the atmosphere), so that an equal and opposite force will be developed as illustrated in the figure. Thus, we conclude that the resultant fluid force on the surface is that due only to the gage pressure contribution of the liquid in contact with the surface— the atmospheric pressure does not contribute to this resultant. Taking this study one step further, if the pressure at the liquid-vapor interface is not atmospheric pressure, both the magnitude and the line of action will differ from that predicted using only gage pressure. An analysis almost identical to that depicted in Fig. 2.20 and described in the associated text is required. γ h1 h1 h2 γ h2 ■ Figure 2.21 Pressure variation along an inclined plane area. www.Ebook777.com 2.9 patm patm h p 67 patm patm A FR Pressure Prism patm A FR γh (a) (b) ■ Figure 2.22 Effect of atmospheric pressure on the resultant force acting on a plane vertical wall. E X AMPLE 2.8 Use of the Pressure Prism Concept GIVEN A pressurized tank contains oil (SG = 0.90) and has a square, 0.6-m by 0.6-m plate bolted to its side, as is illustrated in Fig. E2.8a. The pressure gage on the top of the tank reads 50 kPa, and the outside of the tank is at atmospheric pressure. FIND What is the magnitude and location of the resultant force on the attached plate? p = 50 kPa γ h1 ps Oil surface Air h1 = 2 m Oil h2 = 2.6 m 2m 0.6 m F1 F2 0.6 m FR 0.2 m yO 0.3 m O γ (h2 – h1) (a) Plate (b) ■ Figure E2.8 SOLUTION The pressure distribution acting on the inside surface of the plate is shown in Fig. E2.8b. The pressure at a given point on the plate is due to the air pressure, ps, at the oil surface and the pressure due to the oil, which varies linearly with depth as is shown in the figure. The resultant force on the plate (having an area A) is due to the components, F1 and F2, where F1 and F2 are due to the rectangular and triangular portions of the pressure distribution, respectively. Thus, F1 = ( ps + γh1 ) A   = [50 × 10 Nm 3 2 + (0.90) (9.81 × 103 Nm3 ) (2 m)](0.36 m2 ) = 24.4 × 103 N and F2 = γ ( h2 − h1 A 2 ) = (0.90) (9.81 × 103 Nm3 )   0.6 m 2 ( 2 ) (0.36 m ) = 0.954 × 103 N The magnitude of the resultant force, FR, is therefore FR = F1 + F2 = 25.4 × 103 N = 25.4 kN (Ans) The vertical location of FR can be obtained by summing moments around an axis through point O so that FRyO = F1 (0.3 m) + F2 (0.2 m) 68 Chapter 2 ■ Fluid Statics COMMENT Note that the air pressure used in the calculation or yO = of the force was gage pressure. Atmospheric pressure does not affect the resultant force (magnitude or location), since it acts on both sides of the plate, thereby canceling its effect. (24.4 × 10 N) (0.3 m) + (0.954 × 10 N) (0.2 m) 3 3 25.4 × 103 N = 0.296 m (Ans) Thus, the force acts at a distance of 0.296 m above the bottom of the plate along the vertical axis of symmetry. 2.10 VIDEO Hydrostatic Force on a Curved Surface The equations developed in Section 2.8 for the magnitude and location of the resultant force acting on a submerged surface only apply to plane surfaces. However, many surfaces of interest (such as those associated with dams, pipes, and tanks) are nonplanar. The domed bottom of the beverage bottle shown in the figure in the margin shows a typical curved surface example. Although the resultant fluid force can be determined by integration, as was done for the plane surfaces, this is generally a rather tedious process and no simple, general formulas can be developed. As an alternative approach, we will consider the equilibrium of the fluid volume enclosed by the curved surface of interest and the horizontal and vertical projections of this surface. For example, consider a curved portion of the swimming pool shown in Fig. 2.23a. We wish to find the resultant fluid force acting on section BC (which has a unit length perpendicular to the plane of the paper) shown in Fig. 2.23b. We first isolate a volume of fluid that is bounded by the surface of interest, in this instance section BC, the horizontal plane surface AB, and the vertical plane surface AC. The free-body diagram for this volume is shown in Fig. 2.23c. The magnitude and location of forces F1 and F2 can be determined from the relationships for planar surfaces. The weight, 𝒲, is simply the specific weight of the fluid times the enclosed volume and acts through the center of gravity (CG) of the mass of fluid contained within the volume. The forces FH and FV represent the components of the force that the tank exerts on the fluid. In order for this force system to be in equilibrium, the horizontal component FH must be equal in magnitude and collinear with F2, and the vertical component FV equal in magnitude and collinear with the resultant of the vertical forces F1 and 𝒲. This follows since the three forces acting on the fluid mass (F2, the resultant of F1 and 𝒲, and the resultant force that the tank exerts on the mass) must form a concurrent force system. That is, from the principles of statics, it is known that when a body is held in equilibrium by three nonparallel forces, they must be concurrent (their lines of V2.6 Pop bottle F1 A 𝒲 B A F2 O C C FR = √(FH)2 + (FV)2 B B CG FH O C FV (a) (b) (c) ■ Figure 2.23 Hydrostatic force on a curved surface. (Photograph courtesy of Intex Marketing, Ltd.) (d) 2.10 69 Hydrostatic Force on a Curved Surface action intersect at a common point) and coplanar. Thus, FH = F2 FV = F1 + 𝒲 and the magnitude of the resultant is obtained from the equation FR = √(FH ) 2 + (FV ) 2 The resultant FR passes through the point O, which can be located by summing moments about an appropriate axis. The resultant force of the fluid acting on the curved surface BC is equal and opposite in direction to that obtained from the free-body diagram of Fig. 2.23c. The desired fluid force is shown in Fig. 2.23d. EX AMPLE 2.9 Hydrostatic Pressure Force on a Curved Surface GIVEN A 6-ft-diameter drainage conduit of the type shown in Fig. E2.9a is half full of water at rest, as shown in Fig. E2.9b. FIND Determine the magnitude and line of action of the resul- tant force that the water exerts on a 1-ft length of the curved section BC of the conduit wall. 1.27 ft 3 ft A A B A B CG F1 𝒲 1 ft FR = 523 lb 32.5° FH O 1 ft C C (a) FV (c) (b) (d) ■ Figure E2.9 (Photograph courtesy of CONTECH Construction Products, Inc.) SOLUTION We first isolate a volume of fluid bounded by the curved section BC, the horizontal surface AB, and the vertical surface AC, as shown in Fig. E2.9c. The volume has a length of 1 ft. The forces acting on the volume are the horizontal force, F1, which acts on the vertical surface AC, the weight, 𝒲, of the fluid contained within the volume, and the horizontal and vertical components of the force of the conduit wall on the fluid, FH and FV, respectively. The magnitude of F1 is found from the equation 3 F1 = γhc A = (62.4 lbft3 ) ( 2 ft) (3 ft2 ) = 281 lb and this force acts 1 ft above C as shown. The weight 𝒲 = γV , where V is the fluid volume, is 𝒲 = γ V = (62.4 lbft3 ) (9π4 ft2 ) (1 ft) = 441 lb and acts through the center of gravity of the mass of fluid, which according to Fig. 2.18 is located 1.27 ft to the right of AC as shown. Therefore, to satisfy equilibrium FH = F1 = 281 lb  FV = 𝒲 = 441 lb and the magnitude of the resultant force is FR = √(FH ) 2 + (FV ) 2 = √(281 lb) 2 + (441 lb) 2 = 523 lb (Ans) The force the water exerts on the conduit wall is equal, but opposite in direction, to the forces FH and FV shown in Fig. E2.9c. Thus, the resultant force on the conduit wall is shown in Fig. E2.9d. This force acts through the point O at the angle shown. COMMENT An inspection of this result will show that the line of action of the resultant force passes through the center of the conduit. In retrospect, this is not a surprising result since at each point on the curved surface of the conduit the elemental force due to the pressure is normal to the surface, and each line of action must pass through the center of the conduit. It therefore follows that the resultant of this concurrent force system must also pass through the center of concurrence of the elemental forces that make up the system. Free ebooks ==> www.Ebook777.com 70 Chapter 2 ■ Fluid Statics This same general approach can also be used for determining the force on curved surfaces of pressurized, closed tanks. If these tanks contain a gas, the weight of the gas is usually negligible in comparison with the forces developed by the pressure. Thus, the forces (such as F1 and F2 in Fig. 2.23c) on horizontal and vertical projections of the curved surface of interest can simply be expressed as the internal pressure times the appropriate projected area. THE WIDE WORLD OF FLUIDS Miniature, exploding pressure vessels Our daily lives are the special, impervious hull (pressure vessel) which, when safer because of the effort put forth by engineers to design heated to a proper temperature, turns to steam, causing the safe, lightweight pressure vessels such as boilers, propane kernel to explode and turn itself inside out. Not all popcorn tanks, and pop bottles. Without proper design, the large kernels have the proper properties to make them pop well. hydrostatic pressure forces on the curved surfaces of such First, the kernel must be quite close to 13.5% water. With too containers could cause the vessel to explode with disastrous little moisture, not enough steam will build up to pop the consequences. On the other hand, the world is a friendlier kernel; too much moisture causes the kernel to pop into a place because of miniature pressure vessels that are designed dense sphere rather than the light fluffy delicacy expected. to explode under the proper conditions—popcorn kernels. Second, to allow the pressure to build up, the kernels must not Each grain of popcorn contains a small amount of water within be cracked or damaged. 2.11 Buoyancy, Flotation, and Stability 2.11.1 Archimedes’ Principle (Photograph courtesy of Cameron Balloons.) When a stationary body is completely submerged in a fluid (such as the hot air balloon shown in the figure in the margin), or floating so that it is only partially submerged, the resultant fluid force acting on the body is called the buoyant force. A net upward vertical force results because pressure increases with depth and the pressure forces acting from below are larger than the pressure forces acting from above, as shown by the figure in the margin. This force can be determined through an approach similar to that used in the previous section for forces on curved surfaces. Consider a body of arbitrary shape, having a volume V, that is immersed in a fluid as illustrated in Fig. 2.24a. We enclose the body in a parallelepiped and draw a free-body diagram of the parallelepiped with the body removed as shown in Fig. 2.24b. Note that the forces F1, F2, F3, and F4 are simply the forces exerted on the plane surfaces of the parallelepiped (for simplicity the forces in the x direction are not shown), 𝒲 is the weight of the shaded fluid volume (parallelepiped minus body), and FB is the force the body is exerting on the fluid. The forces on the vertical surfaces, such as F3 and F4, are all equal and cancel, so the equilibrium equation of interest is in the z direction and can be expressed as FB = F2 − F1 − 𝒲 (2.21) If the specific weight of the fluid is constant, then F2 − F1 = γ(h2 − h1 )A where A is the horizontal area of the upper (or lower) surface of the parallelepiped, and Eq. 2.21 can be written as FB = γ(h2 − h1 )A − γ[(h2 − h1 )A − V ] Simplifying, we arrive at the desired expression for the buoyant force FB = γV (2.22) where γ is the specific weight of the fluid and V is the volume of the body. The effects of the specific weight (or density) of the body as compared to that of the surrounding fluid are illustrated by the figure in the margin. The direction of the buoyant force, which is the force of the fluid on the body, is www.Ebook777.com 2.11 Buoyancy, Flotation, and Stability 71 c Centroid of displaced volume h1 h2 A FB z B (d) y x D C Centroid ( a) c VIDEO F1 y1 V2.7 Cartesian Diver y2 A FB B 𝒲 F3 yc F4 FB D (c) C F2 (b) γ1 < γ ■ Figure 2.24 Buoyant force on submerged and floating bodies. γ2 = γ γ γ3 > γ Archimedes’ principle states that the buoyant force has a magnitude equal to the weight of the fluid displaced by the body and is directed vertically upward. opposite to that shown on the free-body diagram. Therefore, the buoyant force has a magnitude equal to the weight of the fluid displaced by the body and is directed vertically upward. This result is commonly referred to as Archimedes’ principle in honor of Archimedes (287–212 B.C.), a Greek mechanician and mathematician who first enunciated the basic ideas associated with hydrostatics. The location of the line of action of the buoyant force can be determined by summing moments of the forces shown on the free-body diagram in Fig. 2.24b with respect to some convenient axis. For example, summing moments about an axis perpendicular to the paper through point D we have FByc = F2 y1 − F1y1 − 𝒲y2 and on substitution for the various forces V yc = V Ty1 − (V T − V) y2   (2.23) where VT is the total volume (h2 − h1 )A. The right-hand side of Eq. 2.23 is the first moment of the displaced volume V with respect to the x–z plane so that yc is equal to the y coordinate of the centroid of the volume V. In a similar fashion, it can be shown that the x coordinate of the buoyant force coincides with the x coordinate of the centroid. Thus, we conclude that the buoyant force passes through the centroid of the displaced volume as shown in Fig. 2.24c. The point through which the buoyant force acts is called the center of buoyancy. THE WIDE WORLD OF FLUIDS Concrete canoes A solid block of concrete thrown into a pond sponsored by the American Society of Civil Engineers and or lake will obviously sink. But if the concrete is formed into the Master Builders Inc. The canoes must be 90% concrete and are shape of a canoe it can be made to float. Of, course the reason typically designed with the aid of a computer by civil engineering the canoe floats is the development of the buoyant force due to students. Final scoring depends on four components: a design the displaced volume of water. With the proper design, this verti- report, an oral presentation, the final product, and racing. In 2011 cal force can be made to balance the weight of the canoe plus California Polytechnic State University, San Luis Obispo, won the passengers—the canoe floats. Each year since 1988 a National national championship held at the University of Evansville with Concrete Canoe Competition for university teams is jointly its 208-pound, 20-foot-long canoe. (See Problem 2.147.) Free ebooks ==> www.Ebook777.com 72 VIDEO Chapter 2 ■ Fluid Statics These same results apply to floating bodies that are only partially submerged, as illustrated in Fig. 2.24d, if the specific weight of the fluid above the liquid surface is very small compared with the liquid in which the body floats. Since the fluid above the surface is usually air, for practical purposes this condition is satisfied. In the derivations presented above, the fluid is assumed to have a constant specific weight, γ. If a body is immersed in a fluid in which γ varies with depth, such as in a layered fluid, the magnitude of the buoyant force remains equal to the weight of the displaced fluid. However, the buoyant force does not pass through the centroid of the displaced volume, but rather, it passes through the center of gravity of the displaced volume. V2.8 Hydrometer EXAMPLE 2.10 Buoyant Force on a Submerged Object GIVEN A Type I offshore life jacket (personal flotation device) of the type worn by commercial fishermen is shown in Fig. E2.10a. It is designed for extended survival in rough, open water. According to U.S. Coast Guard regulations, the life jacket must provide a minimum 22-lb net upward force on the user. Consider such a life jacket that uses a foam material with a specific weight of 2.0 lb/ft3 for the main flotation material. The remaining material (cloth, straps, fasteners, etc.) weighs 1.3 lb and is of negligible volume. FIND Determine the minimum volume of foam needed for this life jacket. SOLUTION A free-body diagram of the life jacket is shown in Fig. E2.10b, where FB is the buoyant force acting on the life jacket, 𝒲F is the weight of the foam, 𝒲S = 1.3 lb is the weight of the remaining material, and FU = 22 lb is the required force on the user. For equilibrium it follows that F B = 𝒲 F + 𝒲 S + FU (1) where from Eq. 2.22 FB = γwater V Here γwater = 64.0 lb/ft3 is the specific weight of seawater and V is the volume of the foam. Also 𝒲foam = γfoam V , where γfoam = 2.0 lb/ft3 is the specific weight of the foam. Thus, from Eq. 1 γwater V = γfoam V + 𝒲S + FU FB ᐃF FU or V = (𝒲S + FU )(γwater − γfoam ) = (1.3 lb + 22 lb)(64.0 lb/ft3 − 2.0 lb/ft3 ) = 0.376 ft3 (Ans) COMMENTS In this example, rather than using difficult-to- ᐃS ■ Figure E2.10b ■ Figure E2.10a calculate hydrostatic pressure force on the irregularly shaped life jacket, we have used the buoyant force. The net effect of the pressure forces on the surface of the life jacket is equal to the upward buoyant force. Do not include both the buoyant force and the hydrostatic pressure effects in your calculations—use one or the other. There is more to the proper design of a life jacket than just the volume needed for the required buoyancy. According to regulations, a Type I life jacket must also be designed so that it provides proper protection to the user by turning an unconscious person in the water to a face-up position as shown in Fig. E2.10a. This involves the concept of the stability of a floating object (see Section 2.11.2). The life jacket should also provide minimum interference under ordinary working conditions so as to encourage its use by commercial fishermen. www.Ebook777.com 2.11 VIDEO V2.9 Atmospheric buoyancy Buoyancy, Flotation, and Stability 73 The effects of buoyancy are not limited to the interaction between a solid body and a fluid. Buoyancy effects can also be seen within fluids alone, as long as a density difference exists. Consider the shaded portion of Fig. 2.24c to be a volume of fluid instead of a solid. This volume of fluid is submerged in the surrounding fluid and therefore has a buoyant force due to the fluid it displaces (like the solid). If this volume contains fluid with a density of ρ1, then the downward force due to weight is 𝒲 = ρ1gV. In addition, if the surrounding fluid has the same density, then the buoyant force on the volume due to the displaced fluid will be FB = γV = ρ1gV. As expected, in this case the weight of the volume is exactly balanced by the buoyant force acting on the volume so there is no net force. However, if the density of fluid in the volume is ρ2, then 𝒲 and FB will not balance and there will be a net force in the upward or downward direction depending on whether the density in the volume (ρ2) is less than or greater than, respectively, the density of the surrounding fluid. Note that this difference can develop from two different fluids with different densities or from temperature differences within the same fluid causing density variations in space. For example, smoke from a fire rises because it is lighter (due to its higher temperature) than the surrounding air. THE WIDE WORLD OF FLUIDS Explosive lake In 1986 a tremendous explosion of carbon dioxide bubbles to nucleate, they rise, grow, and cause other bubbles to (CO2) from Lake Nyos, west of Cameroon, killed more than 1700 form, feeding a chain reaction. A related phenomenon often people and livestock. The explosion resulted from a buildup of CO2 occurs when a pop bottle is shaken and then opened. The pop that seeped into the high-pressure water at the bottom of the shoots from the container rather violently. When this set of events lake from warm springs of CO2-bearing water. The CO2-rich water is occurred in Lake Nyos, the entire lake overturned through a col- heavier than pure water and can hold a volume of CO2 more than umn of rising and expanding buoyant bubbles. The heavier-than- five times the water volume. As long as the gas remains dissolved air CO2 then flowed through the long, deep valleys surrounding in the water, the stratified lake (i.e., pure water on top, CO2 water the lake and asphyxiated human and animal life caught in the gas on the bottom) is stable. But if some mechanism causes the gas cloud. One victim was 27 km downstream from the lake. 2.11.2 Stability VIDEO V2.10 Density differences in fluids Stable Another interesting and important problem associated with submerged or floating bodies is concerned with the stability of the bodies. As illustrated by the figure in the margin, a body is said to be in a stable equilibrium position if, when displaced, it returns to its equilibrium position. Conversely, it is in an unstable equilibrium position if, when displaced (even slightly), it moves to a new equilibrium position. Stability considerations are particularly important for submerged or floating bodies because the centers of buoyancy and gravity do not necessarily coincide. A small rotation can result in either a restoring or overturning couple. For example, for the completely submerged body shown in Fig. 2.25, which has a center of gravity below the center of buoyancy, a rotation from its equilibrium position will create a restoring couple formed by the weight, 𝒲, and the buoyant force, FB, which causes the body to rotate back to its original position. Thus, for this configuration the body is stable. It is to be noted that as long as the center of gravity is below the center of buoyancy, this will always be true; that is, the body is in a stable equilibrium position with respect to small rotations. However, as is illustrated in Fig. 2.26, if the center of gravity of the Unstable The stability of a body can be determined by considering what happens when it is displaced from its equilibrium position. c FB CG FB c CG FB 𝒲 Stable CG c CG 𝒲 𝒲 Restoring couple 𝒲 c FB Unstable Overturning couple ■ Figure 2.25 Stability of a ■ Figure 2.26 Stability of a completely immersed body—center of gravity below centroid. completely immersed body—center of gravity above centroid. 74 Chapter 2 ■ Fluid Statics VIDEO V2.11 Stability of a floating cube 𝒲 𝒲 CG CG c c' FB FB c = centroid of original c' = centroid of new displaced volume displaced volume Restoring couple Stable ■ Figure 2.27 Stability of a floating body—stable configuration. 𝒲 𝒲 CG CG c FB Marginally stable c = centroid of original displaced volume c' FB c' = centroid of new displaced volume Unstable Very stable © RiverNorthPhotography/ iStockphoto VIDEO 2.12 V2.12 Stability of a model barge Overturning couple ■ Figure 2.28 Stability of a floating body—unstable configuration. completely submerged body is above the center of buoyancy, the resulting couple formed by the weight and the buoyant force will cause the body to overturn and move to a new equilibrium position. Thus, a completely submerged body with its center of gravity above its center of buoyancy is in an unstable equilibrium position. For floating bodies the stability problem is more complicated, because as the body rotates the location of the center of buoyancy (which is the centroid of the displaced volume) may change. As is shown in Fig. 2.27, a floating body such as a barge that rides low in the water can be stable even though the center of gravity lies above the center of buoyancy. This is true since as the body rotates the buoyant force, FB, shifts to pass through the centroid of the newly formed displaced volume and, as illustrated, combines with the weight, 𝒲, to form a couple that will cause the body to return to its original equilibrium position. However, for the relatively tall, slender body shown in Fig. 2.28, a small rotational displacement can cause the buoyant force and the weight to form an overturning couple as illustrated. We see from even these relatively simple examples that although the concept of floating body stability is not difficult to understand, determining the stability of specific floating bodies can become quite tedious. For even simple shapes, determination of the centroid of irregular displaced volumes can require many calculations as can determination of the center of mass of inhomogeneous bodies. Although both the relatively narrow kayak and the wide houseboat shown in the figures in the margin are both stable, the kayak will overturn much more easily than the houseboat. The problem can be further complicated by the necessary inclusion of other types of external forces such as those induced by wind gusts or currents. Stability considerations are obviously of great importance in the design of ships, submarines, bathyscaphes, and so forth; such considerations play a significant role in the work of naval architects (see, for example, Ref. 6). Pressure Variation in a Fluid with Rigid-Body Motion Although in this chapter we have been primarily concerned with fluids at rest, the general equation of motion (Eq. 2.2) −∇p − γkˆ = ρa Free ebooks ==> www.Ebook777.com 2.12 Even though a fluid may be in motion, if it moves as a rigid body there will be no shearing stresses present. There are no shearing stresses in a body of fluid experiencing rigid body translation or rotation. Pressure Variation in a Fluid with Rigid-Body Motion 75 was developed for both fluids at rest and fluids in motion, with the only stipulation being that there are no shearing stresses present. Equation 2.2 can be expressed in rectangular coordinate components, with the gravity vector in the negative z-direction, as − ∂p ∂p ∂p = ρax  − = ρay  − = γ + ρaz ∂x ∂y ∂z (2.24) A general class of problems involving fluid motion in which there are no shearing stresses occurs when a mass of fluid undergoes rigid-body motion. For example, if a container of fluid accelerates along a straight path, the fluid will move as a rigid mass (after the initial sloshing motion has died out) with each particle having the same acceleration. Since there is no deformation, there will be no shearing stresses and, therefore, Eq. 2.2 applies. Similarly, if a fluid is contained in a tank that rotates about a fixed axis, the fluid will simply rotate with the tank as a rigid body, and again Eq. 2.2 can be applied to obtain the pressure distribution throughout the moving fluid. Specific results for these two cases (rigid-body uniform motion and rigid-body rotation) are developed in the following two sections. Although problems relating to fluids having rigid-body motion are not, strictly speaking, “fluid statics” problems, they are included in this chapter because, as we will see, the analysis and resulting pressure relationships are similar to those for fluids at rest. 2.12.1 Linear Motion We first consider an open container of a liquid that is translating along a straight path with a constant acceleration a as illustrated in Fig. 2.29. Since ax = 0, it follows from the first of Eqs. 2.24 that the pressure gradient in the x direction is zero (∂p∂x = 0). In the y and z directions ∂p = −ρay ∂y (2.25) ∂p = −ρ(g + az ) ∂z (2.26) The change in pressure between two closely spaced points located at y, z, and y + dy, z + dz can be expressed as dp = g + az ∂p ∂p dy + dz ∂y ∂z Free surface slope = dz/dy ay p1 Constant p2 pressure p3 lines z x (a) y (b) ■ Figure 2.29 Linear acceleration of a liquid with a free surface. www.Ebook777.com az a ay (c) 76 Chapter 2 ■ Fluid Statics z or in terms of the results from Eqs. 2.25 and 2.26 dp = −ρay dy − ρ(g + az ) dz dy Along a line of constant pressure, dp = 0, and therefore from Eq. 2.27 it follows that the slope of this line is given by the relationship dz ay g + az ay dz =− dy g + az y The pressure distribution in a fluid mass that is accelerating along a straight path is not hydrostatic. (2.27) (2.28) This relationship is illustrated by the figure in the margin. Along a free surface the pressure is constant, so that for the accelerating mass shown in Fig. 2.29 the free surface will be inclined if ay ≠ 0. In addition, all lines of constant pressure will be parallel to the free surface as illustrated. For the special circumstance in which ay = 0, az ≠ 0, which corresponds to the mass of fluid accelerating in the vertical direction, Eq. 2.28 indicates that the fluid surface will be horizontal. However, from Eq. 2.26 we see that the pressure distribution is not hydrostatic, but is given by the equation dp = −ρ (g + az ) dz   For fluids of constant density this equation shows that the pressure will vary linearly with depth, but the variation is due to the combined effects of gravity and the externally induced acceleration, ρ(g + az ), rather than simply the specific weight ρg. Thus, for example, the pressure along the bottom of a liquid-filled tank that is resting on the floor of an elevator that is accelerating upward will be increased over that which exists when the tank is at rest (or moving with a constant velocity). It is to be noted that for a freely falling fluid mass (az = −g), the pressure gradients in all three coordinate directions are zero, which means that if the pressure surrounding the mass is zero, the pressure throughout will be zero. The pressure throughout a “blob” of orange juice floating in an orbiting space shuttle (a form of free fall) is uniform and is only slightly higher than ambient due to the surface tension holding it together (see Section 1.9). EXAMPLE 2.11 Pressure Variation in an Accelerating Tank ay GIVEN The cross section for the fuel tank of an experimental Vent vehicle is shown in Fig. E2.11. The rectangular tank is vented to the atmosphere, and the specific gravity of the fuel is SG = 0.65. A pressure transducer is located in its side as illustrated. During testing of the vehicle, the tank is subjected to a constant linear acceleration, ay. FIND (a) Determine an expression that relates ay and the pres- sure (in lbft2 ) at the transducer. (b) What is the maximum acceleration that can occur before the fuel level drops below the transducer? z Air y z1 Fuel (2) (1) 0.75 ft 0.5 ft Transducer 0.75 ft ■ Figure E2.11 SOLUTION (a) For a constant horizontal acceleration the fuel will move as a rigid body, and from Eq. 2.28 the slope of the fuel surface can be expressed as ay dz =− g dy since az = 0. Thus, for some arbitrary ay, the change in depth, z1, of liquid on the right side of the tank can be found from the equation − ay z1 =− g 0.75 ft or z1 = (0.75 ft)( ay g) Since there is no acceleration in the vertical, z, direction, the pressure along the wall varies hydrostatically as shown by Eq. 2.26. Thus, the pressure at the transducer is given by the relationship p = γh 2.12 where h is the depth of fuel above the transducer, and therefore p = (0.65) (62.4 lbft3 )[0.5 ft − (0.75 ft) (ayg)] ay (Ans) = 20.3 − 30.4 g or (ay ) max = 2 (ay ) max = 3 (32.2 fts2 ) = 21.5 fts2 (b) The limiting value for (ay ) max (when the fuel level reaches the transducer) can be found from the equation (Ans) COMMENT Note that the pressure in horizontal layers is not (ay ) max g 2g 3 and for standard acceleration due to gravity for z1 ≤ 0.5 ft. As written, p would be given in lbft2. 0.5 ft = (0.75 ft) [ 77 Pressure Variation in a Fluid with Rigid-Body Motion constant in this example since ∂p∂y = −ρay ≠ 0. Thus, for example, p1 ≠ p2. ] 2.12.2 Rigid-Body Rotation A fluid contained in a tank that is rotating with a constant angular velocity about an axis will rotate as a rigid body. After an initial “start-up” transient, a fluid contained in a tank that rotates with a constant angular velocity ω about an axis as is shown in Fig. 2.30 will rotate with the tank as a rigid body. It is known from elementary particle dynamics that the acceleration of a fluid particle located at a distance r from the axis of rotation is equal in magnitude to rω2, and the direction of the acceleration is toward the axis of rotation, as is illustrated in the figure. Since the paths of the fluid particles are circular, it is convenient to use cylindrical polar coordinates r, θ, and z, defined in the insert in Fig. 2.30. It will be shown in Chapter 6 that in terms of cylindrical coordinates the pressure gradient ∇p can be expressed as ∇p = ∂p ∂p 1 ∂p eˆ r + eˆ θ + eˆ r ∂r ∂θ ∂z z (2.29) Thus, in terms of this coordinate system ar = −rω2 eˆ r  aθ = 0  az = 0 and from Eq. 2.2 ∂p ∂p ∂p = ρrω2   = 0   = −γ ∂r ∂θ ∂z (2.30) Axis of rotation z ω r ar = rω 2 r θ ω y ez eθ x er (a) (b) ■ Figure 2.30 Rigid-body rotation of a liquid in a tank. (Photograph courtesy of Geno Pawlak.) (c) 78 Chapter 2 ■ Fluid Statics These results show that for this type of rigid-body rotation, the pressure is a function of two variables r and z, and therefore the differential pressure is dp = ∂p ∂p dr + dz ∂r ∂z or dp = ρrω2 dr − γ dz p (2.31) On a horizontal plane (dz = 0), it follows from Eq. 2.31 that dp/dr = ρω2r, which is greater than zero. Hence, as illustrated in the figure in the margin, because of centrifugal acceleration, the pressure increases in the radial direction. Along a surface of constant pressure, such as the free surface, dp = 0, so that from Eq. 2.31 (using γ = ρg) z = constant dp dr dp ––– = ρω 2r dr dz rω2 = g dr r Integration of this result gives the equation for surfaces of constant pressure as z= The free surface in a rotating liquid is curved rather than flat. ω2r2 + constant 2g (2.32) This equation reveals that these surfaces of constant pressure are parabolic, as illustrated in Fig. 2.31. Integration of Eq. 2.31 yields ∫ dp = ρω ∫ r dr − γ ∫ dz 2 or p= ρω2r2 − γz + constant 2 (2.33) where the constant of integration can be expressed in terms of a specified pressure at some arbitrary point r0, z0. This result shows that the pressure varies with the distance from the axis of rotation, but at a fixed radius, the pressure varies hydrostatically in the vertical direction as shown in Fig. 2.31. z p1 p1 Constant pressure lines p2 p2 p3 p3 p4 p4 ω 2r 2 ____ 2g r y x ■ Figure 2.31 Pressure distribution in a rotating liquid. 2.13 E X AMPLE 2.12 Chapter Summary and Study Guide 79 Free Surface Shape of Liquid in a Rotating Tank r GIVEN It has been suggested that the angular velocity, ω, of a rotating body or shaft can be measured by attaching an open cylinder of liquid, as shown in Fig. E2.12a, and measuring with some type of depth gage the change in the fluid level, H − h0, caused by the rotation of the fluid. R Depth gage Initial depth r FIND Determine the relationship between this change in fluid level and the angular velocity. h SOLUTION h0 z The height, h, of the free surface above the tank bottom can be determined from Eq. 2.32, and it follows that h 0 ω ω2r2 + h0 h= 2g The initial volume of fluid in the tank, Vi, is equal to H dr (a) (b) ■ Figure E2.12 Vi = πR2H The volume of the fluid with the rotating tank can be found with the aid of the differential element shown in Fig. E2.12b. This cylindrical shell is taken at some arbitrary radius, r, and its volume is dV = 2πrh dr The total volume is, therefore, V = 2π ∫ 0 R r(        ω2r2 πω2R4 + h0) dr = + πR2h0 2g 4g or H − h0 = ω2R2 4g (Ans) COMMENT This is the relationship we were looking for. It shows that the change in depth could indeed be used to determine the rotational speed, although the relationship between the change in depth and speed is not a linear one. Since the volume of the fluid in the tank must remain constant (assuming that none spills over the top), it follows that πR2H = πω2R4 + πR2h0 4g THE WIDE WORLD OF FLUIDS Rotating mercury mirror telescope A telescope mirror has the entered service it was the third largest optical telescope in North same shape as the parabolic free surface of a liquid in a rotating America. With a diameter of 6 ft and a rotation rate of 7 rpm, this tank. The liquid mirror telescope (LMT) consists of a pan of liquid mirror uses 30 liters of mercury for its 1-mm thick, parabolic- (normally mercury because of its excellent reflectivity) rotating to shaped mirror. One of the major benefits of a LMT (compared to a produce the required parabolic shape of the free surface mirror. normal glass mirror telescope) is its low cost. Perhaps the main With recent technological advances, it is possible to obtain the disadvantage is that a LMT can look only straight up, although vibration-free rotation and the constant angular velocity neces- there are many galaxies, supernova explosions, and pieces of sary to produce a liquid mirror surface precise enough for astro- space junk to view in any part of the sky. The next-generation nomical use. Construction of the largest LMT, located at the LMTs may have movable secondary mirrors to allow a larger por- University of British Columbia, was completed in 2003. When it tion of the sky to be viewed. (See Problem 2.163.) 2.13 Chapter Summary and Study Guide In this chapter the pressure variation in a fluid at rest is considered, along with some important consequences of this type of pressure variation. It is shown that for incompressible fluids at rest the pressure varies linearly with depth. This type of variation is commonly referred to as a hydrostatic pressure distribution. For compressible fluids at rest the pressure distribution will not generally be Free ebooks ==> www.Ebook777.com 80 Chapter 2 ■ Fluid Statics Pascal’s law surface force body force incompressible fluid hydrostatic pressure distribution pressure head compressible fluid U.S. standard atmosphere absolute pressure gage pressure vacuum pressure barometer manometer Bourdon pressure gage center of pressure buoyant force Archimedes’ principle center of buoyancy hydrostatic, but Eq. 2.4 remains valid and can be used to determine the pressure distribution if additional information about the variation of the specific weight is specified. The distinction between absolute and gage pressure is discussed along with a consideration of barometers for the measurement of atmospheric pressure. Pressure-measuring devices called manometers, which utilize static liquid columns, are analyzed in detail. A brief discussion of mechanical and electronic pressure gages is also included. Equations for determining the magnitude and location of the resultant fluid force acting on a plane surface in contact with a static fluid are developed. A general approach for determining the magnitude and location of the resultant fluid force acting on a curved surface in contact with a static fluid is developed. For submerged or floating bodies the concept of the buoyant force and the use of Archimedes’ principle are reviewed. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ ■ ■ ■ ■ ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. calculate the pressure at various locations within an incompressible fluid at rest. calculate the pressure at various locations within a compressible fluid at rest using Eq. 2.4 if the variation in the specific weight is specified. use the concept of a hydrostatic pressure distribution to determine pressures from measurements using various types of manometers. determine the magnitude, direction, and location of the resultant hydrostatic force acting on a plane surface. determine the magnitude, direction, and location of the resultant hydrostatic force acting on a curved surface. use Archimedes’ principle to calculate the resultant hydrostatic force acting on floating or submerged bodies. analyze, based on Eq. 2.2, the motion of fluids moving with simple rigid-body linear motion or simple rigid-body rotation. Some of the important equations in this chapter are: Pressure gradient in a stationary fluid Pressure variation in a stationary incompressible fluid Hydrostatic force on a plane surface Location of hydrostatic force on a plane surface dp = −γ dz p1 = γh + p2 FR = γhcA Ixc + yc yR = ycA xR = Ixyc + xc ycA Buoyant force FB = γV ∂p ∂p ∂p Pressure gradient in rigid-body motion − = ρax, − = ρay, − = γ + ρaz ∂x ∂y ∂z ∂p ∂p ∂p Pressure gradient in rigid-body rotation = ρrω2,  = 0,  = −γ ∂r ∂θ ∂z (2.4) (2.7) (2.18) (2.19) (2.20) (2.22) (2.24) (2.30) References 1. The U.S. Standard Atmosphere, 1962, U.S. Government Printing Office, Washington, DC, 1962. 2. The U.S. Standard Atmosphere, 1976, U.S. Government Printing Office, Washington, DC, 1976. 3. Benedict, R. P., Fundamentals of Temperature, Pressure, and Flow Measurements, 3rd Ed., Wiley, New York, 1984. www.Ebook777.com Problems 81 4. Dally, J. W., Riley, W. F., and McConnell, K. G., Instrumentation for Engineering Measurements, 2nd Ed., Wiley, New York, 1993. 5. Holman, J. P., Experimental Methods for Engineers, 4th Ed., McGraw-Hill, New York, 1983. 6. Comstock, J. P., ed., Principles of Naval Architecture, Society of Naval Architects and Marine Engineers, New York, 1967. 7. Hasler, A. F., Pierce, H., Morris, K. R., and Dodge, J., “Meteorological Data Fields ‘In Perspective’,” Bulletin of the American Meteorological Society, Vol. 66, No. 7, July 1985. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are no unique answers to these problems. Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. 2.7 A 30-ft-high downspout of a house is clogged at the bottom. Find the pressure at the bottom if the downspout is filled with 60 °F rainwater. Section 2.3 Pressure Variation in a Fluid at Rest 2.8 How high a column of SAE 30 oil would be required to give the same pressure as 700 mm Hg? 2.1 Obtain a photograph/image of a situation in which the fact that in a static fluid the pressure increases with depth is important. Print this photo and write a brief paragraph that describes the situation involved. 2.9 Bathyscaphes are capable of submerging to great depths in the ocean. What is the pressure at a depth of 5 km, assuming that seawater has a constant specific weight of 10.1 kN/m3? Express your answer in pascals and psi. 2.2 The deepest known spot in the oceans is the Challenger Deep in the Mariana Trench of the Pacific Ocean and is approximately 11,000 m below the surface. Assume that the salt water density is constant at 1025 kg/m3 and determine the pressure at this depth. 2.10 The deepest known spot in the oceans is the Challenger Deep in the Mariana Trench of the Pacific Ocean and is approximately 11,000 m below the surface. For a surface density of 1030 kg/m3, a constant water temperature, and an isothermal bulk modulus of elasticity of 2.3 × 109 N/m2, find the pressure at this depth. 2.3 A closed tank is partially filled with glycerin. If the air pressure in the tank is 6 lb/in.2 and the depth of glycerin is 10 ft, what is the pressure in lb/ft2 at the bottom of the tank? 2.4 A 3-m-diameter vertical cylindrical tank is filled with water to a depth of 11 m. The rest of the tank is filled with air at atmospheric pressure. What is the absolute pressure at the bottom of the tank? Blood pressure is usually given as a ratio of the maximum 2.5 pressure (systolic pressure) to the minimum pressure (diastolic pressure). As shown in Video V2.3, such pressures are commonly measured with a mercury manometer. A typical value for this ratio for a human would be 120/70, where the pressures are in mm Hg. (a) What would these pressures be in pascals? (b) If your car tire was inflated to 120 mm Hg, would it be sufficient for normal driving? 2.6 An unknown immiscible liquid seeps into the bottom of an open oil tank. Some measurements indicate that the depth of the unknown liquid is 1.5 m and the depth of the oil (specific weight = 8.5 kN/m3) floating on top is 5.0 m. A pressure gage connected to the bottom of the tank reads 65 kPa. What is the specific gravity of the unknown liquid? 2.11 A submarine submerges by admitting seawater (S = 1.03) into its ballast tanks. The amount of water admitted is controlled by air pressure, because seawater will cease to flow into the tank when the internal pressure (at the hull penetration) is equal to the hydrostatic pressure at the depth of the submarine. Consider a ballast tank, which can be modeled as a vertical half-cylinder (R = 8 ft, L = 20 ft) for which the air pressure control valve has failed shut. The failure occurred at the beginning of a dive from 60 ft to 1000 ft. The tank was initially filled with seawater to a depth of 2 ft and the air was at a temperature of 40 °F. As the weight of water in the tank is important in maintaining the boat’s attitude, determine the weight of water in the tank as a function of depth during the dive. You may assume that tank internal pressure is always in equilibrium with the ocean’s hydrostatic pressure and that the inlet pipe to the tank is at the bottom of the tank and penetrates the hull at the “depth” of the submarine. 82 Chapter 2 ■ Fluid Statics 2.12 Determine the pressure at the bottom of an open 5-m-deep tank in which a chemical process is taking place that causes the density of the liquid in the tank to vary as ρ = ρsurf B 1 + sin2( h π , hbot 2 ) where h is the distance from the free surface and ρsurf = 1700 kg/m3. *2.13 In a certain liquid at rest, measurements of the specific weight at various depths show the following variation: h (ft) 0 10 20 30 40 50 60 70 80 90 100 γ (lb/ft3 ) 6m †2.14 Because of elevation differences, the water pressure in the second floor of your house is lower than it is in the first floor. For tall buildings this pressure difference can become unacceptable. Discuss possible ways to design the water distribution system in very tall buildings so that the hydrostatic pressure difference is within acceptable limits. *2.15 Under normal conditions the temperature of the atmosphere decreases with increasing elevation. In some situations, however, a temperature inversion may exist so that the air temperature increases with elevation. A series of temperature probes on a mountain give the elevation–temperature data shown in the table below. If the barometric pressure at the base of the mountain is 12.1 psia, determine by means of numerical integration the pressure at the top of the mountain. 5000 5500 6000 6400 7100 7400 8200 8600 9200 9900 2.17 (See The Wide World of Fluids article titled “Giraffe’s blood pressure,” Section 2.3.1.) (a) Determine the change in hydrostatic pressure in a giraffe’s head as it lowers its head from eating leaves 6 m above the ground to getting a drink of water at ground level as shown in Fig. P2.17. Assume the specific gravity of blood is SG = 1. (b) Compare the pressure change calculated in part (a) to the normal 120 mm of mercury pressure in a human’s heart. 70 76 84 91 97 102 107 110 112 114 115 The depth h = 0 corresponds to a free surface at atmospheric pressure. Determine, through numerical integration of Eq. 2.4, the corresponding variation in pressure and show the results on a plot of pressure (in psf) versus depth (in feet). Elevation (ft) child can use to empty a milk container, assuming that the child can develop 75 mm Hg of suction, and use this answer to determine if you think this is a reasonable estimate of the suction that a child can develop. Temperature (°F) 50.1 (base) 55.2 60.3 62.6 67.0 68.4 70.0 69.5 68.0 67.1 (top) 2.16 Often young children drink milk ( ρ = 1030 kg/m3) through a straw. Determine the maximum length of a vertical straw that a Water ■ Figure P2.17 Section 2.4 Standard Atmosphere 2.18 What would be the barometric pressure reading, in mm Hg, at an elevation of 4 km in the U.S. standard atmosphere? (Refer to Table C.2 in Appendix C.) 2.19 Denver, Colorado, is called the “mile-high city” because its state capitol stands on land 1 mi above sea level. Assuming that the Standard Atmosphere exists, what is the pressure and temperature of the air in Denver? The temperature follows the lapse rate (T = T0 − Bz). 2.20 Assume that a person skiing high in the mountains at an altitude of 15,000 ft takes in the same volume of air with each breath as she does while walking at sea level. Determine the ratio of the mass of oxygen inhaled for each breath at this high altitude compared to that at sea level. 2.21 Pikes Peak near Denver, Colorado, has an elevation of 14,110 ft. (a) Determine the pressure at this elevation, based on Eq. 2.12. (b) If the air is assumed to have a constant specific weight of 0.07647 lb/ft3, what would the pressure be at this altitude? (c) If the air is assumed to have a constant temperature of 59 °F, what would the pressure be at this elevation? For all three cases assume standard atmospheric conditions at sea level (see Table 2.1). 2.22 Equation 2.12 provides the relationship between pressure and elevation in the atmosphere for those regions in which the temperature varies linearly with elevation. Derive this equation and verify the value of the pressure given in Table C.2 in Appendix C for an elevation of 5 km. 2.23 As shown in Fig. 2.6 for the U.S. standard atmosphere, the troposphere extends to an altitude of 11 km where the pressure is 22.6 kPa (abs). In the next layer, called the stratosphere, the Problems temperature remains constant at −56.5 °C. Determine the pressure and density in this layer at an altitude of 15 km. Assume g = 9.77 m/s2 in your calculations. Compare your results with those given in Table C.2 in Appendix C. 2.24 (See The Wide World of Fluids article titled “Weather, barometers, and bars,” Section 2.5.) The record low sea-level barometric pressure ever recorded is 25.8 in. of mercury. At what altitude in the standard atmosphere is the pressure equal to this value? 83 2.31 A U-tube manometer is used to check the pressure of natural gas entering a furnace. One side of the manometer is connected to the gas inlet line, and the water level in the other side open to atmospheric pressure rises 3 in. What is the gage pressure of the natural gas in the inlet line in in. H2O and in lb/in2 gage? 2.32 A barometric pressure of 29.4 in. Hg corresponds to what value of atmospheric pressure in psia, and in pascals? 2.25 On a given day, a barometer at the base of the Washington Monument reads 29.97 in. of mercury. What would the barometer reading be when you carry it up to the observation deck 500 ft above the base of the monument? 2.33 For an atmospheric pressure of 101 kPa (abs) determine the heights of the fluid columns in barometers containing one of the following liquids: (a) mercury, (b) water, and (c) ethyl alcohol. Calculate the heights including the effect of vapor pressure and compare the results with those obtained neglecting vapor pressure. Do these results support the widespread use of mercury for barometers? Why? 2.26 Aneroid barometers can be used to measure changes in altitude. If a barometer reads 30.1 in. Hg at one elevation, what has been the change in altitude in meters when the barometer reading is 28.3 in. Hg? Assume a standard atmosphere and that Eq. 2.12 is applicable over the range of altitudes of interest. 2.34 The closed tank of Fig. P.2.34 is filled with water and is 5 ft long. The pressure gage on the tank reads 7 psi. Determine: (a) the height, h, in the open water column, (b) the gage pressure acting on the bottom tank surface AB, and (c) the absolute pressure of the air in the top of the tank if the local atmospheric pressure is 14.7 psia. Section 2.5 Measurement of Pressure Bourdon gages (see Video V2.4 and Fig. 2.13) are com2.27 monly used to measure pressure. When such a gage is attached to the closed water tank of Fig. P2.27 the gage reads 5 psi. What is the absolute air pressure in the tank? Assume standard atmospheric pressure of 14.7 psi. Open 7 psi Air Air h 2 ft Water 2 ft A B ■ Figure P2.34 12 in. Bourdon gage 15 20 10 25 5 30 0 Water 2.35 A mercury manometer is connected to a large reservoir of water as shown in Fig. P2.35. Determine the ratio, hw /hm, of the distances hw and hm indicated in the figure. 35 6 in. Water hw ■ Figure P2.27 hm hm 2.28 On the suction side of a pump, a Bourdon pressure gage reads 40 kPa vacuum. What is the corresponding absolute pressure if the local atmospheric pressure is 100 kPa (abs)? 2.29 A Bourdon pressure gage attached to the outside of a tank containing air reads 77.0 psi when the local atmospheric pressure is 760 mm Hg. What will be the gage reading if the atmospheric pressure increases to 773 mm Hg? Section 2.6 Manometry 2.30 Obtain a photograph/image of a situation in which the use of a manometer is important. Print this photo and write a brief paragraph that describes the situation involved. Mercury ■ Figure P2.35 2.36 The U-tube manometer shown in Fig. P2.36 has two fluids, water and oil (S = 0.80). Find the height difference between the free water surface and the free oil surface with no applied pressure difference. 84 Chapter 2 ■ Fluid Statics pA pA a specific gravity of 0.8, and the remaining parts of the system are filled with water. If the pressure gage reading at A is 60 kPa, determine (a) the pressure in pipe B, and (b) the pressure head, in millimeters of mercury, at the top of the dome (point C). g 10 cm Oil Hemispherical dome SG = 0.8 C Water pA = 60 kPa 4m 3m A 3m Water ■ Figure P2.36 2.37 A U-tube manometer is connected to a closed tank containing air and water as shown in Fig. P2.37. At the closed end of the manometer the air pressure is 16 psia. Determine the reading on the pressure gage for a differential reading of 4 ft on the manometer. Express your answer in psi (gage). Assume standard atmospheric pressure and neglect the weight of the air columns in the manometer. 2m B Water ■ Figure P2.39 2.40 Two pipes are connected by a manometer as shown in Fig. P2.40. Determine the pressure difference, pA − pB, between the pipes. Closed valve Air pressure = 16 psia A 0.5 m 4 ft Water Air 0.6 m 1.3 m 2 ft Water Gage fluid ( γ = 90 lb / ft 3 ) Pressure gage Gage fluid (SG = 2.6) Water B ■ Figure P2.37 ■ Figure P2.40 2.38 The container shown in Fig. P2.38 holds 60 °F water and 60 °F air as shown. Find the absolute pressures at locations A, B, and C. patm = 14.7 lb/in.2 Air B 8″ 2.41 Find the percentage difference in the readings of the two identical U-tube manometers shown in Fig. P2.41. Manometer 90 uses 90 °C water and manometer 30 uses 30 °C water. Both have the same applied pressure difference. Does this percentage change with the magnitude of the applied pressure difference? Can the difference between the two readings be ignored? 10″ 14″ A h90 h30 g Water Water C ■ Figure P2.38 2.39 A closed cylindrical tank filled with water has a hemispherical dome and is connected to an inverted piping system as shown in Fig. P2.39. The liquid in the top part of the piping system has Manometer 90 ■ Figure P2.41 Manometer 30 85 Problems 2.42 A U-tube manometer is connected to a closed tank as shown in Fig. P2.42. The air pressure in the tank is 0.50 psi and the liquid in the tank is oil (γ = 54.0 lb/ft3). The pressure at point A is 2.00 psi. Determine: (a) the depth of oil, z, and (b) the differential reading, h, on the manometer. pL pR h ρA Open Air z ρA H Oil ρB h A 2 ft SG = 3.05 ■ Figure P2.45 ■ Figure P2.42 2.43 For the inclined-tube manometer of Fig. P2.43, the pressure in pipe A is 0.6 psi. The fluid in both pipes A and B is water, and the gage fluid in the manometer has a specific gravity of 2.6. What is the pressure in pipe B corresponding to the differential reading shown? Water B Water A 3 in. 8 in. Vapor 1m Open 1m h Liquid 1m 3 in. Mercury SG = 2.6 30° ■ Figure P2.46 ■ Figure P2.43 2.47 Determine the elevation difference, Δh, between the water levels in the two open tanks shown in Fig. P2.47. 2.44 A flowrate measuring device is installed in a horizontal pipe through which water is flowing. A U-tube manometer is connected to the pipe through pressure taps located 3 in. on either side of the device. The gage fluid in the manometer has a specific weight of 112 lb/ft3. Determine the differential reading of the manometer corresponding to a pressure drop between the taps of 0.5 lb/in.2 SG = 0.90 0.4 m 2.45 The sensitivity Sen of the micromanometer shown in Fig. P2.45 is defined as Sen = H . pL − pR Δh 1m Find the sensitivity of the micromanometer in terms of the densities ρA and ρB. How can the sensitivity be increased? 2.46 The cylindrical tank with hemispherical ends shown in Fig. P2.46 contains a volatile liquid and its vapor. The liquid density is 800 kg/m3, and its vapor density is negligible. The pressure in the vapor is 120 kPa (abs) and the atmospheric pressure is 101 kPa (abs). Determine: (a) the gage pressure reading on the pressure gage, and (b) the height, h, of the mercury, in the manometer. Water ■ Figure P2.47 2.48 What is the specific gravity of the liquid in the left leg of the U-tube manometer shown in Fig. P2.48? 86 Chapter 2 ■ Fluid Statics patm patm g Water Unknown fluid 15 cm Water (S = 1) H 1.00 m 20 cm 10 cm ■ Figure P2.48 Hg 2.49 For the configuration shown in Fig. P2.49 what must be the value of the specific weight of the unknown fluid? Express your answer in lb/ft3. 0.2 m ■ Figure P2.51 Open Open *2.52 Both ends of the U-tube mercury manometer of Fig. P2.52 are initially open to the atmosphere and under standard atmospheric pressure. When the valve at the top of the right leg is open, the level of mercury below the valve is h1. After the valve is closed, air pressure is applied to the left leg. Determine the relationship between the differential reading on the manometer and the applied gage pressure, pg. Show on a plot how the differential reading varies with pg for h1 = 25, 50, 75, and 100 mm over the range 0 ≤ pg ≤ 300 kPa. Assume that the temperature of the trapped air remains constant. Water 5.5 in. 4.9 in. 1.4 in. Unknown fluid 3.3 in. Valve pg ■ Figure P2.49 hi 2.50 The manometer shown in Fig. P2.50 has an air bubble eiter in (a) the right horizontal line or (b) the left vertical leg. Find h1 − h2 for both cases if pA = pB. pA pA = pB = patm = 14.7 psia Mercury pB ■ Figure P2.52 Water p = 16.1 psia h1 2.53 The inverted U-tube manometer of Fig. P2.53 contains oil (SG = 0.9) and water as shown. The pressure differential between pipes A and B, pA − pB, is −5 kPa. Determine the differential reading h. Water h2 h Air bubble Oil 12″ h 0.2 m Water Air bubble ■ Figure P2.50 2.51 The U-tube manometer shown in Fig. P2.51 has legs that are 1.00 m long. When no pressure difference is applied across the manometer, each leg has 0.40 m of mercury. What is the maximum pressure difference that can be indicated by the manometer? A 0.3 m Water B ■ Figure P2.53 Problems 2.54 An inverted U-tube manometer containing oil (SG = 0.8) is located between two reservoirs as shown in Fig. P2.54. The reservoir on the left, which contains carbon tetrachloride, is closed and pressurized to 8 psi. The reservoir on the right contains water and is open to the atmosphere. With the given data, determine the depth of water, h, in the right reservoir. B 0.4 m Oil A 0.3 m Gasoline Carbon tetrachloride 87 Mercury 8 psi 0.7 ft 3 ft Oil ■ Figure P2.56 Water 1 ft h 1 ft 2.57 The mercury manometer of Fig. P2.57 indicates a differential reading of 0.30 m when the pressure in pipe A is 30-mm Hg vacuum. Determine the pressure in pipe B. ■ Figure P2.54 2.55 The sensitivity Sen of the manometer shown in Fig. P2.55 can be defined as 0.50 m Water A h . Sen = pR − pL Three manometer fluids with the listed specific gravities S are available: Kerosene, S = 0.82; SAE 10 oil, S = 0.87; and Normal octane, S = 0.71. Which fluid gives the highest sensitivity? The areas AR and AL are much larger than the cross-sectional area of the manometer tube, so H << h. pL Oil B 0.15 m Mercury 0.30 m ■ Figure P2.57 2.58 Consider the cistern manometer shown in Fig. P2.58. The scale is set up on the basis that the cistern area A1 is infinite. However, A1 is actually 50 times the internal cross-sectional area A2 of the inclined tube. Find the percentage error (based on the scale reading) involved in using this scale. pR A2 pH H HL pL x Fluid hL HR Air hR Oil, S = 0.85 0 Air θ = 30° A1 h ■ Figure P2.58 Water γW Water γW 2.59 The cistern shown in Fig. P2.59 has a diameter D that is 4 times the diameter d of the inclined tube. Find the drop in the fluid pA AL AR d ■ Figure P2.55 2.56 In Fig. P2.56 pipe A contains gasoline (SG = 0.7), pipe B contains oil (SG = 0.9), and the manometer fluid is mercury. Determine the new differential reading if the pressure in pipe A is decreased 25 kPa, and the pressure in pipe B remains constant. The initial differential reading is 0.30 m as shown. ℓ 0 θ D ■ Figure P2.59 pB 88 Chapter 2 ■ Fluid Statics level in the cistern and the pressure difference (pA − pB) if the liquid in the inclined tube rises ℓ = 20 in. The angle θ is 20°. The fluid’s specific gravity is 0.85. 2.60 The inclined differential manometer of Fig. P2.60 contains carbon tetrachloride. Initially the pressure differential between pipes A and B, which contain a brine (SG = 1.1), is zero as illustrated in the figure. It is desired that the manometer give a differential reading of 12 in. (measured along the inclined tube) for a pressure differential of 0.1 psi. Determine the required angle of inclination, θ. A A B Oil (SG = 0.8) Water Area = A1 Area = A2 Mercury ■ Figure P2.63 B 2.64 Determine the change in the elevation of the mercury in the left leg of the manometer of Fig. P2.64 as a result of an increase in pressure of 5 psi in pipe A while the pressure in pipe B remains constant. Brine Brine 12 in. Water Carbon tetrachloride θ 1 _ -in.2 diameter 2.61 Determine the new differential reading along the inclined leg of the mercury manometer of Fig. P2.61, if the pressure in pipe A is decreased 10 kPa and the pressure in pipe B remains unchanged. The fluid in A has a specific gravity of 0.9 and the fluid in B is water. SG = 0.9 30° 12 in. 1 _ -in.-diameter 4 30° Mercury ■ Figure P2.64 2.65 The U-shaped tube shown in Fig. 2.65 initially contains water only. A second liquid with specific weight, γ, less than water is placed on top of the water with no mixing occurring. Can the height, h, of the second liquid be adjusted so that the left and right levels are at the same height? Provide proof of your answer. γ Water B 6 in. 18 in. ■ Figure P2.60 A Oil (SG = 0.9) A h 100 mm B 80 mm 50 mm Mercury D1 = 1.5 D2 D2 Water ■ Figure P2.65 ■ Figure P2.61 *2.66 An inverted hollow cylinder is pushed into the water as is shown in Fig. P2.66. Determine the distance, ℓ, that the water rises in the cylinder as a function of the depth, d, of the lower edge of the cyl2.62 A student needs to measure the air pressure inside a com- inder. Plot the results for 0 ≤ d ≤ H, when H is equal to 1 m. Assume pressed air tank but does not have ready access to a pressure gage. the temperature of the air within the cylinder remains constant. Using materials already in the lab, she builds a U-tube manometer using two clear 3-ft-long plastic tubes, flexible hoses, and a tape measure. The only readily available liquids are water from a tap and a bottle of corn syrup. She selects the corn syrup because it has a larger density (SG = 1.4). What is the maximum air pressure, H Water in psia, that can be measured? d 2.63 Determine the ratio of areas, A 1/A 2, of the two manometer legs of Fig. P2.63 if a change in pressure in pipe B of 0.5 psi gives a corresponding change of 1 in. in the level of the mercury in the right leg. The pressure in pipe A does not change. D Open end ■ Figure P2.66 ℓ Problems Bourdon gage Section 2.8 Hydrostatic Force on a Plane Surface 2.67 Obtain a photograph/image of a situation in which the hydrostatic force on a plane surface is important. Print this photo and write a brief paragraph that describes the situation involved. 2.68 The basic elements of a hydraulic press are shown in Fig. P2.68. The plunger has an area of 1 in.2, and a force, F1, can be applied to the plunger through a lever mechanism having a mechanical advantage of 8 to 1. If the large piston has an area of 150 in.2, what load, F2, can be raised by a force of 30 lb applied to the lever? Neglect the hydrostatic pressure variation. 89 𝒲 0.4-in.-diameter θ Liquid (b) (a) ■ Figure P2.70 2.71 A bottle jack allows an average person to lift one corner of a 4000-lb automobile completely off the ground by exerting less than 20 lb of force. Explain how a 20-lb force can be converted into hundreds or thousands of pounds of force, and why this does not violate our general perception that you can’t get something for nothing (a somewhat loose paraphrase of the first law of thermodynamics). Hint: Consider the work done by each force. F2 2.72 Suction is often used in manufacturing processes to lift 1 objects to be moved to a new location. A 4-ft by 8-ft sheet of 2 -in. plywood weighs approximately 36 lb. If the machine’s end effector has a diameter of 5 in., determine the suction pressure required to lift the sheet, expressed in inches of H2O suction. Plunger F1 Hydraulic fluid ■ Figure P2.68 2.69 The hydraulic cylinder shown in Fig. P2.69, with a 4-in.diameter piston, is advertised as being capable of providing a force of F = 20 tons. If the piston has a design pressure (the maximum pressure at which the cylinder should safely operate) of 2500 lbin2, gage, can the cylinder safely provide the advertised force? P Piston h pmax = 2500 Ib/in.2 gage patm 2.73 A piston having a cross-sectional area of 0.07 m2 is located in a cylinder containing water as shown in Fig. P2.73. An open U-tube manometer is connected to the cylinder as shown. For h1 = 60 mm and h = 100 mm, what is the value of the applied force, P, acting on the piston? The weight of the piston is negligible. Water h1 Mercury F ■ Figure P2.73 ■ Figure P2.69 *2.70 A Bourdon gage (see Fig. 2.13 and Video V2.4) is often used to measure pressure. One way to calibrate this type of gage is to use the arrangement shown in Fig. P2.70a. The container is filled with a liquid and a weight, 𝒲, placed on one side with the gage on the other side. The weight acting on the liquid through a 0.4-in.-diameter opening creates a pressure that is transmitted to the gage. This arrangement, with a series of weights, can be used to determine what a change in the dial movement, θ, in Fig. P2.70b, corresponds to in terms of a change in pressure. For a particular gage, some data are given below. Based on a plot of these data, determine the relationship between θ and the pressure, p, where p is measured in psi. 𝒲 (lb) 0 1.04 2.00 3.23 4.05 5.24 6.31 θ (deg.) 0 20 40 60 80 100 120 2.74 A 6-in.-diameter piston is located within a cylinder that is 1 connected to a 2-in.-diameter inclined-tube manometer as shown in Fig. P2.74. The fluid in the cylinder and the manometer is oil (specific weight = 59 lbft3 ). When a weight, 𝒲, is placed on the top of the cylinder, the fluid level in the manometer tube rises from point (1) to (2). How heavy is the weight? Assume that the change in position of the piston is negligible. 𝒲 6 in. (1) Piston Oil ■ Figure P2.74 (2) 30° 90 Chapter 2 ■ Fluid Statics 2.75 The container shown in Fig. P2.75 has square cross sections. Find the vertical force on the horizontal surface, ABCD. 2 pressure. Calculate the force required to open the door if the force is applied 24 in. from the hinge line. See Fig. P2.78. (The driver did not have the presence of mind to open the window to escape.) 1 _″ 22 4′ A B C 36′′ D 1 _″ Water 24′′ 42 E 40′′ F 4 ■ Figure P2.78 ■ Figure P2.75 2.76 Find the weight 𝒲 needed to hold the wall shown in Fig. P2.76 upright. The wall is 10 m wide. 𝒲 2.79 Consider the gate shown in Fig. P2.79. The gate is massless and has a width b (perpendicular to the paper). The hyprostatic pressure on the vertical side creates a counterclockwise moment about the hinge, and the hydrostatic pressure on the horizontal side (or bottom) creates a clockwise moment about the hinge. Show that the net clockwise moment is 3m ∑u = ρwghb( ℓ2 h2 − ). 2 6 Wall Water 4m Pinned ℓ ■ Figure P2.76 h Hinge Water, pw 2.77 Determine the magnitude and direction of the force that must be applied to the bottom of the gate shown in Fig. P2.77 to keep the gate closed. ■ Figure P2.79 2.80 Will the gate in Problem 44 ever open? 2.81 A tank contains 6 in. of oil (S = 0.82) above 6 in. of water (S = 1.00). Find the force on the bottom of the tank. See Fig. P2.81. Water A = 1 ft2 2.1 m Hinge 2-m-wide gate 0.8 m ■ Figure P2.77 2.78 An automobile has just dropped into a river. The car door is approximately a rectangle, measures 36 in. wide and 40 in. high, and hinges on a vertical side. The water level inside the car is up to the midheight of the door, and the air inside the car is at atmospheric Oil (S = 0.82) Water (S = 1.0) ■ Figure P2.81 2.82 A structure is attached to the ocean floor as shown in Fig. P2.82. A 2-m-diameter hatch is located in an inclined wall and Problems hinged on one edge. Determine the minimum air pressure, p1, within the container that will open the hatch. Neglect the weight of the hatch and friction in the hinge. Free surface 2.86 While building a high, tapered concrete wall, builders used the wooden forms shown in Fig. P2.86. If concrete has a specific gravity of about 2.5, find the total force on each of the three side sections (A, B, and C) of the wooden forms (neglect any restraining force of the two ends of the forms). Freshly poured concrete Seawater 10 m 8 ft 30° Hatch 91 A Wooden forms B Hinge Air pressure, p1 6 ft C 5 ft 12 ft ■ Figure P2.82 2.83 Concrete is poured into the forms as shown in Fig. P2.83 to produce a set of steps. Determine the weight of the sandbag needed to keep the bottomless forms from lifting off the ground. The weight of the forms is 85 lb, and the specific weight of the concrete is 150 lb/ft3. Open top 8 in. risers 3 ft Sand 10 in. tread 8 ft Hardened concrete ■ Figure P2.86 2.87 A homogeneous, 4-ft-wide, 8-ft-long rectangular gate weighing 800 lb is held in place by a horizontal flexible cable as shown in Fig. P2.87. Water acts against the gate, which is hinged at point A. Friction in the hinge is negligible. Determine the tension in the cable. Cable Open bottom 60° ■ Figure P2.83 2.84 A long, vertical wall separates seawater from fresh water. If the seawater stands at a depth of 7 m, what depth of freshwater is required to give a zero resultant force on the wall? When the resultant force is zero, will the moment due to the fluid forces be zero? Explain. 2.85 Forms used to make a concrete basement wall are shown in Fig. P2.85. Each 4-ft-long form is held together by four ties—two at the top and two at the bottom as indicated. Determine the tension in the upper and lower ties. Assume concrete acts as a fluid with a weight of 150 lb /ft3. Water Gate 6 ft 8 ft Hinge A ■ Figure P2.87 2.88 A gate having the shape shown in Fig. P2.88 is located in the vertical side of an open tank containing water. The gate is mounted on a horizontal shaft. (a) When the water level is at the top of the gate, determine the magnitude of the fluid force on the rectangular 10 in. 1 ft Tie Concrete 10 ft Form Water 6m Shaft 3m Side view of gate 1 ft ■ Figure P2.85 ■ Figure P2.88 92 Chapter 2 ■ Fluid Statics portion of the gate above the shaft and the magnitude of the fluid force on the semicircular portion of the gate below the shaft. (b) For this same fluid depth determine the moment of the force acting on the semicircular portion of the gate with respect to an axis that coincides with the shaft. 2.89 A pump supplies water under pressure to a large tank as shown in Fig. P2.89. The circular-plate valve fitted in the short discharge pipe on the tank pivots about its diameter A–A and is held shut against the water pressure by a latch at B. Show that the force on the latch is independent of the supply pressure, p, and the height of the tank, h. 2.94 Figure P2.94 is a representation of the Keswick gravity dam in California. Find the magnitudes and locations of the hydrostatic forces acting on the headwater vertical wall of the dam and on the tailwater inclined wall of the dam. Note that the slope given is the ratio of the run to the rise. Consider a unit length of the dam (b = 1 ft). Headwater 21.5′ El. 595.00′ El. 565.40′ Supply Pressure p 100′ Slope – 0.7:1 Tailwater 30′ El. 491.00′ Water h ■ Figure P2.94 A 2.95 The Keswick dam in Problem 2.94 is made of concrete and has a specific weight of 150 lb/ft3. The hydrostatic forces and the weight of the dam produce a total vertical force of the dam on the foundation. Find the magnitude and location of this total vertical force. Consider a unit length of the dam (b = 1 ft). A B ■ Figure P2.89 †2.90 Sometimes it is difficult to open an exterior door of a building because the air distribution system maintains a pressure difference between the inside and outside of the building. Estimate how big this pressure difference can be if it is “not too difficult” for an average person to open the door. 2.91 Find the center of pressure of an elliptical area of minor axis 2a and major axis 2b where axis 2a is vertical and axis 2b is horizontal. The center of the ellipse is a vertical distance h below the surface of the water (h > a). The fluid density is constant. Will the center of pressure of the ellipse change if the fluid is replaced by another constant-density fluid? Will the center of pressure of the ellipse change if the vertical axis is tilted back an angle α from the vertical about its horizontal axis? Explain. 2.92 The dam shown in Fig. P2.92 is 200 ft long and is made of concrete with a specific gravity of 2.2. Find the magnitude and y coordinate of the line of action of the net horizontal force. 2.96 The Keswick dam in Problem 2.94 is made of concrete and has a specific weight of 150 lb/ft3. The coefficient of friction μ between the base of the dam and the foundation is 0.65. Is the dam likely to slide downstream? Consider a unit length of the dam (b = 1 ft). 2.97 Figure P2.97 is a representation of the Altus gravity dam in Oklahoma. Find the magnitudes and locations of the horizontal and vertical hydrostatic force components acting on the headwater wall of the dam and on the tailwater wall of the dam. Note that the slope given is the ratio of the run to the rise. Consider a unit length of the dam (b = 1 ft). 10′ EI. 1564.00′ EI. 1555.00′ EI. 1553.00′ 20′ X Headwater g y 60′ Tailwater Headwater 20′ 40′ ■ Figure P2.92 2.93 Repeat Problem 2.92 but find the magnitude and x coordinate of the line of action of the vertical force on the dam resulting from the water. 87′ Slope = 0.1:1 Slope = 0.6:1 EI. 1475.00 ■ Figure P2.97 Tailwater 4′ 93 Problems 2.98 The Altus dam in Problem 2.97 is made of concrete with a density of 150 lbm/ft3. The coefficient of friction μ between the base of the dam and the foundation is 0.65. Is the dam likely to slide downstream? Consider a unit length of the dam (b = 1 ft). 2.99 Find the magnitude and location of the net horizontal force on the gate shown in Fig. P2.99. The gate width is 5.0 m. with water as shown in Fig. P2.102. The gate is hinged at its bottom and held closed by a horizontal force, FH, located at the center of the gate. The maximum value for FH is 3500 kN. (a) Determine the maximum water depth, h, above the center of the gate that can exist without the gate opening. (b) Is the answer the same if the gate is hinged at the top? Explain your answer. 2.103 A gate having the cross section shown in Fig. P2.103 is 4 ft wide and is hinged at C. The gate weighs 18,000 lb, and its mass center is 1.67 ft to the right of the plane BC. Determine the vertical reaction at A on the gate when the water level is 3 ft above the base. All contact surfaces are smooth. 5 ft 1.0 m 4.0 m Hinge A Water 45° B 2.0 m Water C 3.0 m 9 ft 2.0 m Water surface ■ Figure P2.99 3 ft B *2.100 Find the magnitude and location of the net vertical force on the gate in Problem 2.99. A 2.101 Find the total vertical force on the cylinder shown in Fig. P2.101. ■ Figure P2.103 patm 2.104 The massless, 4-ft-wide gate shown in Fig. P2.104 pivots about the frictionless hinge O. It is held in place by the 2000 lb counterweight, 𝒲. Determine the water depth, h. D = 6 cm Water Gate 18 cm h 3 cm Pivot O 5 cm d= Width = 4 ft 3 cm 2 ft 3 ft Water 𝒲 ■ Figure P2.101 ■ Figure P2.104 2.102 A 3-m-wide, 8-m-high rectangular gate is located at the end of a rectangular passage that is connected to a large open tank filled *2.105 A 200-lb homogeneous gate 10 ft wide and 5 ft long is hinged at point A and held in place by a 12-ft-long brace as shown in Fig. P2.105. As the bottom of the brace is moved to the right, the water level remains at the top of the gate. The line of action of the force that the brace exerts on the gate is along the brace. (a) Plot h 4m FH 4m Hinge ■ Figure P2.102 5 ft Gate Water A θ ■ Figure P2.105 12 ft Brace Moveable stop 94 Chapter 2 ■ Fluid Statics the magnitude of the force exerted on the gate by the brace as a function of the angle of the gate, θ, for 0 ≤ θ ≤ 90°. (b) Repeat the calculations for the case in which the weight of the gate is negligible. Comment on the result as θ → 0. 2.106 An open tank has a vertical partition and on one side contains gasoline with a density ρ = 700 kg/m3 at a depth of 4 m, as shown in Fig. P2.106. A rectangular gate that is 4 m high and 2 m wide and hinged at one end is located in the partition. Water is slowly added to the empty side of the tank. At what depth, h, will the gate start to open? drain pipe that contains air at atmospheric pressure. Determine the minimum water depth, h, at which the gate will pivot to allow water to flow into the pipe. 2.110 The closed vessel of Fig. P2.110 contains water with an air pressure of 10 psi at the water surface. One side of the vessel contains a spout that is closed by a 6-in.-diameter circular gate that is hinged along one side as illustrated. The horizontal axis of the hinge is located 10 ft below the water surface. Determine the minimum torque that must be applied at the hinge to hold the gate shut. Neglect the weight of the gate and friction at the hinge. Partition 10 psi Stop Air 10 ft 4m h Water Gasoline Water Hinge Axis 6-in.-diameter gate 3 4 ■ Figure P2.106 2.107 A horizontal 2-m-diameter conduit is half filled with a liquid (SG = 1.6) and is capped at both ends with plane vertical surfaces. The air pressure in the conduit above the liquid surface is 200 kPa. Determine the resultant force of the fluid acting on one of the end caps, and locate this force relative to the bottom of the conduit. 2.108 A 4-ft by 3-ft massless rectangular gate is used to close the end of the water tank shown in Fig. P2.108. A 200-lb weight attached to the arm of the gate at a distance ℓ from the frictionless hinge is just sufficient to keep the gate closed when the water depth is 2 ft, that is, when the water fills the semicircular lower portion of the tank. If the water were deeper, the gate would open. Determine the distance ℓ. ℓ Hinge Hinge 1 ft 200 lb Gate Water 2 ft radius 3 ft 4 ft ■ Figure P2.108 2.109 A thin 4-ft-wide, right-angle gate with negligible mass is free to pivot about a frictionless hinge at point O, as shown in Fig. P2.109. The horizontal portion of the gate covers a 1-ft-diameter Width = 4 ft ■ Figure P2.110 2.111 (See The Wide World of Fluids article titled “The Three Gorges Dam,” Section 2.8.) (a) Determine the horizontal hydrostatic force on the 2309-m-long Three Gorges Dam when the average depth of the water against it is 175 m. (b) If all of the 6.4 billion people on Earth were to push horizontally against the Three Gorges Dam, could they generate enough force to hold it in place? Support your answer with appropriate calculations. Section 2.10 Hydrostatic Force on a Curved Surface 2.112 Obtain a photograph/image of a situation in which the hydrostatic force on a curved surface is important. Print this photo and write a brief paragraph that describes the situation involved. 2.113 A 2-ft-diameter hemispherical plexiglass “bubble” is to be used as a special window on the side of an above-ground swimming pool. The window is to be bolted onto the vertical wall of the pool and faces outward, covering a 2-ft-diameter opening in the wall. The center of the opening is 4 ft below the surface. Determine the horizontal and vertical components of the force of the water on the hemisphere. 2.114 Consider the curved surface shown in Fig. P2.114(a) and (b). The two curved surfaces are identical. How are the vertical forces on the two surfaces alike? How are they different? Oil Oil Right-angle gate h Water R R Hinge O 1-ft-diameter pipe 3 ft ■ Figure P2.109 (a) (b) ■ Figure P2.114 2.115 Figure P2.115 shows a cross section of a submerged tunnel used by automobiles to travel under a river. Find the magnitude and Problems 95 1-in. diameter location of the resultant hydrostatic force on the circular roof of the tunnel. The tunnel is 4 mi long. pair = 40 psi 12 in. 10 in. g 50′ 4.3-in. diameter ■ Figure P2.118 R = 20′ ■ Figure P2.115 2.116 The container shown in Fig. P2.116 has circular cross sections. Find the vertical force on the inclined surface. Also find the net vertical force on the bottom, EF. Is the vertical force equal to the weight of the water in the container? 2′ B 8.0 in. diameter 2′ C A D Water E 2.119 In drilling for oil in the Gulf of Mexico, some divers have to work at a depth of 1300 ft. (a) Assume that seawater has a constant density of 64 lb/ft3 and compute the pressure at this depth. The divers breathe a mixture of helium and oxygen stored in cylinders, as shown in Fig. P2.119, at a pressure of 3000 psia. (b) Calculate the force, which trends to blow the end cap off, that the weld must resist while the diver is using the cylinder at 1300 ft. (c) After emptying a tank, a diver releases it. Will the tank rise or fall, and what is its initial acceleration? Weld 1′ 30 in. 4′ Spherical cap Steel p = 489 lbm/ft3 F 1.0 in. thickness ■ Figure P2.119 4′ ■ Figure P2.116 2.117 The 18-ft-long lightweight gate of Fig. P2.117 is a quarter circle and is hinged at H. Determine the horizontal force, P, required to hold the gate in place. Neglect friction at the hinge and the weight of the gate. 2.120 Hoover Dam (see Video 2.5) is the highest arch-gravity type of dam in the United States. A cross section of the dam is shown in Fig. P2.120(a). The walls of the canyon in which the dam is located are sloped, and just upstream of the dam the vertical plane shown in Figure P2.120(b) approximately represents the cross section of the water acting on the dam. Use this vertical cross section to estimate the resultant horizontal force of the water on the dam, and show where this force acts. P 45 ft 6 ft 880 ft Water Gate H Hinge 727 ft 715 ft. ■ Figure P2.117 2.118 The air pressure in the top of the 2-liter pop bottle shown in Video V2.6 and Fig. P2.118 is 40 psi, and the pop depth is 10 in. The bottom of the bottle has an irregular shape with a diameter of 4.3 in. (a) If the bottle cap has a diameter of 1 in. what is the magnitude of the axial force required to hold the cap in place? (b) Determine the force needed to secure the bottom 2 in. of the bottle to its cylindrical sides. For this calculation assume the effect of the weight of the pop is negligible. (c) By how much does the weight of the pop increase the pressure 2 in. above the bottom? Assume the pop has the same specific weight as that of water. 660 ft (a) 290 ft (b) ■ Figure P2.120 2.121 A plug in the bottom of a pressurized tank is conical in shape, as shown in Fig. P2.121. The air pressure is 40 kPa, and the liquid in the tank has a specific weight of 27 kN/m3. Determine the magnitude, direction, and line of action of the force exerted on 96 Chapter 2 ■ Fluid Statics the curved surface of the cone within the tank due to the 40-kPa pressure and the liquid. 40 kPa 2.124 A step-in viewing window having the shape of a half-cylinder is built into the side of a large aquarium. See Fig. P2.124. Find the magnitude, direction, and location of the net horizontal forces on the viewing window. Air x z Liquid 3m y 25′ 1m p = 64 lbm/ft3 60° R = 5′ ■ Figure P2.121 10′ 2.122 The homogeneous gate shown in Fig. P2.122 consists of one quarter of a circular cylinder and is used to maintain a water depth of 4 m. That is, when the water depth exceeds 4 m, the gate opens slightly and lets the water flow under it. Determine the weight of the gate per meter of length. 2′ Steps up to window ■ Figure P2.124 2.125 Find the magnitude, direction, and location of the net vertical force acting on the viewing window in Problem 2.124. Pivot 4m 2.126 A l0-m-long log is stuck against a dam, as shown in Fig. P2.126. Find the magnitudes and locations of both the horizontal force and the vertical force of the water on the log in terms of the diameter D. The center of the log is at the same elevation as the top of the dam. 1m D Water ■ Figure P2.122 2.123 The concrete (specific weight = 150 lb/ft3) seawall of Fig. P2.123 has a curved surface and restrains seawater at a depth of 24 ft. The trace of the surface is a parabola as illustrated. Determine the moment of the fluid force (per unit length) with respect to an axis through the toe (point A). ■ Figure P2.126 2.127 Find the net horizontal force on the 4.0-m-long log shown in Fig. P2.127. 1.0 m Seawater D = 1.0 m 0.5 m y = 0.2x2 24 ft Figure P2.127 y A 15 ft ■ Figure P2.123 x 2.128 An open tank containing water has a bulge in its vertical side that is semicircular in shape as shown in Fig. P2.128. Determine the horizontal and vertical components of the force that the water exerts on the bulge. Base your analysis on a 1-ft length of the bulge. Problems h 2 Hinge 6 ft h h Block Bulge Water 97 3 ft (a) (b) h 2 h ■ Figure P2.128 2.129 A closed tank is filled with water and has a 4-ft-diameter hemispherical dome as shown in Fig. 2.129. A U-tube manometer is connected to the tank. Determine the vertical force of the water on the dome if the differential manometer reading is 7 ft and the air pressure at the upper end of the manometer is 12.6 psi. pA Air 4-ft diameter 20 psi 5 ft ■ Figure P2.131 Section 2.11 Buoyancy, Flotation, and Stability 2.132 Obtain a photograph/image of a situation in which Archimedes’s principle is important. Print this photo and write a brief paragraph that describes the situation involved. 2.133 An iceberg (specific gravity 0.917) floats in the ocean (specific gravity 1.025). What percent of the volume of the iceberg is under water? 2 ft Water 2 ft (c) Gage fluid (SG = 3.0) 2.134 A floating 40-in.-thick piece of ice sinks 1 in. with a 500-lb polar bear in the center of the ice. What is the area of the ice in the plane of the water level? For seawater, S = 1.03. ■ Figure P2.129 2.135 A spherical balloon filled with helium at 40 °F and 20 psia has a 25-ft diameter. What load can it support in atmospheric air at 40 °F and 14.696 psia? Neglect the balloon’s weight. 2.130 A 3-m-diameter open cylindrical tank contains water and has a hemispherical bottom as shown in Fig. P2.130. Determine the magnitude, line of action, and direction of the force of the water on the curved bottom. 2.136 A river barge, whose cross section is approximately rectangular, carries a load of grain. The barge is 28 ft wide and 90 ft long. When unloaded, its draft (depth of submergence) is 5 ft. and with the load of grain the draft is 7 ft. Determine: (a) the unloaded weight of the barge, and (b) the weight of the grain. 2.137 A barge is 40 ft wide by 120 ft long. The weight of the barge and its cargo is denoted by W. When in salt-free riverwater, it floats 0.25 ft deeper than when in seawater (γ = 64 lb/ft3 ). Find the weight W. 8m Water 3m ■ Figure P2.130 2.138 When the Tucurui Dam was constructed in northern Brazil, the lake that was created covered a large forest of valuable hardwood trees. It was found that even after 15 years underwater the trees were perfectly preserved and underwater logging was started. During the logging process a tree is selected, trimmed, and anchored with ropes to prevent it from shooting to the surface like a missile when cut. Assume that a typical large tree can be approximated as a truncated cone with a base diameter of 8 ft, a top diameter of 2 ft, and a height of 100 ft. Determine the resultant vertical force that the ropes must resist when the completely submerged tree is cut. The specific gravity of the wood is approximately 0.6. †2.139 Estimate the minimum water depth needed to float a canoe carrying two people and their camping gear. List all assumptions 2.131 Three gates of negligible weight are used to hold back water and show all calculations. in a channel of width b as shown in Fig. P2.131. The force of the An inverted test tube partially filled with air floats in a gate against the block for gate (b) is R. Determine (in terms of R) 2.140 plastic water-filled soft drink bottle as shown in Video V2.7 and the force against the blocks for the other two gates. 98 Chapter 2 ■ Fluid Statics 2′ Fig. P2.140. The amount of air in the tube has been adjusted so that it just floats. The bottle cap is securely fastened. A slight squeezing of the plastic bottle will cause the test tube to sink to the bottom of the bottle. Explain this phenomenon. 16′ Air d Test tube Lead Water 0.5′ ■ Figure P2.144 Plastic bottle ■ Figure P2.140 2.141 A child’s balloon is a sphere 1 ft. in diameter. The balloon is filled with helium (ρ = 0.014 lbm/ft3). The balloon material weighs 0.008 lbf/ft2 of surface area. If the child releases the balloon, how high will it rise in the Standard Atmosphere. (Neglect expansion of the balloon as it rises.) 2.142 A 1-ft-diameter, 2-ft-long cylinder floats in an open tank containing a liquid having a specific weight γ. A U-tube manometer is connected to the tank as shown in Fig. P2.142. When the pressure in pipe A is 0.1 psi below atmospheric pressure, the various fluid levels are as shown. Determine the weight of the cylinder. Note that the top of the cylinder is flush with the fluid surface. 2.145 When a hydrometer (see Fig. P2.145 and Video V2.8) having a stem diameter of 0.30 in. is placed in water, the stem protrudes 3.15 in. above the water surface. If the water is replaced with a liquid having a specific gravity of 1.10, how much of the stem would protrude above the liquid surface? The hydrometer weighs 0.042 lb. Hydrometer Fluid surface pa = – 0.1 psi 1 ft A 2 ft Cylinder Sp. wt. = γ 1 ft 2 ft Water Gage fluid SG = 1.5 0.5 ft ■ Figure P2.145 2.146 A 2-ft-thick block constructed of wood (SG = 0.6) is submerged in oil (SG = 0.8) and has a 2-ft-thick aluminum (specific weight = 168 lb/ft3) plate attached to the bottom as indicated in Fig. P2.146. Determine completely the force required to hold the block in the position shown. Locate the force with respect to point A. ■ Figure P2.142 6 ft *2.143 A not-too-honest citizen is thinking of making bogus gold bars by first making a hollow iridium (S = 22.5) ingot and plating it with a thin layer of gold (S = 19.3) of negligible weight and volume. The bogus bar is to have a mass of 100 lbm. What must be the volumes of the bogus bar and of the air space inside the iridium so that an inspector would conclude it was real gold after weighing it in air and water to determine its density? Could lead (S = 11.35) or platinum (S = 21.45) be used instead of iridium? Would either be a good idea? 2.144 A solid cylindrical pine (S = 0.50) spar buoy has a cylindrical lead (S = 11.3) weight attached, as shown in Fig. P2.144. Determine the equilibrium position of the spar buoy in seawater (i.e., find d). Is this spar buoy stable or unstable? For seawater, S = 1.03. Oil 4 ft 0.5 ft A Aluminum 10 ft ■ Figure P2.146 2.147 (See The Wide World of Fluids article titled “Concrete canoes,” Section 2.11.1.) How much extra water does a 147-lb concrete canoe displace compared to an ultralightweight 38-lb Kevlar canoe of the same size carrying the same load? 99 Problems *2.148 A submarine is modeled as a cylinder with a length of 300 ft, a diameter of 50 ft, and a conning tower as shown in Fig. P2.148. The submarine can dive a distance of 50 ft from the floating position in about 30 sec. Diving is accomplished by taking water into the ballast tank so the submarine will sink. When the submarine reaches the desired depth, some of the water in the ballast tank is discharged leaving the submarine in “neutral buoyancy” (i.e., it will neither rise nor sink). For the conditions illustrated, find (a) the weight of the submarine and (b) the volume (or mass) of the water that must be in the ballast tank when the submarine is in neutral buoyancy. For seawater, S = 1.03. 3% of cylinder volume 7% of cylinder volume pwater = 64 lbm/ft3 3′ 25′ 50′ Sfluid = 1.0 g Srod = 2.0 String 8.0 cm acyl = 9.8 m/s2 ■ Figure P2.154 2.155 A closed cylindrical tank that is 8 ft in diameter and 24 ft long is completely filled with gasoline. The tank, with its long axis horizontal, is pulled by a truck along a horizontal surface. Determine the pressure difference between the ends (along the long axis of the tank) when the truck undergoes an acceleration of 5 ft/s2. 2.156 The cart shown in Fig. P2.156 measures 10.0 cm long and 6.0 cm high and has rectangular cross sections. It is half-filled with water and accelerates down a 20° incline plane at a = 1.0 m/s2. Find the height h. Ballast tank ℓ = 10.0 cm y Partially submerged position 1.0 cm g Water Totally submerged position h z a ■ Figure P2.148 20° Section 2.12 Pressure Variation in a Fluid with Rigid-Body Motion 2.149 Obtain a photograph/image of a situation in which the pressure variation in a fluid with rigid-body motion is involved. Print this photo and write a brief paragraph that describes the situation involved. 2.150 When an automobile brakes, the fuel gage indicates a fuller tank than when the automobile is traveling at a constant speed on a level road. Is the sensor for the fuel gage located near the front or rear of the fuel tank? Assume a constant deceleration. 2.151 An open container of oil rests on the flatbed of a truck that is traveling along a horizontal road at 55 mi/hr. As the truck slows uniformly to a complete stop in 5 s, what will be the slope of the oil surface during the period of constant deceleration? 2.152 A 5-gal, cylindrical open container with a bottom area of 120 in.2 is filled with glycerin and rests on the floor of an elevator. (a) Determine the fluid pressure at the bottom of the container when the elevator has an upward acceleration of 3 ft/s2. (b) What resultant force does the container exert on the floor of the elevator during this acceleration? The weight of the container is negligible. (Note: 1 gal = 231 in.3) x x ■ Figure P2.156 2.157 The U-tube manometer in Fig. P2.157 is used to measure the acceleration of the cart on which it sits. Develop an expression for the acceleration of the cart in terms of the liquid height h, the liquid density ρ, the local acceleration of gravity g, and the length ℓ. ℓ p h g a ■ Figure P2.157 2.153 A plastic glass has a square cross section measuring 2½ in. *2.158 A tank has a height of 5.0 cm and a square cross section on a side and is filled to within ½ in. of the top with water. The measuring 5.0 cm on a side. The tank is one third full of water and glass is placed in a level spot in a car with two opposite sides is rotated in a horizontal plane with the bottom of the tank 100 cm parallel to the direction of travel. How fast can the driver of the from the center of rotation and two opposite sides parallel to the car accelerate along a level road without spilling any of the water? ground. What is the maximum rotational speed that the tank of water can be rotated with no water coming out of the tank? 2.154 The cylinder in Fig. P2.154 accelerates to the left at the rate 2 of 9.80 m/s . Find the tension in the string connecting at rod of 2.159 An open 1-m-diameter tank contains water at a depth of circular cross section to the cylinder. The volume between the rod 0.7 m when at rest. As the tank is rotated about its vertical axis the and the cylinder is completely filled with water at 10 °C. center of the fluid surface is depressed. At what angular velocity 100 Chapter 2 ■ Fluid Statics will the bottom of the tank first be exposed? No water is spilled from the tank. *2.160 The U-tube in Fig. P2.160 rotates at 2.0 rev/sec. Find the absolute pressures at points C and B if the atmospheric pressure is 14.696 psia. Recall that 70 °F water evaporates at an absolute pressure of 0.363 psia. Determine the absolute pressures at points C and B if the U-tube rotates at 2.0 rev/sec. ω A C B 1′ 2.5′ 70°F water ■ Figure P2.160 2.161 A child riding in a car holds a string attached to a floating, helium-filled balloon. As the car decelerates to a stop, the balloon tilts backwards. As the car makes a right-hand turn, the balloon tilts to the right. On the other hand, the child tends to be forced forward as the car decelerates and to the left as the car makes a right-hand turn. Explain these observed effects on the balloon and child. 2.162 A closed, 0.4-m-diameter cylindrical tank is completely filled with oil (SG = 0.9) and rotates about its vertical longitudinal axis with an angular velocity of 40 rad/s. Determine the difference in pressure just under the vessel cover between a point on the circumference and a point on the axis. 2.163 (See The Wide World of Fluids article titled “Rotating mercury mirror telescope,” Section 2.12.2.) The largest liquid mirror telescope uses a 6-ft-diameter tank of mercury rotating at 7 rpm to produce its parabolic-shaped mirror as shown in Fig. P2.163. Determine the difference in elevation of the mercury, Δh, between the edge and the center of the mirror. Receiver Light rays 6 ft Δh Mercury ω = 7 rpm ■ Figure P2.163 ■ Lifelong Learning Problems 2.1 LL Although it is relatively easy to calculate the net hydrostatic pressure force on a dam, it is not necessarily easy to design and construct an appropriate, long-lasting, inexpensive dam. In fact, inspection of older dams has revealed that many of them are in peril of collapse unless corrective action is soon taken. Obtain information about the severity of the poor conditions of older dams throughout the country. Summarize your findings in a brief report. 2.2 LL Over the years the demand for high-quality, first-growth timber has increased dramatically. Unfortunately, most of the trees that supply such lumber have already been harvested. Recently, however, several companies have started to reclaim the numerous high-quality logs that sank in lakes and oceans during the logging boom times many years ago. Many of these logs are still in excellent condition. Obtain information about the use of fluid mechanics concepts in harvesting sunken logs. Summarize your findings in a brief report. 2.3 LL Liquid-filled manometers and Bourdon tube pressure gages have been the mainstay for measuring pressure for many, many years. However, for many modern applications these triedand-true devices are not sufficient. For example, various new uses need small, accurate, inexpensive pressure transducers with digital outputs. Obtain information about some of the new concepts used for pressure measurement. Summarize your findings in a brief report. 3 Elementary Fluid Dynamics— The Bernoulli Equation Learning Objectives After completing this chapter, you should be able to: ■ discuss the application of Newton’s second law to fluid flows. ■ explain the development, uses, and limitations of the Bernoulli equation. ■ use the Bernoulli equation (stand-alone or in combination with the simple continuity equation) to solve flow problems. ■ apply the concepts of static, stagnation, dynamic, and total pressures. ■ calculate various flow properties using the energy and hydraulic grade lines. The Bernoulli equation may be the most used and abused equation in fluid mechanics. 3.1 In this chapter we investigate some typical fluid motions (fluid dynamics) in an elementary way. We will discuss in some detail the use of Newton’s second law (F = ma) as it is applied to fluid particle motion that is “ideal” in some sense. We will obtain the celebrated Bernoulli equation and apply it to various flows. Although this equation is one of the oldest in fluid mechanics and the assumptions involved in its derivation are numerous, it can be used effectively to predict, understand, and analyze a variety of flow situations. However, if the equation is applied without proper respect for its restrictions, serious errors can arise. Indeed, the Bernoulli equation is appropriately called the most used and the most abused equation in fluid mechanics. A thorough understanding of the elementary approach to fluid dynamics involved in this chapter will be useful on its own. It also provides a good foundation for the material in the following chapters where some of the present restrictions are removed and “more nearly exact” results are presented. Newton’s Second Law Consider a really tiny volume of fluid, which is still large enough to contain a significant number of molecules. This volume is called a fluid particle (with further description in Section 4.1). As the fluid particle moves from one location to another, it usually experiences an acceleration or deceleration. According to Newton’s second law of motion, the net force acting on the fluid particle under consideration must equal its mass times its acceleration, F = ma In this chapter we consider fluid motion while neglecting frictional effects. In a flowing fluid, friction is associated with shear stress, and shear stress is associated with viscosity. The simplest way 101 102 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation to assure a frictionless flow is to imagine that the fluid has no viscosity. We call this imaginary substance an inviscid fluid and the resulting flow an inviscid flow. In practice there are no inviscid fluids, since every fluid supports shear stresses when it is subjected to a rate of strain displacement. For many flow situations the viscous effects are relatively small compared with other effects. As a first approximation for such cases it is often possible to ignore viscous effects. For example, often the viscous forces developed in flowing water may be several orders of magnitude smaller than forces due to other influences, such as gravity or pressure differences. For other water flow situations, however, the viscous effects may be the dominant ones. Similarly, the viscous effects associated with the flow of a gas are often negligible, although in some circumstances they are very important. With the assumption of inviscid (or frictionless) flow, the fluid motion is governed by pressure and gravity forces only, and Newton’s second law as it applies to a fluid particle is Inviscid fluid flow is governed by pressure and gravity forces. (Net pressure force on particle) + (net gravity force on particle) = (particle mass) × (particle acceleration) z x y Rectangular z θ r Cylindrical The results of the interaction between the pressure, gravity, and acceleration provide numerous useful applications in fluid mechanics. To apply Newton’s second law to a fluid (or any other object), we must define an appropriate coordinate system in which to describe the motion. In general the motion will be three-dimensional and unsteady so that three space coordinates and time are needed to describe it. There are numerous coordinate systems available, including the most often used rectangular (x, y, z) and cylindrical (r, θ, z) systems shown by the figures in the margin. Usually the specific flow geometry dictates which system would be most appropriate. In this chapter we will be concerned with two-dimensional motion like that confined to the x–z plane as is shown in Fig. 3.1a. Clearly we could choose to describe the flow in terms of the components of acceleration and forces in the x and z coordinate directions. The resulting equations are frequently referred to as a two-dimensional form of the Euler equations of motion in rectangular Cartesian coordinates. This approach will be discussed in Chapter 6. As is done in the study of dynamics (Ref. 1), the motion of each fluid particle is described in terms of its velocity vector, V, which is defined as the time rate of change of the position of the particle. The particle’s velocity is a vector quantity with a magnitude (the speed, V = V) and direction. As the particle moves about, it follows a particular path, the shape of which is governed by the velocity of the particle. The location of the particle along the path is a function of where the particle started at the initial time and its velocity along the path. If it is steady flow (i.e., nothing changes with time at a given location in the flow field), each successive particle that passes through a given point [such as point (1) in Fig. 3.1a] will follow the same path. For such cases the path is a fixed line in the x–z plane. Neighboring particles that pass on either side of point (1) follow their own paths, which may be of a different shape than the one passing through (1). The entire x–z plane is filled with such paths. For steady flows each particle slides along its path, and its velocity vector is everywhere tangent to the path. The lines that are tangent to the velocity vectors throughout the flow field are called streamlines. For many situations it is easiest to describe the flow in terms of the “streamline” coordinates based on the streamlines as are illustrated in Fig. 3.1b. The particle motion is described in terms of its distance, s = s(t), along the streamline from some convenient origin and the local radius of curvature of the streamline, ℛ = ℛ(s). The distance along the streamline is related to z z Fluid particle V (2) s (1) ℛ = ℛ(s) n = n1 n n=0 V Streamlines (a) x (b) ■ Figure 3.1 (a) Flow in the x–z plane. (b) Flow in terms of streamline and normal coordinates. x 3.1 Fluid particles accelerate normal to and along streamlines. VIDEO V3.1 Streamlines past an airfoil as = an = 0 as > 0 as < 0 an > 0 as > 0, an > 0 Newton’s Second Law 103 the particle’s speed by V = dsdt, and the radius of curvature is related to the shape of the streamline. In addition to the coordinate along the streamline, s, the coordinate normal to the streamline, n, as is shown in Fig. 3.1b, will be of use. To apply Newton’s second law to a particle flowing along its streamline, we must write the particle acceleration in terms of the streamline coordinates. By definition, the acceleration is the time rate of change of the velocity of the particle, a = dVdt. For two-dimensional flow in the x–z plane, the acceleration has two components—one along the streamline, as, the streamwise acceleration, and one normal to the streamline, an, the normal acceleration. Streamwise acceleration results from two facts. First, the speed of the particle may change with time at a point, V(t). Since we are limiting ourselves to steady flow, we will ignore velocity changes with time. (More information is in Section 3.8.2.) Second, the speed of the particle generally varies along the streamline, V(s). For example, in Fig. 3.1a the speed may be 50 fts at point (1) and 100 fts at point (2). Thus, by use of the chain rule of differentiation, the s component of the acceleration is given by as = dVdt = ( ∂V∂s)(dsdt) = (∂V∂s)V. We have used the fact that speed is the time rate of change of distance, V = dsdt. Note that the streamwise acceleration is the product of the rate of change of speed with distance along the streamline, ∂V∂s, and the speed, V. Since ∂V∂s can be positive, negative, or zero, the streamwise acceleration can, therefore, be positive (acceleration), negative (deceleration), or zero (constant speed). The normal component of acceleration, the centrifugal acceleration, is given in terms of the particle speed and the radius of curvature of its path. Thus, an = V 2 ℛ, where both V and ℛ may vary along the streamline. These equations for the acceleration should be familiar from the study of particle motion in physics (Ref. 2) or dynamics (Ref. 1). A more complete derivation and discussion of these topics can be found in Chapter 4. Thus, the (steady flow) components of acceleration in the s and n directions, as and an, are given by as = V ∂V V2 ,  an = ∂s ℛ (3.1) where ℛ is the local radius of curvature of the streamline, and s is the distance measured along the streamline from some arbitrary initial point. In general there is acceleration along the streamline (because the particle speed changes along its path, ∂V∂s ≠ 0) and acceleration normal to the streamline (because the particle does not flow in a straight line, ℛ ≠ ∞). Various flows and the accelerations associated with them are shown in the figure in the margin. As discussed in Section 3.6.2, for incompressible flow the velocity is inversely proportional to the streamline spacing. Hence, converging streamlines produce positive streamwise acceleration. To produce this acceleration there must be a net, nonzero force on the fluid particle. To determine the forces necessary to produce a given flow (or conversely, what flow results from a given set of forces), we consider the free-body diagram of a small fluid particle as is shown in Fig. 3.2. The particle of interest is removed from its surroundings, and the reactions of the surroundings on the particle are indicated by the appropriate forces present, F1, F2, and so forth. For the present case, the important forces are assumed to be gravity and pressure. Other forces, such as viscous forces and surface tension effects, are assumed negligible. The acceleration of gravity, g, is assumed to be constant and acts vertically, in the negative z direction, at an angle θ relative to the normal to the streamline. z Fluid particle F4 F5 θ Streamline F1 F3 g F2 x ■ Figure 3.2 Isolation of a small fluid particle in a flow field. (Photo courtesy of Diana Sailplanes.) 104 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 3.2 F = ma along a Streamline Consider the small fluid particle of size δs by δn in the plane of the figure and δy normal to the figure as shown in the free-body diagram of Fig. 3.3. Unit vectors along and normal to the streamline are denoted by sˆ and nˆ, respectively. For steady flow, the component of Newton’s second law along the streamline direction, s, can be written as ∑ δFs = δm as = δm V ∂V ∂V = ρ δV V ∂s ∂s (3.2) where ∑ δFs represents the sum of the s components of all the forces acting on the particle, which has mass δm = ρ δV, and V ∂V∂s is the acceleration in the s direction. Here, δV = δs δn δy is the particle volume. Equation 3.2 is valid for both compressible and incompressible fluids. That is, the density need not be constant throughout the flow field. The gravity force (weight) on the particle can be written as δ 𝒲 = γ δV, where γ = ρg is the specific weight of the fluid (lbft3 or Nm3 ). Hence, the component of the weight force in the direction of the streamline is δ 𝒲s = −δ 𝒲 sin θ = −γ δV sin θ In a flowing fluid the pressure varies from one location to another. If the streamline is horizontal at the point of interest, then θ = 0, and there is no component of particle weight along the streamline to contribute to its acceleration in that direction. As is indicated in Chapter 2, the pressure is not constant throughout a stationary fluid (∇p ≠ 0) because of the fluid weight. Likewise, in a flowing fluid the pressure is usually not constant. In general, for steady flow, p = p(s, n). If the pressure at the center of the particle shown in Fig. 3.3 is denoted as p, then the pressure at either end may be higher or lower by some amount δps. Thus the pressures on the two end faces that are perpendicular to the streamline are p + δps and p − δps. Since the particle is “small,” we can use a one-term Taylor series expansion for the pressure field (as was done in Chapter 2 for the pressure forces in static fluids) to obtain δps ≈ The net pressure force on a particle is determined by the pressure gradient. ∂p δs ∂s 2 Thus, if δFps is the net pressure force on the particle in the streamline direction, it follows that δFps = ( p − δps ) δn δy − ( p + δps ) δn δy = −2 δps δn δy =− ∂p ∂p δs δn δy = − δV ∂s ∂s g ℛ (p + δ pn) δs δ y τ δs δ y = 0 Particle thickness = δ y θ n δs (p + δ ps) δ n δ y s δn δ 𝒲n δ𝒲 θ δ 𝒲s (p – δ ps)δ n δ y δs θ τ δs δ y = 0 θ δz δz Along streamline δn (p – δ pn) δ s δ y Normal to streamline ■ Figure 3.3 Free-body diagram of a fluid particle for which the important forces are those due to pressure and gravity. 3.2 F = ma along a Streamline 105 Note that the actual level of the pressure, p, is not important. What produces a net pressure force is the fact that the pressure is not constant throughout the fluid. The nonzero pressure gradient, ∇p = ∂p∂s sˆ + ∂p∂n nˆ, is what provides a net pressure force on the particle. Viscous forces, represented by τ δs δy, are zero, since the fluid is inviscid. Thus, the net force acting in the streamline direction on the particle shown in Fig. 3.3 is given by ∑ δFs = δ 𝒲s + δFps = (−γ sin θ − ∂p δV ∂s ) (3.3) By combining Eqs. 3.2 and 3.3, we obtain the following equation of motion for steady inviscid flow along the streamline direction: −γ sin θ − ∂p ∂V = ρV = ρas ∂s ∂s (3.4) We have divided out the common particle volume factor, δV, that appears in both the force and the acceleration portions of the equation. This is a representation of the fact that it is the fluid density (mass per unit volume), not the mass, per se, of the fluid particle that is important. The physical interpretation of Eq. 3.4 is that a change in fluid particle speed is accomplished by the appropriate combination of pressure gradient and particle weight along the streamline. For fluid static situations this balance between pressure and gravity forces is such that no change in particle speed is produced—the right-hand side of Eq. 3.4 is zero, and the particle remains stationary. In a flowing fluid the pressure and weight forces do not necessarily balance—the force unbalance provides the appropriate acceleration and, hence, particle motion. E X AMPLE 3.1 Pressure Variation along a Streamline GIVEN Consider the inviscid, incompressible, steady flow along the horizontal streamline A–B in front of the soccer ball (sphere) of radius a, as shown in Fig. E3.1a. From a more advanced theory of flow past a sphere, the fluid velocity along this streamline is a3 V = V0 1 + 3 ( x ) FIND Determine the pressure variation along the streamline from point A far in front of the sphere (xA = −∞ and VA = V0 ) to point B on the sphere (xB = −a and VB = 0). 1 Vo as shown in Fig. E3.1b. 0.75 Vo z V VA = VO ˆi V = V ˆi VB = 0 B A 0.25 Vo x a 0.5 Vo 0 –3a –2a (a) –2a –a (c) –1a 0 (b) ∂__ p ∂x –3a x p 0.610 ρV02/a 0 x 0.5ρV02 –3a –2a –a 0 (d) ■ Figure E3.1 106 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation SOLUTION Since the flow is steady and inviscid, Eq. 3.4 is valid. In addition, since the streamline is horizontal, sin θ = sin 0° = 0 and the equation of motion along the streamline reduces to ∂p ∂V = −ρV ∂s ∂s (1) With the given velocity variation along the streamline, the acceleration term is V 3V0 a3 ∂V ∂V a3 =V = V0 1 + 3 − 4 ( )( ∂s ∂x x x ) This variation is indicated in Fig. E3.1c. It is seen that the pressure increases in the direction of flow ( ∂p∂x > 0) from point A to point B. The maximum pressure gradient (0.610 ρV 20 a) occurs just slightly ahead of the sphere (x = −1.205a). It is the pressure gradient that slows the fluid down from VA = V0 to VB = 0 as shown in Fig. E3.1b. The pressure distribution along the streamline can be obtained by integrating Eq. 2 from p = 0 (gage) at x = −∞ to pressure p at location x. The result, plotted in Fig. E3.1d, is a 3 (ax)6 p = −ρV 20 [( x ) + 2 ] a3 a3 = −3V 20 (1 + 3 ) 4 x x where we have replaced s by x since the two coordinates are identical (within an additive constant) along streamline A–B. It follows that V ∂V∂s < 0 along the streamline. The fluid slows down from V0 far ahead of the sphere to zero velocity on the “nose” of the sphere (x = −a). Thus, according to Eq. 1, to produce the given motion the pressure gradient along the streamline is ∂p 3ρa3V 20 (1 + a3x3 ) = ∂x x4 (Ans) COMMENT The pressure at B, a stagnation point since VB = 0, is the highest pressure along the streamline ( pB = ρV 20 2). As shown in Chapter 9, this excess pressure on the front of the sphere (i.e., pB > 0) contributes to the net drag force on the sphere. Note that the pressure gradient and pressure are directly proportional to the density of the fluid, a representation of the fact that the fluid inertia is proportional to its mass. (2) THE WIDE WORLD OF FLUIDS Incorrect raindrop shape The incorrect representation that drop size) wins over the increased pressure, ρV 20/2, caused by raindrops are teardrop shaped is found nearly everywhere— the motion of the drop and exerted on its bottom. With increas- from children’s books to weather maps on the Weather Channel. ing size, the drops fall faster and the increased pressure About the only time raindrops possess the typical teardrop causes the drops to flatten. A 2-mm drop, for example, is flat- shape is when they run down a windowpane. The actual shape tened into a hamburger bun shape. Slightly larger drops are of a falling raindrop is a function of the size of the drop and actually concave on the bottom. When the radius is greater results from a balance between surface tension forces and the than about 4 mm, the depression of the bottom increases and air pressure exerted on the falling drop. Small drops with a the drop takes on the form of an inverted bag with an annular radius less than about 0.5 mm have a spherical shape because ring of water around its base. This ring finally breaks up into the surface tension effect (which is inversely proportional to smaller drops. Streamline n = constant n s p = p(s) Equation 3.4 can be rearranged and integrated as follows. First, we note from Fig. 3.3 that along 1 the streamline sin θ = dzds. Also we can write V dVds = 2d(V 2 )ds. Finally, along the streamline the value of n is constant (dn = 0) so that dp = (∂p∂s) ds + (∂p∂n) dn = (∂p∂s) ds. Hence, as indicated by the figure in the margin, along a given streamline p(s, n) = p(s) and ∂p∂s = dpds. These ideas combined with Eq. 3.4 give the following result valid along a streamline −γ dp dz 1 d(V 2 ) − = ρ ds ds 2 ds This simplifies to For steady, inviscid flow the sum of certain pressure, velocity, and elevation effects is constant along a streamline. dp + 1 ρd(V 2 ) + γ dz = 0  (along a streamline) 2 (3.5) which, for constant acceleration of gravity, can be integrated to give ∫ dp 1 + V 2 + gz = C  (along a streamline) ρ 2 (3.6) where C is a constant of integration to be determined by the conditions at some point on the streamline. 3.2 VIDEO V3.2 Balancing ball F = ma along a Streamline In general it is not possible to integrate the pressure term because the density may not be constant and, therefore, cannot be removed from under the integral sign. To carry out this integration we must know specifically how the density varies with pressure. This is not always easily determined. For example, for an ideal gas the density, pressure, and temperature are related according to ρ = p/RT, where R is the gas constant. To know how the density varies with pressure, we must also know the temperature variation. For now we will assume that the density and specific weight are constant (incompressible flow). The justification for this assumption and the consequences of compressibility will be considered further in Section 3.8.1 and more fully in Chapter 11. With the additional assumption that the density remains constant (a very good assumption for liquids and also for gases if the speed is “not too high”), Eq. 3.6 assumes the following simple representation for steady, inviscid, incompressible flow. 1 p + 2 ρV 2 + γz = constant along streamline VIDEO V3.3 Flow past a biker 107 (3.7) This is the celebrated Bernoulli equation—a very powerful tool in fluid mechanics. In 1738 Daniel Bernoulli (1700–1782) published his Hydrodynamics in which an equivalent of this famous equation first appeared. To use it correctly we must constantly remember the basic assumptions used in its derivation: (1) viscous effects are assumed negligible, (2) the flow is assumed to be steady, (3) the flow is assumed to be incompressible, and (4) the equation is applicable along a streamline. In the derivation of Eq. 3.7, we assume that the flow takes place in a plane (the x–z plane). In general, this equation is valid for both planar and nonplanar (three-dimensional) flows, provided it is applied along the streamline. We will provide many examples to illustrate the correct use of the Bernoulli equation and will show how a violation of the basic assumptions used in the derivation of this equation can lead to erroneous conclusions. The constant of integration in the Bernoulli equation can be evaluated if sufficient information about the flow is known at one location along the streamline. E X AMPLE 3.2 The Bernoulli Equation GIVEN Consider the flow of air around a bicyclist moving through still air with velocity V0, as is shown in Fig. E3.2. V2 = 0 (2) V1 = V0 (1) FIND Determine the difference in the pressure between points (1) and (2). SOLUTION In a coordinate fixed to the ground, the flow is unsteady as the bicyclist rides by. However, in a coordinate system fixed to the bike, it appears as though the air is flowing steadily toward the bicyclist with speed V0. Since use of the Bernoulli equation is restricted to steady flows, we select the coordinate system fixed to the bike. If the assumptions of Bernoulli’s equation are valid (steady, incompressible, inviscid flow), Eq. 3.7 can be applied as follows along the streamline that passes through (1) and (2) 1 1 p1 + 2 ρV 21 + γz1 = p2 + 2 ρV 22 + γz2 We consider (1) to be in the free stream so that V1 = V0 and (2) to be at the tip of the bicyclist’s nose and assume that z1 = z2 and V2 = 0 (both of which, as is discussed in Section 3.4, are reasonable assumptions). It follows that the pressure at (2) is greater than that at (1) by an amount 1 1 p2 − p1 = 2 ρV 21 = 2 ρV 20 (Ans) ■ Figure E3.2 COMMENTS A similar result was obtained in Example 3.1 by integrating the pressure gradient, which was known because the velocity distribution along the streamline, V(s), was known. The Bernoulli equation is a general integration of F = ma. To determine p2 − p1, knowledge of the detailed velocity distribution is not needed—only the “boundary conditions” at (1) and (2) are required. Of course, knowledge of the value of V along the streamline is needed to determine the pressure at points between (1) and (2). Note that if we measure p2 − p1 we can determine the speed, V0. As discussed in Section 3.5, this is the principle on which many velocity-measuring devices are based. If the bicyclist were accelerating or decelerating, the flow would be unsteady (i.e., V0 ≠ constant) and the above analysis would be incorrect since Eq. 3.7 is restricted to steady flow. 108 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation The difference in fluid velocity between two points in a flow field, V1 and V2, can often be controlled by appropriate geometric constraints of the fluid. For example, a garden hose nozzle is designed to give a much higher velocity at the exit of the nozzle than at its entrance where it is attached to the hose. As is shown by the Bernoulli equation, the pressure within the hose must be larger than that at the exit (for constant elevation, an increase in velocity requires a decrease in pressure if Eq. 3.7 is valid). It is this pressure drop that accelerates the water through the nozzle. Similarly, an airfoil is designed so that the fluid velocity over its upper surface is greater (on the average) than that along its lower surface. From the Bernoulli equation, therefore, the average pressure on the lower surface is greater than that on the upper surface. A net upward force, the lift, results. F = ma Normal to a Streamline 3.3 VIDEO V3.4 Hydrocyclone separator In this section we will consider application of Newton’s second law in a direction normal to the streamline. In many flows the streamlines are relatively straight, the flow is essentially one-dimensional, and variations in parameters across streamlines (in the normal direction) can often be neglected when compared to the variations along the streamline. However, in numerous other situations valuable information can be obtained from considering F = ma normal to the streamlines. For example, the devastating low-pressure region at the center of a tornado can be explained by applying Newton’s second law across the nearly circular streamlines of the tornado. We again consider the force balance on the fluid particle shown in Fig. 3.3 and the figure in the margin. This time, however, we consider components in the normal direction, nˆ, and write Newton’s second law in this direction as ∑ δFn = ℛ n V δm ρ δV V2 δm V 2 = ℛ ℛ (3.8) where ∑ δFn represents the sum of n components of all the forces acting on the particle and δm is particle mass. We assume the flow is steady with a normal acceleration an = V 2 ℛ, where ℛ is the local radius of curvature of the streamlines. This acceleration is produced by the change in direction of the particle’s velocity as it moves along a curved path. We again assume that the only forces of importance are pressure and gravity. The component of the weight (gravity force) in the normal direction is δ 𝒲n = −δ 𝒲 cos θ = −γ δV cos θ To apply F = ma normal to streamlines, the normal components of force are needed. If the streamline is vertical at the point of interest, θ = 90°, and there is no component of the particle weight normal to the direction of flow to contribute to its acceleration in that direction. If the pressure at the center of the particle is p, then its values on the top and bottom of the particle are p + δpn and p − δpn, where δpn = (∂p∂n)(δn2). Thus, if δFpn is the net pressure force on the particle in the normal direction, it follows that δFpn = ( p − δpn ) δs δy − ( p + δpn ) δs δy = −2 δpn δs δy =− ∂p ∂p δs δn δy = − δV ∂n ∂n Hence, the net force acting in the normal direction on the particle shown in Fig 3.3 is given by VIDEO V3.5 Aircraft wing tip vortex ∑ δFn = δ 𝒲n + δFpn = (−γ cos θ − ∂p δV ∂n ) (3.9) By combining Eqs. 3.8 and 3.9 and using the fact that along a line normal to the streamline cos θ = dzdn (see Fig. 3.3), we obtain the following equation of motion along the normal direction: −γ ∂p ρV 2 dz − = dn ∂n ℛ (3.10a) The physical interpretation of Eq. 3.10 is that a change in the direction of flow of a fluid particle (i.e., a curved path, ℛ < ∞) is accomplished by the appropriate combination of pressure 3.3 gradient and particle weight normal to the streamline. A larger speed or density or a smaller radius of curvature of the motion requires a larger force unbalance to produce the motion. For example, if gravity is neglected (as is commonly done for gas flows) or if the flow is in a horizontal (dzdn = 0) plane, Eq. 3.10 becomes Weight and/or pressure can produce curved streamlines. ∂p ρV 2 =− ∂n ℛ VIDEO 109 F = ma Normal to a Streamline V3.6 Free vortex (3.10b) This indicates that the pressure increases with distance away from the center of curvature (∂p∂n is negative since ρV 2ℛ is positive—the positive n direction points toward the “inside” of the curved streamline). Thus, the pressure outside a tornado (typical atmospheric pressure) is larger than it is near the center of the tornado (where an often dangerously low partial vacuum may occur). This pressure difference is needed to balance the centrifugal acceleration associated with the curved streamlines of the fluid motion. (See Fig. E6.6a in Section 6.5.3.) E X AMPLE 3.3 Pressure Variation Normal to a Streamline Shown in Figs. E3.3a,b are two flow fields with circular streamlines. The velocity distributions are V(r) = (V0 /r0 )r y y GIVEN V = (V0 /r0)r for case (a) and (V0 r0 ) V(r) = r r=ℛ V = (V0 r0)/r x n x for case (b) (a) where V0 is the velocity at r = r0. ( b) 6 Determine the pressure distributions, p = p(r), for each, given that p = p0 at r = r0. FIND SOLUTION We assume the flows are steady, inviscid, and incompressible with streamlines in the horizontal plane (dz/dn = 0). Because the streamlines are circles, the coordinate n points in a direction opposite that of the radial coordinate, ∂/∂n = −∂/∂r, and the radius of curvature is given by ℛ = r. Hence, Eq. 3.10b becomes 2 p – p0 ρV 02/2 0 (b) −2 −4 −6 ∂p ρV 2 = r ∂r (a) 4 0 0.5 1 1.5 2 2.5 r/r0 For case (a) this gives ( c) ∂p = ρ(V0 /r0 ) 2r ∂r ■ Figure E3.3 whereas for case (b) it gives ρ(V0 r0 ) 2 ∂p = ∂r r3 COMMENT For either case the pressure increases as r increases since ∂p/∂r > 0. Integration of these equations with respect to r, starting with a known pressure p = p0 at r = r0, gives p − p0 = (ρV022)[(r/r0 ) 2 − 1] (Ans) p − p0 = (ρV022)[1 − (r0 /r) 2] (Ans) for case (a) and for case (b). These pressure distributions are shown in Fig. E3.3c. The pressure distributions needed to balance the centrifugal accelerations in cases (a) and (b) are not the same because the velocity distributions are different. In fact, for case (a) the pressure increases without bound as r → ∞, whereas for case (b) the pressure approaches a finite value as r → ∞. The streamline patterns are the same for each case, however. Physically, case (a) represents rigid-body rotation (as obtained in a can of water on a turntable after it has been “spun up”) and case (b) represents a free vortex (an approximation to a tornado, a hurricane, or the swirl of water in a drain, the “bathtub vortex”). See Fig. E6.6 for an approximation of this type of flow. 110 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation The sum of pressure, elevation, and velocity effects is constant across streamlines. If we multiply Eq. 3.10 by dn, use the fact that ∂p∂n = dpdn if s is constant, and integrate across the streamline (in the n direction) we obtain ∫ dp + ρ ∫ V2 dn + gz = constant across the streamline ℛ (3.11) To complete the indicated integrations, we must know how the density varies with pressure and how the fluid speed and radius of curvature vary with n. For incompressible flow the density is constant and the integration involving the pressure term gives simply pρ. We are still left, however, with the integration of the second term in Eq. 3.11. Without knowing the n dependence in V = V(s, n) and ℛ = ℛ(s, n) this integration cannot be completed. Thus, the final form of Newton’s second law applied across the streamlines for steady, inviscid, incompressible flow is p+ρ ∫ V2 dn + γz = constant across the streamline ℛ (3.12) As with the Bernoulli equation, we must be careful that the assumptions involved in the derivation of this equation are not violated when it is used. 3.4 Physical Interpretations and Alternate Forms of the Bernoulli Equation In the previous two sections, we developed the basic equations governing fluid motion under a fairly stringent set of restrictions. In spite of the numerous assumptions imposed on these flows, a variety of flows can be readily analyzed with them. A physical interpretation of the equations will be of help in understanding the processes involved. To this end, we rewrite Eqs. 3.7 and 3.12 here and interpret them physically. Application of F = ma along and normal to the streamline results in 1 p + 2 ρV 2 + γz = constant along the streamline (3.13) and p+ρ z p + 1 ρV2 + γ z 2 = constant p + ρ∫ V dn + γ z ℛ 2 = constant ∫ V2 dn + γz = constant across the streamline ℛ (3.14) as indicated by the figure in the margin. The following basic assumptions were made to obtain these equations: The flow is steady, and the fluid is inviscid and incompressible. In practice none of these assumptions is exactly true. A violation of one or more of the above assumptions is a common cause for obtaining an incorrect match between the “real world” and solutions obtained by use of the Bernoulli equation. Fortunately, many “real-world” situations are adequately modeled by the use of Eqs. 3.13 and 3.14 because the flow is nearly steady and incompressible and the fluid behaves as if it were nearly inviscid. One example is flow outside the thin viscous region adjacent to a solid body. The Bernoulli equation was obtained by integration of the equation of motion along the “natural” coordinate direction of the streamline. To produce an acceleration, there must be an unbalance of the resultant forces, of which only pressure and gravity were considered to be important. Thus, there are three processes involved in the flow—mass times acceleration (the ρV 2 2 term), pressure (the p term), and weight (the γz term). Integration of the equation of motion to give Eq. 3.13 actually corresponds to the work– energy principle often used in the study of dynamics [see any standard dynamics text (Ref. 1)]. This principle results from a general integration of the equations of motion for an object in a way very similar to that done for the fluid particle in Section 3.2. With certain assumptions, a statement of the work–energy principle may be written as follows:   The work done on a particle by all forces acting on the particle is equal to the change of the kinetic energy of the particle. 3.4 Physical Interpretations and Alternate Forms of the Bernoulli Equation 111 The Bernoulli equation is a mathematical statement of this principle. Dividing each term by density, we get an alternate but equivalent form of the Bernoulli equation p V2 + + gz = constant along a streamline ρ 2 The Bernoulli equation can be written in terms of heights called heads. Each term in this equation has units of energy per unit mass. As the fluid particle moves, both gravity and pressure forces do work on the particle. Recall that the work done by a force is equal to the product of the distance the particle travels times the component of force in the direction of travel (i.e., work = F ∙ d). The terms gz and p/ρ are related to the work done by the weight and pressure forces, respectively. The remaining term, V2/2, is obviously related to the kinetic energy of the particle. In fact, an alternate method of deriving the Bernoulli equation is to use the first and second laws of thermodynamics (the energy and entropy equations), rather than Newton’s second law. With the appropriate restrictions, the general energy equation reduces to the Bernoulli equation. This approach is discussed in Sections 5.3.3 and 5.4. An alternate but equivalent form of the Bernoulli equation is obtained by dividing each term of Eq. 3.7 by the specific weight, γ, to obtain p V2 + z = constant on a streamline + γ 2g Each of the terms in this equation has the units of energy per weight (LFF = L) or length (feet, meters) and represents a certain type of head. The elevation term, z, is related to the potential energy of the particle and is called the elevation head. The pressure term, pγ, is called the pressure head and represents the height of a column of the fluid that is needed to produce the pressure p. The velocity term, V 22g, is the velocity head and represents the vertical distance needed for the fluid to fall freely (neglecting friction) if it is to reach velocity V from rest. The Bernoulli equation states that the sum of the pressure head, the velocity head, and the elevation head is constant along a streamline. The constant is called the total head.   E X AMPLE 3 .4 Kinetic, Potential, and Pressure Energy GIVEN Consider the flow of water from the syringe shown in Fig. E3.4a. As indicated in Fig. E3.4b, a force, F, applied to the (3) plunger will produce a pressure greater than atmospheric at point (1) within the syringe. The water flows from the needle, point (2), with relatively high velocity and coasts up to point (3) at the top of its trajectory. FIND Discuss the energy of the fluid at points (1), (2), and (3) g by using the Bernoulli equation. Energy Type (2) (1) Point 1 2 3 F (a) ■ Figure E3.4 (b) Kinetic V2/2 Potential gz Pressure p/ ρ Small Large Zero Zero Small Large Large Zero Zero 112 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation SOLUTION If the assumptions (steady, inviscid, incompressible flow) of the Bernoulli equation are approximately valid, it then follows that the flow can be explained in terms of the partition of the total energy of the water. According to Eq. 3.13, the sum of the three types of energy (kinetic, potential, and pressure) or heads (velocity, elevation, and pressure) must remain constant. The table above indicates the relative magnitude of each of these energies at the three points shown in the figure. The motion results in (or is due to) a change in the magnitude of each type of energy as the fluid flows from one location to another. An alternate way to consider this flow is as follows. The pressure gradient between (1) and (2) produces an acceleration to eject the water from the needle. Gravity acting on the particle between (2) and (3) produces a deceleration to cause the water to come to a momentary stop at the top of its flight. COMMENT If friction (viscous) effects were important, there would be an energy loss between (1) and (3) and for the given p1 the water would not be able to reach the height indicated in the figure. Such friction may arise in the needle (see Chapter 8 on pipe flow) or between the water stream and the surrounding air (see Chapter 9 on external flow). Keep in mind that pressure is not energy. It is a stress. Pressure does work between points, not at a point. It is, however, convenient to think of pressure as a "potential" to do work as we do with gravity. THE WIDE WORLD OF FLUIDS Armed with a water jet for hunting Archerfish, known for their with the tip of their tongue. The archerfish can produce a pres- ability to shoot down insects resting on foliage, are like subma- sure head within their gills large enough so that the jet can reach rine water pistols. With their snout sticking out of the water, they 2 to 3 m. However, it is accurate to only about 1 m. Recent eject a high-speed water jet at their prey, knocking it onto the research has shown that archerfish are very adept at calculating water surface where they snare it for their meal. The barrel of where their prey will fall. Within 100 milliseconds (a reaction time their water pistol is formed by placing their tongue against a twice as fast as a human’s), the fish has extracted all the informa- groove in the roof of their mouth to form a tube. By snapping tion needed to predict the point where the prey will hit the water. shut their gills, water is forced through the tube and directed Without further visual cues it charges directly to that point. The pressure variation across straight streamlines is hydrostatic. A net force is required to accelerate any mass. For steady flow the acceleration can be interpreted as arising from two distinct occurrences—a change in speed along the streamline and a change in direction if the streamline is not straight. Integration of the equation of motion along the streamline accounts for the change in speed (kinetic energy change) and results in the Bernoulli equation. Integration of the equation of motion normal to the streamline accounts for the centrifugal acceleration (V 2 ℛ) and results in Eq. 3.14. When a fluid particle travels along a curved path, a net force directed toward the center of curvature is required. Under the assumptions valid for Eq. 3.14, this force may be either gravity or pressure, or a combination of both. In many instances the streamlines are nearly straight (ℛ = ∞) so that centrifugal effects are negligible and the pressure variation across the streamlines is merely hydrostatic (because of gravity alone), even though the fluid is in motion. EXAMPLE 3.5 Pressure Variation in a Flowing Stream GIVEN Water flows in a curved, undulating waterslide as z shown in Fig. E3.5a. As an approximation to this flow, consider (4) Free surface (p = 0) g (2) (3) h4-3 ^ n h2-1 C (1) A D ℛ B ■ Figure E3.5b the inviscid, incompressible, steady flow shown in Fig. E3.5b. From section A to B the streamlines are straight, while from C to D they follow circular paths. FIND Describe the pressure variation between points (1) and ■ Figure E3.5a (Photo courtesy of Schlitterbahn® Waterparks.) (2) and points (3) and (4). 3.5 SOLUTION With the above assumptions and the fact that ℛ = ∞ for the portion from A to B, Eq. 3.14 becomes Static, Stagnation, Dynamic, and Total Pressure With p4 = 0 and z4 − z3 = h4–3 this becomes p3 = γh4–3 − ρ p + γz = constant p1 = p2 + γ(z2 − z1 ) = p2 + γh2–1 (Ans) Note that since the radius of curvature of the streamline is infinite, the pressure variation in the vertical direction is the same as if the fluid were stationary. However, if we apply Eq. 3.14, between points (3) and (4), we obtain (using dn = −dz) p4 + ρ ∫ z3 3.5 Each term in the Bernoulli equation can be interpreted as a form of pressure. z4 ∫ z4 z3 The constant can be determined by evaluating the known variables at the two locations using p2 = 0 (gage), z1 = 0, and z2 = h2–1 to give V2 (−dz) + γz4 = p3 + γz3 ℛ 113 V2 dz ℛ (Ans) To evaluate the integral, we must know the variation of V and ℛ with z. Even without this detailed information we note that the integral has a positive value. Thus, the pressure at (3) is less than the hydrostatic value, γh4–3, by an amount equal to ρ ʃzz34 (V 2ℛ) dz. This lower pressure, caused by the curved streamline, is necessary to accelerate the fluid around the curved path. COMMENT Note that we did not apply the Bernoulli equa- tion (Eq. 3.13) across the streamlines from (1) to (2) or (3) to (4). Rather we used Eq. 3.14. As is discussed in Section 3.8, application of the Bernoulli equation across streamlines (rather than along them) may lead to serious errors. Static, Stagnation, Dynamic, and Total Pressure A useful concept associated with the Bernoulli equation deals with the stagnation and dynamic pressures. These pressures arise from the conversion of kinetic energy in a flowing fluid into a “pressure rise” as the fluid is brought to rest (as in Example 3.2). In this section we explore various results of this process. Each term of the Bernoulli equation, Eq. 3.13, has the dimensions of force per unit area—psi, lbft2, Nm2. The first term, p, is the actual thermodynamic pressure of the fluid as it flows. To measure its value, one could move along with the fluid, thus being “static” relative to the moving fluid. Hence, it is normally termed the static pressure. Another way to measure the static pressure would be to drill a hole in a flat surface and fasten a piezometer tube as indicated by the location of point (3) in Fig. 3.4. As we saw in Example 3.5, the pressure in the flowing fluid of Fig. 3.4 at (1) is p1 = γh3–1 + p3, the same as if the fluid were static. From the manometer considerations of Chapter 2, we know that p3 = γh4–3. Thus, since h3–1 + h4–3 = h it follows that p1 = γh. The third term in Eq. 3.13, γz, is termed the hydrostatic pressure, in obvious regard to the hydrostatic pressure variation discussed in Chapter 2. It is not actually a pressure but does represent the change in pressure possible due to potential energy variations of the fluid as a result of elevation changes. The second term in the Bernoulli equation, ρV 2 2, is termed the dynamic pressure. Its interpretation can be seen in Fig. 3.4 by considering the pressure at the end of a small tube inserted into the flow and pointing upstream. After the initial transient motion has died out, the liquid will fill the tube to a height of H as shown. The fluid in the tube, including that at its tip, (2), will be stationary. That is, V2 = 0, or point (2) is a stagnation point. If we apply the Bernoulli equation between points (1) and (2), using V2 = 0 and assuming that z1 = z2, we find that 1 p2 = p1 + 2 ρV 21 Open H (4) h V h3-1 (3) h4-3 ρ (1) (2) V1 = V V2 = 0 ■ Figure 3.4 Measurement of static and stagnation pressures. 114 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation Stagnation streamline Stagnation point Stagnation point (b) (a) ■ Figure 3.5 Stagnation points. VIDEO V3.7 Stagnation point flow Hence, the pressure at the stagnation point is greater than the static pressure, p1, by an amount ρV 21 2, the dynamic pressure. It can be shown that there is a stagnation point on any stationary body that is placed into a flowing fluid. Some of the fluid flows “over” and some “under” the object. The dividing line (or surface for three-dimensional flows) is termed the stagnation streamline and terminates at the stagnation point on the body. (See the photograph at the beginning of the chapter.) For symmetrical nonrotating objects (such as a baseball) the stagnation point is clearly at the tip or front of the object as shown in Fig. 3.5a. For other flows such as a water jet against a car as shown in Fig. 3.5b, there is also a stagnation point on the car. If elevation effects are neglected, the stagnation pressure, p + ρV 2 2, is the largest pressure obtainable along a given streamline. It represents the conversion of all of the kinetic energy into a pressure rise. The sum of the static pressure, hydrostatic pressure, and dynamic pressure is termed the total pressure, pT. The Bernoulli equation is a statement that the total pressure remains constant along a streamline. That is,   1 p + 2 ρV 2 + γz = pT = constant along a streamline (3.15) Again, we must be careful that the assumptions used in the derivation of this equation are appropriate for the flow being considered. THE WIDE WORLD OF FLUIDS Pressurized eyes Our eyes need a certain amount of internal ments, all of which measure the slight deformation of the eyeball pressure in order to work properly, with the normal range being when a force is put on it. Some methods use a physical probe that between 10 and 20 mm of mercury. The pressure is determined by makes contact with the front of the eye, applies a known force, a balance between the fluid entering and leaving the eye. If the and measures the deformation. One noncontact method uses a pressure is above the normal level, damage may occur to the calibrated “puff” of air that is blown against the eye. The stagna- optic nerve where it leaves the eye, leading to a loss of the visual tion pressure resulting from the air blowing against the eyeball field termed glaucoma. Measurement of the pressure within the causes a slight deformation, the magnitude of which is correlated eye can be done by several different noninvasive types of instru- with the pressure within the eyeball. (See Problem 3.29.) Knowledge of the values of the static and stagnation pressures in a fluid implies that the fluid speed can be calculated. This is the principle on which the Pitot-static tube is based [H. de Pitot (1695–1771)]. As shown in Fig. 3.6, two concentric tubes are attached to two pressure gages (or a differential gage) so that the values of p3 and p4 (or the difference p3 − p4 ) can be determined. The center tube measures the stagnation pressure at its open tip. If elevation changes are negligible, 1 p3 = p + 2 ρV 2 3.5 115 Static, Stagnation, Dynamic, and Total Pressure (3) (2) (1) (4) (1) V p (2) (a) (b) ■ Figure 3.6 The Pitot-static tube. where p and V are the pressure and velocity of the fluid upstream of point (2). The outer tube is made with several small holes at an appropriate distance from the tip so that they measure the static pressure. If the effect of the elevation difference between (1) and (4) is negligible, then p4 = p1 = p Pitot-static tubes measure fluid velocity by converting velocity into pressure. By combining these two equations we see that 1 p3 − p4 = 2 ρV 2 which can be rearranged to give V = √2( p3 − p4 )ρ (3.16) The actual shape and size of Pitot-static tubes vary considerably. A typical Pitot-static probe used to determine aircraft airspeed is shown in Fig. 3.7. (See Fig. E3.6a also.) THE WIDE WORLD OF FLUIDS Bugged and plugged Pitot tubes Although a Pitot tube is a built their nest within the tube where the standard visual check simple device for measuring aircraft airspeed, many airplane acci- cannot detect it. The most serious accident (in terms of number dents have been caused by inaccurate Pitot tube readings. Most of fatalities) caused by a blocked Pitot tube involved a Boeing of these accidents are the result of having one or more of the 757 and occurred shortly after takeoff from Puerto Plata in the holes blocked and, therefore, not indicating the correct pressure Dominican Republic. Incorrect airspeed data were automatically (speed). Usually this is discovered during takeoff when time to fed to the computer, causing the autopilot to change the angle resolve the issue is short. The two most common causes for such of attack and the engine power. The flight crew became con- a blockage are either that the pilot (or ground crew) has forgotten fused by the false indications; the aircraft stalled and then to remove the protective Pitot tube cover or that insects have plunged into the Caribbean Sea, killing all aboard. Four static pressure ports Heated outer case VIDEO V3.8 Airspeed indicator Stagnation pressure port Mounting flange 2 in. Stagnation pressure fitting Static pressure fitting Heater leads (a) ■ Figure 3.7 Airplane Pitot-static probe. (a) Schematic, (b) Photograph. (Photograph courtesy of Aero-Instruments Co., LLC.) (b) 116 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation EXAMPLE 3 .6 Pitot-Static Tube GIVEN An airplane flies 200 mph at an elevation of 10,000 ft in a standard atmosphere as shown in Fig. E3.6a. FIND Determine the pressure at point (1) far ahead of the air- plane, the pressure at the stagnation point on the nose of the airplane, point (2), and the pressure difference indicated by a Pitot-static probe attached to the fuselage. SOLUTION From Table C.1 we find that the static pressure at the altitude given is p1 = 1456 lbft (abs) = 10.11 psia 2 (2) (1) V1 = 200 mph ■ Figure E3.6a (Photo courtesy of Hawker Beechcraft Corporation.) (Ans) Also the density is ρ = 0.001756 slugft . If the flow is steady, inviscid, and incompressible and elevation changes are neglected, Eq. 3.13 becomes 1 3 0.6 0.4 0.2 With V1 = 200 mph = 293 fts and V2 = 0 (since the coordinate system is fixed to the airplane), we obtain p2 = 1456 lbft2 + (0.001756 slugsft3 ) (2932 ft2s2 )2 = (1456 + 75.4) lbft2 (abs) Hence, in terms of gage pressure p2 = 75.4 lbft2 = 0.524 psi (Ans) Thus, the pressure difference indicated by the Pitot-static tube is ρV 21 = 0.524 psi 2 (Ans) COMMENTS Note that it is very easy to obtain incorrect results by using improper units. Do not add lbin.2 and lbft2. Recall that (slugft3 ) (ft2s2 ) = (slug fts2 )(ft2 ) = lbft2. ˙ Accurate measurement of static pressure requires great care. (200 mph, 0.951) 0.8 p1/p2 ρV 21 p2 = p1 + 2 p2 − p1 = Pitot-static tube 0 0 100 200 300 400 500 600 V1, mph ■ Figure E3.6b It was assumed that the flow is incompressible—the density remains constant from (1) to (2). However, since ρ = pRT, a change in pressure (or temperature) will cause a change in density. For this relatively low speed, the ratio of the absolute pressures is nearly unity [i.e., p1p2 = (10.11 psia)(10.11 + 0.524 psia) = 0.951], so that the density change is negligible. However, by repeating the calculations for various values of the speed, V1, the results shown in Fig. E3.6b are obtained. Clearly at the 500- to 600-mph speeds normally flown by commercial airliners, the pressure ratio is such that density changes are important. In such situations it is necessary to use compressible flow concepts to obtain accurate results. (See Section 3.8.1 and Chapter 11.) The Pitot-static tube provides a simple, relatively inexpensive way to measure fluid speed. Its use depends on the ability to measure the static and stagnation pressures. Care is needed to obtain these values accurately. For example, an accurate measurement of static pressure requires that none of the fluid’s kinetic energy be converted into a pressure rise at the point of measurement. This requires a smooth hole with no burrs or imperfections. As indicated in Fig. 3.8, such imperfections can cause the measured pressure to be greater or less than the actual static pressure. Also, the pressure along the surface of an object varies from the stagnation pressure at its stagnation point to values that may be less than the free stream static pressure. A typical pressure variation for a Pitot-static tube is indicated in Fig. 3.9. Clearly it is important that the pressure taps be properly located to ensure that (for all practical purposes) the pressure measured is actually the static pressure. V p V p V p (1) (1) (1) p1 > p p1 < p p1 = p ■ Figure 3.8 Incorrect and correct design of static pressure taps. 3.6 Examples of Use of the Bernoulli Equation 117 p V (2) (1) Stagnation pressure at tip Tube Stagnation pressure on stem Stem (1) (2) 0 Static pressure ■ Figure 3.9 Typical pressure distribution along a Pitotstatic tube. β (3) If θ = 0 p1 = p3 = p (2) V p θ β _ ρ V2 p2 = p + 1 (1) 2 ■ Figure 3.10 Cross section of a directional-finding Pitot-static tube, sometimes called a Fechheimer probe. In practice it is often difficult to align the Pitot-static tube directly into the flow direction. Any misalignment will produce a nonsymmetrical flow field that may introduce errors. Typically, yaw angles up to 12 to 20° (depending on the particular probe design) give results that are less than 1% in error from the perfectly aligned results. Generally it is more difficult to measure static pressure than stagnation pressure. One method of determining the flow direction and its speed (thus the velocity) is to use a directional-finding Pitot tube as is illustrated in Fig. 3.10. Three pressure taps are drilled into a small circular cylinder, fitted with small tubes, and connected to three pressure transducers. The cylinder is rotated until the pressures in the two side holes are equal, thus indicating that the center hole points directly upstream. The center tap then measures the stagnation pressure. The two side holes are located at an angle (β) specific to the flow conditions, usually between about 33° and about 38°, so that they measure the static pressure. The speed is then obtained from V = [2( p2 − p1 )ρ]12. The above discussion is valid for incompressible flows. At high speeds, compressibility becomes important (the density is not constant) and other phenomena occur. Some of these ideas are discussed in Section 3.8, while others (such as shockwaves for supersonic Pitot-tube applications) are discussed in Chapter 11. The concepts of static, dynamic, stagnation, and total pressure are useful in a variety of flow problems. These ideas are used more fully in the remainder of the book. 3.6 Examples of Use of the Bernoulli Equation In this section we illustrate various additional applications of the Bernoulli equation. Between any two points, (1) and (2), on a streamline with steady, inviscid, incompressible flow the Bernoulli equation can be applied in the form 1 1 p1 + 2 ρV 21 + γz1 = p2 + 2 ρV 22 + γz2 (3.17) Obviously, if five of the six variables are known, the remaining one can be determined. In many instances it is necessary to introduce other equations, such as the conservation of mass. Such considerations will be discussed briefly in this section and in more detail in Chapter 5. 118 VIDEO Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation (1) V3.9 Flow from a tank h z (3) ℓ d H (4) (2) (2) V ■ Figure 3.11 Vertical flow (5) from a tank. 3.6.1 Free Jets One of the oldest equations in fluid mechanics deals with the flow of a liquid from a large reservoir. A modern version of this type of flow involves the flow of coffee from a coffee urn as indicated by the figure in the margin. The basic principles of this type of flow are shown in Fig. 3.11 where a jet of liquid of diameter d flows from the nozzle with velocity V. (A nozzle is a device shaped to accelerate a fluid.) Application of Eq. 3.17 between points (1) and (2) on the streamline shown gives 1 γh = 2 ρV 2 We have used the facts that z1 = h, z2 = 0, the reservoir is large (V1 ≅ 0) and open to the atmosphere ( p1 = 0 gage), and the fluid leaves as a “free jet” ( p2 = 0). Thus, we obtain V V= The exit pressure for an incompressible fluid jet is equal to the surrounding pressure. γh = √2gh B ρ 2 (3.18) which is the modern version of a result obtained in 1643 by Torricelli (1608–1647), an Italian physicist. The fact that the exit pressure equals the surrounding pressure ( p2 = 0) can be seen by applying F = ma, as given by Eq. 3.14, across the streamlines between (2) and (4). If the streamlines at the tip of the nozzle are straight (ℛ = ∞), it follows that p2 = p4. Since (4) is on the surface of the jet, in contact with the atmosphere, we have p4 = 0. Thus, p2 = 0 also. Since (2) is an arbitrary point in the exit plane of the nozzle, it follows that the pressure is atmospheric across this plane. Physically, since there is no component of the weight force or acceleration in the normal (horizontal) direction, the pressure is constant in that direction. Once outside the nozzle, the stream continues to fall as a free jet with zero pressure throughout ( p5 = 0) and as seen by applying Eq. 3.17 between points (1) and (5), the speed increases according to V = √2g (h + H) V=0 h V = 2gh where, as shown in Fig. 3.11, H is the distance the fluid has fallen outside the nozzle. Equation 3.18 could also be obtained by writing the Bernoulli equation between points (3) and (4) using the fact that z4 = 0, z3 = ℓ. Also, V3 = 0 since it is far from the nozzle, and from hydrostatics, p3 = γ(h − ℓ). As learned in physics or dynamics and illustrated in the figure in the margin, any object dropped from rest that falls through a distance h in a vacuum will obtain the speed V = √2gh, the same as the water leaving the spout of the watering can shown in the figure in the margin on the next page. This is consistent with the fact that all of the particle’s potential energy is converted to kinetic energy, provided viscous (friction) effects are negligible. In terms of heads, the elevation head at point (1) is converted into the velocity head at point (2). Recall that for the case shown in Fig. 3.11 the pressure is the same (atmospheric) at points (1) and (2). For the horizontal nozzle of Fig. 3.12a, the velocity of the fluid at the centerline, V2, will be slightly greater than that at the top, V1, and slightly less than that at the bottom, V3, due to the differences in elevation. In general, d ⪡ h as shown in Fig. 3.12b and we can safely use the centerline velocity as a reasonable “average velocity.” 3.6 h Examples of Use of the Bernoulli Equation d 119 h (1) (2) d (3) (a) (b) ■ Figure 3.12 Horizontal flow from a tank. (1) dh a dj (2) (3) a ■ Figure 3.13 Vena contracta effect for a sharp-edged orifice. (1) h V = √2gh The diameter of a fluid jet is often smaller than that of the hole from which it flows. (2) If the exit is not a smooth, well-contoured nozzle, but rather a flat plate as shown in Fig. 3.13, the diameter of the jet, dj, will be less than the diameter of the hole, dh. This phenomenon, called a vena contracta effect, is a result of the inability of the fluid to turn the sharp 90° corner indicated by the dotted lines in the figure. Since the streamlines in the exit plane are curved (ℛ < ∞), the pressure across them is not constant. It would take an infinite pressure gradient across the streamlines to cause the fluid to turn a “sharp” corner (ℛ = 0). The highest pressure occurs along the centerline at (2) and the lowest pressure, p1 = p3 = 0, is at the edge of the jet. Thus, the assumption of uniform velocity with straight streamlines and constant pressure is not valid at the exit plane. It is valid, however, in the plane of the vena contracta, section a–a. The uniform velocity assumption is valid at this section provided dj ⪡ h, as is discussed for the flow from the nozzle shown in Fig. 3.12. The vena contracta effect is a function of the geometry of the outlet. Some typical configurations are shown in Fig. 3.14 along with typical values of the experimentally obtained contraction coefficient, Cc = Aj Ah, where Aj and Ah are the areas of the jet at the vena contracta and the area of the hole, respectively. THE WIDE WORLD OF FLUIDS Cotton candy, glass wool, and steel wool Although cotton somewhat more complex, but the basic process is similar. Liquid candy and glass wool insulation are made of entirely different glass is forced through tiny orifices and emerges as very fine materials and have entirely different uses, they are made by glass streams that quickly solidify. The resulting intertwined flex- similar processes. Cotton candy, invented in 1897, consists of ible fibers, glass wool, form an effective insulation material sugar fibers. Glass wool, invented in 1938, consists of glass because the tiny air “cavities” between the fibers inhibit air fibers. In a cotton candy machine, sugar is melted and then motion. Although steel wool looks similar to cotton candy or forced by centrifugal action to flow through numerous tiny glass wool, it is made by an entirely different process. Solid steel orifices in a spinning “bowl.” Upon emerging, the thin streams of wires are drawn over special cutting blades that have grooves liquid sugar cool very quickly and become solid threads that are cut into them so that long, thin threads of steel are peeled off to collected on a stick or cone. Making glass wool insulation is form the matted steel wool. 120 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation dj dh (a) Knife edge CC = 0.61 CC = 1.0 CC = A j /A h = (dj /dh)2 (b) Well rounded CC = 0.50 CC = 0.61 (c) Sharp edge ■ (d) Re-entrant Figure 3.14 Typical flow patterns and contraction coefficients for various round exit configurations. (a) Knife edge, (b) Well rounded, (c) Sharp edge, (d) Re-entrant. 3.6.2 Confined Flows In many cases the fluid is physically constrained within a device so that its pressure cannot be prescribed a priori as was done for the free jet examples above. Such cases include nozzles and pipes of variable diameter for which the fluid velocity changes because the flow area is different from one section to another. For these situations it is necessary to use the concept of conservation of mass (the continuity equation) along with the Bernoulli equation. The complete derivation and use of this equation are discussed in detail in Chapters 4 and 5. For the needs of this chapter we can use a simplified form of the continuity equation obtained from the following intuitive arguments. Consider a fluid flowing through a fixed volume (such as a syringe) that has one inlet and one outlet as shown in Fig. 3.15a. If the flow is steady so that there is no additional accumulation of fluid within the volume, the rate at which the fluid flows into the volume must equal the rate at which it flows out of the volume (otherwise, mass would not be conserved). The mass flowrate from an outlet, m ˙ (slugs/s or kg/s), is given by m˙ = ρQ, where Q (ft3s or m3s) is the volume flowrate. If the outlet area is A and the fluid flows across this area (normal to the area) with an average velocity V, then the volume of the fluid crossing this area in a time interval δt is VA δt, equal to that in a volume of length V δt and cross-sectional area A (see Fig. 3.15b). Hence, the volume flowrate (volume per unit time) is Q = VA. Thus, m ˙ = ρVA. To conserve mass, the inflow rate must equal the outflow rate. If the inlet is designated as (1) and the outlet as (2), it follows that m ˙ 1 = m˙ 2. Thus, conservation of mass requires The continuity equation states that mass cannot be created or destroyed. V2 = 2V1 A2 (2) ρ1A1V1 = ρ2A2V2 If the density remains constant, then ρ1 = ρ2, and the above becomes the continuity equation for incompressible, one-dimensional flow Q A1V1 = A2V2, or Q1 = Q2 A1 = 2A2 V1 (1) (3.19) For example, if as shown by the figure in the margin the outlet flow area is one-half the size of the inlet flow area, it follows that the outlet velocity is twice that of the inlet velocity, since 3.6 121 Examples of Use of the Bernoulli Equation V1 V2 (2) (1) V1 δ t V2 δ t Volume = V1 δ t A1 V2 V1 (2) Volume = V2 δ t A2 (1) Same parcel at t = δ t Fluid parcel at t = 0 ■ Figure 3.15 (a) Flow through a syringe. (b) Steady flow into and out of a volume. V2 = A1V1 A2 = 2V1. Use of the Bernoulli equation and the flowrate equation (continuity equation) is demonstrated by Example 3.7. E X AMPLE 3 .7 Flow from a Tank—Gravity Driven GIVEN A stream of refreshing beverage of diameter d = 0.01 m flows steadily from the cooler of diameter D = 0.20 m as shown in Figs. E3.7a and b. FIND Determine the flowrate, Q, from the bottle into the cooler if the depth of beverage in the cooler is to remain constant at h = 0.20 m. 1.10 Q (1) Q/Q0 1.05 D = 0.20 m h = 0.20 m (2) (3) d = 0.01 m (a) ■ Figure E3.7 (b) (0.05, 1.000003) 1.00 0 0.2 0.4 d/D (c) 0.6 0.8 122 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation SOLUTION For steady, inviscid, incompressible flow, the Bernoulli equation applied between points (1) and (2) is p1 + 1 2 2 ρV 1 + γz1 = p2 + 1 2 2 ρV 2 + γz2 (1) With the assumptions that p1 = p2 = 0, z1 = h, and z2 = 0, Eq. 1 becomes 1 2 2V1 1 + gh = 2 V 22 (2) Although the liquid level remains constant (h = constant), there is an average velocity, V1, across section (1) because of the flow from the tank. From Eq. 3.19 for steady incompressible flow, conservation of mass requires Q1 = Q2, where Q = AV. Thus, A1V1 = A2V2, or π π 2 D V1 = d 2V2 4 4 Hence, d 2 V1 = ( ) V2 D (3) Equations 1 and 3 can be combined to give V2 = 2gh B 1 − (dD) 4 = 2(9.81 ms2 ) (0.20 m) = 1.98 ms B 1 − (0.01 m0.20 m) 4 Thus, π (0.01 m) 2 (1.98 ms) 4 = 1.56 × 10 −4 m3s Q = A1V1 = A2V2 = (Ans) COMMENTS In this example we have not neglected the kinetic energy of the water in the tank (V1 ≠ 0). If the tank diameter is large compared to the jet diameter (D ⪢ d), Eq. 3 indicates that V1 ⪡ V2 and the assumption that V1 ≈ 0 would be reasonable. The error associated with this assumption can be seen by calculating the ratio of the flowrate assuming V1 ≠ 0, denoted Q, to that assuming V1 = 0, denoted Q0. This ratio, written as √2gh[1 − (dD) 4] V2 Q 1 = = = Q0 V2  D=∞ √2gh √1 − (dD) 4 is plotted in Fig. E3.7c. With 0 < dD < 0.4 it follows that 1 < QQ0 ≲ 1.01, and the error in assuming V1 = 0 is less than 1%. For this example with d/D = 0.01 m/0.20 m = 0.05, it follows that Q/Q0 = 1.000003. Thus, it is often reasonable to assume V1 = 0. Another method to observe that V1 ≈ 0 is to calculate the velocity of the fluid surface in the cooler using Q and A1. The result is V1 ≈ 0.001 m/s. Note that this problem was solved using points (1) and (2) located at the free surface and the exit of the pipe, respectively. Although this was convenient (because most of the variables are known at those points), other points could be selected and the same result would be obtained. For example, consider points (1) and (3) as indicated in Fig. E3.7b. At (3), located sufficiently far from the tank exit, V3 = 0 and z3 = z2 = 0. Also, p3 = γh since the pressure is hydrostatic sufficiently far from the exit. Use of this information in the Bernoulli equation applied between (3) and (2) gives the exact same result as obtained using it between (1) and (2). The only difference is that the elevation head, z1 = h, has been interchanged with the pressure head at (3), p3/λ = h. The fact that a kinetic energy change is often accompanied by a change in pressure is shown by Example 3.8. EXAMPLE 3.8 Flow from a Tank—Pressure Driven GIVEN Air flows steadily from a tank, through a hose of diameter D = 0.03 m, and exits to the atmosphere from a nozzle of diameter d = 0.01 m as shown in Fig. E3.8. The pressure in the tank remains constant at 3.0 kPa (gage) and the atmospheric conditions are standard temperature and pressure. p1 = 3.0 kPa (1) Air D = 0.03 m d = 0.01 m Q (2) (3) ■ Figure E3.8a FIND Determine (a) the flowrate and (b) the pressure in the hose. SOLUTION (a) If the flow is assumed steady, inviscid, and incompressible, we can apply the Bernoulli equation along the streamline from (1) to (2) to (3) as 1 1 p1 + 2 ρV 21 + γz1 = p2 + 2 ρV 22 + γz2 1 = p3 + 2 ρV 23 + γz3 With the assumption that z1 = z2 = z3 (horizontal hose), V1 = 0 (large tank), and p3 = 0 (free jet), this becomes V3 = 2p1 B ρ and 1 p2 = p1 − 2 ρV 22 (1) The density of the air in the tank is obtained from the ideal gas law, using standard absolute pressure and temperature, as 3.6 = p1 RT1 3000 (3.0 + 101) kN/m2 × 103 NkN (286.9 N ∙ mkg ∙ K) (15 + 273)K 2000 = 1.26 kgm3 1000 Thus, we find that V3 = 2(3.0 × 103 Nm2 ) B (0.01 m, 2963 N/m2) p2, N/m2 ρ= 123 Examples of Use of the Bernoulli Equation 1.26 kgm3 = 69.0 ms 0 or 0 0.01 0.02 0.03 d, m π 2 π d V3 = (0.01 m) 2 (69.0 ms) 4 4 = 0.00542 m3s ■ Figure E3.8b Q = A 3 V3 = 0.05 COMMENT Note that the value of V3 is determined strictly by 0.04 the value of p1 (and the assumptions involved in the Bernoulli equation), independent of the “shape” of the nozzle. The pressure head within the tank, p1γ = (3.0 kPa)(9.81 ms2)(1.26 kgm3) = 243 m, is converted to the velocity head at the exit, V 222g = (69.0 ms) 2(2 × 9.81 ms2 ) = 243 m. Although we used gage pressure in the Bernoulli equation ( p3 = 0), we had to use absolute pressure in the ideal gas law when calculating the density. Q, m3/s (Ans) 0.03 0.02 0.01 (0.01 m, 0.00542 m3/s) (b) The pressure within the hose can be obtained from Eq. 1 and the continuity equation (Eq. 3.19) 0 A2V2 = A3V3 ■ Figure E3.8c Hence, 2 d V2 = A3V3 A2 = ( ) V3 D =( 0.01 m 2 (69.0 ms) = 7.67 ms 0.03 m ) and from Eq. 1 1 p2 = 3.0 × 103 Nm2 − 2 (1.26 kgm3 ) (7.67 ms) 2 = (3000 − 37.1)Nm2 = 2963 Nm2 (Ans) COMMENTS In the absence of viscous effects, the pressure throughout the hose is constant and equal to p2. Physically, the decreases in pressure from p1 to p2 to p3 accelerate the air and increase its kinetic energy from zero in the tank to an interme- 0 0.01 0.02 0.03 d, m diate value in the hose and finally to its maximum value at the nozzle exit. Since the air velocity in the nozzle exit is nine times that in the hose, most of the pressure drop occurs across the nozzle ( p1 = 3000 Nm2, p2 = 2963 Nm2, and p3 = 0). Since the pressure change from (1) to (3) is not too great [i.e., in terms of absolute pressure (p1 − p3)p1 = 3.0101 = 0.03], it follows from the ideal gas law that the density change is also not significant. Hence, the incompressibility assumption is reasonable for this problem. If the tank pressure were considerably larger or if viscous effects were important, application of the Bernoulli equation to this situation would be incorrect. By repeating the calculations for various nozzle diameters, d, the results shown in Figs. E3.8b,c are obtained. The flowrate increases as the nozzle is opened (i.e., larger d). Note that if the nozzle diameter is the same as that of the hose (d = 0.03 m), the pressure throughout the hose is atmospheric (zero gage). THE WIDE WORLD OF FLUIDS Hi-tech inhaler The term inhaler often brings to mind a treat- divide into a mist of micron-scale droplets. The device fits the ment for asthma or bronchitis. Work is underway to develop a hand and accepts a disposable strip that contains the medicine family of inhalation devices that can do more than treat respira- solution sealed inside a blister of laminated plastic and the tory ailments. They will be able to deliver medication for diabetes nozzle. An electrically actuated piston drives the liquid from its and other conditions by spraying it to reach the bloodstream reservoir through the nozzle array and into the respiratory sys- through the lungs. The concept is to make the spray droplets tem. To take the medicine, the patient breathes through the fine enough to penetrate to the lungs’ tiny sacs, the alveoli, device and a differential pressure transducer in the inhaler where exchanges between blood and the outside world take senses when the patient’s breathing has reached the best con- place. This is accomplished by use of a laser-machined nozzle dition for receiving the medication. At that point, the piston is containing an array of very fine holes that cause the liquid to automatically triggered. 124 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation In many situations the combined effects of kinetic energy, pressure, and gravity are important. Example 3.9 illustrates this. EXAMPLE 3.9 Flow in a Variable Area Pipe GIVEN Water flows through a pipe reducer as is shown in Fig. SG E3.9. The static pressures at (1) and (2) are measured by the inverted U-tube manometer containing oil of specific gravity, SG, less than one. h FIND Determine the manometer reading, h. SOLUTION With the assumptions of steady, inviscid, incompressible flow, the Bernoulli equation can be written as p1 + 1 2 2 ρV 1 + γz1 = p2 + 1 2 2 ρV 2 + γz2 Q = A1V1 = A2V2 By combining these two equations we obtain p1 − p2 = γ(z2 − z1 ) + (2) z2 – z1 The continuity equation (Eq. 3.19) provides a second relationship between V1 and V2 if we assume the velocity profiles are uniform at those two locations and the fluid incompressible: 1 2 2 ρV 2 [1 D2 ℓ γ (1) Water θ D1 − (A2A1 ) ] 2 (1) This pressure difference is measured by the manometer and can be determined by using the pressure–depth ideas developed in Chapter 2. Thus, p1 − γ (z2 − z1) − γℓ − γh + SG γh + γℓ = p2 ■ Figure E3.9 or since V2 = QA2 h = (QA2 ) 2 1 − (A2A1 ) 2 2g(1 − SG) (Ans) or p1 − p2 = γ(z2 − z1 ) + (1 − SG)γh (2) As discussed in Chapter 2, this pressure difference is neither merely γh nor γ(h + z1 − z2 ), because the manometer fluid is not water. Equations 1 and 2 can be combined to give the desired result as follows: (1 − SG)γh = VIDEO V3.10 Venturi channel Cavitation occurs when the pressure is reduced to the vapor pressure. A2 2 1 ρV22[1 − ( ) ] 2 A1 COMMENT The difference in elevation, z1 − z2, was not needed because the change in elevation term in the Bernoulli equation exactly cancels the elevation term in the manometer equation. However, the pressure difference, p1 − p2, depends on the angle θ, because of the elevation, z1 − z2, in Eq. 1. Thus, for a given flowrate, the pressure difference, p1 − p2, as measured by a pressure gage would vary with θ, but the manometer reading, h, would be independent of θ. In general, an increase in velocity is accompanied by a decrease in pressure. For example, the velocity of the air flowing over the top surface of an airplane wing is, on the average, faster than that flowing under the bottom surface. Thus, the net pressure force is greater on the bottom than on the top—the wing generates a lift. If the differences in velocity are considerable, the differences in pressure can also be considerable. For flows of gases, this may introduce compressibility effects as discussed in Section 3.8 and Chapter 11. For flows of liquids, this may result in cavitation, a potentially dangerous situation that results when the liquid pressure is reduced to the vapor pressure and the liquid “boils.” As discussed in Chapter 1, the vapor pressure, pv, is the pressure at which vapor bubbles form in a liquid. It is the pressure at which the liquid starts to boil. Obviously this pressure depends on the type of liquid and its temperature. For example, water, which boils at 212 °F at standard 3.6 Examples of Use of the Bernoulli Equation 125 Q (1) (2) (3) p (Absolute pressure) Small Q Moderate Q pv Large Q Incipient cavitation 0 x ■ Figure 3.16 Pressure variation and cavitation in a ■ Figure 3.17 Tip cavitation from a propeller. (Photograph courtesy of The variable area pipe. Pennsylvania State University, Applied Research Laboratory, Garfield Thomas Water Tunnel.) Cavitation can cause damage to equipment. atmospheric pressure, 14.7 psia, boils at 80 °F if the pressure is 0.507 psia. That is, pυ = 0.507 psia at 80 °F and pυ = 14.7 psia at 212 °F. (See Tables B.1 and B.2.) One way to produce cavitation in a flowing liquid is noted from the Bernoulli equation. If the fluid velocity is increased (for example, by a reduction in flow area as shown in Fig. 3.16), the pressure will decrease. This pressure decrease (needed to accelerate the fluid through the constriction) can be large enough so that the pressure in the liquid is reduced to its vapor pressure. A simple example of cavitation can be demonstrated with an ordinary garden hose. If the hose is “kinked,” a restriction in the flow area in some ways analogous to that shown in Fig. 3.16 will result. The water velocity through this restriction will be relatively large. With a sufficient amount of restriction the sound of the flowing water will change—a definite “hissing” sound is produced. This sound is a result of cavitation. In such situations boiling occurs (though the temperature need not be high), vapor bubbles form, and then they collapse as the fluid moves into a region of higher pressure (lower velocity). This process can produce dynamic effects (imploding) that cause very large pressure transients in the vicinity of the bubbles. Pressures as large as 100,000 psi (690 MPa) are believed to occur. If the bubbles collapse close to a physical boundary they can, over a period of time, cause damage to the surface in the cavitation area. It is common to find pits or holes in the metal surfaces of old propellers or pump impellers. Tip cavitation from a propeller is shown in Fig. 3.17. In this case the high-speed rotation of the propeller produced a corresponding low pressure on the propeller. Obviously, proper design and use of equipment are needed to eliminate cavitation damage. E X AMPLE 3 .10 Siphon and Cavitation GIVEN A liquid can be siphoned from a container as shown in Fig. E3.10a, provided the end of the tube, point (3), is below the free surface in the container, point (1), and the maximum elevation of the tube, point (2), is “not too great.” Consider water at 60 °F being siphoned from a large tank through a constant-diameter hose as shown in Fig. E3.10b. The end of the siphon is 5 ft below the bottom of the tank, and the atmospheric pressure is 14.7 psia. FIND Determine the maximum height of the hill, H, over which the water can be siphoned without cavitation occurring. 126 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation SOLUTION If the flow is steady, inviscid, and incompressible we can apply the Bernoulli equation along the streamline from (1) to (2) to (3) as follows: 1 (2) 1 p1 + 2 ρV 21 + γz1 = p2 + 2 ρV 22 + γz2 1 = p3 + 2 ρV 23 + γz3 (1) (1) With the tank bottom as the datum, we have z1 = 15 ft, z2 = H, and z3 = −5 ft. Also, V1 = 0 (large tank), p1 = 0 (open tank), p3 = 0 (free jet), and from the continuity equation A2V2 = A3V3, or because the hose is constant diameter, V2 = V3. Thus, the speed of the fluid in the hose is determined from Eq. 1 to be (3) V3 = √2g(z1 − z3 ) = √2(32.2 fts2 )[15 − (−5)] ft = 35.9 fts = V2 Use of Eq. 1 between points (1) and (2) then gives the pressure p2 at the top of the hill as 1 ■ Figure E3.10a 1 p2 = p1 + 2 ρV 21 + γz1 − 2 ρV 22 − γz2 1 = γ(z1 − z2 ) − 2 ρV 22 (2) (2) From Table B.1, the vapor pressure of water at 60 °F is 0.256 psia. Hence, for incipient cavitation the lowest pressure in the system will be p = 0.256 psia. Careful consideration of Eq. 2 and Fig. E3.10b will show that this lowest pressure will occur at the top of the hill. Since we have used gage pressure at point (1) ( p1 = 0), we must use gage pressure at point (2) also. Thus, p2 = 0.256 − 14.7 = −14.4 psi and Eq. 2 gives (1) H 15 ft Water 5 ft (3) (−14.4 lbin. ) (144 in. ft ) 1 = (62.4 lbft3 ) (15 − H)ft − 2 (1.94 slugsft3 ) (35.9 fts) 2 2 ■ Figure E3.10b (Ans) For larger values of H, vapor bubbles will form at point (2) and the siphon action may stop. By using the fluid properties listed in Table 1.5 and repeating the calculations for various fluids, the results shown in Fig. E3.10c are obtained. The value of H is a function of both the specific weight of the fluid, γ, and its vapor pressure, pυ . COMMENTS Note that we could have used absolute pressure 40 35 30 25 H, ft throughout ( p2 = 0.256 psia and p1 = 14.7 psia) and obtained the same result. The lower the elevation of point (3), the larger the flowrate and, therefore, the smaller the value of H allowed. We could also have used the Bernoulli equation between (2) and (3), with V2 = V3, to obtain the same value of H. In this case it would not have been necessary to determine V2 by use of the Bernoulli equation between (1) and (3). The above results are independent of the diameter and length of the hose (provided viscous effects are not important). Proper design of the hose (or pipe) is needed to ensure that it will not collapse due to the large pressure difference (vacuum) between the inside and outside of the hose. 20 15 10 5 0 Carbon tet H = 28.2 ft Gasoline or Water 2 Alcohol 2 Fluid ■ Figure E3.10c 3.6.3 Flowrate Measurement Many types of devices using principles involved in the Bernoulli equation have been developed to measure fluid velocities and flowrates. The Pitot-static tube discussed in Section 3.5 is an example. Other examples discussed below include devices to measure flowrates in pipes and conduits and 3.6 (1) Examples of Use of the Bernoulli Equation 127 (2) Orifice Nozzle Venturi (1) ■ Figure 3.18 Typical devices for measuring (2) flowrate in pipes. devices to measure flowrates in open channels. In this chapter we will consider “ideal” flowmeters— those devoid of viscous, compressibility, and other “real-world” effects. Corrections for these effects are discussed in Chapters 8 and 10. Our goal here is to understand the basic operating principles of these simple flowmeters. An effective way to measure the flowrate through a pipe is to place some type of restriction within the pipe as shown in Fig. 3.18 and to measure the pressure difference between the lowvelocity, high-pressure upstream section (1) and the high-velocity, low-pressure downstream section (2). Three commonly used types of flowmeters are illustrated: the orifice meter, the nozzle meter, and the Venturi meter. The operation of each is based on the same physical principles—an increase in velocity causes a decrease in pressure. The difference between them is a matter of cost, accuracy, and how closely their actual operation obeys the idealized flow assumptions. We assume the flow is horizontal (z1 = z2 ), steady, inviscid, and incompressible between points (1) and (2). The Bernoulli equation becomes 1 1 p1 + 2 ρV 21 = p2 + 2 ρV 22 (The effect of nonhorizontal flow can be incorporated easily by including the change in elevation, z1 − z2, in the Bernoulli equation.) If we assume the velocity profiles are uniform at sections (1) and (2), the continuity equation (Eq. 3.19) can be written as The flowrate varies as the square root of the pressure difference across the flowmeter. Q Q ~ Δp Δ p = p1 – p2 Q = A1V1 = A2V2 where A2 is the small (A2 < A1 ) flow area at section (2). Combination of these two equations results in the following theoretical flowrate Q = A2 2(p1 − p2 ) B ρ[1 − (A2 A1 ) 2] (3.20) Thus, as shown by the figure in the margin, for a given flow geometry (A1 and A2 ) the flowrate can be determined if the pressure difference, p1 − p2, is measured. The actual measured flowrate, Qactual, will be smaller than this theoretical result because of various differences between the “real world” and the assumptions used in the derivation of Eq. 3.20. These differences (which are quite consistent and may be as small as 1 to 2% or as large as 40%, depending on the geometry used) can be accounted for by using an empirically obtained discharge coefficient as discussed in Section 128 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 8.6.1. A significant portion of the difference between ideal and actual flowrate is due to the vena contracta effect discussed above in Sec. 3.6.1. EXAMPLE 3.11 Venturi Meter GIVEN Kerosene (SG = 0.85) flows through the Venturi meter shown in Fig. E3.11a with flowrates between 0.005 and 0.050 m3s. Kerosene, SG = 0.85 D1 = 0.1 m FIND Determine the range in pressure difference, p1 − p2, needed to measure these flowrates. (1) (2) D2 = 0.06 m Q 0.005 m3/s < Q < 0.050 m3/s ■ Figure E3.11a SOLUTION If the flow is assumed to be steady, inviscid, and incompressible, the relationship between flowrate and pressure is given by Eq. 3.20. This can be rearranged to give p1 − p2 = Q2ρ[1 − (A2A1 ) 2] 2 A22 With the density of the flowing fluid ρ = SG ρH2O = 0.85(1000 kgm3 ) = 850 kgm3 and the area ratio A2 /A1 = (D2/D1 ) 2 = (0.06 m/0.10 m) 2 = 0.36 the pressure difference for the smallest flowrate is p1 − p2 = (0.005 m3s) 2 (850 kgm3 ) ideal results presented here are independent of the particular flowmeter geometry—an orifice, nozzle, or Venturi meter (see Fig. 3.18). It is seen from Eq. 3.20 that the flowrate varies as the square root of the pressure difference. Hence, as indicated by the numerical results and shown in Fig. E3.11b, a 10-fold increase in flowrate requires a 100-fold increase in pressure difference. This nonlinear relationship can cause difficulties when measuring flowrates over a wide range of values. Such measurements would require pressure transducers with a wide range of operation. An alternative is to use two flowmeters in parallel—one for the larger and one for the smaller flowrate ranges. (1 − 0.362 ) 2[(π4) (0.06 m) 2]2 120 = 1160 Nm2 = 1.16 kPa 100 p1− p2, kPa Likewise, the pressure difference for the largest flowrate is p1 − p2 = (0.05) 2 (850) (1 − 0.362 ) 2[(π4) (0.06) 2]2 = 1.16 × 10 Nm = 116 kPa 5 2 Thus, (0.05 m3/s, 116 kPa) 80 60 40 (0.005 m3/s, 1.16 kPa) 20 1.16 kPa ≤ p1 − p2 ≤ 116 kPa (Ans) 0 0 0.01 These values represent the pressure differences for inviscid, steady, incompressible conditions. The 0.02 0.03 0.04 0.05 Q, m3/s COMMENTS ■ Figure E3.11b Other flowmeters based on the Bernoulli equation are used to measure flowrates in open channels such as flumes and irrigation ditches. Two of these devices, the sluice gate and the sharpcrested weir, are discussed below under the assumption of steady, inviscid, incompressible flow. These and other open-channel flow devices are discussed in more detail in Chapter 10. Sluice gates like those shown in Fig. 3.19a are often used to regulate and measure the flowrate in open channels. As indicated in Fig. 3.19b, the flowrate, Q, is a function of the water depth upstream, z1, the width of the gate, b, and the gate opening, a. Application of the Bernoulli equation and continuity equation between points (1) and (2) can provide a good approximation to the actual flowrate obtained. We assume the velocity profiles are uniform sufficiently far upstream and downstream of the gate. 3.6 129 Examples of Use of the Bernoulli Equation V1 (1) Sluice gate width = b Sluice gates b z1 a V2 Q (3) a z2 (2) (4) (a) (b) ■ Figure 3.19 Sluice gate geometry. (Photograph courtesy of Plasti-Fab, Inc.) Thus, we apply the Bernoulli equation between points on the free surfaces at (1) and (2) to give 1 1 p1 + 2 ρV 21 + γz1 = p2 + 2 ρV 22 + γz2 Also, if the gate is the same width as the channel so that A1 = bz1 and A2 = bz2, the continuity equation gives Q = A1V1 = bV1z1 = A2V2 = bV2z2 The flowrate under a sluice gate depends on the water depths on either side of the gate. With the fact that p1 = p2 = 0, these equations can be combined and rearranged to give the flowrate as Q = z2 b 2g(z1 − z2 ) B 1 − (z2 z1 ) 2 (3.21) In the limit of z1 ⪢ z2 this result simply becomes Q = z2b√2gz1 This limiting result represents the fact that if the depth ratio, z1 z2, is large, the kinetic energy of the fluid upstream of the gate is negligible and the fluid velocity after it has fallen a distance (z1 − z2 ) ≈ z1 is approximately V2 = √2gz1. The results of Eq. 3.21 could also be obtained by using the Bernoulli equation between points (3) and (4) and the fact that p3 = γz1 and p4 = γz2 since the streamlines at these sections are straight. In this formulation, rather than the potential energies at (1) and (2), we have the pressure contributions at (3) and (4). The downstream depth, z2, not the gate opening, a, was used to obtain the result of Eq. 3.21. As was discussed relative to flow from an orifice (Fig. 3.14), the fluid cannot turn a sharp 90° corner. A vena contracta results with a contraction coefficient, Cc = z2 a, less than 1. Typically Cc is approximately 0.61 over the depth ratio range of 0 < az1 < 0.2. For larger values of az1 the value of Cc increases rapidly. E X AMPLE 3 .12 Sluice Gate GIVEN Water flows under the sluice gate shown in Fig. FIND Determine the approximate flowrate per unit width of E3.12a. the channel. Free ebooks ==> www.Ebook777.com 130 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation SOLUTION Under the assumptions of steady, inviscid, incompressible flow, we can apply Eq. 3.21 to obtain Qb, the flowrate per unit width, as 2g(z1 − z2 ) Q = z2 B 1 − (z2z1 ) 2 b In this instance z1 = 5.0 m and a = 0.80 m so the ratio az1 = 0.16 < 0.20, and we can assume that the contraction coefficient is approximately Cc = 0.61. Thus, z2 = Cc a = 0.61 (0.80 m) = 0.488 m and we obtain the flowrate 6.0 m 5.0 m Q 0.8 m ■ Figure E3.12a 9 Q 2(9.81 ms ) (5.0 m − 0.488 m) = (0.488 m) b B 1 − (0.488 m5.0 m) 2 8 2 (Ans) COMMENT If we consider z1 ⪢ z2 and neglect the kinetic Q/b, m2/s = 4.61 m2s 7 6 5 (5m, 4.61 m2/s) 4 3 2 energy of the upstream fluid, we would have 1 Q = z2 √2gz1 = 0.488 m √2(9.81 ms2 ) (5.0 m) b = 4.83 m2s 0 In this case the difference in Q with or without including V1 is not too significant because the depth ratio is fairly large (z1z2 = 5.00.488 = 10.2). Thus, it is often reasonable to neglect the kinetic energy upstream from the gate compared to that downstream of it. By repeating the calculations for various flow depths, z1 , the results shown in Fig. E3.12b are obtained. Note that the 0 5 z1, m 10 15 ■ Figure E3.12b flowrate is not directly proportional to the flow depth. Thus, for example, if during flood conditions the upstream depth doubled from z1 = 5 m to z1 = 10 m, the flowrate per unit width of the channel would not double but would increase only from 4.61 m2/s to 6.67 m2/s. Another device used to measure flow in an open channel is a weir. A typical rectangular, sharp-crested weir is shown in Fig. 3.20. For such devices the flowrate of liquid over the top of the weir plate is dependent on the weir height, Pw, the width of the channel, b, and the head, H, of the water above the top of the weir. Application of the Bernoulli equation can provide a simple approximation of the flowrate expected for these situations, even though the actual flow is quite complex. Between points (1) and (2) the pressure and gravitational fields cause the fluid to accelerate from velocity V1 to velocity V2. At (1) the pressure is p1 = γh, while at (2) the pressure is essentially atmospheric, p2 = 0. Across the curved streamlines directly above the top of the weir plate (section a–a), the pressure changes from atmospheric on the top surface to some maximum value within the fluid stream and then to atmospheric again at the bottom surface. This distribution is indicated in Fig. 3.20. Such a pressure distribution, combined with the streamline curvature and gravity, produces a rather nonuniform velocity profile across this section. This velocity distribution can be obtained from experiments or a more advanced theory. b a Pressure distribution Width = b H h Q a V1 (1) Weir plate (2) V2 Pw Weir plate ■ Figure 3.20 Rectangular, sharpcrested weir geometry. www.Ebook777.com 3.7 The Energy Line and the Hydraulic Grade Line 131 For now, we will take a very simple approach and assume that the weir flow is similar in many respects to an orifice-type flow with a free streamline. In this instance we would expect the average velocity across the top of the weir to be proportional to √2gH and the flow area for this rectangular weir to be proportional to Hb. Hence, it follows that Q Q = C1Hb √2gH = C1b √2g H32 Q ~ H3/2 H where C1 is a constant to be determined. Simple use of the Bernoulli equation has provided a method to analyze the relatively complex flow over a weir. The correct functional dependence of Q on H has been obtained (Q ∼ H 32, as indicated by the figure in the margin), but the value of the coefficient C1 is unknown. Even a more advanced analysis cannot predict its value accurately. As is discussed in Chapter 10, experiments are used to determine the value of C1. E X AMPLE 3.13 GIVEN Weir _θ H tan 2 Water flows over a triangular weir, as is shown in Fig. E3.13. θ FIND Based on a simple analysis using the Bernoulli equation, determine the dependence of the flowrate on the depth H. If the flowrate is Q0 when H = H0, estimate the flowrate when the depth is increased to H = 3H0. SOLUTION With the assumption that the flow is steady, inviscid, and incompressible, it is reasonable to assume from Eq. 3.18 that the average speed of the fluid over the triangular notch in the weir plate is proportional to √2gH. Also, the flow area for a depth of H is H[H tan (θ2)]. The combination of these two ideas gives θ θ Q = AV = H 2 tan (C2 √2gH) = C2 tan √2g H 52 2 2 The hydraulic grade line and energy line are graphical forms of the Bernoulli equation. H ■ Figure E3.13 Q3H0 QH0 = C2 tan(θ2) √2g (3H0 ) 52 C2 tan(θ2) √2g (H0 ) 52 = 15.6 (Ans) (Ans) where C2 is an unknown constant to be determined experimentally. Thus, an increase in the depth by a factor of three (from H0 to 3H0 ) results in an increase of the flowrate by a factor of 3.7 H COMMENT Note that for a triangular weir the flowrate is proportional to H 52, whereas for the rectangular weir discussed above, it is proportional to H 32. The triangular weir can be accurately used over a wide range of flowrates. The Energy Line and the Hydraulic Grade Line As was discussed in Section 3.4, the Bernoulli equation can be considered a form of the energy equation representing the partitioning of energy for an inviscid, incompressible, steady flow. The sum of the various energies of the fluid remains constant as the fluid flows from one section to another. A useful interpretation of the Bernoulli equation can be obtained through use of the concepts of the hydraulic grade line (HGL) and the energy line (EL). These ideas represent a geometrical interpretation of a flow and can often be effectively used to better grasp the fundamental processes involved. For steady, inviscid, incompressible flow the total energy remains constant along a streamline. The concept of “head” was introduced by dividing each term in Eq. 3.7 by the specific weight, γ = ρg, to give the Bernoulli equation in the following form p V2 + z = constant on a streamline = H + γ 2g (3.22) 132 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation V12/2g 2 V ___ 2 2g V 2/2g p2/γ Energy line (EL) Hydraulic grade line (HGL) p/γ p1/ γ H Static z2 z Q z1 Datum Stagnation ■ Figure 3.21 Representation of the energy line and the hydraulic grade line. V 2/2g p/γ H z Under the assumptions of the Bernoulli equation, the energy line is horizontal. Each of the terms in this equation has the units of length (feet or meters) and represents a certain type of head. The Bernoulli equation states that the sum of the pressure head, the velocity head, and the elevation head is constant along a streamline. This constant is called the total head, H, and is shown in the figure in the margin. The energy line is a line that represents the total head available to the fluid. As shown in Fig. 3.21, the elevation of the energy line can be obtained by measuring the stagnation pressure with a Pitot tube. (A Pitot tube is the portion of a Pitot-static tube that measures the stagnation pressure. See Section 3.5.) The stagnation point at the end of the Pitot tube provides a measurement of the total head (or energy) of the flow. The static pressure tap connected to the piezometer tube shown, on the other hand, measures the sum of the pressure head and the elevation head, pγ + z. This sum is often called the piezometric head. The static pressure tap does not measure the velocity head. According to Eq. 3.22, the total head remains constant along the streamline (provided the assumptions of the Bernoulli equation are valid). Thus, a Pitot tube at any other location in the flow will measure the same total head, as is shown in the figure. The elevation head, velocity head, and pressure head may vary individually along the streamline, however. The locus of elevations provided by a series of Pitot tubes is termed the energy line, EL. The locus provided by a series of piezometer taps is termed the hydraulic grade line, HGL. Under the assumptions of the Bernoulli equation, the energy line is horizontal. If the fluid velocity changes along the streamline, the hydraulic grade line will not be horizontal. If viscous effects are important (as they often are in pipe flows), the total head does not remain constant due to a loss in useful energy as the fluid flows along its streamline. This means that the energy line is no longer horizontal. Such viscous effects are discussed in Chapters 5 and 8. The energy line and hydraulic grade line for flow from a large tank are shown in Fig. 3.22. If the flow is steady, incompressible, and inviscid, the energy line is horizontal and at the elevation of the liquid in the tank (since the fluid velocity in the tank and the pressure on the surface are zero). V1 = p1 = 0 H = z1 EL 2 V ___ 2 2g (1) p__2 γ HGL 2 V ___ 3 2g p3 = 0 (2) z2 z3 (3) ■ Figure 3.22 The energy line and hydraulic grade line for flow from a tank. 3.7 The Energy Line and the Hydraulic Grade Line Q EL 2 V __ 2g p<0 p/γ HGL p/γ p>0 133 z z ■ Figure 3.23 Use of the energy line and the hydraulic grade line. For flow below (above) the hydraulic grade line, the pressure is positive (negative). The hydraulic grade line lies a distance of one velocity head, V 2 2g, below the energy line. Thus, a change in fluid velocity due to a change in the pipe diameter results in a change in the elevation of the hydraulic grade line. At the pipe outlet the pressure head is zero (gage), so the pipe elevation and the hydraulic grade line coincide. The distance from the pipe to the hydraulic grade line indicates the pressure within the pipe, as is shown in Fig. 3.23. If the pipe lies below the hydraulic grade line, the pressure within the pipe is positive (above atmospheric). If the pipe lies above the hydraulic grade line, the pressure is negative (below atmospheric). Thus, a scale drawing of a pipeline and the hydraulic grade line can be used to readily indicate regions of positive or negative pressure within a pipe. E X AMPLE 3 .14 Energy Line and Hydraulic Grade Line GIVEN Water is siphoned from the tank shown in Fig. E3.14 through a hose of constant diameter. A small hole is found in the hose at location (1) as indicated. FIND When the siphon is used, will water leak out of the hose, or will air leak into the hose, thereby possibly causing the siphon to malfunction? SOLUTION Whether air will leak into or water will leak out of the hose depends on whether the pressure within the hose at (1) is less than or greater than atmospheric. Which happens can be easily determined by using the energy line and hydraulic grade line concepts. With the assumption of steady, incompressible, inviscid flow it follows that the total head is constant—thus, the energy line is horizontal. Since the hose diameter is constant, it follows from the continuity equation (AV = constant) that the water velocity in the hose is constant throughout. Thus, the hydraulic grade line is a constant distance, V 22g, below the energy line as shown in Fig. E3.14. Since the pressure at the end of the hose is atmospheric, it follows that the hydraulic grade line is at the same elevation as the end of the hose outlet. The fluid within the hose at any point above the hydraulic grade line will be at less than atmospheric pressure. Thus, air will leak into the hose through the hole at point (1). (Ans) (1) HGL with valve closed and EL with valve open or closed 2 V __ 2g z p_ γ HGL with valve open Valve ■ Figure E3.14 COMMENT In practice, viscous effects may be quite important, making this simple analysis (horizontal energy line) incorrect. However, if the hose is “not too small diameter,” “not too long,” the fluid “not too viscous,” and the flowrate “not too large,” the above result may be very accurate. If any of these assumptions are relaxed, a more detailed analysis is required (see Chapter 8). If the end of the hose were closed so that the flowrate were zero, the hydraulic grade line would coincide with the energy line (V 22g = 0 throughout), the pressure at (1) would be greater than atmospheric, and water would leak through the hole at (1). The above discussion of the hydraulic grade line and the energy line is restricted to ideal situations involving inviscid, incompressible flows. Another restriction is that there are no “sources” or “sinks” of energy within the flow field. That is, there are no pumps or turbines involved. Alterations in the energy line and hydraulic grade line concepts due to these devices are discussed in Chapters 5 and 8. 134 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 3.8 Restrictions on Use of the Bernoulli Equation Proper use of the Bernoulli equation requires close attention to the assumptions used in its derivation. In this section we review some of these assumptions and consider the consequences of incorrect use of the equation. 3.8.1 Compressibility Effects Δp Δp ~ V 2 V One of the main assumptions is that the fluid is incompressible. Although this is reasonable for most liquid flows, it can, in certain instances, introduce considerable errors for gases. In the previous section, we saw that the stagnation pressure, pstag, is greater than the static pressure, pstatic, by an amount Δp = pstag − pstatic = ρV2/2 provided that the density remains constant. If this dynamic pressure is not too large compared with the static pressure, the density change between two points is not very large and the flow can be considered incompressible. However, since the dynamic pressure varies as V2, the error associated with the assumption that a fluid is incompressible increases with the square of the velocity of the fluid, as indicated by the figure in the margin. To account for compressibility effects, we could return to Eq. 3.6 and properly integrate the term ʃ dp/ρ when ρ is not constant. That method can be more complicated than necessary and will likely introduce new restrictions. Instead we will develop a criterion that allows us to check the validity of the incompressible flow assumption. From Eq. 1.13, dp dρ = ρ Ev which for a finite change is Δp Δρ = ρ Ev If we accept a rather arbitrary density change of 3% as insignificant, we can model the fluid as incompressible when Δρ ≤ 0.03 ρ The Bernoulli equation can help us determine a velocity limit for incompressible flow. Of course, this criterion can be decreased or increased. A simple, although specialized, case of compressible flow occurs when the temperature of an ideal gas remains constant along the streamline—isothermal flow. As indicated in Section 1.7.2, Eq. 1.16, for an isothermal flow, Ev = p, so Δp Δρ = ≤ 0.03 ρ p This criterion should be used with caution because most isothermal flows are accompanied by viscous effects. A much more common compressible flow condition is that of isentropic (constant entropy) flow of an ideal gas. Such flows are reversible adiabatic processes—“no friction or heat transfer”— and are closely approximated in many physical situations. As indicated in Section 1.7.2, Eq. 1.17, for an isentropic flow, Ev = kp, so Δp Δρ = ρ kp For air, k = 1.4, so applying our criterion, Δp Δρ = ≤ 0.03 ρ 1.4p or Δp ≤ 0.042 p 3.8 Restrictions on Use of the Bernoulli Equation 135 The Bernoulli equation as presented in this chapter applies to an isentropic process. Therefore, we can relate pressure to the fluid velocity with the Bernoulli equation (with negligible elevation change) as Δp = ρV2/2. Thus, we can model a flow as incompressible if ρV2 ≤ 0.084 p Actually a value of 0.09 will suffice. For air at standard conditions, this quantity corresponds to a velocity of 102 m/s (335 ft/s). As discussed in Chapter 11, this velocity corresponds to about 30% of the speed of sound in standard air (Ma = 0.3). At higher speeds, compressibility may become important. 3.8.2 Unsteady Effects The Bernoulli equation can be modified for unsteady flows. Another restriction of the Bernoulli equation (Eq. 3.7) is the assumption that the flow is steady. For such flows, on a given streamline the velocity is a function of only s, the location along the streamline. That is, along a streamline V = V(s). For unsteady flows the velocity is also a function of time, so that along a streamline V = V(s, t). Thus when taking the time derivative of the velocity to obtain the streamwise acceleration, we obtain as = ∂V∂t + V ∂V∂s rather than just as = V ∂V∂s as is true for steady flow. For steady flows the acceleration is due to the change in velocity resulting from a change in position of the particle (the V ∂V∂s term), whereas for unsteady flow there is an additional contribution to the acceleration resulting from a change in velocity with time at a fixed location (the ∂V∂t term). These effects are discussed in detail in Chapter 4. The net effect is that the inclusion of the unsteady term, ∂V∂t, does not allow the equation of motion to be easily integrated (as was done to obtain the Bernoulli equation) unless additional assumptions are made. The Bernoulli equation was obtained by integrating the component of Newton’s second law (Eq. 3.5) along the streamline. When integrated, the acceleration contribution to this equation, the 1 2 2 ρd(V ) term, gave rise to the kinetic energy term in the Bernoulli equation. If the steps leading to Eq. 3.5 are repeated with the inclusion of the unsteady effect ( ∂V∂t ≠ 0), the following is obtained: ρ ∂V 1 ds + dp + ρd(V 2 ) + γ dz = 0  (along a streamline) ∂t 2 For incompressible flow this can be integrated between points (1) and (2) to give p1 + VIDEO V3.11 Oscillations in a U-tube 1 2 ρV 1 + γz1 = ρ 2 ∫ S2 S1 ∂V 1 ds + p2 + ρV 22 + γz2  (along a streamline) ∂t 2 Equation 3.23 is an unsteady form of the Bernoulli equation valid for unsteady, incompressible, inviscid flow. Except for the integral involving the local acceleration, ∂V∂t, it is identical to the steady Bernoulli equation. In general, it is not possible to evaluate this integral because the variation of ∂V∂t along the streamline is not known. In some situations the concepts of “irrotational flow” and the “velocity potential” can be used to simplify this integral. These topics are discussed in Chapter 6. E X AMPLE 3 .15 Unsteady Flow—U-Tube Open tube GIVEN An incompressible, inviscid liquid is placed in a ver- tical, constant diameter U-tube as indicated in Fig. E3.15. When released from the nonequilibrium position shown, the liquid column will oscillate at a specific frequency. FIND (3.23) V2 (2) g z Determine this frequency. z= 0 Equilibrium position (1) V1 ℓ ■ Figure E3.15 136 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation SOLUTION The frequency of oscillation can be calculated by use of Eq. 3.23 as follows. Let points (1) and (2) be at the air–water interfaces of the two columns of the tube and z = 0 correspond to the equilibrium position of these interfaces. Hence, p1 = p2 = 0 and if z2 = z, then z1 = −z. In general, z is a function of time, z = z(t). For a constant diameter tube, at any instant in time the fluid speed is constant throughout the tube, V1 = V2 = V, and the integral representing the unsteady effect in Eq. 3.23 can be written as ∫ s1 s2 ∂V dV ds = ∂t dt ∫ s2 ds = ℓ s1 dV dt where ℓ is the total length of the liquid column as shown in the figure. Thus, Eq. 3.23 can be written as dV γ(−z) = ρℓ + γz dt Since V = dzdt and γ = ρg, this can be written as the secondorder differential equation describing simple harmonic motion 2g d 2z + z=0 2 ℓ dt   which has the solution z(t) = C1 sin( √2gℓ t) + C2 cos ( √2gℓ t). The values of the constants C1 and C2 depend on the initial state (velocity and position) of the liquid at t = 0. Thus, the liquid oscillates in the tube with a frequency ω = √2gℓ (Ans) COMMENT This frequency depends on the length of the column and the acceleration of gravity (in a manner very similar to the oscillation of a pendulum). The period of this oscillation (the time required to complete an oscillation) is t0 = 2π √ℓ2g. In a few unsteady flow cases, the flow can be made steady by an appropriate selection of the coordinate system. Example 3.16 illustrates this. EXAMPLE 3 .16 Unsteady or Steady Flow GIVEN A submarine moves through seawater (SG = 1.03) at a depth of 50 m with velocity V0 = 5.0 ms as shown in Fig. E3.16. FIND Determine the pressure at the stagnation point (2). SOLUTION In a coordinate system fixed to the ground, the flow is unsteady. For example, the water velocity at (1) is zero with the submarine in its initial position, but at the instant when the nose, (2), reaches point (1) the velocity there becomes V1 = −V0 ˆi . Thus, ∂V1∂t ≠ 0 and the flow is unsteady. Application of the steady Bernoulli equation between (1) and (2) would give the incorrect result that p1 = p2 + ρV 20 2. According to this result, the static pressure is greater than the stagnation pressure—an incorrect use of the Bernoulli equation. We can either use an unsteady analysis for the flow (which is outside the scope of this text) or redefine the coordinate system so that it is fixed on the submarine, giving steady flow with respect to this system. The correct method would be p2 = γ h = 50 m (1) (2) V0 = 5 m/s x ■ Figure E3.16 = (12,900 + 505,000) Nm2 = 518 kPa (Ans) similar to that discussed in Example 3.2. COMMENT If the submarine were accelerating, ∂V0 ∂t ≠ 0, the flow would be unsteady in either of the above coordinate systems and we would be forced to use an unsteady form of the Bernoulli equation. ρV 21 + γh = [(1.03) (1000) kgm3] (5.0 ms) 22 2 + (9.80 × 103 Nm3 ) (1.03) (50 m) Some unsteady flows may be treated as “quasisteady” and solved approximately by using the steady Bernoulli equation. In these cases the unsteadiness is “not too great” (in some sense), and the steady flow results can be applied at each instant in time as though the flow were steady. The slow draining of a tank filled with liquid provides an example of this type of flow. 3.8 Restrictions on Use of the Bernoulli Equation 137 3.8.3 Rotational Effects Another of the restrictions of the Bernoulli equation is that it is applicable along the streamline. Application of the Bernoulli equation across streamlines (i.e., from a point on one streamline to a point on another streamline) can lead to considerable errors, depending on the particular flow conditions involved. In general, the Bernoulli constant varies from streamline to streamline. However, under certain restrictions this constant is the same throughout the entire flow field. Example 3.17 illustrates this fact. Care must be used in applying the Bernoulli equation across streamlines. E X AMPLE 3.17 Use of Bernoulli Equation across Streamlines Fluid particles spin GIVEN Consider the uniform flow in the channel shown in Fig. E3.17a. The liquid in the vertical piezometer tube is stationary. p5 = 0 FIND Discuss the use of the Bernoulli equation between points (1) and (2), points (3) and (4), and points (4) and (5). SOLUTION If the flow is steady, inviscid, and incompressible, Eq. 3.7 written between points (1) and (2) gives 1 1 p1 + 2 ρV 21 + γz1 = p2 + 2 ρV 22 + γz2 = constant = C12 Since V1 = V2 = V0 and z1 = z2 = 0, it follows that p1 = p2 = p0 and the Bernoulli constant for this streamline, C12, is given by (5) H V0 (3) h (1) p1 = p0 (4) z=h (2) z=0 (b) V0 (a) ■ Figure E3.17 1 C12 = 2 ρV 20 + p0 Along the streamline from (3) to (4) we note that V3 = V4 = V0 and z3 = z4 = h. As was shown in Example 3.5, application of F = ma across the streamline (Eq. 3.12) gives p3 = p1 − γh because the streamlines are straight and horizontal. The above facts combined with the Bernoulli equation applied between (3) and (4) show that p3 = p4 and that the Bernoulli constant along this streamline is the same as that along the streamline between (1) and (2). That is, C34 = C12, or 1 1 p3 + 2 ρV 23 + γz3 = p4 + 2 ρV 24 + γz4 = C34 = C12 Similar reasoning shows that the Bernoulli constant is the same for any streamline in Fig. E3.17. Hence, 1 p + 2 ρV 2 + γz = constant throughout the flow Again from Example 3.5 we recall that p4 = p5 + γH = γH VIDEO V3.12 Flow over a cavity If we apply the Bernoulli equation across streamlines from (4) to (5), we obtain the incorrect result H = p4γ + V 24 2g. The correct result is H = p4γ. From the above we see that we can apply the Bernoulli equation across streamlines (1)–(2) and (3)–(4) (i.e., C12 = C34 ) but not across streamlines from (4) to (5). The reason for this is that while the flow in the channel is “irrotational,” it is “rotational” between the flowing fluid in the channel and the stationary fluid in the piezometer tube. Because of the uniform velocity profile across the channel, it is seen that the fluid particles do not rotate or “spin” as they move. The flow is “irrotational.” However, as seen in Fig. E3.17b, there is a very thin shear layer between (4) and (5) in which adjacent fluid particles interact and rotate or “spin.” This produces a “rotational” flow. A more complete analysis would show that the Bernoulli equation cannot be applied across streamlines if the flow is “rotational” (see Chapter 6). As is suggested by Example 3.17, if the flow is “irrotational” (i.e., the fluid particles do not “spin” as they move), it is appropriate to use the Bernoulli equation across streamlines. However, if the flow is “rotational” (fluid particles “spin”), use of the Bernoulli equation is restricted to flow along a streamline. The distinction between irrotational and rotational flow is often a very subtle and confusing one. These topics are discussed in more detail in Chapter 6. A thorough discussion can be found in more advanced texts (Ref. 3). 138 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 3.8.4 Other Restrictions The Bernoulli equation is not valid for flows that involve pumps or turbines. 3.9 steady flow streamline Bernoulli equation elevation head pressure head velocity head static pressure dynamic pressure stagnation point stagnation pressure total pressure Pitot-static tube free jet volume flowrate continuity equation cavitation flowmeter hydraulic grade line energy line Another restriction on the Bernoulli equation is that the flow is inviscid. As is discussed in Section 3.4, the Bernoulli equation is actually a first integral of Newton’s second law along a streamline. This general integration was possible because, in the absence of viscous effects, the fluid system considered was a conservative system. The total energy of the system remains constant. If viscous effects are important the system is nonconservative (dissipative) and useful energy losses occur. A more detailed analysis is needed for these cases. Such material is presented in Chapter 5. The final basic restriction on use of the Bernoulli equation is that there are no mechanical devices (pumps or turbines) in the system between the two points for which the equation is applied. These devices necessarily cause unsteady flow or interrupt the streamlines (Ref. 4), if they are present. Analysis of flows with pumps and turbines is covered in Chapters 5 and 12. In this chapter we have spent considerable time investigating fluid dynamic situations governed by a relatively simple analysis for steady, inviscid, incompressible flows. Many flows can be adequately analyzed by use of these ideas. However, because of the rather severe restrictions imposed, many others cannot. An understanding of these basic ideas will provide a firm foundation for the remainder of the topics in this book. Chapter Summary and Study Guide In this chapter, several aspects of the steady flow of an inviscid, incompressible fluid are discussed. Newton’s second law, F = ma, is applied to flows for which the only important forces are those due to pressure and gravity (weight)—viscous effects are assumed negligible. The result is the often-used Bernoulli equation, which provides a simple relationship among pressure, elevation, and velocity variations along a streamline. A similar but less often used equation is also obtained to describe the variations in these parameters normal to a streamline. The concept of a stagnation point and the corresponding stagnation pressure is introduced, as are the concepts of static, dynamic, and total pressure and their related heads. Several applications of the Bernoulli equation are discussed. In some flow situations, such as the use of a Pitot-static tube to measure fluid velocity or the flow of a liquid as a free jet from a tank, a Bernoulli equation alone is sufficient for the analysis. In other instances, such as confined flows in tubes and flowmeters, it is necessary to use both the Bernoulli equation and the continuity equation, which is a statement of the fact that mass is conserved as fluid flows. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ ■ ■ ■ ■ ■ ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. explain the origin of the pressure, elevation, and velocity terms in the Bernoulli equation and how they are related to Newton’s second law of motion. apply the Bernoulli equation to simple flow situations, including Pitot-static tubes, free jet flows, confined flows, and flowmeters. use the concept of conservation of mass (the continuity equation) in conjunction with the Bernoulli equation to solve simple flow problems. apply Newton’s second law across streamlines for appropriate steady, inviscid, incompressible flows. use the concepts of pressure, elevation, velocity, and total heads to solve various flow problems. explain and use the concepts of static, stagnation, dynamic, and total pressures. use the energy line and the hydraulic grade line concepts to solve various flow problems. explain the various restrictions on use of the Bernoulli equation. References 139 Some of the important equations in this chapter are: Streamwise and normal acceleration Force balance along a streamline for steady inviscid flow Bernoulli equation as = V ∫ dpρ + 12 V 2 + gz = C V2 ℛ (3.1) (along a streamline) (3.6) an = 1 p + 2ρV2 + γz = constant along streamline Pressure gradient normal to streamline for inviscid flow in absence of gravity Force balance normal to a streamline for steady, inviscid, incompressible flow ∂V , ∂s ∂p ρV2 =− ∂n ℛ p+ρ (3.7) (3.10b) 2 ∫ Vℛ dn + γz = constant across the streamline (3.12) Velocity measurement for a Pitot-static tube V = √2 (p3 − p4 )/ρ (3.16) Free jet V= γh 2 = √2gh B ρ (3.18) Continuity equation A1V1 = A2V2, or Q1 = Q2 2(p1 − p2 ) B ρ[1 − (A2 A1 ) 2] Flowmeter equation (ideal) Q = A2 Sluice gate equation (ideal) Q = z2b Total head 2g(z1 − z2 ) B 1 − (z2 z1 ) 2 p V2 + z = constant on a streamline = H + γ 2g (3.19) (3.20) (3.21) (3.22) References 1. 2. 3. 4. * † Riley, W. F., and Sturges, L. D., Engineering Mechanics: Dynamics, 2nd Ed., Wiley, New York, 1996. Tipler, P. A., Physics, Worth, New York, 1982. Panton, R. L., Incompressible Flow, Wiley, New York, 1984. Dean, R. C., “On the Necessity of Unsteady Flow in Fluid Machines,” ASME Journal of Basic Engineering 81D; 24–28, March 1959. Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are not unique answers to these problems. Free ebooks ==> www.Ebook777.com 140 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. Section 3.2 F = ma along a Streamline 3.1 Obtain a photograph/image of a situation that can be analyzed by use of the Bernoulli equation. Print this photo and write a brief paragraph that describes the situation involved. 3.2 Air flows steadily along a streamline from point (1) to point (2) with negligible viscous effects. The following conditions are measured: At point (1) z1 = 2 m and p1 = 0 kPa; at point (2) z2 = 10 m, p2 = 20 N/m2, and V2 = 0. Determine the velocity at point (1). 3.3 Water flows steadily through the variable area horizontal pipe shown in Fig. P3.3. The centerline velocity is given by V = 10(1 + x) ˆi fts, where x is in feet. Viscous effects are neglected. (a) Determine the pressure gradient, ∂p∂x (as a function of x), needed to produce this flow. (b) If the pressure at section (1) is 50 psi, determine the pressure at (2) by (i) integration of the pressure gradient obtained in (a), (ii) application of the Bernoulli equation. ℓ = 3 ft x=0 V0 p0 Stagnation point x a ■ Figure P3.8 3.9 Consider a compressible liquid that has a constant bulk modulus. Integrate “F = ma” along a streamline to obtain the equivalent of the Bernoulli equation for this flow. Assume steady, inviscid flow. Section 3.3 F = ma Normal to a Streamline 3.11 Air flows along a horizontal, curved streamline with a 20 ft radius with a speed of 100 ft/s. Determine the pressure gradient normal to the streamline. (2) (1) Dividing streamline 3.10 Obtain a photograph/image of a situation in which Newton’s second law applied across the streamlines (as given by Eq. 3.12) is important. Print this photo and write a brief paragraph that describes the situation involved. V(x) Q the pressure p(x) for −∞ ≤ x ≤ −a. (c) Show from the result of part (b) that the pressure at the stagnation point (x = −a) is p0 + ρV 20 2, as expected from the Bernoulli equation. x 3.12 Water flows around the vertical two-dimensional bend with circular streamlines and constant velocity as shown in Fig. P3.12. If the pressure is 40 kPa at point (1), determine the pressures at points (2) and (3). Assume that the velocity profile is uniform as indicated. ■ Figure P3.3 3.4 What pressure gradient along the streamline, dpds, is required to accelerate water in a horizontal pipe at a rate of 30 ms2? g 3.5 At a given location the airspeed is 20 m/s and the pressure gradient along the streamline is 100 N/m3. Estimate the airspeed at a point 0.5 m farther along the streamline. 3.6 What pressure gradient along the streamline, dp/ds, is required to accelerate water upward in a vertical pipe at a rate of 30 ft/s2? What is the answer if the flow is downward? 4m (3) 2m 1m 3.7 The Bernoulli equation is valid for steady, inviscid, incompressible flows with constant acceleration of gravity. Consider flow on a planet where the acceleration of gravity varies with height so that g = g0 − cz, where g0 and c are constants. Integrate “F = ma” along a streamline to obtain the equivalent of the Bernoulli equation for this flow. ■ Figure P3.12 3.8 An incompressible fluid flows steadily past a circular cylinder as shown in Fig. P3.8. The fluid velocity along the dividing streamline (−∞ ≤ x ≤ −a) is found to be V = V0 (1 − a2x2 ), where a is the radius of the cylinder and V0 is the upstream velocity. (a) Determine the pressure gradient along this streamline. (b) If the upstream pressure is p0, integrate the pressure gradient to obtain 3.13 Water flows around the vertical two-dimensional bend with circular streamlines as is shown in Fig. P3.13. The pressure at point (1) is measured to be p1 = 25 psi and the velocity across section a–a is as indicated in the table. Calculate and plot the pressure across section a–a of the channel [p = p(z) for 0 ≤ z ≤ 2 ft]. (2) V = 10m/s (1) www.Ebook777.com Problems z (ft) V (ft/s) 0 0.2 0.4 0.6 0.8 1.0 0 8.0 14.3 20.0 19.5 15.6 1.2 1.4 1.6 1.8 2.0 8.3 6.2 3.7 2.0 0 141 3.17 At a given point on a horizontal streamline in flowing air, the static pressure is −2.0 psi (i.e., a vacuum) and the velocity is 150 ft/s. Determine the pressure at a stagnation point on that streamline. †3.18 A drop of water in a zero-g environment (as in the International Space Station) will assume a spherical shape as shown in Fig. P3.18a. A raindrop in the cartoons is typically drawn as in Fig. P3.18b. The shape of an actual raindrop is more nearly like that shown in Fig. 3.18c. Discuss why these shapes are as indicated. (a) (b) (c) ■ Figure P3.18 g 20 ft P1 = 25 psi 3.20 Air flows over the airfoil shown in Fig. P3.20. Sensors give the pressures shown at points A, B, C, and D. Find the air velocities just above points A, B, C, and D. The air density is 0.0020 slug/ft3. a (1) 3.19 When an airplane is flying 200 mph at 5000-ft altitude in a standard atmosphere, the air velocity at a certain point on the wing is 273 mph relative to the airplane. (a) What suction pressure is developed on the wing at that point? (b) What is the pressure at the leading edge (a stagnation point) of the wing? 2 ft V = V (z) z V0 = 100 mph a y V r x B C D ■ Figure P3.13 3.14 Water in a container and air in a tornado flow in horizontal circular streamlines of radius r and speed V as shown in Video V3.6 and Fig. P3.14. Determine the radial pressure gradient, ∂p/∂r, needed for the following situations: (a) The fluid is water with r = 3 in. and V = 0.8 ft/s. (b) The fluid is air with r = 300 ft and V = 200 mph. A p0 = 13.20 psia pA = 13.35 psia pB = 12.5 psia pC = 12.8 psia pD = 13.0 psia ■ Figure P3.20 3.21 Some animals have learned to take advantage of the Bernoulli effect without having read a fluid mechanics book. For example, a typical prairie dog burrow contains two entrances—a flat front door and a mounded back door as shown in Fig. P3.21. When the wind blows with velocity V0 across the front door, the average velocity across the back door is greater than V0 because of the mound. Assume the air velocity across the back door is 1.07V0. For a wind velocity of 6 m/s, what pressure differences, p1 − p2, are generated to provide a fresh airflow within the burrow? 1.07 V0 V0 ■ Figure P3.14 3.15 Air flows smoothly over the hood of your car and up past the windshield. However, a bug in the air does not follow the same path; it becomes splattered against the windshield. Explain why this is so. Section 3.5 Static, Stagnation, Dynamic, and Total Pressure 3.16 Obtain a photograph/image of a situation in which the concept of the stagnation pressure is important. Print this photo and write a brief paragraph that describes the situation involved. (2) (1) Q ■ Figure P3.21 †3.22 Estimate the pressure on your hand when you hold it in the stream of air coming from the air hose at a filling station. List all assumptions and show calculations. Warning: Do not try this experiment; it can be dangerous! 142 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 3.23 2013 Indianapolis 500 champion Tony Kanaan holds his †3.31 Estimate the pressure needed at the pumper truck in order hand out of his IndyCar while driving through still air with stan- to shoot water from the street level onto a fire on the roof of a dard atmospheric conditions. (a) For safety, the pit lane speed five-story building. List all assumptions and show all calculations. limit is 60 mph. At this speed, what is the maximum pressure 3.32 The tank shown in Fig. P3.32 contains air at atmospheric on his hand? (b) Back on the race track, what is the maximum pressure above the water surface. The velocity of the water flowing pressure when he is driving his IndyCar at 225 mph? (c) On the from the tank is 22.5 ft/sec. Determine the water level (h). straightaways, the IndyCar reaches speeds in excess of 235 mph. For this speed, is your solution method for parts (a) and (b) reasonable? Explain. D = 12 in. 3.24 What is the minimum height for an oil (SG = 0.75) manometer to measure airplane speeds up to 30 m/s at altitudes up to 1500 m? The manometer is connected to a Pitot-static tube as shown in Fig. P3.24. Air p = 14.7 psia V = 30 m/s h d2 V2 = 22.5 ft/sec ■ Figure P3.32 h SG = 0.75 3.33 Water flows from the faucet on the first floor of the building shown in Fig. P3.33 with a maximum velocity of 20 ft/s. For steady inviscid flow, determine the maximum water velocity from the basement faucet and from the faucet on the second floor (assume each floor is 12 ft tall). ■ Figure P3.24 3.25 A Pitot-static tube is used to measure the velocity of helium in a pipe. The temperature and pressure are 40 °F and 25 psia. A water manometer connected to the Pitot-static tube indicates a reading of 2.3 in. Determine the helium velocity. Is it reasonable to consider the flow as incompressible? Explain. 4 ft 3.26 A Bourdon-type pressure gage is used to measure the pressure from a Pitot tube attached to the leading edge of an airplane wing. The gage is calibrated to read in miles per hour at standard sea level conditions (rather than psi). If the airspeed meter indicates 150 mph when flying at an altitude of 10,000 ft, what is the true airspeed? V = 20 ft/s †3.27 Estimate the force of a hurricane strength wind against the side of your house. List any assumptions and show all calculations. 12 ft 3.28 A 40-mph wind blowing past your house speeds up as it flows up and over the roof. If elevation effects are negligible, determine (a) the pressure at the point on the roof where the speed is 60 mph if the pressure in the free stream blowing toward your house is 14.7 psia. Would this effect tend to push the roof down against the house, or would it tend to lift the roof? (b) Determine the pressure on a window facing the wind if the window is assumed to be a stagnation point. 3.29 (See The Wide World of Fluids article titled “Pressurized eyes,” Section 3.5.) Determine the air velocity needed to produce a stagnation pressure equal to 10 mm of mercury. Section 3.6.1 Free Jets 3.30 Water flows through a hole in the bottom of a large, open tank with a speed of 8 m/s. Determine the depth of water in the tank. Viscous effects are negligible. 8 ft 4 ft 4 ft ■ Figure P3.33 3.34 Laboratories containing dangerous materials are often kept at a pressure slightly less than ambient pressure so that contaminants can be filtered through an exhaust system rather than leaked through cracks around doors, etc. If the pressure in such a room is 0.1 in. of water below that of the surrounding rooms, with what velocity will air enter the room through an opening? Assume viscous effects are negligible. †3.35 The “supersoaker” water gun shown in Fig. P3.35 can shoot more than 30 ft in the horizontal direction. Estimate the minimum pressure, p1 , needed in the chamber in order to accomplish this. List all assumptions and show all calculations. 143 Problems 3.39 The piston shown in Fig. P3.39 is forcing 70 °F water out the exit at 12 ft/sec. The exit pressure has been measured as pe = 40 psig. Determine the force on the piston for a piston diameter Dp = 3.0 in. (1) F ■ Figure P3.35 V = 2 ft/sec 3.36 Streams of water from two tanks impinge upon each other as shown in Fig. P3.36. If viscous effects are negligible and point A is a stagnation point, determine the height h. h = 8.5 ft g pe = 40 psig h Free jets A p1 = 25 psi ■ Figure P3.39 Air 20 ft 8 ft ■ Figure P3.36 3.37 Several holes are punched into a tin can as shown in Fig. P3.37. Which of the figures represents the variation of the water velocity as it leaves the holes? Justify your choice. Section 3.6.2 Confined Flows 3.40 Obtain a photograph/image of a situation that involves a confined flow for which the Bernoulli and continuity equations are important. Print this photo and write a brief paragraph that describes the situation involved. 3.41 Air flows steadily through a horizontal 4-in.-diameter pipe and exits into the atmosphere through a 3-in.-diameter nozzle. The velocity at the nozzle exit is 150 ft/s. Determine the pressure in the pipe if viscous effects are negligible. 3.42 Figure P3.42 shows a tube for siphoning water from an aquarium. Determine the rate at which the water leaves the aquarium for the conditions shown. Is there an advantage to having the large-diameter section? The water flow is inviscid. (a) (b) ¼ in. I.D. tube (c) ■ Figure P3.37 3.38 Water flows from a pressurized tank, through a 6-in.-diameter pipe, exits from a 2-in.-diameter nozzle, and rises 20 ft above the nozzle as shown in Fig. P3.38. Determine the pressure in the tank if the flow is steady, frictionless, and incompressible. h = 5 ft 2 A = 6 ft ■ Figure P3.42 3.43 For the pipe enlargement shown in Fig. P3.43, the pressures at sections (1) and (2) are 56.3 and 58.2 psi, respectively. Determine the weight flowrate (lb/s) of the gasoline in the pipe. 20 ft 2 in. Air 2 ft 2.05 in. Q Gasoline 3.71 in. (1) 6 in. ■ Figure P3.38 (2) ■ Figure P3.43 144 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 1 3.44 A fire hose nozzle has a diameter of 18 in. According to some fire codes, the nozzle must be capable of delivering at least 250 gal/min. If the nozzle is attached to a 3-in.-diameter hose, what pressure must be maintained just upstream of the nozzle to deliver this flowrate? 3.45 Water flowing from the 0.75-in.-diameter outlet shown in Video V8.15 and Fig. P3.45 rises 2.8 in. above the outlet. Determine the flowrate. 60 °F water h D A 5 ft d = 1 in.-diameter plastic pipe B patm 2.8 in. ■ Figure P3.49 0.75 in. Q 3.50 Water (assumed inviscid and incompressible) flows steadily in the vertical variable-area pipe shown in Fig. P3.50. Determine the flowrate if the pressure in each of the gages reads 50 kPa. 2m ■ Figure P3.45 10 m p = 50 kPa 3.46 A fire hose has a nozzle outlet velocity of 30 mph. What is the maximum height the water can reach? 3.47 At what rate does oil (SG = 0.85) flow from the tank shown in Fig. P3.47? 1m Q D = 14 cm ■ Figure P3.50 13.5 cm 10.5 cm 3.51 Air is drawn into a wind tunnel used for testing automobiles as shown in Fig. P3.51. (a) Determine the manometer reading, h, when the velocity in the test section is 60 mph. Note that there is a 1-in. column of oil on the water in the manometer. (b) Determine the difference between the stagnation pressure on the front of the automobile and the pressure in the test section. dT = 3.0 cm Oil dB = 3.0 cm ■ Figure P3.47 Wind tunnel 3.48 Find the water height hB in tank B shown in Fig. P3.48 for steady-state conditions. 60 mph Fan Open h Water hA = 10.0 cm dA = 0.01 m ■ Figure P3.51 Tank A hB 1 in. Oil (SG = 0.9) dB = 0.02 m Tank B 3.52 Figure P3.52 shows a duct for testing a centrifugal fan. Air is drawn from the atmosphere (patm = 14.7 psia, Tatm = 70 °F). The inlet box is 4 ft × 2 ft. At section 1, the duct is 2.5 ft square. At section 2, the duct is circular and has a diameter of 3 ft. A water manometer in 2 in. water Inlet box ■ Figure P3.48 3.49 The pressure and average velocity at point A in the pipe shown in Fig. P3.49 are 16.0 psia and 4.0 ft/sec, respectively. Find the height h and the pressure and average velocity at point B. Fluid fills the 1-in.-diameter discharge pipe. Fan (1) ■ Figure P3.52 Throttling plug (2) Problems 145 the inlet box measures a static pressure of −2.0 in. of water. Calculate the volume flow rate of air into the fan and the average fluid velocity at both sections 1 and 2. Assume constant density. 50 kN/m2. Determine (a) the height h, to which the water rises, (b) the water velocity in the pipe, and (c) the pressure in the horizontal part of the pipe. 3.53 Natural gas (methane) flows from a 3-in.-diameter gas main, through a 1-in.-diameter pipe, and into the burner of a furnace at a rate of 100 ft3/hour. Determine the pressure in the gas main if the pressure in the 1-in. pipe is to be 6 in. of water greater than atmospheric pressure. Neglect viscous effects. 3.58 Water (assumed inviscid and incompressible) flows steadily with a speed of 10 ft/s from the large tank shown in Fig. P3.50. Determine the depth, H, of the layer of light liquid (specific weight = 50 lb/ft3 ) that covers the water in the tank. 3.54 Air flows radially outward between the two parallel circular plates shown in Fig. P3.54. The pressure at the outer radius Ro = 5.0 cm is atmospheric. Find the pressure at the inner radius Ri = 0.5 cm if the air density is constant, the air flow is inviscid, the volume flow rate is 4.0 L/s, and the plate spacing is h = 0.125 cm. 10 ft/s 50 lb/ft3 H Water po = patm = 101.3 kPa T = 20°C Vi h Ri Vo Ro g 4 ft 5 ft ■ Figure P3.58 3.59 Water flows through the pipe contraction shown in Fig. P3.59. For the given 0.2-m difference in the manometer level, determine the flowrate as a function of the diameter of the small pipe, D. ■ Figure P3.54 0.2 m 3.55 Repeat Problem 3.54 for air flowing radially inward. 3.56 Find the water mass flow rate at the nozzle outlet O shown in Fig. P3.56, and calculate the maximum height to which the water stream will rise. The water density is 1.9 slugs/ft3, and the flow is inviscid. 0.1 m Q D ■ Figure P3.59 h1 = 21 ft O 3.60 Carbon tetrachloride flows in a pipe of variable diameter with negligible viscous effects. At point A in the pipe the pressure and velocity are 20 psi and 30 ft/s, respectively. At location B the pressure and velocity are 23 psi and 14 ft/s. Which point is at the higher elevation and by how much? d = 1.0 in. h2 = 6 ft *3.61 Water flows from a 20-mm-diameter pipe with a flowrate Q as shown in Fig. P3.61. Plot the diameter of the water stream, A = 20 ft d, as a function of distance below the faucet, h, for values of 0 ≤ h ≤ 1 m and 0 ≤ Q ≤ 0.004 m3/s. Discuss the validity of the ■ Figure P3.56 one-dimensional assumption used to calculate d = d(h), noting, in 3.57 Water (assumed frictionless and incompressible) flows steadi- particular, the conditions of small h and small Q. ly from a large tank and exits through a vertical, constant diameter Q pipe as shown in Fig. P3.57. The air in the tank is pressurized to 2 20 mm h Pipe exit d 50 kN/m2 h 4m Air Water ■ Figure P3.57 ■ Figure P3.61 2m 3.62 A liquid stream directed vertically upward leaves a nozzle with a steady velocity V0 and cross-sectional area A0. Find the velocity V and cross-sectional area A as a function of the vertical position z. 3.63 Water leaves a pump at 200 kPa and a velocity of 12 m/s. It then enters a diffuser to increase its pressure to 250 kPa. What must be the ratio of the outlet area to the inlet area of the diffuser? 146 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 3.64 Water flows upward through a variable area pipe with a constant flowrate, Q, as shown in Fig. P3.64. If viscous effects are negligible, determine the diameter, D(z), in terms of D1 if the pressure is to remain constant throughout the pipe. That is, p(z) = p1. 3.68 A 50-mm-diameter plastic tube is used to siphon water from the large tank shown in Fig. P3.68. If the pressure on the outside of the tube is more than 35 kPa greater than the pressure within the tube, the tube will collapse and siphon will stop. If viscous effects are negligible, determine the minimum value of h allowed without the siphon stopping. 2m D(z) z 4m D1 h (1) ■ Figure P3.68 Q ■ Figure P3.64 3.65 The circular stream of water from a faucet is observed to taper from a diameter of 20 mm to 10 mm in a distance of 50 cm. Determine the flowrate. 3.66 Water is siphoned from the tank shown in Fig. P3.66. The water barometer indicates a reading of 30.2 ft. Determine the maximum value of h allowed without cavitation occurring. Note that the pressure of the vapor in the closed end of the barometer equals the vapor pressure. 3.69 Water is siphoned from the tank shown in Fig. P3.69. Determine the flowrate from the tank and the pressure at points (1), (2), and (3) if viscous effects are negligible. 2-in.-diameter hose 2 ft (3) 3 ft 8 ft Closed end (2) 3-in. diameter 30.2 ft (1) ■ Figure P3.69 6 ft 3.70 Redo Problem 3.69 if a 1-in.-diameter nozzle is placed at the end of the tube. 3.71 Water exits a pipe as a free jet and flows to a height h above the exit plane as shown in Fig. P3.71. The flow is steady, incompressible, and frictionless. (a) Determine the height h. (b) Determine the velocity and pressure at section (1). h 5-in. diameter ■ Figure P3.66 h V = 16 ft/s 3.67 Water is siphoned from a tank as shown in Fig. P3.67. Determine the flowrate and the pressure at point A, a stagnation point. 0.04-m diameter 6-in. diameter 8 ft (1) 3m 4-in. diameter ■ Figure P3.71 A ■ Figure P3.67 3.72 Water flows steadily from a large, closed tank as shown in Fig. P3.72. The deflection in the mercury manometer is 1 in. and viscous effects are negligible. (a) Determine the volume flowrate. Problems (b) Determine the air pressure in the space above the surface of the water in the tank. inviscid and incompressible and the flowing fluid is (a) water, (b) gasoline, or (c) air at standard conditions. Q Air 0.09-m diameter 0.05 m 1-ft diameter 8 ft 147 3-in. diameter 20 mm 10 mm 1 in. ■ Figure P3.76 Mercury ■ Figure P3.72 3.73 Carbon dioxide flows at a rate of 1.5 ft3/s from a 3-in. pipe in which the pressure and temperature are 20 psi (gage) and 120 °F into a 1.5-in. pipe. If viscous effects are neglected and incompressible conditions are assumed, determine the pressure in the smaller pipe. 3.77 Water flows steadily with negligible viscous effects through the pipe shown in Fig. P3.77. It is known that the 4-in.-diameter section of thin-walled tubing will collapse if the pressure within it becomes less than 10 psi below atmospheric pressure. Determine the maximum value that h can have without causing collapse of the tubing. 3.74 Oil of specific gravity 0.83 flows in the pipe shown in Fig. P3.74. If viscous effects are neglected, what is the flowrate? 4 ft 4-in.-diameter thin-walled tubing h 6 in. 4 in. ■ Figure P3.77 Water SG = 0.83 3 in. Q 4 in. ■ Figure P3.74 3.75 Water flows steadily through the variable area pipe shown in Fig. P3.75 with negligible viscous effects. Determine the manometer reading, H, if the flowrate is 0.4 m3/s and the density of the manometer fluid is 500 kg/m3. Density = 500 kg/m3 H 3.78 Helium flows through a 0.30-m-diameter horizontal pipe with a temperature of 20 °C and a pressure of 200 kPa (abs) at a rate of 0.30 kg/s. If the pipe reduces to 0.25-m-diameter, determine the pressure difference between these two sections. Assume incompressible, inviscid flow. 3.79 Water is pumped from a lake through an 8-in. pipe at a rate of 10 ft3s. If viscous effects are negligible, what is the pressure in the suction pipe (the pipe between the lake and the pump) at an elevation 6 ft above the lake? 3.80 Air is drawn into a small open-circuit wing tunnel as shown in Fig. P3.80. Atmospheric pressure is 98.7 kPa (abs) and the temperature is 27 °C. If viscous effects are negligible, determine the pressure at the stagnation point on the nose of the airplane. Also determine the manometer reading, h, for the manometer attached to the static pressure tap within the test section of the wind tunnel if the air velocity within the test section is 50 m/s. Test section Inlet nozzle Area = 0.05 m2 Area = 0.07 m2 Diffuser Q Fan ■ Figure P3.75 Inlet 3.76 The specific gravity of the manometer fluid shown in Fig. P3.76 is 1.07. Determine the volume flowrate, Q, if the flow is ■ Figure P3.80 h Water 148 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 3.81 Air flows through the device shown in Fig. P3.81. If the flowrate is large enough, the pressure within the constriction will be low enough to draw the water up into the tube. Determine the flowrate, Q, and the pressure needed at section (1) to draw the water into section (2). Neglect compressibility and viscous effects. Free jet 25 mm (2) (1) 50 mm Air 3.85 Air flows steadily through a converging–diverging rectangular channel of constant width as shown in Fig. 3.85 and Video V3.10. The height of the channel at the exit and the exit velocity are H0 and V0 respectively. The channel is to be shaped so that the distance, d, that water is drawn up into tubes attached to static pressure taps along the channel wall is linear with distance along the channel. That is d = (dmax/L) x, where L is the channel length and dmax is the maximum water depth (at the minimum channel height: x = L). Determine the height, H(x), as a function of x and the other important parameters. Q x=L x L 0.3 m x=0 V0 50 mm Water Q Air H(x) H0 ■ Figure P3.81 3.82 Water flows steadily from the large open tank shown in Fig. 3.82. If viscous effects are negligible, determine (a) the flowrate, Q, and (b) the manometer reading, h. dmax d Water ■ Figure P3.85 2m Q 0.10 m 0.08 m 4m h *3.86 Water flows from a large tank and through a pipe of variable area as shown in Fig. 3.86. The area of the pipe is given by A = A0[1 − x(1 − x/ℓ)/2ℓ], where A0 is the area at the beginning (x = 0) and end (x = ℓ) of the pipe. Plot graphs of the pressure within the pipe as a function of distance along the pipe for water depths of h = 1, 4, 10, and 25 m. Mercury ■ Figure P3.82 3.83 Water from a faucet fills a 16-oz glass (volume = 28.9 in.3) in 20 s. If the diameter of the jet leaving the faucet is 0.60 in., what is the diameter of the jet when it strikes the water surface in the glass which is positioned 14 in. below the faucet? h ℓ 3.84 The nozzle shown in Fig. P3.84 has two water manometers to indicate the static pressures at sections 1 and 2. The diameters D1 and D2 are 8 in. and 2 in., respectively. Air flows through the nozzle, and the air and water temperatures are 60 °F. Find the air volume flow rate. Assume constant density and inviscid flow. Neglect elevation changes. Free jet x x=0 ■ Figure P3.86 patm = 14.7 psia 3.87 If viscous effects are neglected and the tank is large, determine the flowrate from the tank shown in Fig. P3.87. D2 D1 Oil, SG = 0.81 2m h2 = 8.0 in. h1 = 0.7 m 1 in. 32 50-mm diameter ■ Figure P3.84 ■ Figure P3.87 Water Problems 3.88 Water flows steadily downward in the pipe shown in Fig. P.3.88 with negligible losses. Determine the flowrate. Open Oil SG = 0.7 149 3.91 Water flows into the sink shown in Fig. P3.91 and Video V5.1 at a rate of 2 gal/min. If the drain is closed, the water will eventually flow through the overflow drain holes rather than over the edge of the sink. How many 0.4-in.-diameter drain holes are needed to ensure that the water does not overflow the sink? Neglect viscous effects. 1 in. 1.5 m Q = 2 gal/min 0.4-in.-diameter holes 1.2 m Stopper 2m 1m ■ Figure P3.88 ■ Figure P3.91 3.89 Water flows steadily from a large open tank and discharges into the atmosphere though a 3-in.-diameter pipe as shown in Fig. P3.89. Determine the diameter, d, in the narrowed section of the pipe at A if the pressure gages at A and B indicate the same pressure. 3.92 What pressure, p1, is needed to produce a flowrate of 0.09 ft3/s from the tank shown in Fig. P3.92? p1 B 9 ft Air 3-in. diameter Gasoline 2.0 ft Salt water SG = 1.1 3.6 ft 8 ft A 16 ft diameter = d 0.06-ft diameter ■ Figure P3.89 ■ Figure P3.92 3.90 Water flows from a large tank as shown in Fig. P3.90. Atmospheric pressure is 14.5 psia, and the vapor pressure is 1.60 psia. If viscous effects are neglected, at what height, h, will cavitation begin? To avoid cavitation, should the value of D1 be increased or decreased? To avoid cavitation, should the value of D2 be increased or decreased? Explain. 3.93 The vent on the tank shown in Fig. P3.93 is closed and the tank pressurized to increase the flowrate. What pressure, p1, is needed to produce twice the flowrate of that when the vent is open? p1 Vent 4 ft h D3 = 4 in. Air 10 ft Water Q D1 = 1 in. ■ Figure P3.90 D2 = 2 in. ■ Figure P3.93 150 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 3.94 A sump pump is submerged in 60 °F ordinary water that vaporizes at a pressure of 0.256 psia. The pump inlet has an inside diameter of 2.067 in. and is 15 ft below the water surface. Find the maximum possible flow rate before cavitation (vaporization) occurs. 3.98 JP-4 fuel (SG = 0.77) flows through the Venturi meter shown in Fig. P3.98 with a velocity of 15 ft/s in the 6-in. pipe. If viscous effects are negligible, determine the elevation, h, of the fuel in the open tube connected to the throat of the Venturi meter. 3.95 Water is siphoned from a large tank and discharges into the atmosphere through a 2-in.-diameter tube as shown in Fig. P3.95. The end of the tube is 3 ft below the tank bottom, and viscous effects are negligible. (a) Determine the volume flowrate from the tank. (b) Determine the maximum height, H, over which the water can be siphoned without cavitation occurring. Atmospheric pressure is 14.7 psia, and the water vapor pressure is 0.26 psia. 6 ft 6 in. JP-4 fuel 8 in. 20° 4 in. H h 6 in. 2-in. diameter 9 ft V = 15 ft/s ■ Figure P3.98 3 ft 3.99 Water, considered an inviscid, incompressible fluid, flows steadily as shown in Fig. P3.99. Determine h. ■ Figure P3.95 h 3.96 Water flows in the system shown in Fig. P3.96. Assume frictionless flow. (a) Calculate the rate Q at which water must be added at the inlet to maintain the 16-ft height. (b) Calculate the height h in feet of water in the static-pressure tube. Air Q = 4 ft3/s Water 0.5-ft diameter 1-ft diameter 3 ft 1 16 ft A = 0.4 ft 2 h A = 0.1 ft 2 ■ Figure P3.99 Q 2 3 4 ■ Figure P3.96 3.100 Determine the flowrate through the submerged orifice shown in Fig. P3.100 if the contraction coefficient is Cc = 0.63. 3.97 Water flows steadily from the pipe shown in Fig. P3.97 with negligible viscous effects. Determine the maximum flowrate if the water is not to flow from the open vertical tube at A. A 6 ft 4 ft 3-in. diameter 3 ft 2 ft ■ Figure P3.100 Q 0.15-ft diameter 0.1-ft diameter End of pipe ■ Figure P3.97 3.101 The water clock (clepsydra) shown in Fig. P3.101 is an ancient device for measuring time by the falling water level in a large glass container. The water slowly drains out through a small hole in the bottom. Determine the approximate shape R (z) of a container of circular cross section required for the water level to Problems 151 fall at a constant rate of 5 cm/hr if the drain hole is 1.5 mm in *3.104 A spherical tank of diameter D has a drain hole of diameter diameter. What size container (h and R0) permits the clock to run d at its bottom. A vent at the top of the tank maintains atmospheric pressure at the liquid surface within the tank. The flow is quasifor 24 hr without refilling? steady and inviscid and the tank is full of water initially. Determine the water depth as a function of time, h = h(t), and plot graphs of R0 h(t) for tank diameters of 1, 5, 10, and 20 ft if d = 1 in. Water R (z) level z h ■ Figure P3.105 ■ Figure P3.101 3.102 A long water trough of triangular cross section is formed from two planks as is shown in Fig. P3.102. A gap of 0.1 in. remains at the junction of the two planks. If the water depth initially was 2 ft, how long a time does it take for the water depth to reduce to 1 ft? 90° 2 ft †3.105 A small hole develops in the bottom of the stationary rowboat shown in Fig. P3.105. Estimate the amount of time it will take for the boat to sink. List all assumptions and show all calculations. 3.106 When the drain plug is pulled, water flows from a hole in the bottom of a large, open cylindrical tank. Show that if viscous effects are negligible and if the flow is assumed to be quasisteady, then it takes 3.41 times longer to empty the entire tank than it does to empty the first half of the tank. Explain why this is so. 3.107 Someone siphoned 15 gal of gasoline from a gas tank in the middle of the night. The gas tank measures 12 in. wide × 24 in. long × 18 in. high and was full when the thief started. If the siphon1 ing tube has an inside diameter of 2 in., find the minimum amount of time needed to siphon the 15 gal from the tank. Assume that at any instant of time the steady-state equations are adequate to predict the gasoline velocity in the siphon tube. Also assume that the end of the siphon tube outside the gas tank is at the same level as the bottom of the tank. You may consider the gasoline to be inviscid. *3.108 The surface area, A, of the pond shown in Fig. P3.108 varies with the water depth, h, as shown in the table. At time t = 0 a valve is opened and the pond is allowed to drain through a pipe of diameter D. If viscous effects are negligible and quasisteady ■ Figure P3.102 conditions are assumed, plot the water depth as a function of time from when the valve is opened (t = 0) until the pond is drained for 3.103 Pop (with the same properties as water) flows from a pipe diameters of D = 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 ft. Assume 4-in.-diameter pop container that contains three holes as shown h = 18 ft at t = 0. in Fig. P3.103 (see Video 3.9). The diameter of each fluid stream Area A is 0.15 in., and the distance between holes is 2 in. If viscous effects are negligible and quasi-steady conditions are assumed, determine the time at which the pop stops draining from the top h 3 ft hole. Assume the pop surface is 2 in. above the top hole when t = 0. Compare your results with the time you measure from the video. 0.1 in. D ■ Figure P3.108 Surface at t = 0 2 in. 2 in. 2 in. 4 in. ■ Figure P3.103 0.15 in. h (ft) A [acres (1 acre = 43,560 ft2)] 0 2 4 6 8 10 12 14 16 18 0 0.3 0.5 0.8 0.9 1.1 1.5 1.8 2.4 2.8 152 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation 3.109 Water flows through a horizontal branching pipe as shown in Fig. P3.109. Determine the pressure at section (3). A2 = 0.07 ft2 p2 = 5.0 psi (2) V3 A3 = 0.07 m2 (3) (2) A3 = 0.2 ft2 V3 = 20 ft /s (1) (3) A1 = 1 ft2 Q1 = 10 ft3/s p1 = 10 psi V2 p2 = 350 kPa A2 = 0.02 m2 (4) ■ Figure P3.112 (1) 3.113 Water flows from a large tank through a large pipe that splits into two smaller pipes as shown in Fig. P3.113. If viscous effects are negligible, determine the flowrate from the tank and the pressure at point (1). V1 = 4 m/s p1 = 400 kPa A1 = 0.1 m2 ■ Figure P3.109 3.110 The “wye” fitting shown in Fig. P.3.110 lies in a horizontal plane. The fitting splits the inlet flow into two equal parts. At section 1, the water velocity is 12 ft/sec and the pressure is 20 psig. Calculate the water pressure at sections 2 and 3. Assume inviscid flow and a water temperature of 60 °F. 3m 7m 0.03-m diameter 0.05-m diameter ½Q1 (1) 0.02-m diameter p2 D2 D1 ■ Figure P3.113 D 2 = 3 in. 3.114 A gutter running along the side of a house is 6 in. wide and 40 ft long. During a hard downpour, water is 1 in. deep in the gutter. The gutter has only one downspout, and it is 3 in. in diameter. What is the velocity of the water entering the downspout? The pressure at the downspout entrance is atmospheric pressure. Q1 1 D 1 = 6 in. D3 3.115 Air, assumed incompressible and inviscid, flows into the outdoor cooking grill through nine holes of 0.40-in. diameter as shown in Fig. P3.115. If a flowrate of 40 in.3/s into the grill is required to maintain the correct cooking conditions, determine the pressure within the grill near the holes. D 3 = 4 in. p3 ½Q1 ■ Figure P3.110 3.111 Water flows through the branching pipe shown in Fig. P3.111. If viscous effects are negligible, determine the pressure at section (2) and the pressure at section (3). Q1 = 1 m3/s A1 = 0.1 m2 p1 = 300 kPa z1 = 0 (1) A3 = 0.035 m2 z3 = 10 m 9 holes, each 0.40-in. diameter (3) V2 = 14 m/s A2 = 0.03 m2 z2 = 0 (2) ■ Figure P3.111 3.112 Water flows through the horizontal branching pipe shown in Fig. P3.112 at a rate of 10 ft3/s. If viscous effects are negligible, determine the water speed at section (2), the pressure at section (3), and the flowrate at section (4). ■ Figure P3.115 3.116 An air cushion vehicle is supported by forcing air into the chamber created by a skirt around the periphery of the vehicle as shown in Fig. P3.116. The air escapes through the 3-in. clearance between the lower end of the skirt and the ground (or water). Assume the vehicle weighs 10,000 lb and is essentially Problems rectangular in shape, 30 by 65 ft. The volume of the chamber is large enough so that the kinetic energy of the air within the chamber is negligible. Determine the flowrate, Q, needed to support the vehicle. If the ground clearance were reduced to 2 in., what flowrate would be needed? If the vehicle weight were reduced to 5000 lb and the ground clearance maintained at 3 in., what flowrate would be needed? 3.119 Figure P3.119 shows two tall towers. Air at 10 °C is blowing toward the two towers at V0 = 30 km/hr. If the two towers are 10 m apart and half the air flow approaching the two towers passes between them, find the minimum air pressure between the two towers. Assume constant air density, inviscid flow, and steady-state conditions. The atmospheric pressure is 101 kPa. d = 10 m Q Fan 153 Vehicle Skirt D = 60 m D = 60 m 3 in. ■ Figure P3.116 Top view 3.117 Water flows from the pipe shown in Fig. P3.117 as a free jet and strikes a circular flat plate. The flow geometry shown is axisymmetrical. Determine the flowrate and the manometer reading, H. H 0.1-m diameter V Elevation view 0.4 mm ■ Figure P3.119 0.2 m 3.120 Water flows steadily from a nozzle into a large tank as shown in Fig. P3.120. The water then flows from the tank as a jet of diameter d. Determine the value of d if the water level in the tank remains constant. Viscous effects are negligible. 0.01-m diameter Pipe Q ■ Figure P3.117 4 ft 0.1-ft diameter 3.118 A conical plug is used to regulate the airflow from the pipe shown in Fig. P3.118. The air leaves the edge of the cone with a uniform thickness of 0.02 m. If viscous effects are negligible and the flowrate is 0.50 m3/s, determine the pressure within the pipe. 0.15-ft diameter 3 ft V 1 ft Free jet Pipe 0.20 m Q = 0.50 m3/s 0.23 m Cone 0.02 m V ■ Figure P3.118 d ■ Figure P3.120 3.121 A small card is placed on top of a spool as shown in Fig. P3.121. It is not possible to blow the card off the spool by blowing air through the hole in the center of the spool. The harder one blows, the harder the card “sticks” to the spool. In fact by blowing hard enough it is possible to keep the card against the spool with the spool turned upside down. Give this experiment a try. (Note: It may be necessary to use a thumb tack to prevent the card from sliding from the spool.) Explain this phenomenon. 154 Chapter 3 ■ Elementary Fluid Dynamics—The Bernoulli Equation Card Section 3.6.3 Flowrate Measurement 3.125 Obtain a photograph/image of a situation that involves some type of flowmeter. Print this photo and write a brief paragraph that describes the situation involved. 3.126 A Venturi meter with a minimum diameter of 3 in. is to be used to measure the flowrate of water through a 4-in.-diameter pipe. Determine the pressure difference indicated by the pressure gage attached to the flowmeter if the flowrate is 0.5 ft3/s and viscous effects are negligible. Spool Q 3.127 Determine the flowrate through the Venturi meter shown in Fig. P3.127 if ideal conditions exist. ■ Figure P3.121 p1 = 735 kPa 3.122 Observations show that it is not possible to blow the table tennis ball from the funnel shown in Fig. P3.122a. In fact, the ball can be kept in an inverted funnel, Fig. P3.122b, by blowing through it. The harder one blows through the funnel, the harder the ball is held within the funnel. Try this experiment on your own. Explain this phenomenon. Q p2 = 550 kPa 19 mm 31 mm γ = 9.1 kN/m3 ■ Figure P3.127 Q 3.128 For what flowrate through the Venturi meter of Problem 3.127 will cavitation begin if p1 = 275 kPa gage, atmospheric pressure is 101 kPa (abs), and the vapor pressure is 3.6 kPa (abs)? 3.129 What diameter orifice hole, d, is needed if under ideal conditions the flowrate through the orifice meter of Fig. P3.129 is to be 30 gal/min of seawater with p1 − p2 = 2.37 lb/in.2? The contraction coefficient is assumed to be 0.63. Q (a) p1 (b) ■ Figure P3.122 Q 3.123 Water flows down the sloping ramp shown in Fig. P3.123 with negligible viscous effects. The flow is uniform at sections (1) and (2). For the conditions given show that three solutions for the downstream depth, h2, are obtained by use of the Bernoulli and continuity equations. However, show that only two of these solutions are realistic. Determine these values. V1 = 10 ft/s h1 = 1 ft h2 H = 2 ft p2 d 2-in. diameter ■ Figure P3.129 3.130 A weir (see Video V10.13) of trapezoidal cross section is used to measure the flowrate in a channel as shown in Fig. P3.130. If the flowrate is Q0 when H = ℓ/2, what flowrate is expected when H = ℓ? V2 ■ Figure P3.123 H 3.124 Water flows in a rectangular channel that is 2.0 m wide as shown in Fig. P3.124. The upstream depth is 70 mm. The water surface rises 40 mm as it passes over a portion where the channel bottom rises 10 mm. If viscous effects are negligible, what is the flowrate? 30° ℓ ■ Figure P3.130 100 mm Q 70 mm ■ Figure P3.124 10 mm 3.131 The flowrate in a water channel is sometimes determined by use of a device called a Venturi flume. As shown in Fig. P3.131, this device consists simply of a bump on the bottom of the channel. If the water surface dips a distance of 0.07 m for the conditions shown, what is the flowrate per width of the channel? Assume the velocity is uniform and viscous effects are negligible. Problems 0.07 m V1 Flight speed, km/hr V2 1.2 m 155 Aircraft Cruise Landing approach Boeing 787 F-22 fighter 913 1960 214 250 0.2 m ■ Figure P3.131 3.132 Water flows under the inclined sluice gate shown in Fig. P3.132. Determine the flowrate if the gate is 8 ft wide. 3.138 A meteorologist uses a Pitot-static tube to measure the wind speed in a tornado. Based on the damage caused by the storm, the tornado is rated as EF5 on the Enhanced Fujita Scale. This means that the wind speed is estimated to be in the range of 261 to 318 mph. Is it reasonable to use the incompressible Pitot-tube equation (Eq. 3.16) to determine the actual wind speed, or must compressible effects to taken into account? Explain. ■ 30° 6 ft 1.6 ft 1 ft ■ Figure P3.132 Section 3.7 The Energy Line and the Hydraulic Grade Line 3.133 Water flows in a vertical pipe of 0.15-m diameter at a rate of 0.2 m3/s and a pressure of 200 kPa at an elevation of 25 m. Determine the velocity head and pressure head at elevations of 20 and 55 m. 3.134 Draw the energy line and the hydraulic grade line for the flow of Problem 3.90. 3.135 Draw the energy line and hydraulic grade line for the flow shown in Problem 3.77. Section 3.8 Restrictions on Use of the Bernoulli Equation 3.136 Obtain a photograph/image of a flow in which it would not be appropriate to use the Bernoulli equation. Print this photo and write a brief paragraph that describes the situation involved. 3.137 The following table lists typical flight speeds for two aircraft. For which of these conditions would it be reasonable to use the incompressible Bernoulli equation to study the aerodynamics associated with their flight? Explain. Lifelong Learning Problems 3.1 LL The concept of the use of a Pitot-static tube to measure the airspeed of an airplane is rather straightforward. However, the design and manufacture of reliable, accurate, inexpensive Pitotstatic tube airspeed indicators is not necessarily simple. Obtain information about the design and construction of modern Pitotstatic tubes. Summarize your findings in a brief report. 3.2 LL In recent years damage due to hurricanes has been significant, particularly in the southeastern United States. The low barometric pressure, high winds, and high tides generated by hurricanes can combine to cause considerable damage. According to some experts, in the coming years hurricane frequency may increase because of global warming. Obtain information about the fluid mechanics of hurricanes. Summarize your findings in a brief report. 3.3 LL Orifice, nozzle, or Venturi flowmeters have been used for a long time to accurately measure the flowrate in pipes. However, recently there have been several new concepts suggested or used for such flowrate measurements. Obtain information about new methods to obtain pipe flowrate information. Summarize your findings in a brief report. 3.4 LL Ultra-high-pressure, thin jets of liquids can be used to cut various materials ranging from leather to steel and beyond. Obtain information about new methods and techniques proposed for liquid jet cutting and investigate how they may alter various manufacturing processes. Summarize your findings in a brief report. 4 Fluid Kinematics Learning Objectives After completing this chapter, you should be able to: ■ discuss the differences between the Eulerian and Lagrangian descriptions of fluid motion. ■ identify various flow characteristics based on the velocity field. ■ determine the streamline pattern and acceleration field given a velocity field. ■ discuss the differences between a system and a control volume. ■ apply the Reynolds transport theorem and the material derivative. In this chapter we will discuss various aspects of fluid motion without being concerned with the forces necessary to produce the motion. That is, we will consider the kinematics of the motion—the velocity and acceleration of the fluid, and the description and visualization of its motion. The analysis of the specific forces necessary to produce the motion (the dynamics of the motion) will be discussed in detail in the following chapters. A wide variety of useful information can be gained from a thorough understanding of fluid kinematics. Such an understanding of how to describe and observe fluid motion is an essential step to the complete understanding of fluid dynamics. V4.1 Streaklines VIDEO 4.1 The Velocity Field V u v z ^ j ^ k ^ x 156 i y w Fluids flow. That is, there is a net motion of molecules from one point in space to another as a function of time. As is discussed in Chapter 1, a typical portion of fluid contains so many molecules that it becomes totally unrealistic (except in special cases) for us to attempt to account for the motion of individual molecules. Rather, we employ the continuum hypothesis and consider fluids to be made up of fluid particles that interact with each other and with their surroundings. Although it is very tiny, each particle contains numerous molecules. Thus, we can describe the flow of a fluid in terms of the motion of fluid particles rather than individual molecules. This motion can be described in terms of the velocity and acceleration of the fluid particles. The infinitesimal particles of a fluid are tightly packed together (as is implied by the continuum assumption). Thus, at a given instant, a description of any fluid property (such as density, pressure, velocity, and acceleration) may be given as a function of the fluid’s location. This representation of fluid parameters as functions of the spatial coordinates is termed a field representation of the flow. Of course, the specific field representation may be different at different times, so that to describe a fluid flow we must determine the various parameters not only 4.1 The Velocity Field 157 z Particle path Particle A at time t rA(t) VIDEO Particle A at time t + δ t rA(t + δ t) V4.2 Velocity field y x ■ Figure 4.1 Particle location in terms of its position vector. as a function of the spatial coordinates (x, y, z, for example) but also as a function of time, t. Thus, to completely specify the temperature, T, in a room we must specify the temperature field, T = T (x, y, z, t), throughout the room (from floor to ceiling and wall to wall) at any time of the day or night. Shown in the margin figure is one of the most important fluid variables, the velocity vector,   V = u(x, y, z, t)iˆ + υ(x, y, z, t)jˆ + w(x, y, z, t)kˆ Flow (Photograph courtesy of Alvaro Valencia) VIDEO V4.3 Cylindervelocity vectors where u, υ, and w are the x, y, and z components of the velocity vector. By definition, the velocity of a particle is the time rate of change of the position vector for that particle. As is illustrated in Fig. 4.1, the position of particle A relative to the coordinate system is given by its position vector, rA, which (if the particle is moving) is a function of time. The time derivative of this position gives the velocity of the particle, drAdt = VA. By writing the velocity for all of the particles, we can obtain the field description of the velocity vector V = V(x, y, z, t). The figure in the margin shows the velocity field (i.e., velocity vectors) at a given instant for flow past two square bars. It is possible to obtain much qualitative and quantitative information for complex flows by using plots such as this. Because the velocity is a vector, it has both a direction and a magnitude. The magnitude of V, denoted V =  V = (u2 + υ2 + w2 ) 12, is the speed of the fluid. (It is very common in practical situations to call V velocity rather than speed, i.e., “the velocity of the fluid is 12 ms.”) As is discussed in the next section, a change in velocity results in an acceleration. This acceleration may be due to a change in speed and/or direction. THE WIDE WORLD OF FLUIDS Follow those particles Superimpose two photographs of a The flow is illuminated with a light sheet from a double-pulsed bouncing ball taken a short time apart and draw an arrow laser. A digital camera captures both light pulses on the same between the two images of the ball. This arrow represents an image frame, allowing the movement of the particles to be approximation of the velocity (displacement/time) of the ball. The tracked. By using appropriate computer software to carry out a particle image velocimeter (PIV) uses this technique to provide pixel-by-pixel interrogation of the double image, it is possible to the instantaneous velocity field for a given cross section of a track the motion of the particles and determine the two compo- flow. The flow being studied is seeded with numerous micron- nents of velocity in the given cross section of the flow. By using sized particles that are small enough to follow the flow yet big two cameras in a stereoscopic arrangement, it is possible to enough to reflect enough light to be captured by the camera. determine all three components of velocity. (See Problem 4.18.) 158 Chapter 4 ■ Fluid Kinematics EXAMPLE 4.1 Velocity Field Representation GIVEN A velocity field is given by V = (V0 ℓ) (−xiˆ + yjˆ ) where V0 and ℓ are constants. FIND At what location in the flow field is the speed equal to V0? Make a sketch of the velocity field for x ≥ 0 by drawing arrows representing the fluid velocity at representative locations. SOLUTION The x, y, and z components of the velocity are given by u = −V0 xℓ, υ = V0 yℓ, and w = 0 so that the fluid speed, V, is V = (u2 + υ2 + w2 ) 12 = V0 2 (x + y2 ) 12 ℓ (1) The speed is V = V0 at any location on the circle of radius ℓ centered at the origin [(x2 + y2 ) 12 = ℓ] as shown in Fig. E4.1a. (Ans) The direction of the fluid velocity relative to the x axis is given in terms of θ = arctan (υu) as shown in Fig. E4.1b. For this flow tan θ = V0 yℓ y υ = = u −x −V0 xℓ Thus, along the x axis (y = 0) we see that tan θ = 0, so that θ = 0° or θ = 180°. Similarly, along the y axis (x = 0) we obtain tan θ = ±∞ so that θ = 90° or θ = 270°. Also, for y = 0 we find V = (−V0 xℓ)iˆ, while for x = 0 we have V = (V0yℓ)jˆ, indicating (if V0 > 0) that the flow is directed away from the origin along the y axis and toward the origin along the x axis as shown in Fig. E4.1a. By determining V and θ for other locations in the x–y plane, the velocity field can be sketched as shown in the figure. For example, on the line y = x the velocity is at a 45° angle relative to the x axis (tan θ = υu = −yx = −1). At the origin x = y = 0 so that V = 0. This point is a stagnation point. The farther from the origin the fluid is, the faster it is flowing (as seen from Eq. 1). By careful consideration of the velocity field it is possible to determine considerable information about the flow. COMMENT The velocity field given in this example approxi- mates the flow in the vicinity of the center of the sign shown in Fig. E4.1c. When wind blows against the sign, some air flows over the sign, some under it, producing a stagnation point as indicated. 2V0 y 2ℓ V0 ℓ V θ v u (b) V0/2 0 2V0 V0 x y −ℓ V0 V=0 −2ℓ 2V0 (c) (a) ■ Figure E4.1 x 4.1 The Velocity Field 159 4.1.1 Eulerian and Lagrangian Flow Descriptions Flow (Courtesy of Chris Griffin) Either Eulerian or Lagrangian methods can be used to describe flow fields. There are two general approaches for analyzing fluid mechanics problems (or problems in other branches of the physical sciences, for that matter). The first method, called the Eulerian method, uses the field concept introduced above. In this case, the fluid motion is given by completely prescribing the necessary properties (pressure, density, velocity, etc.) as functions of space and time. From this method we obtain information about the flow in terms of what happens at fixed points in space as the fluid flows through those points. A typical Eulerian representation of the flow is shown by the figure in the margin, which involves flow past an airfoil at angle of attack. The pressure field is indicated by using a contour plot showing lines of constant pressure, with gray shading indicating the intensity of the pressure. The second method, called the Lagrangian method, involves following individual fluid particles as they move about and determining how the fluid properties associated with these particles change as a function of time. That is, the fluid particles are “tagged” or identified, and their properties determined as they move. The difference between the two methods of analyzing fluid flow problems can be seen in the example of smoke discharging from a chimney, as is shown in Fig. 4.2. In the Eulerian method one may attach a temperature-measuring device to the top of the chimney (point 0) and record the temperature at that point as a function of time. At different times there are different fluid particles passing by the stationary device. Thus, one would obtain the temperature, T, for that location (x = x0, y = y0, and z = z0 ) as a function of time. That is, T = T(x0, y0, z0, t). The use of numerous temperature-measuring devices fixed at various locations would provide the temperature field, T = T(x, y, z, t). The temperature of a specific particle as a function of time would not be known unless the location of the particle were known as a function of time. In the Lagrangian method, one would attach the temperature-measuring device to a particular fluid particle (particle A) and record that particle’s temperature as it moves about. Thus, one would obtain that particle’s temperature as a function of time, TA = TA (t). The use of many such measuring devices moving with various fluid particles would provide the temperature of these fluid particles as a function of time. The temperature would not be known as a function of position unless the location of each particle were known as a function of time. If enough information in Eulerian form is available, Lagrangian information can be derived from the Eulerian data—and vice versa. Example 4.1 provides an Eulerian description of the flow. For a Lagrangian description we would need to determine the velocity as a function of time for each particle as it flows from one point to another. In fluid mechanics it is usually easier to use the Eulerian method to describe a flow—in either experimental or analytical investigations. There are, however, certain instances in which the Lagrangian method is more convenient. For example, some numerical fluid mechanics calculations are based on determining the motion of individual fluid particles (based on the appropriate interactions among the particles), thereby describing the motion in Lagrangian terms. Similarly, in some experiments individual fluid particles are “tagged” and are followed throughout their motion, providing a Lagrangian description. Oceanographic measurements obtained from devices that flow with the ocean currents provide this information. Similarly, by using X-ray opaque dyes, it is y Location 0: T = T(x0, y0, t) 0 y0 Particle A: TA = TA(t) x x0 ■ Figure 4.2 Eulerian and Lagrangian descriptions of temperature of a flowing fluid. Free ebooks ==> www.Ebook777.com 160 Chapter 4 ■ Fluid Kinematics VIDEO V4.4 Follow the particles (experiment) VIDEO V4.5 Follow the particles (computer) possible to trace blood flow in arteries and to obtain a Lagrangian description of the fluid motion. A Lagrangian description may also be useful in describing fluid machinery (such as pumps and turbines) in which fluid particles gain or lose energy as they move along their flow paths. Another illustration of the difference between the Eulerian and Lagrangian descriptions can be seen in the following biological example. Each year thousands of birds migrate between their summer and winter habitats. Ornithologists study these migrations to obtain various types of important information. One set of data obtained is the rate at which birds pass a certain location on their migration route (birds per hour). This data may be obtained by training a video camera on the fixed location. This corresponds to an Eulerian description—“flowrate” at a given location as a function of time. Individual birds need not be followed to obtain this information. Another type of information is obtained by “tagging” certain birds with radio transmitters and following their motion along the migration route. This corresponds to a Lagrangian description—“position” of a given particle as a function of time. 4.1.2 One-, Two-, and Three-Dimensional Flows Most flow fields are actually threedimensional. VIDEO V4.6 Flow past a wing Generally, a fluid flow is a rather complex three-dimensional, time-dependent phenomenon— V = V(x, y, z, t) = uiˆ + υjˆ + wkˆ . In many situations, however, it is possible to make simplifying assumptions that allow a much easier understanding of the problem without sacrificing too much accuracy. One of these simplifications involves approximating a real flow as a simpler one- or twodimensional flow. In almost any flow situation, the velocity field actually contains all three velocity components (u, υ, and w, for example). In many situations the three-dimensional flow characteristics are important in terms of the physical effects they produce. For these situations it is necessary to analyze the flow in its complete three-dimensional character. Neglect of one or two of the velocity components in these cases would lead to considerable misrepresentation of the effects produced by the actual flow. The flow of air past an airplane wing provides an example of a complex three-dimensional flow. A feel for the three-dimensional structure of such flows can be obtained by studying Fig. 4.3, which is a photograph of the flow past a model wing; the flow has been made visible by using a flow visualization technique. In many situations one of the velocity components may be small (in some sense) relative to the two other components. In situations of this kind it may be reasonable to neglect the smaller component and assume two-dimensional flow. That is, V = uiˆ + υjˆ, where u and υ are functions of x and y (and possibly time, t). It is sometimes possible to further simplify a flow analysis by assuming that two of the velocity components are negligible, leaving the velocity field to be approximated as a one-dimensional flow field. That is, V = uiˆ. As we will learn from examples throughout the remainder of the book, although there are very few, if any, flows that are truly one-dimensional, there are many flow fields ■ Figure 4.3 Flow visualization of the complex three-dimensional flow past a delta wing. (Photograph courtesy of T. T. Lim and Y. D. Cui, National University of Singapore.) www.Ebook777.com 4.1 The Velocity Field 161 for which the one-dimensional flow assumption provides a reasonable approximation. There are also many flow situations for which use of a one-dimensional flow field assumption will give completely erroneous results. 4.1.3 Steady and Unsteady Flows Solenoid off,valve closed Spring Solenoid Diaphragm Outlet Inlet Solenoid on,valve open VIDEO V4.7 Flow types In the previous discussion we have assumed steady flow—the velocity at a given point in space does not vary with time, ∂V∂t = 0. In reality, almost all flows are unsteady in some sense. That is, the velocity does vary with time. Of course unsteady flows are usually more difficult to analyze (and to investigate experimentally) than are steady flows. Hence, considerable simplicity often results if one can make the assumption of steady flow without compromising the usefulness of the results. Among the various types of unsteady flows are nonperiodic flow, periodic flow, and truly random flow. Whether or not unsteadiness of one or more of these types must be included in an analysis is not always immediately obvious. An example of a nonperiodic, unsteady flow is that produced by turning off a faucet to stop the flow of water. Usually this unsteady flow process is quite mundane, and the forces developed as a result of the unsteady effects need not be considered. However, if the water is turned off suddenly (as with the electrically operated valve in a dishwasher shown in the figure in the margin), the unsteady effects can become important [as in the “water hammer” effects made apparent by the loud banging of the pipes under such conditions (Ref. 1)]. In other flows the unsteady effects may be periodic, occurring time after time in basically the same manner. The periodic injection of the air–gasoline mixture into the cylinder of an automobile engine is such an example. The unsteady effects are quite regular and repeatable and the engine could not operate without them. THE WIDE WORLD OF FLUIDS New pulsed liquid-jet scalpel High-speed liquid-jet cutters are Researchers have developed a new pulsed jet cutting tool that used for cutting a wide variety of materials such as leather may allow surgeons to perform microsurgery on tissues that are goods, jigsaw puzzles, plastic, ceramic, and metal. Typically, com- immersed in water. Rather than using a steady water jet, the sys- pressed air is used to produce a continuous stream of water that tem uses unsteady flow. A high-energy electrical discharge inside is ejected from a tiny nozzle. As this stream impacts the material the nozzle momentarily raises the temperature of the microjet to to be cut, a high pressure (the stagnation pressure) is produced approximately 10,000 °C. This creates a rapidly expanding vapor on the surface of the material, thereby cutting the material. Such bubble in the nozzle and expels a tiny fluid jet from the nozzle. liquid-jet cutters work well in air but are difficult to control if the Each electrical discharge creates a single, brief jet, which makes jet must pass through a liquid as often happens in surgery. a small cut in the material. VIDEO V4.8 Jupiter red spot In many situations the unsteady character of a flow is seemingly random. That is, there is no repeatable sequence or regular variation to the unsteadiness. This behavior occurs in turbulent flow and is absent from laminar flow. The “smooth” flow of highly viscous syrup onto a pancake represents a “deterministic” laminar flow. It is quite different from the turbulent flow observed in the “irregular” splashing of water from a faucet onto the sink below it. The “irregular” gustiness of the wind represents another turbulent flow. The differences between these types of flows are discussed in considerable detail in Chapters 8 and 9. It must be understood that the definition of steady or unsteady flow pertains to the behavior of a fluid property as observed at a fixed point in space. For steady flow, the values of all fluid properties (velocity, temperature, density, etc.) at any fixed point are independent of time. However, the value of those properties for a given fluid particle may change with time as the particle flows along, even in steady flow. Thus, the temperature of the exhaust at the exit of a car’s exhaust pipe may be constant for several hours, but the temperature of a fluid particle that left the exhaust pipe five minutes ago is lower now than it was when it left the pipe, even though the flow is steady. 4.1.4 Streamlines, Streaklines, and Pathlines Although fluid motion can be quite complicated, there are various concepts that can be used to help in the visualization and analysis of flow fields. To this end we discuss the use of streamlines, 162 Chapter 4 ■ Fluid Kinematics y v dy dx u x VIDEO streaklines, and pathlines in flow analysis. The streamline is often used in analytical work, while the streakline and pathline are often used in experimental work. A streamline is a line that is everywhere tangent to the velocity vector. If the flow is steady, nothing at a fixed point (including the velocity direction) changes with time, so the streamlines are fixed lines in space. For unsteady flows the streamlines may change shape with time. Streamlines are obtained analytically by integrating the equations defining lines tangent to the velocity field. As illustrated in the margin figure, for two-dimensional flows the slope of the streamline, dydx, must be equal to the tangent of the angle that the velocity vector makes with the x axis or dy υ = u dx V4.9 Streamlines (4.1) If the velocity field is known as a function of x and y (and t if the flow is unsteady), this equation can be integrated to give the equation of the streamlines. For unsteady flow there is no easy way to visualize streamlines experimentally in the laboratory. As discussed below, the observation of dye, smoke, or some other tracer injected into a flow can provide useful information, but for unsteady flows it is not necessarily information about the streamlines. EXAMPLE 4. 2 Streamlines for a Given Velocity Field GIVEN Consider the two-dimensional steady flow discussed FIND Determine the streamlines for this flow. in Example 4.1, V = (V0ℓ) (−xiˆ + yjˆ ). SOLUTION y Since u = (−V0ℓ)x and υ = (V0ℓ)y (1) it follows that streamlines are given by solution of the equation (V0ℓ)y dy y υ = = =− u x dx −(V0ℓ)x in which variables can be separated and the equation integrated to give ∫ dy =− y ∫ 4 2 C=9 dx x C=4 or ln y = −ln x + constant 0 Thus, along the streamline xy = C,  where C is a constant C=1 2 4 C = –1 C = –4 (Ans) By using different values of the constant C, we can plot various lines in the x–y plane—the streamlines. The streamlines for x ≥ 0 are plotted in Fig. E4.2. A comparison of this figure with Fig. E4.1a illustrates the fact that streamlines are lines tangent to the velocity vector. COMMENT Note that a flow is not completely specified by the shape of the streamlines alone. For example, the streamlines for the flow with V0ℓ = 10 have the same shape as those for the flow with V0ℓ = −10. However, the direction of the flow is opposite for these two cases. The arrows in Fig. E4.2 representing the flow direction are correct for V0ℓ = 10 since, from Eq. 1, u = −10x and υ = 10y. That is, the flow is from right to left. For V0ℓ = −10 the arrows are reversed. The flow is from left to right. –2 –4 ■ Figure E4.2 C = –9 x 4.1 VIDEO V4.10 Streaklines For steady flow, streamlines, streaklines, and pathlines are the same. The Velocity Field 163 A streakline consists of all particles in a flow that have previously passed through a common point. Streaklines are a laboratory tool rather than an analytical tool. They can be obtained by taking instantaneous photographs of marked particles that all passed through a given location in the flow field at some earlier time. Such a line can be produced by continuously injecting marked fluid (neutrally buoyant smoke in air or dye in water) at a given location (Ref. 2). (See Fig. 9.1.) If the flow is steady, each successively injected particle follows precisely behind the previous one, forming a steady streakline that is exactly the same as the streamline through the injection point. For unsteady flows, particles injected at the same point at different times need not follow the same path. An instantaneous photograph of the marked fluid would show the streakline at that instant, but it would not necessarily coincide with the streamline through the point of injection at that particular time nor with the streamline through the same injection point at a different time (see Example 4.3). The third method used for visualizing and describing flows involves the use of pathlines. A pathline is the line traced out by a given particle as it flows from one point to another. The pathline is a Lagrangian concept that can be produced in the laboratory by marking a fluid particle (dyeing a small fluid element) and taking a time exposure photograph of its motion. For steady flow, a pathline is the same as a streamline; for unsteady flow, they are different. If the flow is steady, the path taken by a marked particle (a pathline) will be the same as the line formed by all other particles that previously passed through the point of injection (a streakline). For such cases these lines are tangent to the velocity field. Hence, pathlines, streamlines, and streaklines are the same for steady flows. For unsteady flows none of these three types of lines need be the same (Ref. 3). Often one sees pictures of “streamlines” made visible by the injection of smoke or dye into a flow as is shown in Fig. 4.3. Actually, such pictures show streaklines rather than streamlines. However, for steady flows the two are identical; only the nomenclature is incorrectly used. THE WIDE WORLD OF FLUIDS Air bridge spanning the oceans It has long been known that materials and organisms are also carried along these literal par- large quantities of material are transported from one location to ticle paths. Satellite images reveal the amazing rate by which another by airborne dust particles. It is estimated that 2 billion desert soils and other materials are transformed into air-borne metric tons of dust are lifted into the atmosphere each year. particles as a result of storms that produce strong winds. Once Most of these particles settle out fairly rapidly, but significant the tiny particles are aloft, they may travel thousands of miles, amounts travel large distances. Scientists are beginning to crossing the oceans and eventually being deposited on other understand the full impact of this phenomenon—it is not only the continents. For the health and safety of all, it is important that tonnage transported, but the type of material transported that we obtain a better understanding of the air bridges that span is significant. In addition to the mundane inert material we all the oceans and also understand the ramification of such mate- term “dust,” it is now known that a wide variety of hazardous rial transport. E X AMPLE 4.3 Comparison of Streamlines, Pathlines, and Streaklines GIVEN Water flowing from the oscillating slit shown in Fig. E4.3a produces a velocity field given by V = u0 sin[ω(t − yυ0 )]iˆ + υ0 ˆj, where u0, υ0, and ω are constants. Thus, the y component of velocity remains constant (υ = υ0 ), and the x component of velocity at y = 0 coincides with the velocity of the oscillating sprinkler head [u = u0 sin(ωt) at y = 0]. FIND (a) Determine the streamline that passes through the origin at t = 0; at t = π2ω. (b) Determine the pathline of the particle that was at the origin at t = 0; at t = π2. (c) Discuss the shape of the streakline that passes through the origin. SOLUTION (a) Since u = u0 sin[ω(t − yυ0 )] and υ = υ0, it follows from Eq. 4.1 that streamlines are given by the solution of υ0 dy υ = = u dx u0 sin[ω(t − yυ0 )] in which the variables can be separated and the equation integrated (for any given time t) to give u0 ∫ sin [ω (t − υy )] dy = υ ∫ dx 0 0 164 Chapter 4 ■ Fluid Kinematics This can be integrated to give the x component of the pathline as or y u0 (υ0ω) cos[ω(t − )] = υ0 x + C υ0 where C is a constant. For the streamline at t = 0 that passes through the origin (x = y = 0), the value of C is obtained from Eq. 1 as C = u0υ0ω. Hence, the equation for this streamline is x= u0 ωy − 1] cos ω [ ( υ0 ) x = 0 and y = υ0 t (5) (6) (Ans) Similarly, for the particle that was at the origin at t = π2ω, Eqs. 4 and 5 give C1 = −πυ02ω and C2 = −πu02ω. Thus, the pathline for this particle is u0 u0 y ωy π π − )] = cos( − cos [ω( ω υ0 ω υ0 ) 2ω 2 π π x = u0 t −  and y = υ0(t − ( 2ω ) 2ω ) (7) The pathline can be drawn by plotting the locus of x(t), y(t) values for t ≥ 0 or by eliminating the parameter t from Eq. 7 to give or x= C1ω t + C2 υ0 )] where C2 is a constant. For the particle that was at the origin (x = y = 0) at time t = 0, Eqs. 4 and 5 give C1 = C2 = 0. Thus, the pathline is (2) (Ans) Similarly, for the streamline at t = π2ω that passes through the origin, Eq. 1 gives C = 0. Thus, the equation for this streamline is x= x = −[u0 sin ( (1)   u0 ωy sin( ) ω υ0 (3) (Ans) COMMENT These two streamlines, plotted in Fig. E4.3b, are not the same because the flow is unsteady. For example, at the origin (x = y = 0) the velocity is V = υ0 ˆj at t = 0 and V = u0ˆi + υ0 jˆ at t = π2ω. Thus, the angle of the streamline passing through the origin changes with time. Similarly, the shape of the entire streamline is a function of time. y= υ0 x u0 (8) (Ans) COMMENT The pathlines given by Eqs. 6 and 8, shown in Fig. E4.3c, are straight lines from the origin (rays). The pathlines and streamlines do not coincide because the flow is unsteady. The y equation can be integrated (since υ0 = constant) to give the y coordinate of the pathline as (c) The streakline through the origin at time t = 0 is the locus of particles at t = 0 that previously (t < 0) passed through the origin. The general shape of the streaklines can be seen as follows. Each particle that flows through the origin travels in a straight line (pathlines are rays from the origin), the slope of which lies between ±υ0u0 as shown in Fig. E4.3d. Particles passing through the origin at different times are located on different rays from the origin and at different distances from the origin. The net result is that a stream of dye continually injected at the origin (a streakline) would have the shape shown in Fig. E4.3d. Because of the unsteadiness, the streakline will vary with time, although it will always have the oscillating, sinuous character shown. y = υ0 t + C1 (4) COMMENT Similar streaklines are given by the stream of where C1 is a constant. With this known y = y(t) dependence, the x equation for the pathline becomes water from a garden hose nozzle that oscillates back and forth in a direction normal to the axis of the nozzle. In this example neither the streamlines, pathlines, nor streaklines coincide. If the flow were steady, all of these lines would be the same. (b) The pathline of a particle (the location of the particle as a function of time) can be obtained from the velocity field and the definition of the velocity. Since u = dxdt and υ = dydt we obtain dy y dx = u0 sin ω t −  and  = υ0 [ ( υ0 )] dt dt υ0 t + C1 C1 ω dx = u0 sin[ω (t − = −u0 sin( υ0 )] υ0 ) dt y y 2π v0/ω Streamlines through origin x 0 t=0 π v0/ω t = π /2ω Oscillating sprinkler head Q (a) –2u0/ω 0 (b) 2u0/ω x ■ Figure E4.3(a), (b) 4.2 The Acceleration Field 165 y y t=0 v 0 / u0 Pathlines of particles at origin at time t v0 t = π /2ω –1 0 1 Pathline u0 x VIDEO x 0 (c) 4.2 Streaklines through origin at time t (d) ■ Figure E4.3(c), (d ) The Acceleration Field V4.11 Pathlines Acceleration is the time rate of change of velocity for a given particle. As indicated in the previous section, we can describe fluid motion by either (1) following individual particles (Lagrangian description) or (2) remaining fixed in space and observing different particles as they pass by (Eulerian description). In either case, to apply Newton’s second law (F = ma) we must be able to describe the particle acceleration. For the infrequently used Lagrangian method, we describe the fluid acceleration just as is done in solid body dynamics—a = a(t) for each particle. For the Eulerian description we describe the acceleration field as a function of position and time without actually following any particular particle. This is analogous to describing the flow in terms of the velocity field, V = V (x, y, z, t), rather than the velocity for particular particles. In this section we will discuss how to obtain the acceleration field if the velocity field is known. The acceleration of a particle is the time rate of change of its velocity. For unsteady flows the velocity at a given point in space (occupied by different particles at different times) may vary with time, giving rise to a portion of the fluid acceleration. In addition, a fluid particle may experience an acceleration because its velocity changes as it flows from one point to another in space. For example, water flowing through a garden hose nozzle under steady conditions (constant number of gallons per minute from the hose) will experience an acceleration as it changes from its relatively low velocity in the hose to its relatively high velocity at the tip of the nozzle. 4.2.1 Acceleration and the Material Derivative Consider a fluid particle moving along its pathline as is shown in Fig. 4.4. In general, the particle’s velocity, denoted VA for particle A, is a function of its location and the time. That is, VA = VA (rA, t) = VA[xA (t), yA (t), zA (t), t] z VA(rA, t) Particle A at time t Particle path vA(rA, t) rA wA(rA, t) uA(rA, t) y zA(t) x yA(t) xA(t) ■ Figure 4.4 Velocity and position of particle A at time t. 166 Chapter 4 ■ Fluid Kinematics where xA = xA (t), yA = yA (t), and zA = zA (t) define the location of the moving particle. By definition, the acceleration of a particle is the time rate of change of its velocity. Since the velocity may be a function of both position and time, its value may change because of the change in time as well as a change in the particle’s position. Thus, we use the chain rule of differentiation to obtain the acceleration of particle A, denoted aA, as aA (t) = d VA ∂VA ∂VA dxA ∂VA dyA ∂VA dzA = + + + dt ∂t ∂x dt ∂y dt ∂z dt (4.2) Using the fact that the particle velocity components are given by uA = dxAdt, υA = dyAdt, and wA = dzAdt, Eq. 4.2 becomes aA = ∂VA ∂VA ∂VA ∂VA + uA + υA + wA ∂t ∂x ∂y ∂z Since the above is valid for any particle, we can drop the reference to particle A and obtain the acceleration field from the velocity field as a= ∂V ∂V ∂V ∂V +u +υ +w ∂t ∂x ∂y ∂z (4.3) This is a vector result whose scalar components can be written as ∂u ∂u ∂u ∂u +u +υ +w ∂t ∂x ∂y ∂z ∂υ ∂υ ∂υ ∂υ ay = +u +υ +w ∂t ∂x ∂y ∂z ax = (4.4) and az = The material derivative is used to describe time rates of change for a given particle. ∂w ∂w ∂w ∂w +u +υ +w ∂t ∂x ∂y ∂z where ax, ay, and az are the x, y, and z components of the acceleration. The above result is often written in shorthand notation as a= DV Dt where the operator D( ) ∂( ) ∂( ) ∂( ) ∂( ) ≡ +u +υ +w Dt ∂t ∂x ∂y ∂z (4.5) is termed the material derivative or substantial derivative because it gives the rate of change as we follow a particle of the material (or substance). An often-used shorthand notation for the material derivative operator is D( ) ∂( ) = + (V ∙ ∇ )( ) Dt ∂t (4.6) The dot product of the velocity vector, V, and the gradient operator, ∇ ( ) = ∂ ( )∂x ˆi + ∂ ( )∂y ˆj + ∂ ( )∂z kˆ (a vector operator) provides a convenient notation for the spatial derivative terms appearing in the material derivative. Note that the notation V ∙ ∇ represents the operator V ∙ ∇ ( ) = u∂ ( )∂x + υ∂ ( )∂y + w∂ ( )∂z. (Recall from calculus that the gradient operator has different forms for different coordinate systems.) The material derivative concept is very useful in analysis involving various fluid parameters, not just the acceleration. The material derivative of any variable is the rate at which that variable changes with time for a given particle as seen by one moving along with the fluid. For example, 4.2 167 consider a temperature field T = T(x, y, z, t) associated with a given flow, like the flame shown in the figure in the margin. It may be of interest to determine the time rate of change of temperature of a fluid particle (particle A) as it moves through this temperature field. If the velocity, V = V (x, y, z, t), is known, we can apply the chain rule to determine the rate of change of temperature as T = T (x, y, z, t) dTA ∂TA ∂TA dxA ∂TA dyA ∂TA dzA = + + + dt ∂t ∂x dt ∂y dt ∂z dt V Particle A The Acceleration Field This can be written as DT ∂T ∂T ∂T ∂T ∂T = +u +υ +w = + V ∙ ∇T Dt ∂t ∂x ∂y ∂z ∂t z x y As in the determination of the acceleration, the material derivative operator, D( )Dt, appears. E X AMPLE 4.4 Acceleration along a Streamline y GIVEN An incompressible, inviscid fluid flows steadily past a tennis ball of radius R, as shown in Fig. E4.4a. According to advanced analysis of the flow, the fluid velocity along streamline A–B is given by R A R3 V = u(x)iˆ = V0 1 + 3 ˆi ( x ) x B V0 where V0 is the upstream velocity far ahead of the sphere. (a) FIND Determine the acceleration experienced by fluid particles as they flow along this streamline. ax _______ (V02/R) SOLUTION Along streamline A–B there is only one component of velocity (υ = w = 0) so that from Eq. 4.3 a= –3 –2 –1 x/R B A –0.2 ∂V ∂u ∂u ∂V +u = + u ) ˆi ∂t ∂x ( ∂t ∂x –0.4 –0.6 or ax = (b) ∂u ∂u + u ,  ay = 0,  az = 0 ∂t ∂x ■ Figure E4.4 Since the flow is steady, the velocity at a given point in space does not change with time. Thus, ∂u∂t = 0. With the given velocity distribution along the streamline, the acceleration becomes ∂u R3 ax = u = V0(1 + 3 ) V0[R3 (−3x−4 )] ∂x x or ax = −3(V 20 R) 1 + (Rx) 3 (xR) 4 (Ans) COMMENTS Along streamline A–B (−∞ ≤ x ≤ −R and y = 0) the acceleration has only an x component, and it is negative (a deceleration). Thus, the fluid slows down from its upstream velocity of V = V0ˆi at x = −∞ to its stagnation point velocity of V = 0 at x = −R, the “nose” of the ball. The variation of ax along streamline A–B is shown in Fig. E4.4b. It is the same result as is obtained in Example 3.1 by using the streamwise component of the acceleration, ax = V ∂V∂s. The maximum deceleration occurs at x = −1.205R and has a value of ax,max = −0.610 V 20R. Note that this maximum deceleration increases with increasing velocity and decreasing size. As indicated in the following table, typical values of this deceleration can be quite large. For example, the ax,max = −4.08 × 104 ft/s2 value for a pitched baseball is a deceleration approximately 1500 times that of gravity. (Note that this is not the deceleration of the ball; it is the deceleration of the air relative to the ball.) 168 Chapter 4 ■ Fluid Kinematics Object V0 (ft/s) R (ft) ax,max (ft/s2 ) 1 20 90 100 200 4.0 0.80 0.121 0.104 0.070 −0.153 −305 −4.08 × 104 −5.87 × 104 −3.49 × 105 Rising weather balloon Soccer ball Baseball Tennis ball Golf ball In general, for fluid particles on streamlines other than A–B, all three components of the acceleration (ax, ay, and az ) will be nonzero. 4.2.2 Unsteady Effects The local derivative is a result of the unsteadiness of the flow. VIDEO V4.12 Unsteady flow As is seen from Eq. 4.5, the material derivative formula contains two types of terms—those involving the time derivative [∂ ( )∂t] and those involving spatial derivatives [∂ ( )∂x, ∂ ( )∂y, and ∂ ( )∂z]. The time derivative portion is termed the local derivative. It represents the effects of the unsteadiness of the flow. If the parameter involved is the acceleration, that portion given by ∂V∂t is termed the local acceleration. For steady flow the time derivative is zero throughout the flow field [∂ ( )∂t ≡ 0], and the local effect vanishes. Physically, there is no change in flow parameters at a fixed point in space if the flow is steady. There may be a change of those parameters for a fluid particle as it moves about, however. If a flow is unsteady, its parameter values (velocity, temperature, density, etc.) at any location may change with time. For example, an unstirred (V = 0) cup of coffee will cool over time because of heat transfer to its surroundings. That is, DTDt = ∂T∂t + V ∙ ∇T = ∂T∂t < 0. Similarly, a fluid particle may have nonzero acceleration as a result of the unsteady effect of the flow. Consider flow in a constant diameter pipe as is shown in Fig. 4.5. The flow is assumed to be spatially uniform throughout the pipe. That is, V = V0 (t) ˆi at all points in the pipe. The value of the acceleration depends on whether V0 is being increased, ∂V0 ∂t > 0, or decreased, ∂V0 ∂t < 0. Unless V0 is independent of time (V0 ≡ constant) there will be an acceleration, the local acceleration term. Thus, the acceleration field, a = ∂V0 ∂t ˆi, is uniform throughout the entire flow, although it may vary with time (∂V0 ∂t need not be constant). The acceleration due to the spatial variations of velocity (u ∂u∂x, υ ∂υ∂y, etc.) vanishes automatically for this flow, since ∂u∂x = 0 and υ = w = 0. That is, a= ∂V0 ∂V ∂V ∂V ∂V ∂V ˆi +u +υ +w = = ∂t ∂x ∂y ∂z ∂t ∂t 4.2.3 Convective Effects V2 > V1 V1 The portion of the material derivative (Eq. 4.5) represented by the spatial derivatives is termed the convective derivative. It represents the fact that a flow property associated with a fluid particle may vary because of the motion of the particle from one point in space where the parameter has one value to another point in space where its value is different. For example, the water velocity at the inlet of the garden hose nozzle shown in the figure in the margin is different (both in direction and speed) than it is at the exit. This contribution to the time rate of change of the parameter for the particle can occur whether the flow is steady or unsteady. It is due to the convection, or motion, of the V0(t) x V0(t) ■ Figure 4.5 Uniform, unsteady flow in a constant diameter pipe. 4.2 The Acceleration Field 169 Hot Water heater Tout > Tin Pathline ∂T = 0 ___ ∂t DT ≠ 0 ___ Dt Cold Tin The convective derivative is a result of the spatial variation of the flow. u 0 x ax 0 x ■ Figure 4.6 Steady-state operation of a water heater. (Photograph courtesy of American Water Heater Company.) particle through space in which there is a gradient [∇ ( ) = ∂ ( )∂x ˆi + ∂ ( )∂y jˆ + ∂ ( )∂z kˆ ] in the parameter value. That portion of the acceleration given by the term (V ∙ ∇ )V is termed the convective acceleration. As is illustrated in Fig. 4.6, the temperature of a water particle changes as it flows through a water heater. The water entering the heater is always the same cold temperature, and the water leaving the heater is always the same hot temperature. The flow is steady. However, the temperature, T, of each water particle increases as it passes through the heater—Tout > Tin. Thus, DTDt ≠ 0 because of the convective term in the total derivative of the temperature. That is, ∂T∂t = 0, but u ∂T∂x ≠ 0 (where x is directed along the streamline), since there is a nonzero temperature gradient along the streamline. A fluid particle traveling along this nonconstant temperature path (∂T∂x ≠ 0) at a specified speed (u) will have its temperature change with time at a rate of DTDt = u ∂T∂x even though the flow is steady ( ∂T∂t = 0). The same types of processes are involved with fluid accelerations. Consider flow in a variable area pipe as shown in Fig. 4.7. It is assumed that the flow is steady and one-dimensional with velocity that increases and decreases in the flow direction as indicated. As the fluid flows from section (1) to section (2), its velocity increases from V1 to V2. Thus, even though ∂V∂t = 0 (steady flow), fluid particles experience an acceleration given by ax = u ∂u∂x (convective acceleration). For x1 < x < x2, it is seen that ∂u∂x > 0 so that ax > 0—the fluid accelerates. For x2 < x < x3, it is seen that ∂u∂x < 0 so that ax < 0—the fluid decelerates. This acceleration and deceleration are shown in the figure in the margin. If V1 = V3, the amount of acceleration precisely balances the amount of deceleration even though the distances between x2 and x1 and x3 and x2 are not the same. The concept of the material derivative can be used to determine the time rate of change of any parameter associated with a particle as it moves about. Its use is not restricted to fluid mechanics alone. The basic ingredients needed to use the material derivative concept are the field description of the parameter, P = P(x, y, z, t), and the rate at which the particle moves through that field, V = V (x, y, z, t). u = V1 u = V2 > V1 u = V3 = V1 < V2 x x1 x2 x3 ■ Figure 4.7 Uniform, steady flow in a variable area pipe. 170 Chapter 4 ■ Fluid Kinematics EXAMPLE 4.5 Acceleration from a Given Velocity Field GIVEN Consider the steady, two-dimensional flow field dis- FIND Determine the acceleration field for this flow. cussed in Example 4.2. SOLUTION y In general, the acceleration is given by ∂V DV = + (V ∙ ∇ ) (V) a= Dt ∂t ∂V ∂V ∂V ∂V = +u +υ +w ∂t ∂x ∂y ∂z V a Streamline (1) where the velocity is given by V = (V0ℓ) (−xiˆ + yjˆ ) so that u = −(V0ℓ) x and υ = (V0ℓ)y. For steady [∂( )∂t = 0], two-dimensional [w = 0 and ∂( )∂z = 0] flow, Eq. l becomes ∂V ∂V +υ ∂x ∂y ∂u ∂u ∂υ ∂υ = (u + υ ) ˆi + (u + υ ) jˆ ∂x ∂y ∂x ∂y a=u x ■ Figure E4.5 Hence, for this flow the acceleration is given by V0 V0 V0 (x) − + (y) (0) ] ˆi ℓ) ( ℓ) (ℓ) V0 V0 V0 + [(− (x) (0) + ( ) (y) ( )] ˆj ℓ) ℓ ℓ a = [(− Also, the acceleration vector is oriented at an angle θ from the x axis, where tan θ = or ax = V 20 x ℓ2 ,  ay = V 20 y ℓ2 (Ans) COMMENTS The fluid experiences an acceleration in both the x and y directions. Since the flow is steady, there is no local acceleration—the fluid velocity at any given point is constant in time. However, there is a convective acceleration due to the change in velocity from one point on the particle’s pathline to another. Recall that the velocity is a vector—it has both a magnitude and a direction. In this flow both the fluid speed (magnitude) and flow direction change with location (see Fig. E4.1a). For this flow the magnitude of the acceleration is constant on circles centered at the origin, as is seen from the fact that  a = (a2x + a2y + a2z ) 12 = ( EXAMPLE 4.6 V0 2 2 (x + y2 ) 12 ℓ) ay ax = y x This is the same angle as that formed by a ray from the origin to point (x, y). Thus, the acceleration is directed along rays from the origin and has a magnitude proportional to the distance from the origin. Typical acceleration vectors (from Eq. 2) and velocity vectors (from Example 4.1) are shown in Fig. E4.5 for the flow in the first quadrant. Note that a and V are not parallel except along the x and y axes (a fact that is responsible for the curved pathlines of the flow) and that both the acceleration and velocity are zero at the origin (x = y = 0). An infinitesimal fluid particle placed precisely at the origin will remain there, but its neighbors (no matter how close they are to the origin) will drift away. (2) The Material Derivative GIVEN A fluid flows steadily through a two-dimensional nozzle of length ℓ as shown in Fig. E4.6a. The nozzle shape is given by y/ℓ = ± 0.5[1 + (xℓ)] If viscous and gravitational effects are negligible, the velocity field is approximately u = V0[1 + xℓ], υ = −V0yℓ (1) 4.2 V0 and the pressure field is p − p0 = −(ρV202)[(x2 + y )ℓ + 2xℓ] 2 2 where V0 and p0 are the velocity and pressure at the origin, x = y = 0. Note that the fluid speed increases as it flows through the nozzle. For example, along the centerline (y = 0), V = V0 at x = 0 and V = 2V0 at x = ℓ. 171 The Acceleration Field y 0 0.5 _y = _______ ℓ (1 + x/ℓ) ℓ 0.5 _y = – _______ ℓ (1 + x/ℓ) 2V0 FIND Determine, as a function of x and y, the time rate of change of pressure felt by a fluid particle as it flows through the nozzle. x SOLUTION The time rate of change of pressure at any given, fixed point in this steady flow is zero. However, the time rate of change of pressure felt by a particle flowing through the nozzle is given by the material derivative of the pressure and is not zero. Thus, Dp ∂p ∂p ∂p ∂p ∂p = +u +υ =u +υ Dt ∂t ∂x ∂y ∂x ∂y (2) where the x and y components of the pressure gradient can be written as ρV 20 x ∂p =− + 1) ∂x ℓ (ℓ (3) ρV 20 y ∂p =− ∂y ℓ (ℓ) (4) ■ Figure E4.6a streamline, it moves into areas of lower and lower pressure. Hence, even though the flow is steady, the time rate of change of the pressure for any given particle is negative. and Therefore, by combining Eqs. (1), (2), (3), and (4) we obtain ρV 20 p – p0 ______ /2 ρV02 0.5 1 –0.5 _y ℓ ρV 20 Dp y y x x = V0(1 + )(− + 1) + (−V0 )(− Dt ℓ ℓ )( ℓ ℓ ℓ )( ℓ ) 0 or 2 ρV30 x Dp y 2 =− + 1 − ) (ℓ) ] Dt ℓ [( ℓ (5) –1.0 –1.5 –2.0 –2.5 –3.0 _x 1 ℓ (Ans) COMMENT Lines of constant pressure within the nozzle are indicated in Fig. E4.6b, along with some representative streamlines of the flow. Note that as a fluid particle flows along its ■ Figure E4.6b 4.2.4 Streamline Coordinates VIDEO V4.13 Streamline coordinates In many flow situations, it is convenient to use a coordinate system defined in terms of the streamlines of the flow. An example for steady, two-dimensional flows is illustrated in Fig. 4.8. Such flows can be described either in terms of the usual x, y Cartesian coordinate system (or some other system such as the r, θ polar coordinate system) or the streamline coordinate system. In the streamline coordinate system the flow is described in terms of one coordinate along the streamlines, denoted s, and the second coordinate normal to the streamlines, denoted n. Unit vectors in these two directions are denoted by sˆ and nˆ , as shown in the figure. Care is needed not to confuse the coordinate distance s (a scalar) with the unit vector along the streamline direction, sˆ. 172 Chapter 4 ■ Fluid Kinematics y s = s2 n = n2 n = n1 s = s1 s=0 n=0 Streamlines ^ n ^ s V s x ■ Figure 4.8 Streamline coordinate system for twodimensional flow. The flow plane is therefore covered by an orthogonal curved net of coordinate lines. At any point the s and n directions are perpendicular, but the lines of constant s or constant n are not necessarily straight. Without knowing the actual velocity field (hence, the streamlines) it is not actually possible to construct this flow net. In many situations, appropriate simplifying assumptions can be made so that this lack of information does not present an insurmountable difficulty. One of the major advantages of using the streamline coordinate system is that the velocity is always tangent to the s direction. That is, V = V sˆ V a an as This allows simplifications in describing the fluid particle acceleration and in solving the equations governing the flow. For steady, two-dimensional flow we can determine the acceleration as a= DV = as sˆ + annˆ Dt where as and an are the streamline and normal components of acceleration, respectively, as indicated by the figure in the margin. We use the material derivative because by definition the acceleration is the time rate of change of the velocity of a given particle as it moves about. If the streamlines are curved, both the speed of the particle and its direction of flow may change from one point to another. In general, for steady flow both the speed and the flow direction are a function of location—V = V(s, n) and sˆ = sˆ (s, n). For a given particle, the value of s changes with time, but the value of n remains fixed because the particle flows along a streamline defined by n = constant. (Recall that streamlines and pathlines coincide in steady flow.) Thus, application of the chain rule gives a= D(V sˆ ) DV Dsˆ = sˆ + V Dt Dt Dt 4.3 Control Volume and System Representations 173 O O δθ δθ ℛ ℛ ^ n δs s A B ^s (s + ) δs A ^s (s ^ s (s) B δs ) δs + δθ O' B' ^ δs A' ^ s (s) ■ Figure 4.9 Relationship between the unit vector along the streamline, sˆ, and the radius of curvature of the streamline, ℛ. or a =( ∂V ∂V ds ∂V dn ∂sˆ ∂sˆ ds ∂sˆ dn + + sˆ + V( + + ) ∂t ∂s dt ∂n dt ∂t ∂s dt ∂n dt ) This can be simplified by using the fact that for steady flow nothing changes with time at a given point so that both ∂V∂t and ∂sˆ ∂t are zero. Also, the velocity along the streamline is V = dsdt, and the particle remains on its streamline (n = constant) so that dndt = 0. Hence, a = (V The orientation of the unit vector along the streamline changes with distance along the streamline. ∂V ∂sˆ ˆ+V V s ( ∂s ) ∂s ) The quantity ∂sˆ ∂s represents the limit as δs → 0 of the change in the unit vector along the streamline, δsˆ, per change in distance along the streamline, δs. The magnitude of sˆ is constant ( sˆ  = 1; it is a unit vector), but its direction is variable if the streamlines are curved. From Fig. 4.9 it is seen that the magnitude of ∂sˆ ∂s is equal to the inverse of the radius of curvature of the streamline, ℛ, at the point in question. This follows because the two triangles shown (AOB and A′O′B′) are similar triangles so that δsℛ =  δ sˆ   sˆ  =  δsˆ , or  δsˆ δs = 1ℛ. Similarly, in the limit δs → 0, the direction of δsˆ δs is seen to be normal to the streamline. That is,   ∂sˆ δsˆ nˆ = lim = δs→0 ∂s δs ℛ Hence, the acceleration for steady, two-dimensional flow can be written in terms of its streamwise and normal components in the form a=V ∂V V2 ∂V V2 sˆ + nˆ  or as = V ,  an = ∂s ℛ ∂s ℛ (4.7) The first term, as = V ∂V∂s, represents the convective acceleration along the streamline, and the second term, an = V 2 ℛ, represents centripetal acceleration (one type of convective acceleration) normal to the fluid motion. These components can be noted in Fig. E4.5 by resolving the acceleration vector into its components along and normal to the velocity vector. Note that the unit vector nˆ is directed from the streamline toward the center of curvature. These forms of the acceleration were used in Chapter 3 and are probably familiar from previous dynamics or physics considerations. 4.3 Control Volume and System Representations As is discussed in Chapter 1, a fluid is a type of matter that is relatively free to move, deform, and interact with its surroundings. As with any matter, a fluid’s behavior is governed by fundamental physical laws that are modeled by an appropriate set of equations. The application of laws such as the conservation of mass, Newton’s laws of motion, and the laws of thermodynamics forms the 174 Chapter 4 ■ Fluid Kinematics Both control volume and system concepts can be used to describe fluid flow. (Photograph courtesy of NASA.) foundation of fluid mechanics analyses. There are various ways that these governing laws can be applied to a fluid, including the system approach and the control volume approach. By definition, a system is a collection of matter of fixed identity (always the same atoms or fluid particles), which may move, flow, and interact with its surroundings. A control volume, on the other hand, is a volume in space (a geometric entity, independent of mass) through which fluid may flow. A system is a specific, identifiable quantity of matter. It may consist of a relatively large amount of mass (such as all of the air in the Earth’s atmosphere), or it may be an infinitesimal size (such as a single fluid particle). In any case, the molecules making up the system are “tagged” in some fashion (dyed red, either actually or only in your mind) so that they can be continually identified as they move about. The system may interact with its surroundings by various means (by the transfer of heat or the exertion of a pressure force, for example). It may continually change size and shape, but it always contains the same mass. A mass of air drawn into an air compressor can be considered as a system. It changes shape and size (it is compressed), its temperature may change, and it is eventually expelled through the outlet of the compressor. The matter associated with the original air drawn into the compressor remains as a system, however. The behavior of this material could be investigated by applying the appropriate governing equations to this system. One of the important concepts used in the study of statics and dynamics is that of the freebody diagram. That is, we identify an object, isolate it from its surroundings, replace its surroundings by the equivalent actions that they put on the object, and apply Newton’s laws of motion. The body in such cases is our system—an identified portion of matter that we follow during its interactions with its surroundings. In fluid mechanics, it is often quite difficult to identify and keep track of a specific quantity of matter. A finite portion of a fluid contains an uncountable number of fluid particles that move about quite freely, unlike a solid that may deform but usually remains relatively easy to identify. For example, we cannot as easily follow a specific portion of water flowing in a river as we can follow a branch floating on its surface. We may often be more interested in determining the forces acting on a fan, airplane, or automobile by air flowing past the object than we are in the information obtained by following a given portion of the air (a system) as it flows along. Similarly, for the Space Shuttle launch vehicle shown in the figure in the margin, we may be more interested in determining the thrust produced than we are in the information obtained by following the highly complex, irregular path of the exhaust plume from the rocket engine nozzle. For these situations we often use the control volume approach. We identify a specific volume in space (a volume associated with the fan, airplane, or automobile, for example) and analyze the fluid flow within, through, or around that volume. In general, the control volume can be a moving volume, although for most situations considered in this book we will use only fixed, nondeformable control volumes. The matter within a control volume may change with time as the fluid flows through it. Similarly, the amount of mass within the volume may change with time. The control volume itself is a specific geometric entity, independent of the flowing fluid. Examples of control volumes and control surfaces (the surface of the control volume) are shown in Fig. 4.10. For case (a), fluid flows through a pipe. The fixed control surface consists of the inside surface of the pipe, the outlet end at section (2), and a section across the pipe at (1). One portion of the control surface is a physical surface (the pipe), while the remainder is simply Jet engine Pipe Balloon V (1) (2) (a) Control volume surface (b) System at time t1 (c) System at time t2 > t1 ■ Figure 4.10 Typical control volumes: (a) fixed control volume, (b) fixed or moving control volume, (c) deforming control volume. 4.4 The governing laws of fluid motion are stated in terms of fluid systems, not control volumes. 4.4 The Reynolds Transport Theorem 175 a surface in space (across the pipe). Fluid flows across part of the control surface but not across all of it. Another control volume is the rectangular volume surrounding the jet engine shown in Fig. 4.10b. If the airplane to which the engine is attached is sitting still on the runway, air flows through this control volume because of the action of the engine within it. The air that was within the engine itself at time t = t1 (a system) has passed through the engine and is outside of the control volume at a later time t = t2 as indicated. At this later time other air (a different system) is within the engine. If the airplane is moving, the control volume is fixed relative to an observer on the airplane, but it is a moving control volume relative to an observer on the ground. In either situation air flows through and around the engine as indicated. The deflating balloon shown in Fig. 4.10c provides an example of a deforming control volume. As time increases, the control volume (whose surface is the inner surface of the balloon) decreases in size. If we do not hold onto the balloon, it becomes a moving, deforming control volume as it darts about the room. The majority of the problems we will analyze can be solved by using a fixed, nondeforming control volume. In some instances, however, it will be advantageous, in fact necessary, to use a moving, deforming control volume. In many ways the relationship between a system and a control volume is similar to the relationship between the Lagrangian and Eulerian flow description introduced in Section 4.1.1. In the system or Lagrangian description, we follow the fluid and observe its behavior as it moves about. In the control volume or Eulerian description, we remain stationary and observe the fluid’s behavior at a fixed location. (If a moving control volume is used, it virtually never moves with the system—the system flows through the control volume.) These ideas are discussed in more detail in the next section. All of the laws governing the motion of a fluid can be stated in their basic form in terms of a system approach. For example, “the mass of a system remains constant,” or “the time rate of change of momentum of a system is equal to the sum of all the forces acting on the system.” Note the word system, not control volume, in these statements. To use the governing equations in a control volume approach to problem solving, we must recast the laws in an appropriate manner. To this end we introduce the Reynolds transport theorem in the following section. The Reynolds Transport Theorem V We are sometimes interested in what happens to a particular part of the fluid as it moves about. Other times we may be interested in what effect the fluid has on a particular object or volume in space as fluid interacts with it. Thus, we need to describe the laws governing fluid motion using both system concepts (consider a given mass of the fluid) and control volume concepts (consider a given volume). To do this we need an analytical tool to shift from one representation to the other. The Reynolds transport theorem provides this tool. All physical laws are stated in terms of various physical parameters. Velocity, acceleration, mass, temperature, and momentum are but a few of the more common parameters. Let B represent any of these (or other) fluid parameters and b represent the amount of that parameter per unit mass. That is, B = mb m B b = B/m m 1 mV V 1 _ mV 2 2 1 _V 2 2 where m is the mass of the portion of fluid of interest. For example, as shown by the figure in the margin, if B = m, the mass, it follows that b = 1. The mass per unit mass is unity. If B = mV 2 2, the kinetic energy of the mass, then b = V 2 2, the kinetic energy per unit mass. The parameters B and b may be scalars or vectors. Thus, if B = mV, the momentum of the mass, then b = V. (The momentum per unit mass is the velocity.) The parameter B is termed an extensive property, and the parameter b is termed an intensive property. The value of B is directly proportional to the amount of the mass being considered, whereas the value of b is independent of the amount of mass. The amount of an extensive property that a system possesses at a given instant, Bsys, can be determined by adding up the amount associated with each fluid particle in the system. For infinitesimal fluid particles of volume δ V and mass 176 Chapter 4 ■ Fluid Kinematics ρ δV, this summation (in the limit of δV → 0) takes the form of an integration over all the particles in the system and can be written as Bsys = lim ∑ bi (ρi δVi ) = δV→0 i ∫ ρb dV sys The limits of integration cover the entire system—a (usually) moving volume. We have used the fact that the amount of B in a fluid particle of mass ρ δV is given in terms of b by δB = bρ δV. Most of the laws governing fluid motion involve the time rate of change of an extensive property of a fluid system—the rate at which the momentum of a system changes with time, the rate at which the mass of a system changes with time, and so on. Thus, we often encounter terms such as dBsys dt = d ( ∫ sys ρb dV) (4.8) dt To formulate the laws into a control volume approach, we must obtain an expression for the time rate of change of an extensive property within a control volume, Bcv, not within a system. This can be written as d ( dBcv = dt Differences between control volume and system concepts are subtle but very important. ∫ cv ρb dV) (4.9) dt where the limits of integration, denoted by cv, cover the control volume of interest. Although Eqs. 4.8 and 4.9 may look very similar, the physical interpretation of each is quite different. Mathematically, the difference is represented by the difference in the limits of integration. Recall that the control volume is a volume in space (in most cases stationary, although if it moves it need not move with the system). On the other hand, the system is an identifiable collection of mass that moves with the fluid (indeed it is a specified portion of the fluid). We will learn that even for those instances when the control volume and the system momentarily occupy the same volume in space, the two quantities dBsysdt and dBcv dt need not be the same. The Reynolds transport theorem provides the relationship between the time rate of change of an extensive property for a system and that for a control volume—the relationship between Eqs. 4.8 and 4.9. EXAMPLE 4.7 Time Rate of Change for a System and a Control Volume GIVEN Fluid flows from the fire extinguisher tank shown in Fig. E4.7a. FIND Discuss the differences between dBsys dt and dBcv dt if B represents mass. SOLUTION With B = m, the system mass, it follows that b = 1 and Eqs. 4.8 and 4.9 can be written as dBsys dt ≡ dmsys dt = d( ∫ sys ρ dV) dt and ∫ d( ρ dV) dmcv dBcv cv ≡ = dt dt dt (a) ■ Figure E4.7 (© DNY59/iStockphoto) 4.4 Physically, these represent the time rate of change of mass within the system and the time rate of change of mass within the control volume, respectively. We choose our system to be the fluid within the tank at the time the valve was opened (t = 0) and the control volume to be the tank itself as shown in Fig. E4.7b. A short time after the valve is opened, part of the system has moved outside of the control volume as is shown in Fig. E4.7c. The control volume remains fixed. The limits of integration are fixed for the control volume; they are a function of time for the system. Clearly, if mass is to be conserved (one of the basic laws governing fluid motion), the mass of the fluid in the system is constant, so that d ( ∫ sys ρ d V) dt ∫ cv ρ d V) dt 177 The actual numerical value of the rate at which the mass in the control volume decreases will depend on the rate at which the fluid flows through the nozzle (i.e., the size of the nozzle and the speed and density of the fluid). Clearly the meanings of dBsysdt and dBcvdt are different. For this example, dBcvdt < dBsysdt. Other situations may have dBcvdt ≥ dBsysdt. t=0 t>0 System =0 On the other hand, it is equally clear that some of the fluid has left the control volume through the nozzle on the tank. Hence, the amount of mass within the tank (the control volume) decreases with time, or d( The Reynolds Transport Theorem Control surface (b) (c) ■ Figure E4.7 < 0 4.4.1 Derivation of the Reynolds Transport Theorem A simple version of the Reynolds transport theorem relating system concepts to control volume concepts can be obtained easily for the one-dimensional flow through a fixed control volume such as the variable area duct section shown in Fig. 4.11a. We consider the control volume to be that stationary volume within the duct between sections (1) and (2) as indicated in Fig. 4.11b. The system that we consider is that fluid occupying the control volume at some initial time t. A short time later, at time t + δt, the system has moved slightly to the right. The fluid particles that coincided with section (2) of the control surface at time t have moved a distance δℓ2 = V2 δt to the right, where V2 is the velocity of the fluid as it passes section (2). Similarly, the fluid initially at section (1) has moved a distance δℓ1 = V1 δt, where V1 is the fluid velocity at section (1). We assume the fluid flows across sections (1) and (2) in a direction normal to these surfaces and that V1 and V2 are constant across sections (1) and (2). As is shown in Fig. 4.11c, the outflow from the control volume from time t to t + δt is denoted as volume II, the inflow as volume I, and the control volume itself as CV. Thus, the system at time t consists of the fluid in section CV; that is, “SYS = CV” at time t. At time t + δt the system consists of the same fluid that now occupies sections (CV − I) + II. That is, “SYS = CV − I + II” at time t + δt. The control volume remains as section CV for all time. The moving system flows through the fixed control volume. δ ℓ2 = V2 δ t V1 δ ℓ1 = V1 δ t V2 I CV – I (1) (1) (2) (2) Fixed control surface and system boundary at time t System boundary at time t + δ t (a) II (b) ■ Figure 4.11 Control volume and system for flow through a variable area pipe. (c) 178 Chapter 4 ■ Fluid Kinematics If B is an extensive parameter of the system, then the value of it for the system at time t is Bsys (t) = Bcv (t) since the system and the fluid within the control volume coincide at this time. Its value at time t + δt is Bsys (t + δt) = Bcv (t + δt) − BI (t + δt) + BII (t + δt) Thus, the change in the amount of B in the system, δBsys, in the time interval δt divided by this time interval is given by δBsys δt The time rate of change of a system property is a Lagrangian concept. V2 = Bsys (t + δt) − Bsys (t) δt = Bcv (t + δt) − BI (t + δt) + BII (t + δt) − Bsys (t) δt By using the fact that at the initial time t we have Bsys (t) = Bcv (t), this ungainly expression may be rearranged as follows. δBsys δt = Bcv (t + δt) − Bcv (t) BI (t + δt) BII (t + δt) − + δt δt δt (4.10) In the limit δt → 0, the left-hand side of Eq. 4.10 is equal to the time rate of change of B for the system and is denoted as DBsysDt. We use the material derivative notation, D( )Dt, to denote this time rate of change to emphasize the Lagrangian character of this term. (Recall from Section 4.2.1 that the material derivative, DPDt, of any quantity P represents the time rate of change of that quantity associated with a given fluid particle as it moves along.) Similarly, the quantity DBsysDt represents the time rate of change of property B associated with a system (a given portion of fluid) as it moves along. In the limit δt → 0, the first term on the right-hand side of Eq. 4.10 is seen to be the time rate of change of the amount of B within the control volume ∂ ( Bcv (t + δt) − Bcv (t) ∂Bcv lim = = δt→0 δt ∂t ∫ cv ρb dV) ∂t (4.11) The third term on the right-hand side of Eq. 4.10 represents the rate at which the extensive parameter B flows from the control volume across the control surface. As indicated by the figure in the margin, during the time interval from t = 0 to t = δt the volume of fluid that flows across section (2) is given by δVII = A2 δℓ2 = A2 (V2δt). Thus, the amount of B within region II, the outflow region, is its amount per unit volume, ρb, times the volume (2) t=0 BII (t + δt) = (ρ2b2 )(δ VII ) = ρ2b2A2V2 δt δ VII V2δ t where b2 and ρ2 are the constant values of b and ρ across section (2). Thus, the rate at which this property flows from the control volume, Ḃ out, is given by BII (t + δt) Ḃ out = lim = ρ2 A2V2 b2 δt→0 δt (4.12) Similarly, the inflow of B into the control volume across section (1) during the time interval δt corresponds to that in region I and is given by the amount per unit volume times the volume, δ VI = A1 δℓ1 = A1 (V1 δt). Hence, BI (t + δt) = (ρ1b1 )(δ V1 ) = ρ1b1A1V1 δt where b1 and ρ1 are the constant values of b and ρ across section (1). Thus, the rate of inflow of the property B into the control volume, Ḃ in, is given by (2) t = δt BI (t + δt) Ḃ in = lim = ρ1A1V1b1 δt→0 δt (4.13) 4.4 The Reynolds Transport Theorem 179 If we combine Eqs. 4.10, 4.11, 4.12, and 4.13, we see that the relationship between the time rate of change of B for the system and that for the control volume is given by DBsys ∂Bcv + Ḃ out − Ḃ in δt (4.14) ∂Bcv + ρ2A2V2b2 − ρ1A1V1b1 ∂t (4.15) Dt = or DBsys Dt = This is a version of the Reynolds transport theorem valid under the restrictive assumptions associated with the flow shown in Fig. 4.11—fixed control volume with one inlet and one outlet having uniform properties (density, velocity, and the parameter b) across the inlet and outlet with the velocity normal to sections (1) and (2). Note that the time rate of change of B for the system (the left-hand side of Eq. 4.15 or the quantity in Eq. 4.8) is not necessarily the same as the rate of change of B within the control volume (the first term on the right-hand side of Eq. 4.15 or the quantity in Eq. 4.9). This is true because the inflow rate (b1ρ1V1A1 ) and the outflow rate (b2 ρ2V2 A2 ) of the property B for the control volume need not be the same. The time derivative associated with a system may be different from that for a control volume. E X AMPLE 4.8 Use of the Reynolds Transport Theorem GIVEN Consider again the flow from the fire extinguisher shown in Fig. E4.7. Let the extensive property of interest be the system mass (B = m, the system mass, or b = 1). FIND Write the appropriate form of the Reynolds transport theorem for this flow. SOLUTION Again we take the control volume to be the fire extinguisher and the system to be the fluid within it at time t = 0. For this case there is no inlet, section (1), across which the fluid flows into the control volume (A1 = 0). There is, however, an outlet, section (2). Thus, the Reynolds transport theorem, Eq. 4.15, along with Eq. 4.9 with b = 1 can be written as Dmsys Dt = ∂( ∫ cv ρ d V) ∂t + ρ2 A2V2 (1) (Ans) law of conservation of mass, we may set the left-hand side of this equation equal to zero (the amount of mass in a system is constant) and rewrite Eq. 1 in the form ∫ cv ρ d V) ∂t ∂( ∫ cv ρ d V) ∂t COMMENT If we proceed one step further and use the basic ∂( opposite to the rate of flow of mass from the exit, ρ2 A2V2. Note the units for the two terms of Eq. 2 (kgs or slugss). If there were both an inlet and an outlet to the control volume shown in Fig. E4.7, Eq. 2 would become = −ρ2 A2V2 (2) = ρ1 A1V1 − ρ2 A2V2 (3) In addition, if the flow were steady, the left-hand side of Eq. 3 would be zero (the amount of mass in the control would be constant in time) and Eq. 3 would become ρ1 A1V1 = ρ2 A2V2 This is one form of the conservation of mass principle discussed in Section 3.6.2—the mass flowrates into and out of the control volume are equal. Other more general forms are discussed in Chapter 5. The physical interpretation of this result is that the rate at which the mass in the tank decreases in time is equal in magnitude but Right Atrium Left Atrium Left Ventricle Right Ventricle Equation 4.15 is a simplified version of the Reynolds transport theorem. We will now derive it for much more general conditions. A general, fixed control volume with fluid flowing through it is shown in Fig. 4.12. The flow field may be quite simple (as in the above one-dimensional flow considerations), or it may involve a quite complex, unsteady, three-dimensional situation such as the flow through a human heart as illustrated by the figure in the margin. In any case we again consider the system to be the fluid within the control volume at the initial time t. A short time later a portion of the fluid (region II) has exited from the control volume, and additional fluid (region I, not part of the original system) has entered the control volume. Free ebooks ==> www.Ebook777.com 180 Chapter 4 ■ Fluid Kinematics V2 II Inflow V3 Ib IIc CV–I Ia I IId V1 V5 Outflow II b IIa Fixed control surface and system boundary at time t System boundary at time t + δ t ■ Figure 4.13 Typical control volume with more than one inlet and outlet. flow through an arbitrary, fixed control volume. The simplified Reynolds transport theorem can be easily generalized. V4 V6 ■ Figure 4.12 Control volume and system for We consider an extensive fluid property B and seek to determine how the rate of change of B associated with the system is related to the rate of change of B within the control volume at any instant. By repeating the exact steps that we did for the simplified control volume shown in Fig. 4.11, we see that Eq. 4.14 is valid for the general case also, provided that we give the correct interpretation to the terms Ḃ out and Ḃ in. In general, the control volume may contain more (or less) than one inlet and one outlet. A typical pipe system may contain several inlets and outlets as are shown in Fig. 4.13. In such instances we think of all inlets grouped together (I = Ia + Ib + Ic + … ) and all outlets grouped together (II = IIa + IIb + IIc + … ), at least conceptually. The term Ḃ out represents the net flowrate of the property B from the control volume. Its value can be thought of as arising from the addition (integration) of the contributions through each infinitesimal area element of size δA on the portion of the control surface dividing region II and the control volume. This surface is denoted CSout. As is indicated in Fig. 4.14, in time δt the volume of fluid that passes across each area element is given by δ V = δℓn δA, where δℓn = δℓ cos θ is the height (normal to the base, δA) of the small volume element, and θ is the angle between the velocity vector and the outward pointing normal to the surface, nˆ . Thus, since δℓ = V δt, the amount of the property B carried across the area element δA in the time interval δt is given by δB = bρ δ V = bρ(V cos θ δt) δA The rate at which B is carried out of the control volume across the small area element δA, denoted δḂ out, is ρb δ V (ρbV cos θ δt) δA = lim = ρbV cos θ δA δḂ out = lim δt→0 δt→0 δt δt Outflow portion of control surface CSout δ V = δ ℓn δ A n n δA θ δA V (a) δ ℓn ^ ^ ^ n θ θ δℓ = Vδt V (b) ■ Figure 4.14 Outflow across a typical portion of the control surface. www.Ebook777.com V δℓ (c) 4.4 181 The Reynolds Transport Theorem Inflow portion of control surface CSin δ ℓn δ ℓ = V δt δA ^ n ^ n ^ n θ θ θ V ( a) δℓ V δ V = δ ℓn δ A (b) V (c) ■ Figure 4.15 Inflow across a typical portion of the control surface. By integrating over the entire outflow portion of the control surface, CSout, we obtain Ḃ out = ∫ dḂ out = csout The flowrate of a parameter across the control surface is written in terms of a surface integral. ∫ ρbV cos θ dA csout The quantity V cos θ is the component of the velocity normal to the area element δA. From the definition of the dot product, this can be written as V cos θ = V ∙ nˆ . Hence, an alternate form of the outflow rate is Ḃ out = ∫ ρbV ∙ nˆ dA (4.16) csout In a similar fashion, by considering the inflow portion of the control surface, CSin, as shown in Fig. 4.15, we find that the inflow rate of B into the control volume is Ḃ in = − ∫ ρbV cos θ dA = − csin ∫ ρbV ∙ nˆ dA (4.17) csin We use the standard notation that the unit normal vector to the control surface, nˆ , points out from the control volume. Thus, as is shown in Fig. 4.16, −90° < θ < 90° for outflow regions (the normal component of V is positive; V ∙ nˆ > 0). For inflow regions 90° < θ < 270° (the normal component of V is negative; V ∙ nˆ < 0). The value of cos θ is, therefore, positive on the CVout portions of the control surface and negative on the CVin portions. Over the remainder of the control surface, there is no inflow or outflow, leading to V ∙ nˆ = V cos θ = 0 on those portions. On such portions either V = 0 (the fluid “sticks” to the surface) or cos θ = 0 (the fluid “slides” along the surface without crossing it) (see Fig. 4.16). Therefore, the net flux (flowrate) of parameter B across the entire control surface is Ḃ out − Ḃ in = ∫ csout = ρbV ∙ nˆ dA − (− ∫ csin ∫ ρbV ∙ nˆ dA ρbV ∙ nˆ dA) (4.18) cs where the integration is over the entire control surface. CS CSin CSout V=0 ^ n ^ n ^ n ■ Figure 4.16 Possible velocity ^ • V n>0 V •^ n<0 (a) V •^ n=0 (b) (c) configurations on portions of the control surface: (a) inflow, (b) no flow across the surface, (c) outflow. 182 Chapter 4 ■ Fluid Kinematics By combining Eqs. 4.14 and 4.18 we obtain DBsys Dt = ∂Bcv + ∂t ∫ ρbV ∙ nˆ dA cs This can be written in a slightly different form by using Bcv = ʃcv ρb dV so that DBsys Dt = ∂ ∂t ∫ ρb dV + cv ∫ ρb V ∙ nˆ dA (4.19) cs Equation 4.19 is the general form of the Reynolds transport theorem for a fixed, nondeforming control volume. Its interpretation and use are discussed in the following sections. 4.4.2 Physical Interpretation Control surface V •^ n=0 V •^ n>0 V •^ n<0 The Reynolds transport theorem as given in Eq. 4.19 is widely used in fluid mechanics (and other areas as well). At first it appears to be a rather formidable mathematical expression—perhaps one to avoid if possible. However, a physical understanding of the concepts involved will show that it is a rather straightforward, relatively easy-to-use tool. Its purpose is to provide a link between control volume ideas and system ideas. The left side of Eq. 4.19 is the time rate of change of an arbitrary extensive parameter of a system. This may represent the rate of change of mass, momentum, energy, or angular momentum of the system, depending on the choice of the parameter B. Because the system is moving and the control volume is stationary, the time rate of change of the amount of B within the control volume is not necessarily equal to that of the system. The first term on the right side of Eq. 4.19 represents the rate of change of B within the control volume as the fluid flows through it. Recall that b is the amount of B per unit mass, so that ρb dV is the amount of B in a small volume dV. Thus, the time derivative of the integral of ρb throughout the control volume is the time rate of change of B within the control volume at a given time. If this time derivative is positive, B (whatever it is) is accumulating in the control volume. The last term in Eq. 4.19 (an integral over the control surface) represents the net flowrate of the parameter B across the entire control surface. As illustrated by the figure in the margin, over a portion of the control surface this property is being carried out of the control volume (V ∙ nˆ > 0); over other portions it is being carried into the control volume (V ∙ nˆ < 0). Over the remainder of the control surface there is no transport of B across the surface since bV ∙ nˆ = 0, because either b = 0, V = 0, or V is parallel to the surface at those locations. The mass flowrate through area element δA, given by ρV ∙ nˆ δA, is positive for outflow (efflux) and negative for inflow (influx). Each fluid particle or fluid mass carries a certain amount of B with it, as given by the product of B per unit mass, b, and the mass. The rate at which this B is carried across the control surface is given by the area integral term of Eq. 4.19. This net rate across the entire control surface may be negative, zero, or positive depending on the particular situation involved. 4.4.3 Relationship to Material Derivative The Reynolds transport theorem is the integral counterpart of the material derivative. In Section 4.2.1 we discussed the concept of the material derivative D( )Dt= ∂ ( )∂t +V ∙ ∇ ( ) = ∂ ( )∂t + u ∂ ( ) ∂x + υ ∂ ( )∂y + w ∂ ( )∂z. The physical interpretation of this derivative is that it provides the time rate of change of a fluid property (temperature, velocity, etc.) associated with a particular fluid particle as it flows. The value of that property for that particle may change because of unsteady effects [the ∂ ( )∂t term] or because of effects associated with the particle’s motion [the V ∙ ∇ ( ) term]. Careful consideration of Eq. 4.19 indicates the same type of physical interpretation for the Reynolds transport theorem. The term involving the time derivative of the control volume integral represents unsteady effects associated with the fact that values of the parameter within the control volume may change with time. For steady flow this effect vanishes—fluid flows through the control volume, but the amount of any property, B, within the control volume is constant in time. The term involving the control surface integral represents the convective effects associated with the flow of 4.4 The Reynolds Transport Theorem 183 the system across the fixed control surface. The sum of these two terms gives the rate of change of the parameter B for the system. This corresponds to the interpretation of the material derivative, D( )Dt = ∂ ( )∂t + V ∙ ∇ ( ), in which the sum of the unsteady effect and the convective effect gives the rate of change of a parameter for a fluid particle. As is discussed in Section 4.2, the material derivative operator may be applied to scalars (such as temperature) or vectors (such as velocity). This is also true for the Reynolds transport theorem. The particular parameters of interest, B and b, may be scalars or vectors. Thus, both the material derivative and the Reynolds transport theorem equations represent ways to transfer from the Lagrangian viewpoint (follow a particle or follow a system) to the Eulerian viewpoint (observe the fluid at a given location in space or observe what happens in the fixed control volume). The material derivative (Eq. 4.5) is essentially the infinitesimal (or derivative) equivalent of the finite size (or integral) Reynolds transport theorem (Eq. 4.19). 4.4.4 Steady Effects Consider a steady flow [∂ ( )∂t ≡ 0] so that Eq. 4.19 reduces to DBsys = Dt Control volume Momentum out (Photograph courtesy of NASA.) ∫ ρbV ∙ nˆ dA (4.20) cs In such cases if there is to be a change in the amount of B associated with the system (nonzero lefthand side), there must be a net difference in the rate that B flows into the control volume compared with the rate that it flows out of the control volume. That is, the integral of ρbV ∙ nˆ over the inflow portions of the control surface would not be equal and opposite to that over the outflow portions of the surface. Consider steady flow through the “black box” control volume that is shown in Fig. 4.17. If the parameter B is the mass of the system, the left-hand side of Eq. 4.20 is zero (conservation of mass for the system as discussed in detail in Section 5.1). Hence, the flowrate of mass into the box must be the same as the flowrate of mass out of the box because the right-hand side of Eq. 4.20 represents the net flowrate through the control surface. On the other hand, assume the parameter B is the momentum of the system. The momentum of the system need not be constant. In fact, according to Newton’s second law the time rate of change of the system momentum equals the net force, F, acting on the system. In general, the left-hand side of Eq. 4.20 will therefore be nonzero. Thus, the right-hand side, which then represents the net flux of momentum across the control surface, will be nonzero. The flowrate of momentum into the control volume need not be the same as the flowrate of momentum from the control volume. We will investigate these concepts much more fully in Chapter 5. They are the basic principles describing the operation of such devices as jet or rocket engines like the one shown in the figure in the margin. For steady flows the amount of the property B within the control volume does not change with time. The amount of the property associated with the system may or may not change with time, depending on the particular property considered and the flow situation involved. The difference between that associated with the control volume and that associated with the system is determined by the rate at which B is carried across the control surface—the term ʃcs ρbV ∙ nˆ dA. Control volume Vin Vout F ■ Figure 4.17 Steady flow through a control volume. 184 Chapter 4 ■ Fluid Kinematics y ^ n = –i^ V2 = V0(t) V1 = V0(t) V0^ i (1) x ^ ^ n=i (2) ^ n = –j ^ ■ Figure 4.18 Unsteady flow through a constant diameter pipe. Control surface 4.4.5 Unsteady Effects The Reynolds transport theorem involves both steady and unsteady effects. For some flow situations, certain portions of the Reynolds transport theorem are automatically zero. Consider unsteady flow [∂ ( )∂t ≠ 0] so that all terms in Eq. 4.19 must be retained. When they are viewed from a control volume standpoint, the amount of parameter B within the system may change because the amount of B within the fixed control volume may change with time [the ∂ ( ʃcv ρb dV)∂t term] and because there may be a net nonzero flow of that parameter across the control surface (the ʃcs ρbV ∙ nˆ dA term). For the special unsteady situations in which the rate of inflow of parameter B is exactly balanced by its rate of outflow, it follows that ʃcs ρbV ∙ nˆ dA = 0, and Eq. 4.19 reduces to DBsys Dt = ∂ ∂t ∫ (4.21) ρb d V cv For such cases, any rate of change in the amount of B associated with the system is equal to the rate of change of B within the control volume. This can be illustrated by considering flow through a constant diameter pipe as is shown in Fig. 4.18. The control volume is as shown, and the system is the fluid within this volume at time t0. We assume the flow is one-dimensional with V = V0ˆi, where V0 (t) is a function of time, and that the density is constant. At any instant, all particles in the system have the same velocity. We let B = system momentum = mV = mV0ˆi, where m is the system mass, so that b = Bm = V = V0ˆi, the fluid velocity. The magnitude of the momentum efflux across the outlet [section (2)] is the same as the magnitude of the momentum influx across the inlet [section (1)]. However, the sign of the efflux is opposite to that of the influx since V ∙ nˆ > 0 for the outflow and V ∙ nˆ < 0 for the inflow. Note that V ∙ nˆ = 0 along the sides of the control volume. Thus, with V ∙ nˆ = −V0 on section (1), V ∙ nˆ = V0 on section (2), and A1 = A2 , we obtain ∫ ρbV ∙ nˆ dA = cs ∫ ρ(V0ˆi )(V ∙ nˆ ) dA ∫ ρ(V0ˆi )(−V0 ) dA + cs = (1) ∫ ρ(V0ˆi )(V0 ) dA (2) = −ρV 20 A1ˆi + ρV 20 A2 ˆi = 0 It is seen that for this special case Eq. 4.21 is valid. The rate at which the momentum of the system changes with time is the same as the rate of change of momentum within the control volume. If V0 is constant in time, there is no rate of change of momentum of the system, and for this special case each of the terms in the Reynolds transport theorem is zero by itself. Consider the flow through a variable area pipe shown in Fig. 4.19. In such cases the fluid velocity is not the same at section (1) as it is at (2). Hence, the efflux of momentum from the control y V2 < V1 ^ ^ n = –i x V1 (1) Control surface ^ ^ n=i (2) ■ Figure 4.19 Flow through a variable area pipe. 4.4 The Reynolds Transport Theorem 185 Moving vane VCV = V0 Nozzle Control volume moves with speed V0 V0 V1 ■ Figure 4.20 Example of a moving control volume. volume is not equal to the influx of momentum, so that the convective term in Eq. 4.20 [the integral of ρV(V ∙ nˆ ) over the control surface] is not zero. These topics will be discussed in considerably more detail in Chapter 5. 4.4.6 Moving Control Volumes The relative velocity is the difference between the absolute velocity and the velocity of the control volume. For most problems in fluid mechanics, the control volume may be considered as a fixed volume through which the fluid flows. There are, however, situations for which the analysis is simplified if the control volume is allowed to move or deform. The most general situation would involve a control volume that moves, accelerates, and deforms. As one might expect, the use of these control volumes can become fairly complex. A number of important problems can be most easily analyzed by using a nondeforming control volume that moves with a constant velocity. An example is shown in Fig. 4.20 in which a stream of water with velocity V1 strikes a vane that is moving with constant velocity V0. It may be of interest to determine the force, F, that the water exerts on the vane. This type of problem frequently occurs in turbines where a stream of fluid (water or steam, for example) strikes a series of blades that move past the nozzle. To analyze such problems it is advantageous to use a moving control volume. We will obtain the Reynolds transport theorem for such control volumes. We consider a control volume that moves with a constant velocity as is shown in Fig. 4.21. The shape, size, and orientation of the control volume do not change with time. The control volume merely translates with a constant velocity, Vcv, as shown. In general, the velocity of the control volume and that of the fluid are not the same, so that there is a flow of fluid through the moving control volume just as in the stationary control volume cases discussed in Section 4.4.2. The main difference between the fixed and the moving control volume cases is that it is the relative velocity, W, that carries fluid across the moving control surface, whereas it is the absolute velocity, V, that carries the fluid across the fixed control surface. The relative velocity is the fluid velocity relative to the moving control volume—the fluid velocity seen by an observer riding along on the control volume. The absolute velocity is the fluid velocity as seen by a stationary observer in a fixed coordinate system. The relative velocity is the difference between the absolute velocity and the velocity of the control volume, W = V − Vcv, or V = W + Vcv Particle A at t0 Particle B at t0 VA At t1 (4.22) Control volume and system at time t0 Control volume at time t1 > t0 System at time t1 > t0 At t1 VB VCV = Control volume velocity ■ Figure 4.21 Typical moving control volume and system. 186 Chapter 4 ■ Fluid Kinematics VA = Absolute velocity of A VB VCV VCV The Reynolds transport theorem for a moving control volume involves the relative velocity. WB = Velocity of B relative WA = Velocity of A relative to control volume to control volume ■ Figure 4.22 Relationship between absolute and relative velocities. Since the velocity is a vector, we must use vector addition as is shown in Fig. 4.22 to obtain the relative velocity if we know the absolute velocity and the velocity of the control volume. Thus, if the water leaves the nozzle in Fig. 4.20 with a velocity of V1 = 100iˆ fts and the vane has a velocity of V0 = 20iˆ fts (the same as the control volume), it appears to an observer riding on the vane that the water approaches the vane with a velocity of W = V − Vcv = 80iˆ fts. In general, the absolute velocity, V, and the control volume velocity, Vcv, will not be in the same direction so that the relative and absolute velocities will have different directions (see Fig. 4.22). The Reynolds transport theorem for a moving, nondeforming control volume can be derived in the same manner that it was obtained for a fixed control volume. As is indicated in Fig. 4.23, the only difference that needs be considered is the fact that relative to the moving control volume the fluid velocity observed is the relative velocity, not the absolute velocity. An observer fixed to the moving control volume may or may not even know that he or she is moving relative to some fixed coordinate system. If we follow the derivation that led to Eq. 4.19 (the Reynolds transport theorem for a fixed control volume), we note that the corresponding result for a moving control volume can be obtained by simply replacing the absolute velocity, V, in that equation by the relative velocity, W. Thus, the Reynolds transport theorem for a control volume moving with constant velocity is given by DBsys Dt = ∂ ∂t ∫ ρb d V + cv ∫ ρb W ∙ nˆ dA (4.23) cs where the relative velocity is given by Eq. 4.22. 4.4.7 Selection of a Control Volume Any volume in space can be considered as a control volume. It may be of finite size or it may be infinitesimal in size, depending on the type of analysis to be carried out. In most of our cases, the control volume will be a fixed, nondeforming volume. In some situations, we will consider control volumes that move with constant velocity. In either case, it is important that considerable thought go into the selection of the specific control volume to be used. The selection of an appropriate control volume in fluid mechanics is very similar to the selection of an appropriate free-body diagram in dynamics or statics. In dynamics, we select the body in which we are interested, represent the object in a free-body diagram, and then apply the appropriate governing laws to that body. The ease of solving a given dynamics problem is often very dependent on the specific object that we select for use in our free-body diagram. Similarly, the ease of solving a given fluid mechanics problem is often very dependent on the choice of the control volume used. Control volume and system at time t System at time t + δt Pathlines as seen from the moving control volume W = V – VCV Flow as seen by an observer moving with velocity VCV ■ Figure 4.23 Control volume and system as seen by an observer moving with the control volume. 4.5 Chapter Summary and Study Guide 187 Control surface V (1) V (a) (1) V (1) (b) (c) ■ Figure 4.24 Various control volumes for flow through a pipe. VIDEO V4.14 Control volume ^ n θ V Control surface 4.5 field representation velocity field Eulerian method Lagrangian method one-, two-, and threedimensional flow steady and unsteady flow streamline streakline pathline acceleration field material derivative local acceleration convective acceleration system control volume Reynolds transport theorem Only by practice can we develop skill at selecting the “best” control volume. None are “wrong,” but some are “much better” than others. Solution of a typical problem will involve determining parameters such as velocity, pressure, and force at some point in the flow field. It is best to ensure that this point is located on the control surface, not “buried” within the control volume. The unknown will then appear in the convective term (the surface integral) of the Reynolds transport theorem. If possible, the control surface should be normal to the fluid velocity so that the angle θ (V ∙ nˆ = V cos θ as shown by the figure in the margin) in the flux terms of Eq. 4.19 will be 0 or 180°. This will usually simplify the solution process. Figure 4.24 illustrates three possible control volumes associated with flow through a pipe. If the problem is to determine the pressure at point (1), the selection of the control volume (a) is better than that of (b) because point (1) lies on the control surface. Similarly, control volume (a) is better than (c) because the flow is normal to the inlet and exit portions of the control volume. None of these control volumes are wrong—(a) will be easier to use. Proper control volume selection will become much clearer in Chapter 5 where the Reynolds transport theorem is used to transform the governing equations from the system formulation into the control volume formulation, and numerous examples using control volume ideas are discussed. Chapter Summary and Study Guide This chapter considered several fundamental concepts of fluid kinematics. That is, various aspects of fluid motion are discussed without regard to the forces needed to produce this motion. The concepts of a field representation of a flow and the Eulerian and Lagrangian approaches to describing a flow are introduced, as are the concepts of velocity and acceleration fields. The properties of one-, two-, or three-dimensional flows and steady or unsteady flows are introduced along with the concepts of streamlines, streaklines, and pathlines. Streamlines, which are lines tangent to the velocity vectors, are identical to streaklines and pathlines if the flow is steady. For unsteady flows, they need not be identical. As a fluid particle moves about, its properties (e.g., velocity, density, temperature) may change. The rate of change of these properties can be obtained by using the material derivative, which involves both unsteady effects (time rate of change at a fixed location) and convective effects (time rate of change due to the motion of the particle from one location to another). The concepts of a control volume and a system are introduced, and the Reynolds transport theorem is developed. By using these ideas, the analysis of flows can be carried out using a control volume (a volume, usually fixed, through which the fluid flows), whereas the governing principles are stated in terms of a system (a flowing portion of fluid). The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. explain the concept of the field representation of a flow and the difference between Eulerian and Lagrangian methods of describing a flow. explain the differences among streamlines, streaklines, and pathlines. calculate and plot streamlines for flows with given velocity fields. 188 Chapter 4 ■ Fluid Kinematics ■ ■ ■ ■ use the concept of the material derivative, with its unsteady and convective effects, to determine time rate of change of a fluid property. determine the acceleration field for a flow with a given velocity field. discuss the properties of and differences between a system and a control volume. interpret, physically and mathematically, the concepts involved in the Reynolds transport theorem. Some of the important equations in this chapter are: dy υ = u dx Equation for streamlines a= Acceleration Material derivative Reynolds transport theorem (general form) ∂V ∂V ∂V ∂V +u +υ +w ∂t ∂x ∂y ∂z D( ) ∂( ) = + (V ∙ ∇ )( ) Dt ∂t Streamwise and normal components of acceleration Reynolds transport theorem (restricted form) (4.1) as = V DBsys Dt DBsys Dt Relative and absolute velocities ∂V V2 ,  an = ∂s ℛ (4.3) (4.6) (4.7) = ∂Bcv + ρ2A2V2b2 − ρ1A1V1b1 ∂t (4.15) = ∂ ∂t (4.19) ∫ ρb d V + ∫ ρb V ∙ nˆ dA cv V = W + Vcv cs (4.22) References 1. Streeter, V. L., and Wylie, E. B., Fluid Mechanics, 8th Ed., McGraw-Hill, New York, 1985. 2. Goldstein, R. J., Fluid Mechanics Measurements, Hemisphere, New York, 1983. 3. Homsy, G. M., et al., Multimedia Fluid Mechanics CD-ROM, 2nd Ed., Cambridge University Press, New York, 2007. 4. Magarvey, R. H., and MacLatchy, C. S., The Formation and Structure of Vortex Rings, Canadian Journal of Physics, Vol. 42, 1964. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are no unique answers to these problems. Problems 189 Problems Note: Unless specific values of required fluid properties are *4.7 A car accelerates from rest to a final constant velocity Vf and given in the problem statement, use the values found in the a police officer records the following velocities at various locations tables on the inside of the front cover. x along the highway: x=0 V = 0 mph x = 100 ft V = 34.8 mph Section 4.1 The Velocity Field x = 200 ft V = 47.6 mph 4.1 Obtain a photograph/image of a situation in which a fluid is x = 300 ft V = 52.3 mph flowing. Print this photo and draw in some lines to represent how x = 400 ft V = 54.0 mph you think some streamlines may look. Write a brief paragraph to describe the acceleration of a fluid particle as it flows along one x = 1000 ft V = 55.0 mph = Vf . of these streamlines. Find a mathematical Eulerian expression for the velocity V traveled 4.2 The surface velocity of a river is measured at several locations by the car as a function of the final velocity Vf of the car and x if t = 0 at V = 0. [Hint: Try an exponential fit to the data.] x and can be reasonably represented by V = V0 + ΔV(1 − e−ax ), where V0, ΔV, and a are constants. Find the Lagrangian description of the velocity of a fluid particle flowing along the surface if x = 0 at time t = 0. 4.3 The velocity field of a flow is given by V = 2x2tiˆ + [4y(t − 1) + 2x2t]jˆ m/s, where x and y are in meters and t is in seconds. For fluid particles on the x axis, determine the speed and direction of flow. 4.4 A two-dimensional velocity field is given by u = 1 + y and υ = 1. Determine the equation of the streamline that passes through the origin. On a graph, plot this streamline. 4.5 Streamlines are given in Cartesian coordinates by the equation. y .  x2 + y2 ≥ 1. ψ = U(y − 2 x + y2 ) Plot the streamlines for ψ = 0, ±60.0625, and describe the physical situation represented by this equation. The parameter U is an upstream uniform velocity of 1.0 ft/s. A flow can be visualized by plotting the velocity field as 4.6 velocity vectors at representative locations in the flow as shown in Video V4.2 and Fig. E4.1. Consider the velocity field given in polar coordinates by υr = −10r, and υθ = 10r. This flow approximates a fluid swirling into a sink as shown in Fig. P4.6. Plot the velocity field at locations given by r = 1, 2, and 3 with θ = 0, 30, 60, and 90°. 4.8 The components of a velocity field are given by u = x + y, υ = xy3 + 16, and w = 0. Determine the location of any stagnation points (V = 0) in the flow field. 4.9 A two-dimensional, unsteady velocity field is given by u = 5x(1 + t) and v = 5y(−1 + t), where u is the x-velocity component and v the y-velocity component. Find x(t) and y(t) if x = x0 and y = y0 at t = 0. Do the velocity components represent an Eulerian description or a Lagrangian description? 4.10 The velocity field of a flow is given by u = −V0y(x2 + y2)1/2 and υ = V0x(x2 + y2)½, where V0 is a constant. Where in the flow field is the speed equal to V0? Determine the equation of the streamlines and discuss the various characteristics of this flow. 4.11 A velocity field is given by V = xiˆ + x(x − 1) (y + 1)jˆ , where u and υ are in ft/s and x and y are in feet. Plot the streamline that passes through x = 0 and y = 0. Compare this streamline with the streakline through the origin. 4.12 From time t = 0 to t = 5 hr radioactive steam is released from a nuclear power plant accident located at x = −1 mile and y = 3 miles. The following wind conditions are expected: V = 10iˆ − 5jˆ mph for 0 < t < 3 hr, V = 15iˆ + 8jˆ mph for 3 < t < 10 hr, and V = 5iˆ mph for t > 10 hr. Draw to scale the expected streakline of the steam for t = 3, 10, and 15 hr. 4.13 The x and y components of a velocity field are given by u = x2y and υ = −xy2. Determine the equation for the streamlines of this flow and compare it with those in Example 4.2. Is the flow in this problem the same as that in Example 4.2? Explain. 4.14 In addition to the customary horizontal velocity components of the air in the atmosphere (the “wind”), there often are vertical air currents (thermals) caused by buoyant effects due to uneven heating of the air as indicated in Fig. P4.14. Assume that the vθ y vr r θ u0 0 ■ Figure P4.6 ■ Figure P4.14 x Free ebooks ==> www.Ebook777.com 190 Chapter 4 ■ Fluid Kinematics velocity field in a certain region is approximated by u = u0, υ = υ0 (1 − yh) for 0 < y < h, and u = u0, υ = 0 for y > h. Plot the shape of the streamline that passes through the origin for values of u0υ0 = 0.5, 1, and 2. 4.19 Air flows steadily through a circular, constant-diameter duct. The air is perfectly inviscid, so the velocity profile is flat across each flow area. However, the air density decreases as the air flows down the duct. Is this a one-, two-, or three-dimensional flow? *4.15 A test car is traveling along a level road at 88 km/hr. In order to study the acceleration characteristics of a newly installed engine, the car accelerates at its maximum possible rate. The test crew records the following velocities at various locations along the level road: 4.20 A constant-density fluid flows in the converging, twodimensional channel shown in Fig. P4.20. The width perpendicular to the paper is quite large compared to the channel height. The velocity in the z direction is zero. The channel half-height, Y, and the fluid x velocity, u, are given by x=0 x = 0.1 km x = 0.2 km x = 0.3 km x = 0.4 km x = 0.5 km x = 1.0 km V V V V V V V = 88.5 km/hr = 93.1 km/hr = 98.3 km/hr = 104.0 km/hr = 110.3 km/hr = 117.2 km/hr = 164.5 km/hr. Y= Y0 y 2 x  and u = u0(1 + )[1 − ( ) ], 1 + x/ℓ ℓ Y Where x, y, Y, and ℓ are in meters, u is in m/s, u0 = 1.0 m/s, and Y0 = 1.0 m. (a) Is this flow steady or unsteady? Is it one-dimensional, two-dimensional, or three-dimensional? (b) Plot the velocity distribution u(y) at x/ℓ = 0, 0.5, and 1.0. Use y/Y values of 0, ±0.2, ±0.4, ±0.6, ±0.8, and ±1.0. A preliminary study shows that these data follow an equation of the form V = A(1 + eBx), where A and B are positive constants. Find A and B and a Lagrangian expression V = V(V0, t), where V0 is the car velocity at time t = 0 when x = 0. †4.16 For any steady flow the streamlines and streaklines are the same. For most unsteady flows this is not true. However, there are unsteady flows for which the streamlines and streaklines are the same. Describe a flow field for which this is true. y Y0 = 1.0 m Y(x) u(x, y) + Yℓ = 0.5 m x 4.17 A tornado has the following velocity components in polar coordinates: Vr = − C1 C2  and Vθ = − . r r ℓ = 5.0 m Note that the air is spiraling inward. Find an equation for the streamlines. r and θ are polar coordinates. ■ Figure P4.20 Two-dimensional channel. 4.18 (See The Wide World of Fluids article titled “Follow those †4.21 Pathlines and streaklines provide ways to visualize flows. particles,” Section 4.1.) Two photographs of four particles in a Another technique would be to instantly inject a line of dye across flow past a sphere are superposed as shown in Fig. P4.18. The time streamlines and observe how this line moves as time increases. For interval between the photos is Δt = 0.002 s. The locations of the example, consider the initially straight dye line injected in front of particles, as determined from the photos, are shown in the table. the circular cylinder shown in Fig. P4.21. Discuss how this dye line (a) Determine the fluid velocity for these particles. (b) Plot a graph would appear at later times. How would you calculate the location of to compare the results of part (a) with the theoretical velocity, this line as a function of time? This type of line is called a timeline. which is given by V = V0 (1 + a3/x3 ), where a is the sphere radius and V0 is the fluid speed far from the sphere. Dye at t = 0 Particle x at t = 0 s (ft) x at t = 0.002 s (ft) 1 2 3 4 −0.500 −0.250 −0.140 −0.120 −0.480 −0.232 −0.128 −0.112 V ■ Figure P4.21 y, ft 4.22 Classify the following flows as one-, two-, or three-dimensional. Sketch a few streamlines for each. t=0 t = 0.002 s –0.4 ■ Figure P4.18 –0.2 x, ft a = 0.1 ft (a) Rainwater flow down a wide driveway (b) Flow in a straight horizontal pipe (c) Flow in a straight pipe inclined upward at a 45o angle (d) Flow in a long pipe that follows the ground in hilly country (e) Flow over an airplane (f) Wind blowing past a tall telephone (g) Flow in the impeller of a centrifugal pump www.Ebook777.com 191 Problems Section 4.2 The Acceleration Field V1 = 4.23 The velocity components of u and v of a two-dimensional flow are given by (2) by bx u = ax + 2 2  and v = ay + 2 2 , xy xy (1) ■ Figure P4.29 where a and b are constants. Calculate the acceleration. 4.24 Air is delivered through a constant-diameter duct by a fan. The air is inviscid, so the fluid velocity profile is “flat” across each cross section. During the fan start-up, the following velocities were measured at the time t and axial positions x: t=0s t = 1.0 s t = 2.0 s t = 3.0 s V2 = 1.0t m/s 0.50t m/s x=0 x = 10 m x = 20 m V = 0 m/s V = 1.00 m/s V = 1.70 m/s V = 2.10 m/s V = 0 m/s V = 1.20 m/s V = 1.80 m/s V = 2.15 m/s V = 0 m/s V = 1.40 m/s V = 1.90 m/s V = 2.20 m/s 4.30 A shock wave is a very thin layer (thickness = ℓ) in a highspeed (supersonic) gas flow across which the flow properties (velocity, density, pressure, etc.) change from state (1) to state (2) as shown in Fig. P4.30. If V1 = 1800 fps, V2 = 700 fps, and ℓ = 10−4 in., estimate the average deceleration of the gas as it flows across the shock wave. How many g’s deceleration does this represent? V1 V V1 V2 V2 ℓ Shock wave x ℓ Calculate the local acceleration, the convective acceleration, ■ Figure P4.30 and the total acceleration at t = 1.0 s and x = 10 m. What is the local acceleration after the fan has reached a steady air †4.31 Estimate the average acceleration of water as it travels flow rate? through the nozzle on your garden hose. List all assumptions and 4.25 Water flows through a constant diameter pipe with a uni- show all calculations. ˆ form velocity given by V = (8/t + 5)j m/s, where t is in seconds. †4.32 A stream of water from the faucet strikes the bottom of the Determine the acceleration at time t = 1, 2, and 10 s. sink. Estimate the maximum acceleration experienced by the water 4.26 The velocity of air in the diverging pipe shown in Fig. P4.26 particles. List all assumptions and show calculations. is given by V1 = 4t fts and V2 = 2t fts, where t is in seconds. (a) Determine the local acceleration at points (1) and (2). (b) Is the 4.33 As a valve is opened, water flows through the diffuser shown average convective acceleration between these two points negative, in Fig. P4.33 at an increasing flowrate so that the velocity along the centerline is given by V = uiˆ = V0 (1 − e−ct ) (1 − xℓ) ˆi, where zero, or positive? Explain. u0, c, and ℓ are constants. Determine the acceleration as a function of x and t. If V0 = 10 fts and ℓ = 5 ft, what value of c (other than c = 0) is needed to make the acceleration zero for any x at t = 1 s? Explain how the acceleration can be zero if the flowrate is increasing with time. V1 = 4t ft/s V2 = 2t ft/s y (1) ℓ/2 (2) ■ Figure P4.26 4.27 A certain flow field has the velocity vector −2xyz (x2 − y2 )z y V= 2 i + j+ 2 k. Find the 2 2 2 2 2 (x + y ) (x + y ) x + y2 acceleration vector for this flow. 4.28 Determine the x-component of the acceleration, ax, along the centerline ( y = 0) for the flow of Problem 4.20. Can you determine the acceleration vector at a location not on the centerline? Why or why not? 4.29 The velocity of the water in the pipe shown in Fig. P4.29 is given by V1 = 0.50t m/s and V2 = 1.0t m/s, where t is in seconds. Determine the local acceleration at points (1) and (2). Is the average convective acceleration between these two points negative, zero, or positive? Explain. u u = V0(1 – e–ct) u = 1– V0(1 – e–ct) 2 x ■ Figure P4.33 4.34 The fluid velocity along the x axis shown in Fig. P4.34 changes from 6 m/s at point A to 18 m/s at point B. It is also known that the velocity is a linear function of distance along the streamline. Determine the acceleration at points A, B, and C. Assume steady flow. 192 Chapter 4 ■ Fluid Kinematics VA = 6 m/s VB = 18 m/s A C B x 0.05 m at t = 0 is given approximately by s = 0.6e−0.5t, where t is in seconds and s is in feet. (a) Determine the speed of a fluid particle as a function of time, Vparticle (t), as it flows along the streamline. (b) Determine the speed of the fluid as a function of position along the streamline, V = V(s). (c) Determine the fluid acceleration along the streamline as a function of position, as = as (s). 4.39 A nozzle is designed to accelerate the fluid from V1 to V2 in a linear fashion. That is, V = ax + b, where a and b are constants. If the flow is constant with V1 = 10 ms at x1 = 0 and V2 = 25 ms at x2 = 1 m, determine the local acceleration, the convective acceleration, and the acceleration of the fluid at points (1) and (2). 0.1 m ■ Figure P4.34 4.35 A fluid flows along the x axis with a velocity given by V = (xt) ˆi, where x is in feet and t in seconds. (a) Plot the speed for 0 ≤ x ≤ 10 ft and t = 3 s. (b) Plot the speed for x = 7 ft and 2 ≤ t ≤ 4 s. (c) Determine the local and convective acceleration. (d) Show that the acceleration of any fluid particle in the flow is zero. (e) Explain physically how the velocity of a particle in this unsteady flow remains constant throughout its motion. 4.40 An incompressible fluid flows through the converging duct shown in Fig. P4.40a with velocity V0 at the entrance. Measurements indicate that the actual velocity of the fluid near the wall of the duct along streamline A–F is as shown in Fig. P4.40b. Sketch the component of acceleration along this streamline, a, as a function of s. Discuss the important characteristics of your result. s A 4.36 A constant-density fluid flows through a converging section having an area A given by A= B C D A0 , 1 + (xℓ) where A0 is the area at x = 0. Determine the velocity and acceleration of the fluid in Eulerian form and then the velocity and acceleration of a fluid particle in Lagrangian form. The velocity is V0 at x = 0 when t = 0. 4.37 A hydraulic jump is a rather sudden change in depth of a liquid layer as it flows in an open channel as shown in Fig. P4.37 and Video V10.12. In a relatively short distance (thickness = ℓ) the liquid depth changes from z1 to z2, with a corresponding change in velocity from V1 to V2. If V1 = 1.20 fts, V2 = 0.30 fts, and ℓ = 0.02 ft, estimate the average deceleration of the liquid as it flows across the hydraulic jump. How many g’s deceleration does this represent? Hydraulic jump F V0 1.5 V0 (a) V 1.5 V0 V0 A B C D E F (b) ■ Figure P4.40 V2 ℓ E V1 z2 z1 ■ Figure P4.37 4.38 A fluid particle flowing along a stagnation streamline, as shown in Video V4.9 and Fig. P4.38, slows down as it approaches the stagnation point. Measurements of the dye flow in the video indicate that the location of a particle starting on the stagnation streamline a distance s = 0.6 ft upstream of the stagnation point Stagnation point, s = 0 *4.41 Air flows steadily through a variable area pipe with a velocity of V = u(x)iˆ fts, where the approximate measured values of u(x) are given in the table. Plot the acceleration as a function of x for 0 ≤ x ≤ 12 in. Plot the acceleration if the flowrate is increased by a factor of N (i.e., the values of u are increased by a factor of N) for N = 2, 4, 10. x (in.) u (fts) x (in.) u (fts) 0 1 2 3 4 5 6 10.0 10.2 13.0 20.1 28.3 28.4 25.8 7 8 9 10 11 12 13 20.1 17.4 13.5 11.9 10.3 10.0 10.0 Fluid particle s ■ Figure P4.38 V *4.42 As is indicated in Fig. P4.42, the speed of exhaust in a car’s exhaust pipe varies in time and distance because of the periodic nature of the engine’s operation and the damping effect Problems with distance from the engine. Assume that the speed is given by V = V0[1 + ae−bx sin(ωt)], where V0 = 8 fps, a = 0.05, b = 0.2 ft−1, and ω = 50 rads. Calculate and plot the fluid acceleration at x = 0, 1, 2, 3, 4, and 5 ft for 0 ≤ t ≤ π25 s. 4.46 A fluid flows past a sphere with an upstream velocity of V0 = 40 m/s as shown in Fig. P4.46. From a more advanced theory it is found that the speed of the fluid along the front part of the 3 sphere is V = 2V0 sin θ. Determine the streamwise and normal components of acceleration at point A if the radius of the sphere is a = 0.20 m. 5 ft V V V = V0[1 + ae–bx sin(ω t)] x 193 V0 40° A a θ ■ Figure P4.42 4.43 Water flows down the face of the dam shown in Fig. P4.43. The face of the dam consists of two circular arcs with radii of 10 and 20 ft as shown. If the speed of the water along streamline A–B is approximately V = (2gh)1/2, where the distance h is as indicated, plot the normal acceleration as a function of distance along the streamline, an = an(s). s 4 ft h A ℛ1 = 10 ft ■ Figure P4.46 4.47 Assume that the streamlines for the wingtip vortices from an airplane (see Fig. P4.15 and Video V4.6) can be approximated by circles of radius r and that the speed is V = K/r, where K is a constant. Determine the streamline acceleration, as, and the normal acceleration, an, for this flow. 4.48 The velocity components for steady flow through the nozzle shown in Fig. P4.48 are u = −V0 x/ℓ and υ = V0 [1 + (y/ℓ)], where V0 and ℓ are constants. Determine the ratio of the magnitude of the acceleration at point (1) to that at point (2). 20° C ℛ2 = 20 ft y V B ■ Figure P4.43 4.44 Water flows over the crest of a dam with speed V as shown in Fig. P4.44. Determine the speed if the magnitude of the normal acceleration at point (1) is to equal the acceleration of gravity, g. (1) V (1) (2) ℓ/2 ℛ = 2 ft x ■ Figure P4.48 4.49 Water flows through the curved hose shown in Fig. P4.49 with an increasing speed of V = 10t ft/s, where t is in seconds. For t = 2 s determine (a) the component of acceleration along the streamline, (b) the component of acceleration normal to the streamline, and (c) the net acceleration (magnitude and direction). ■ Figure P4.44 4.45 Water flows under the sluice gate shown in Fig. P4.45. If V1 = 3 m/s, what is the normal acceleration at point (1)? V Sluice gate ℛ = 20 ft ℛ = 0.12 m (1) ■ Figure P4.49 V1 = 3 m/s ■ Figure P4.45 4.50 Water flows though the slit at the bottom of a two-dimensional water trough as shown in Fig. P4.50. Throughout most of the 194 Chapter 4 ■ Fluid Kinematics trough the flow is approximately radial (along rays from O) with a velocity of V = cr, where r is the radial coordinate and c is a constant. If the velocity is 0.04 ms when r = 0.1 m, determine the acceleration at points A and B. A 4.55 Assume the temperature of the exhaust in an exhaust pipe can be approximated by T = T0(1 + ae−bx) [1 + c cos(ωt)], where T0 = 100 °C, a = 3, b = 0.03 m−1, c = 0.05, and ω = 100 rad/s. If the exhaust speed is a constant 3 m/s, determine the time rate of change of temperature of the fluid particles at x = 0 and x = 4 m when t = 0. V r 0.8 m B 0.2 m O ■ Figure P4.50 4.51 Air flows from a pipe into the region between two parallel circular disks as shown in Fig. P4.51. The fluid velocity in the gap between the disks is closely approximated by V = V0Rr, where R is the radius of the disk, r is the radial coordinate, and V0 is the fluid velocity at the edge of the disk. Determine the acceleration for r = 1, 2, or 3 ft if V0 = 5 fts and R = 3 ft. Disks 4.54 A gas flows along the x axis with a speed of V = 5x m/s and a pressure of p = 10x2 N/m2, where x is in meters. (a) Determine the time rate of change of pressure at the fixed location x = 1. (b) Determine the time rate of change of pressure for a fluid particle flowing past x = 1. (c) Explain without using any equations why the answers to parts (a) and (b) are different. 4.56 A bicyclist leaves from her home at 9 A.M. and rides to a beach 40 mi away. Because of a breeze off the ocean, the temperature at the beach remains 60 °F throughout the day. At the cyclist’s home the temperature increases linearly with time, going from 60 °F at 9 A.M. to 80 °F by 1 P.M. The temperature is assumed to vary linearly as a function of position between the cyclist’s home and the beach. Determine the rate of change of temperature observed by the cyclist for the following conditions: (a) as she pedals 10 mph through a town 10 mi from her home at 10 A.M.; (b) as she eats lunch at a rest stop 30 mi from her home at noon; (c) as she arrives enthusiastically at the beach at 1 P.M., pedaling 20 mph. 4.57 The following pressures for the air flow in Problem 4.24 were measured: R r V0 Pipe V t=0s t = 1.0 s t = 2.0 s t = 3.0 s x=0 x = 10 m x = 20 m p = 101 kPa p = 121 kPa p = 141 kPa p = 171 kPa p = 101 kPa p = 116 kPa p = 131 kPa p = 151 kPa p = 101 kPa p = 111 kPa p = 121 kPa p = 131 kPa. ■ Figure P4.51 4.52 Air flows into a pipe from the region between a circular disk and a cone as shown in Fig. P4.52. The fluid velocity in the gap between the disk and the cone is closely approximated by V = V0R2r2, where R is the radius of the disk, r is the radial coordinate, and V0 is the fluid velocity at the edge of the disk. Determine the acceleration for r = 0.5 and 2 ft if V0 = 5 fts and R = 2 ft. Cone Pipe V Disk r Find the local rate of change of pressure ∂p/∂t and the convective rate of change of pressure V ∂p/∂x at t = 2.0 s and x = 10 m. Section 4.4 The Reynolds Transport Theorem 4.58 Obtain a photograph/image of a situation in which a fluid is flowing. Print this photo and draw a control volume through which the fluid flows. Write a brief paragraph that describes how the fluid flows into and out of this control volume. 4.59 In the region just downstream of a sluice gate, the water may develop a reverse flow region as is indicated in Fig. P4.59 and Video V10.9. The velocity profile is assumed to consist of two uniform regions, one with velocity Va = 10 fps and the other with Vb = 3 fps. Determine the net flowrate of water across the portion of the control surface at section (2) if the channel is 20 ft wide. R Control surface Sluice gate ■ Figure P4.52 Vb = 3 ft/s 1.8 ft Section 4.2.1 The Material Derivative 4.53 Fluid flows through a pipe with a velocity of 2.0 ft/s and is being heated, so the fluid temperature T at axial position x increases at a steady rate of 30.0 °F/min. In addition, the fluid temperature is increasing in the axial direction at the rate of 2.0 °F/ft. Find the value of the material derivative DT/Dt at position x. 1.2 ft (1) ■ Figure P4.59 (2) Va = 10 ft/s 195 Problems 4.60 At time t = 0 the valve on an initially empty (perfect vacuum, ρ = 0) tank is opened and air rushes in. If the tank has a volume of V0 and the density of air within the tank increases as ρ = ρ∞ (1 − e−bt ), where b is a constant, determine the time rate of change of mass within the tank. †4.61 From calculus, one obtains the following formula (Leibnitz rule) for the time derivative of an integral that contains time in both the integrand and the limits of the integration: d dt ∫ x2(t) f (x, t)dx = ∫ x2 x1 x1(t) 4.64 Figure P4.64 shows a fixed control volume. It has a volume V0 = 1.0 ft3 , a flow area A = 1.0 ft2, and a length ℓ0 = 1.0 ft. Position x represents the center of the control volume where the fluid velocity V0 = 1.0 ft/s and the density ρ0 = 1.800 slug/ft3. Also, at position x the fluid density does not change locally with time but decreases in the axial direction at the linear rate of 0.25 slug/ft4. Use the system or Lagrangian approach to evaluate dp/dt. Compare this result with that of the material derivative and flux terms. ∂f dx2 dx1 dx + f (x2, t) − f (x1, t) ∂t dt dt Discuss how this formula is related to the time derivative of the total amount of a property in a system and to the Reynolds transport theorem. System and control volume at time t ℓ0 = 1 ft 4.62 Air enters an elbow with a uniform speed of 10 m/s as shown in Fig. P4.62. At the exit of the elbow, the velocity profile is not uniform. In fact, there is a region of separation or reverse flow. The fixed control volume ABCD coincides with the system at time t = 0. Make a sketch to indicate (a) the system at time t = 0.01 s and (b) the fluid that has entered and exited the control volume in that time period. V0, ρ0 A = 1 ft2 + x0 D System at time t + st ■ Figure P4.64 1m 4.65 Find DV/Dt for the system in Problem 4.64. V = 10 m/s A Control volume 1m 5 m/s Reverse B flow C 15 m/s ■ Figure P4.62 4.63 A layer of oil flows down a vertical plate as shown in Fig. P4.63 with a velocity of V = (V0h2 ) (2hx − x2 )jˆ where V0 and h are constants. (a) Show that the fluid sticks to the plate and that the shear stress at the edge of the layer (x = h) is zero. (b) Determine the flowrate across surface AB. Assume the width of the plate is b. (Note: The velocity profile for laminar flow in a pipe has a similar shape. See Video V6.13.) x 4.66 Water enters a 5-ft-wide, 1-ft-deep channel as shown in Fig. P4.66. Across the inlet the water velocity is 6 ft/s in the center portion of the channel and 1 ft/s in the remainder of it. Farther downstream the water flows at a uniform 2 ft/s velocity across the entire channel. The fixed control volume ABCD coincides with the system at time t = 0. Make a sketch to indicate (a) the system at time t = 0.5 s and (b) the fluid that has entered and exited the control volume in that time period. 1 ft/s 6 ft/s 1 ft/s D B 2 ft 1 ft v(x) Plate A 2 ft/s 5 ft 2 ft C Control surface ■ Figure P4.66 A B Oil y ■ Figure P4.63 h 4.67 Figure P4.67 illustrates a system and fixed control volume at time t and the system at a short time δt later. The system temperature is T = 100 °F at time t and T = 103 °F at time t + δt, where δt = 0.1 s. The system mass, m, is 2.0 slugs, and 10 percent of it moves out of the control volume in δt = 0.1 s. The energy per unit mass u ̃ is cvT, where cv = 32.0 Btu/slug ∙ °F. The energy U of the system at any time t is mu ̃ . Use the system or Lagrangian approach to evaluate DU/Dt of the system. Compare this result with the DU/Dt evaluated with that of the material time derivative and flux terms. 196 Chapter 4 ■ Fluid Kinematics System at time t ■ I has 10% of system mass at time t. I Control volume II has 90% of system mass at times t and t + δ t. II III III has 10% of system mass at time t + δ t. System at time t + δ t ■ Figure P4.67 4.68 The wind blows across a field with an approximate velocity profile as shown in Fig. P4.68. Use Eq. 4.16 with the parameter b equal to the velocity to determine the momentum flowrate across the vertical surface A–B, which is of unit depth into the paper. 15 ft/s B 10 ft A ■ Figure P4.68 4.69 Water flows from a nozzle with a speed of V = 10 m/s and is collected in a container that moves toward the nozzle with a speed of Vcv = 2 m/s as shown in Fig. P4.69. The moving control surface consists of the inner surface of the container. The system consists of the water in the container at time t = 0 and the water between the nozzle and the tank in the constant diameter stream at t = 0. At time t = 0.1 s what volume of the system remains outside of the control volume? How much water has entered the control volume during this time period? Repeat the problem for t = 0.3 s. Stream diameter = 0.1 m Container at t = 0 Stream at t = 0 V = 10 m/s Vcv = 2 m/s 3m ■ Figure P4.69 4.1 LL Even for the simplest flows it is often not easy to visually represent various flow field quantities such as velocity, pressure, or temperature. For more complex flows, such as those involving three-dimensional or unsteady effects, it is extremely difficult to “show the data.” However, with the use of computers and appropriate software, novel methods are being devised to more effectively illustrate the structure of a given flow. Obtain information about methods used to present complex flow data. Summarize your findings in a brief report. 4.2 LL For centuries people have obtained qualitative and quantitative information about various flow fields by observing the motion of objects or particles in a flow. For example, the speed of the current in a river can be approximated by timing how long it takes a stick to travel a certain distance. The swirling motion of a tornado can be observed by following debris moving within the tornado funnel. Recently, various high-tech methods using lasers and minute particles seeded within the flow have been developed to measure velocity fields. Such techniques include the laser doppler anemometer (LDA), the particle image velocimeter (PIV), and others. Obtain information about new laser-based techniques for measuring velocity fields. Summarize your findings in a brief report. 4.3 LL Flow visualization is direct observation of the flow field, usually accomplished in a laboratory. Several photographs in this book (see Figs. 3.1, 9.19, and 11.4) and the majority of the accompanying videos were made with the aid of a flow visualization technique. Do some research on flow visualization techniques, including those for both incompressible and compressible flows. Find examples of the use of flow visualization to solve engineering problems. Prepare a paper on your findings; be sure to include several pictures. 20 ft Nozzle Lifelong Learning Problems Moving control volume 5 Finite Control Volume Analysis Learning Objectives After completing this chapter, you should be able to: ■ select an appropriate finite control volume to solve a fluid mechanics problem. ■ apply conservation of mass and energy and Newton’s second law of motion to the contents of a finite control volume to get important answers. ■ explain how velocity changes and energy transfers in fluid flows are related to forces and torques. ■ choose between the Bernoulli equation and the mechanical energy equation for flow along a streamline. Many fluid mechanics problems can be solved by using finite control volume analysis. To solve many practical problems in fluid mechanics, fundamental laws of nature are applied to a finite region in space (a finite control volume). For example, we may estimate the maximum anchoring force required to hold a turbojet engine stationary during a test. Or we may design a propeller to move a boat both forward and backward. Or we may determine how much power it would take to move natural gas from one location to another many miles away. The bases of finite control volume analysis are some fundamental laws of physics, namely, conservation of mass, Newton’s second law of motion, and the first and second laws of thermodynamics. While some simplifying approximations are made for practicality, the engineering answers possible with this powerful analysis method have proven valuable in numerous instances. Conservation of mass is the key to tracking flowing fluid. How much enters and leaves a control volume can be ascertained. Newton’s second law of motion leads to the conclusion that forces can result from or cause changes in a flowing fluid’s velocity magnitude and/or direction. Moment of force (torque) can result from or cause changes in a flowing fluid’s moment-of-momentum. These forces and torques can be associated with work and power transfer as in pumps and turbines. The first law of thermodynamics is a statement of conservation of energy. The second law of thermodynamics clarifies the loss of useful energy associated with every actual process. The mechanical energy equation based on these two laws can be used to analyze a large variety of steady, incompressible flows in terms of changes in pressure, elevation, speed, and of shaft work and loss. Good judgment is required in defining the finite region in space, the control volume, used in solving a problem. What exactly to leave out of and what to leave in the control volume are important considerations. The formulas resulting from applying the fundamental laws to the contents of the control volume are easy to interpret physically and are not difficult to derive and use. 197 198 Chapter 5 ■ Finite Control Volume Analysis Because a finite region of space, a control volume, contains many fluid particles, detailed knowledge of distributions of fluid properties and velocity are not available so the fluid properties and characteristics are often average values. Chapter 6 presents an analysis of fluid flow based on what is happening to the contents of an infinitesimally small region of space or control volume through which fluid particles move one at a time. 5.1 Conservation of Mass—The Continuity Equation 5.1.1 Derivation of the Continuity Equation A system is defined as a collection of unchanging contents, so the conservation of mass principle for a system is simply stated as Time rate of change of the system mass = 0 The amount of mass in a system does not change with time. or DMsys =0 Dt (5.1) where the system mass, Msys, is more generally expressed as ∫ Msys = (5.2) ρ dV sys and the integration is over the volume of the system. In words, Eq. 5.2 states that the system mass is equal to the sum of all the density-volume element products for the contents of the system. For a system and a fixed, nondeforming control volume that are coincident at an instant of time, as illustrated in Fig. 5.1, the Reynolds transport theorem (Eq. 4.19) with B = mass and b = 1 allows us to state that D Dt ∫ ρ dV = sys ∂ ∂t ∫ ρ dV + ∫ cv ρV ∙ nˆ dA (5.3) cs or time rate of change net rate of flow Time rate of change of the mass of the of mass through of the mass of the = + contents of the cointhe control coincident system cident control volume surface In Eq. 5.3, we express the time rate of change of the system mass as the sum of two control volume quantities, the time rate of change of the mass of the contents of the control volume, ∂ ∂t ∫ ρ dV cv and the net rate of mass flow through the control surface, ∫ ρV ∙ nˆ dA cs Control volume V2 (2) V1 System (a) (1) (b) (c) ■ Figure 5.1 System and control volume at three different instances of time. (a) System and control volume at time t − δt. (b) System and control volume at time t, coincident condition. (c) System and control volume at time t + δt. 5.1 Control surface V ^ n V⋅^ n>0 Conservation of Mass—The Continuity Equation When a flow is steady, all field properties (i.e., properties at any specified point) including density remain constant with time and the time rate of change of the mass of the contents of the control volume is zero. That is, ∂ ∂t V ^ n V⋅^ n<0 199 ∫ ρ dV = 0 cv The integrand, V ∙ nˆ dA, in the mass flowrate integral represents the product of the component of velocity, V, perpendicular to the small portion of control surface and the differential area, dA. Thus, V ∙ nˆ dA is the volume flowrate through dA and ρV ∙ nˆ dA is the mass flowrate through dA. Furthermore, as shown in the sketch in the margin, the sign of the dot product V ∙ nˆ is “+” for flow out of the control volume and “−” for flow into the control volume since nˆ is considered positive when it points out of the control volume. When all of the differential quantities, ρV ∙ nˆ dA, are summed over the entire control surface, as indicated by the integral ∫ ρV ∙ nˆ dA cs the result is the net mass flowrate through the control surface, or ∫ ρV ∙ nˆ dA = ∑ m ˙ out − ∑ m˙ in (5.4) cs The continuity equation is a statement that mass is conserved. ˙ is the mass flowrate (lbm/s, slugs, or kgs). If the integral in Eq. 5.4 is positive, the net where m flow is out of the control volume; if the integral is negative, the net flow is into the control volume. The control volume expression for conservation of mass, which is commonly called the continuity equation, for a fixed, nondeforming control volume is obtained by combining Eqs. 5.1, 5.2, and 5.3 to obtain ∂ ∂t ∫ ρ dV + cv ∫ ρV ∙ nˆ dA = 0 (5.5) cs In words, Eq. 5.5 states that because mass is conserved the time rate of change of the mass of the contents of the control volume plus the net rate of mass flow through the control surface must equal zero. Actually, the same result could have been obtained more directly by equating the rates of mass flow into and out of the control volume to the rates of accumulation and depletion of mass within the control volume (see Section 3.6.2). It is reassuring, however, to see that the Reynolds transport theorem works for this simple-to-understand case. This confidence will serve us well as we develop control volume expressions for other important principles. An often-used expression for mass flowrate, m ˙ , through a section of control surface having area A is m ˙ = ρQ = ρAV Mass flowrate equals the product of density and volume flowrate. (5.6) where ρ is the fluid density, Q is the volume flowrate (ft3s or m3s), and V is the component of fluid velocity perpendicular to area A. Since m ˙= ∫ ρV ∙ nˆ dA A application of Eq. 5.6 involves the use of representative or average values of fluid density, ρ, and fluid velocity, V. For incompressible flows, ρ is uniformly distributed over area A. For compressible flows, we will normally assume a uniformly distributed fluid density at each section of flow and allow density changes to occur only from section to section. The appropriate fluid velocity to use Free ebooks ==> www.Ebook777.com 200 Chapter 5 ■ Finite Control Volume Analysis in Eq. 5.6 is the average value of the component of velocity normal to the section area involved. This average value, V, defined as V= V ∫ ρV ∙ nˆ dA A (5.7) ρA is shown in the figure in the margin. If the velocity, V, is assumed to be uniformly distributed (one-dimensional flow) over the section area, A, then V V= V5.1 Sink flow VIDEO ∫ ρV ∙ nˆ dA A ρA =V (5.8) and the bar notation is not necessary (as in Example 5.1). When the flow is not uniformly distributed over the flow cross-sectional area, the bar notation reminds us that an average velocity is being used (as in Examples 5.2 and 5.4). 5.1.2 Fixed, Nondeforming Control Volume In many applications of fluid mechanics, an appropriate control volume to use is fixed and nondeforming. Several example problems that involve the continuity equation for fixed, nondeforming control volumes (Eq. 5.5) follow. EXAMPLE 5.1 Conservation of Mass—Steady, Incompressible Flow GIVEN A great danger to workers in confined spaces involves the consumption of breathable air (oxygen) and its replacement with other gases such as carbon dioxide (CO2). To prevent this from happening, confined spaces need to be ventilated. Although there is no standard for air exchange rates, a complete change of the air every 3 minutes has been accepted by industry as providing effective ventilation. A worker is performing maintenance in a small rectangular tank with a height of 10 ft and a square base 6 ft by 6 ft. Fresh air enters through an 8-in.-diameter hose and exits through a 4-in.diameter port on the tank wall. The flow is assumed steady and incompressible. FIND Determine (a) the exchange rate needed (ft3/min) for this tank and (b) the velocity of the air entering and exiting the tank at this exchange rate. SOLUTION (a) The necessary exchange rate, which will provide the flowrate entering and exiting the space, is based on the volume, V, of the tank, where V = (10 ft) (6 ft) (6 ft) = 360 ft3 Thus, 360 ft3 of air is needed to provide one complete air exchange. As described in the problem statement, to provide effective ventilation this must be done every 3 minutes. Therefore, the required flowrate, Q, is Q= 360 ft3 = 120 cfm 3 min (b) The continuity equation, Eq. 5.5, can be used to calculate the velocities at the inlet and outlet. Thus, ∂ ∂t ∫ ρ dV + ∑ ρ outVout Aout − ∑ ρinVin Ain = 0 (1) cv We consider the volume within the tank to be the control volume, Ain the cross-sectional area of the hose as it protrudes through the tank wall, and Aout the area of the port in the tank wall. Since the flow is assumed steady and incompressible, (Ans) www.Ebook777.com ∂ ∂t ∫ ρdV = 0 cv 5.1 and ρout = ρin Thus, Eq. 1 reduces to Vout Aout − Vin Ain = 0 or (2) which can be rearranged to solve for Vout and Vin . Q 120 ft2/min = = 1380 ft/min = 22.9 ft/s 2 Aout 4 π ft ( 4 )( 12 ) Vout = Vin = Q 120 ft2/min = = 343 ft/min = 5.73 ft/s 2 Ain 8 π ft ( 4 )( 12 ) 25 (Ans) COMMENTS In this example it is quite apparent how much the problem is simplified by assuming the flow is steady and incompressible. It is important for engineers to understand when such assumptions can be made. Also one can see how the velocity through an outlet/inlet is directly related to the flowrate and geometry of the outlet/inlet. For example, if the air velocity E X AMPLE 5.2 through the outlet causes too much dust to be stirred up, the velocity could be decreased by increasing the diameter of the outlet. The relationship between velocity and diameter is shown in Fig. E5.1. As expected, the velocity is inversely proportional to the square of the diameter. As mentioned, there is no written standard for air exchange rates, but changing the air every 3 minutes has been accepted by industry. It is important for engineers to understand that individual industries and companies will have their own established safety precautions, depending on the nature of their work. As an engineer, one must heed these precautions. Safety is always a key component in the design and operation of engineering systems. 15 10 5 0 0 2 Flow Section (1) The average fluid velocity at any section is that velocity which yields the section mass flowrate when multiplied by the section average fluid density and section area (Eq. 5.7). We relate the flows at sections (1) and (2) with the control volume designated with a dashed line in Fig. E5.2. Equation 5.5 is applied to the contents of this control volume to obtain cv ρV ∙ nˆ dA = 0 ∫ ρV ∙ nˆ dA = m ˙ 2 − m˙ 1 = 0 cs or ˙ 1 = m˙ 2 m (1) and from Eqs. 1, 5.6, and 5.7 we obtain ρ1 A1V1 = ρ2 A2V2 cs The time rate of change of the mass of the contents of this control volume is zero because the flow is steady. The control surface integral involves mass flowrates at sections (1) and (2) so that from Eq. 5.4 we get p2 = 18.4 psia T2 = 453 °R V2 = 1000 ft/s ■ Figure E5.2 0 (flow is steady) ∫ Section (2) D1 = D2 = 4 in. p1 = 100 psia T1 = 540 °R SOLUTION ρ dV + 12 Pipe Control volume Calculate the average air velocity at section (1). ∫ 10 Conservation of Mass—Steady, Compressible Flow Air flows steadily between two sections in a long, straight portion of 4-in. inside diameter pipe as indicated in Fig. E5.2. The uniformly distributed temperature and pressure at each section are given. The average air velocity (nonuniform velocity distribution) at section (2) is 1000 ft /s. ∂ ∂t 4 6 8 Outlet diameter (in.) ■ Figure E5.1 GIVEN FIND (4 in., 22.9 ft/s) 20 Vout (ft /s) Vout A out = Vin Ain = Q 201 Conservation of Mass—The Continuity Equation (2) or since A1 = A2 V1 = ρ2 V ρ1 2 (3) 202 Chapter 5 ■ Finite Control Volume Analysis Air at the pressures and temperatures involved in this example problem behaves like an ideal gas. The ideal gas equation of state (Eq. 1.8) is p ρ= (4) RT COMMENT We learn from this example that the continuity equation (Eq. 5.5) is valid for compressible as well as incompressible flows. Also, nonuniform velocity distributions can be handled with the average velocity concept. Significant average velocity changes can occur in pipe flow if the fluid is compressible. Thus, combining Eqs. 3 and 4 we obtain V1 = = p2T1V2 p1T2 (18.4 psia) (540 °R) (1000 ft/s) = 219 ft/s (100 psia) (453 °R) EXAMPLE 5.3 (Ans) Conservation of Mass—Two Fluids GIVEN The inner workings of a dehumidifier are shown in Fig. E5.3a. Moist air (a mixture of dry air and water vapor) enters the dehumidifier at the rate of 600 lbm/hr. Liquid water drains out of the dehumidifier at a rate of 3.0 lbm/hr. A simplified sketch of the process is provided in Fig. E5.3b. FIND Determine the mass flowrate of the dry air and the water vapor leaving the dehumidifier. Fan Section (1) m• 4 m• 5 Control volume Cooling coil Motor • m1 = 600 lbm/hr Fan Cooling coil m• 2 = ? Section (2) Condensate (water) Section (3) m• 3 = 3.0 lbm/hr ■ Figure E5.3a ■ Figure E5.3b SOLUTION The unknown mass flowrate at section (2) is linked with the known flowrates at sections (1) and (3) with the control volume designated with a dashed line in Fig. E5.3b. The contents of the control volume are the air and water vapor mixture and the condensate (liquid water) in the dehumidifier at any instant. Not included in the control volume are the fan and its motor, and the condenser coils and refrigerant. Even though the flow in the vicinity of the fan blade is unsteady, it is unsteady in a cyclical way. Thus, the flowrates at sections (1), (2), and (3) appear steady, and the time rate of change of the mass of the contents of 5.1 the control volume may be considered equal to zero on a timeaverage basis. The application of Eqs. 5.4 and 5.5 to the control volume contents results in ∫ or m ˙ 2 = m˙ 1 − m˙ 3 = 600 lbm/hr − 3.0 lbm/hr (Ans) 203 The answer is the same with a control volume that includes the cooling coils within the control volume. The continuity equation becomes m ˙ 2 = m˙ 1 − m˙ 3 + m˙ 4 − m˙ 5 ρV ∙ nˆ dA = −m ˙ 1 + m˙ 2 + m˙ 3 = 0 cs = 597 lbm/hr Conservation of Mass—The Continuity Equation (1) ˙ 4 is the mass flowrate of the cooling fluid flowing into where m the control volume, and m ˙ 5 is the flowrate out of the control volume through the cooling coil. Since the flow through the coils is steady, it follows that m ˙ 4 = m˙ 5. Hence, Eq. 1 gives the same answer as obtained with the original control volume. COMMENT Note that the continuity equation (Eq. 5.5) can be used when there is more than one stream of fluid flowing through the control volume. E X AMPLE 5.4 Conservation of Mass—Nonuniform Velocity Profile GIVEN Incompressible, laminar water flow develops in a straight pipe having radius R as indicated in Fig. E5.4a. At section (1), the velocity profile is uniform; the velocity is equal to a constant value U and is parallel to the pipe axis everywhere. At section (2), the velocity profile is axisymmetric and parabolic, with zero velocity at the pipe wall and a maximum value of umax at the centerline. Section (1) Control volume R r u1 = U Pipe FIND (a) Section (2) dA2 = 2π r dr [ ()] u2 = umax 1 - _r R How are U and umax related? (b) How are the average velocity at section (2), V2 , and umax related? 2 ■ Figure E5.4a SOLUTION (a) An appropriate control volume is sketched (dashed lines) in Fig. E5.4a. The application of Eq. 5.5 to this control volume yields 0 (flow is steady) ∂ ∂t ∫ ρ dV + cv ∫ ρV ∙ nˆ dA = 0 (1) At the inlet, section (1), the velocity is uniform with V1 = U so that ∫ ∫ ρV ∙ nˆ dA = ρ2 (2) (1) At the outlet, section (2), the velocity is not uniform. However, the net flowrate through this section is the sum of flows through numerous small washer-shaped areas of size dA2 = 2πr dr as shown by the shaded area element in Fig. E5.4b. On each of these dA2 R (3) u22π r dr − ρ1A1U = 0 (4) 0 By combining Eqs. 1, 2, and 3, we get ρ2 ∫ R 0 Since the flow is considered incompressible, ρ1 = ρ2. The parabolic velocity relationship for flow through section (2) is used in Eq. 4 to yield 2πumax ∫ 0 dr ∫ u22πr dr (2) cs ρV ∙ nˆ dA = −ρ1A1U infinitesimal areas the fluid velocity is denoted as u2. Thus, in the limit of infinitesimal area elements, the summation is replaced by an integration, and the outflow through section (2) is given by R r 2 1 − [ ( R ) ]r dr − A1U = 0 Integrating, we get from Eq. 5 r ■ Figure E5.4b 2πumax ( r2 r4 R − 2 ) − πR2U = 0 2 4R 0 (5) 204 Chapter 5 ■ Finite Control Volume Analysis V2 = umax/2 (parabolic) or umax = 2U (Ans) (b) Since this flow is incompressible and the area is constant, we conclude from Eq. 5.7 that U is the average velocity at all sections of the control volume. Thus, the average velocity at section (2), V2 , is one-half the maximum velocity, umax, there or V2 = umax umax 2 (Ans) COMMENT The relationship between the maximum velocity at section (2) and the average velocity is a function of the “shape” of the velocity profile. For the parabolic profile assumed in this example, the average velocity, umax2, is the actual “average” of the maximum velocity at section (2), u2 = umax, and the minimum velocity at that section, u2 = 0. However, as shown in Fig. E5.4c, if the velocity profile is a different shape (nonparabolic), the average velocity is not necessarily one-half of the maximum velocity. EXAMPLE 5.5 V2 = umax/2 (nonparabolic) ■ Figure E5.4c Conservation of Mass—Unsteady Flow GIVEN Construction workers in a trench of the type shown in Fig. E5.5a are installing a new waterline. The trench is 10 ft long, 5 ft wide, and 8 ft deep. As a result of being near an intersection, carbon dioxide from vehicle exhaust enters the trench at a rate of 10 ft3/min. Because carbon dioxide has a greater density than air, it will settle to the bottom of the trench and displace the air the workers need to breathe. Assume that there is negligible mixing between the air and carbon dioxide. FIND (a) Estimate the time rate of change of the depth of carbon dioxide in the trench, ∂h/∂t, in feet per minute at any instant. (b) Calculate the time, th=6, it would take for the level of carbon dioxide to reach 6 ft, the approximate height to fully engulf the utility workers. ■ Figure E5.5a (Photo courtesy of Architect of the Capitol.) SOLUTION (a) We use the fixed, nondeforming control volume outlined with a dashed line in Fig. E5.5b. This control volume includes, at any instant, the carbon dioxide accumulated in the trench, some of the carbon dioxide flowing from the street into the trench, and some air. Application of Eqs. 5.5 and 5.6 to the contents of the control volume results in ∂ ∂t ∫ air volume ρair d Vair + ∂ ∂t ∫ co 2 volume ρco2 d Vco2 −m ˙ co2 + m˙ air = 0 the sum of the first two terms is the time rate of change of mass within the control volume. Note that the time rates of change of air mass and carbon dioxide mass are each not zero. Recognizing, however, that the air mass must be conserved, we know that the time rate of change of the mass of air in the control volume must be equal to the rate of air mass flow out of the control volume. Thus, applying Eqs. 5.5 and 5.6 to the air only and to the carbon dioxide only, we obtain ∂ ∂t (1) ˙ co2 and m˙ air are the mass flowrates of carbon dioxide into where m and air out of the control volume. Recall that the mass, dm, of fluid contained in a small volume dV − is dm = ρdV −. Hence, the two integrals in Eq. 1 represent the total amount of air and carbon dioxide in the control volume, and ∫ air volume ρair d Vair + m ˙ air = 0 for air and ∂ ∂t ∫ ρco2 co2 volume d Vco2 = m ˙ co2 (2) 5.1 Control volume Conservation of Mass—The Continuity Equation 205 or QCO 10 ft3/min ∂h = = 0.20 ft /min ∂t 50 ft2 2 (Ans) (b) To find the time it will take to engulf the workers at a depth of h = 6 ft, we need only to divide the height the level needs to reach by the time rate of change of depth. That is, 8 ft h th=6 = 5 ft 10 ft co2 volume ρco2 d Vco2 = ρco2 [h(10 ft) (5 ft)] Combining Eqs. 2 and 3, we obtain ρco2 (50 ft2 ) ∂h =m ˙ co2 ∂t and, thus, since m ˙ = ρQ, Qco2 ∂h = ∂t 50 ft 2 (Ans) COMMENTS As shown in this example, it would not take ■ Figure E5.5b for carbon dioxide. The volume of carbon dioxide in the control volume is given by ∫ 6 ft = 30 min 0.2 (ft /min) (3) long for the air within the confined space of the trench to be displaced enough to become a danger to the workers. Furthermore, by the time workers feel the effects of an oxygen-deficient atmosphere, they may be unable to remove themselves from the dangerous space. Note that the answer to part (b) can be easily obtained by realizing that the accumulated volume, V , of a flow is the flowrate, Q, times the time it has been flowing, t. That is, V = Qt. For this example, the volume of the carbon dioxide in the trench when it is 6 ft deep (neglecting the volume of the workers and equipment within the trench) is V = 5 ft × 10 ft × 6 ft = 300 ft3. Thus, with a flowrate of Q = 10 ft3/min, th=6 = 300 ft3/10 ft3/min = 30 min in agreement with the foregoing answer. The preceding example problems illustrate some important results of applying the conservation of mass principle to the contents of a fixed, nondeforming control volume. The dot product V ∙ nˆ is “+” for flow out of the control volume and “−” for flow into the control volume. Thus, mass flowrate out of the control volume is “+” and mass flowrate in is “−.” When the flow is steady, the time rate of change of the mass of the contents of the control volume For steady flow, mass flowrate in equals mass flowrate out. ∂ ∂t ∫ ρ dV cv is zero and the net amount of mass flowrate, m ˙ , through the control surface is, therefore, also zero ∑ m˙ out − ∑ m˙ in = 0 VIDEO V5.2 Shop vac filter (5.9) If the steady flow is also incompressible, the net amount of volume flowrate, Q, through the control surface is also zero: ∑ Qout − ∑ Qin = 0 (5.10) An unsteady, but cyclical, flow can be considered steady on a time-average basis. When the flow is unsteady, the instantaneous time rate of change of the mass of the contents of the control volume is not necessarily zero and can be an important variable. When the value of VIDEO V5.3 Flow through a contraction ∂ ∂t ∫ ρ dV cv is “+,” the mass of the contents of the control volume is increasing. When it is “−,” the mass of the contents of the control volume is decreasing. When the flow is uniformly distributed over the opening in the control surface (one-dimensional flow), m ˙ = ρAV 206 Chapter 5 ■ Finite Control Volume Analysis where V is the uniform value of the velocity component normal to the section area A. When the velocity is nonuniformly distributed over the opening in the control surface, Section (1) m ˙ = ρAV – ρ 1,A1,V1 Section (2) – (5.11) where V is the average value of the component of velocity normal to the section area A as defined by Eq. 5.7. For steady flow involving only one stream of a specific fluid flowing through the control volume at sections (1) and (2), as shown by the figure in the margin ρ 2,A2,V2 m ˙ = ρ1A1V1 = ρ2A2V2 (5.12) Q = A1V1 = A2V2 (5.13) and for incompressible flow, For steady flow involving more than one stream of a specific fluid or more than one specific fluid flowing through the control volume, ∑ m˙ in = ∑ m˙ out The variety of example problems solved above should give the correct impression that the fixed, nondeforming control volume is versatile and useful. THE WIDE WORLD OF FLUIDS “Green” 1.6-gpf standards Toilets account for approximately that flushes properly with such a small amount of water is not 40% of all indoor household water use. To conserve water, the simple. Today there are two basic types involved: those that are new standard is 1.6 gallons of water per flush (gpf). Old toilets gravity powered and those that are pressure powered. Gravity use up to 7 gpf; those manufactured after 1980 use 3.5 gpf. toilets (typical of most currently in use) have rather long cycle Neither is considered a low-flush toilet. A typical 3.2-person times. The water starts flowing under the action of gravity and household in which each person flushes a 7-gpf toilet 4 times a the swirling vortex motion initiates the siphon action that builds to day uses 32,700 gallons of water each year; with a 3.5-gpf toilet a point of discharge. In the newer pressure-assisted models, the the amount is reduced to 16,400 gallons. Clearly the new 1.6-gpf flowrate is large, but the cycle time is short and the amount of toilets will save even more water. However, designing a toilet water used is relatively small. (See Problem 5.34.) 5.1.3 Moving, Nondeforming Control Volume Some problems are most easily solved by using a moving control volume. V VCV It is sometimes necessary to use a nondeforming control volume attached to a moving reference frame. Examples include control volumes containing a gas turbine engine on an aircraft in flight, the exhaust stack of a ship at sea, and the gasoline tank of an automobile passing by. As discussed in Section 4.4.6, when a moving control volume is used, the fluid velocity relative to the moving control volume (relative velocity) is an important flow field variable. The relative velocity, W, is the fluid velocity seen by an observer moving with the control volume. The control volume velocity, Vcv, is the velocity of the control volume as seen from a fixed coordinate system. The absolute velocity, V, is the fluid velocity seen by a stationary observer in a fixed coordinate system. These velocities are related to each other by the vector equation V = W + Vcv W (5.14) as illustrated by the figure in the margin. This is the same as Eq. 4.22, introduced earlier. For a system and a moving, nondeforming control volume that are coincident at an instant of time, the Reynolds transport theorem (Eq. 4.23) for a moving control volume leads to DMsys Dt = ∂ ∂t ∫ ρ dV + cv ∫ ρW ∙ nˆ dA (5.15) cs From Eqs. 5.1 and 5.15, we can get the control volume expression for conservation of mass (the continuity equation) for a moving, nondeforming control volume, namely, ∂ ∂t ∫ cv ρ dV + ∫ cs ρW ∙ nˆ dA = 0 (5.16) 5.1 Conservation of Mass—The Continuity Equation 207 Some examples of the application of Eq. 5.16 follow. E X AMPLE 5.6 Conservation of Mass—Compressible Flow with a Moving Control Volume GIVEN 1050 km/hr. The engine exhaust area is 0.558 m2 , and the exhaust gas density is 0.515 kg/m3. An airplane moves forward at a speed of 971 km/hr as shown in Fig. E5.6a. The frontal intake area of the jet engine is 0.80 m2 and the entering air density is 0.736 kg/m3. A stationary observer determines that relative to the Earth, the jet engine exhaust gases move away from the engine with a speed of FIND Estimate the mass flowrate of fuel into the engine in kg/hr. Vplane = 971 km/hr m• fuel in Control volume Vplane = 971 km/hr (a) V2 = 1050 km/hr W1 = 971 km/hr W2 = 1050 + 971 = 2021 km/hr Section (2) Section (1) (b) ■ Figure E5.6 SOLUTION The control volume, which moves with the airplane (see Fig. E5.6b), surrounds the engine and its contents and includes all fluids involved at an instant. The application of Eq. 5.16 to these contents of the control volume yields 0 (flow relative to moving control volume is considered steady on a time-average basis) ∂ ∂t ∫ ρ dV + cv ∫ ρW ∙ nˆ dA = 0 observer noted that the exhaust gases were moving away from the engine at a speed of 1050 km/hr, the speed of the exhaust gases relative to the moving control volume, W2, is determined as follows by using Eq. 5.14 V2 = W2 + Vplane or W2 = V2 − Vplane = 1050 km/hr − (−971 km/hr) = 2021 km/hr (1) cs Assuming one-dimensional flow, we evaluate the surface integral in Eq. 1 and get m ˙ fuel = (0.515 kg/m3 ) (0.558 m2 ) (2021 km/hr) (1000 m/km) in − (0.736 kg/m3 ) (0.80 m2 ) (971 km/hr) (1000 m/km) = (580,800 − 571,700) kg/hr −m ˙ fuel − ρ1A1W1 + ρ2A2W2 = 0       in or ˙ fuel = ρ2A2W2 − ρ1A1W1 m and is shown in Fig. E5.6b. From Eq. 2, (2) in We consider the intake velocity, W1 , relative to the moving control volume, as being equal in magnitude to the speed of the airplane, 971 km/hr. The exhaust velocity, W2, also needs to be measured relative to the moving control volume. Since a fixed m ˙ fuel = 9100 kg/hr (Ans) in COMMENT Note that the fuel flowrate was obtained as the difference of two large, nearly equal numbers. Precise values of W2 and W1 are needed to obtain a modestly accurate value of m ˙ fuel. in 208 Chapter 5 ■ Finite Control Volume Analysis EXAM PLE 5.7 Conservation of Mass—Relative Velocity GIVEN Water enters a rotating lawn sprinkler through its base A2 = 30 mm2 Section (2) at the steady rate of 1000 ml/s as sketched in Fig. E5.7. The exit area of each of the two nozzles is 30 mm2. Control volume FIND Determine the average speed of the water leaving the Section (1) Sprinkler head nozzle, relative to the nozzle, if (a) the rotary sprinkler head is stationary, Q Section (3) (b) the sprinkler head rotates at 600 rpm, and (c) W2 the sprinkler head accelerates from 0 to 600 rpm. Q = 1000 ml/s ■ Figure E5.7 SOLUTION (a) We specify a control volume that contains the water in the rotary sprinkler head at any instant. This control volume is nondeforming, but it moves (rotates) with the sprinkler head. The application of Eq. 5.16 to the contents of this control volume for situation (a), (b), or (c) of the problem results in the same expression, namely, 0 flow is steady or the control volume is filled with an incompressible fluid ∂ ∂t (1) The time rate of change of the mass of water in the control volume is zero because the flow is steady and the control volume is filled with water. Because there is only one inflow [at the base of the rotating arm, section (1)] and two outflows [the two nozzles at the tips of the arm, sections (2) and (3), each with the same area and fluid velocity], Eq. 1 becomes ρ2A2W2 + ρ3 A 3W3 − ρ1A1W1 = 0 W2 = Q 2 A2 (1000 ml/s) (0.001 m3/liter) (106 mm2/m2 ) (1000 ml/liter) (2) (30 mm2 ) = 16.7 m/s or ∑ ρin Ain Win = 0 With Q = A1W1, A2 = A3, and W2 = W3 it follows that W2 = cs ∑ ρout Aout Wout − A2W2 + A3W3 − A1W1 = 0 or ∫ ρ d V + ∫ ρW ∙ nˆ dA = 0 cv Hence, for incompressible flow with ρ1 = ρ2 = ρ3, Eq. 2 becomes (2) (Ans) (b) and (c) The value of W2 is independent of the speed of rotation of the sprinkler head and represents the average velocity of the water exiting from each nozzle with respect to the nozzle for cases (a), (b), and (c). COMMENT The velocity of water discharging from each nozzle, when viewed from a stationary reference (i.e., V2), will vary as the rotation speed of the sprinkler head varies since from Eq. 5.14, V2 = W2 − U where U = ωR is the speed of the nozzle and ω and R are the angular velocity and radius of the sprinkler head, respectively. When a moving, nondeforming control volume is used, the dot product sign convention used earlier for fixed, nondeforming control volume applications is still valid. Also, if the flow within the moving control volume is steady, or steady on a time-average basis, the time rate of change of the mass of the contents of the control volume is zero. Velocities seen from the control volume reference frame (relative velocities) must be used in the continuity equation. Relative and absolute velocities are related by a vector equation (Eq. 5.14), which also involves the control volume velocity. Flowrate is calculated using the velocity relative to the control volume. 5.1.4 Deforming Control Volume Occasionally, a deforming control volume can simplify the solution of a problem. A deforming control volume involves changing volume size and control surface movement. Thus, the Reynolds 5.1 Conservation of Mass—The Continuity Equation 209 transport theorem for a moving control volume can be used for this case, and Eqs. 4.23 and 5.1 lead to DMsys Dt = ∂ ∂t ∫ ρ dV + cv ∫ ρW ∙ nˆ dA = 0 (5.17) cs The time rate of change term in Eq. 5.17, ∂ ∂t ∫ ρ dV cv is usually nonzero and must be carefully evaluated because the geometry and extent of the control volume vary with time. The mass flowrate term in Eq. 5.17, ∫ ρW ∙ nˆ dA cs must be determined with the relative velocity, W, the velocity referenced to the control surface. Since the control volume is deforming, the control surface velocity is not necessarily uniform and identical to the control volume velocity, Vcv, as was true for moving, nondeforming control volumes. For the deforming control volume, The velocity of the surface of a deforming control volume is not the same at all points on the surface. V = W + Vcs (5.18) where Vcs is the velocity of the control surface as seen by a fixed observer. The relative velocity, W, must be ascertained with care wherever fluid crosses the control surface. Two example problems that illustrate the use of the continuity equation for a deforming control volume, Eq. 5.17, follow. E X AMPLE 5.8 Conservation of Mass—Deforming Control Volume GIVEN A syringe (Fig. E5.8a) is used to inoculate a cow. The plunger has a face area of 500 mm2. The liquid in the syringe is to be injected steadily at a rate of 300 cm3min. The leakage rate past the plunger is 0.10 times the volume flowrate out of the needle. FIND (a) (© James McQuillan/iStockphoto.) With what speed should the plunger be advanced? Plunger motion SOLUTION Qleak = 0.1 Q2 ℓ Ap = Vp The control volume selected for solving this problem is the deforming one illustrated in Fig. E5.8b. Section (1) of the control surface moves with the plunger. The surface area of section (1), A1, is considered equal to the circular area of the face of the plunger, Ap , although this is not strictly true, since leakage occurs. The difference is small, however. Thus, A 1 = Ap (1) Liquid also leaves the needle through section (2), which involves fixed area A2. The application of Eq. 5.17 to the contents of this control volume gives ∂ ∂t ∫ ρ dV + m ˙ 2 + ρQleak = 0 (2) cv Even though Qleak and the flow through section area A2 are steady, the time rate of change of the mass of liquid in the shrinking control volume is not zero because the control volume Section (1) 500 mm2 Q2 = 300 cm3/min Section (2) Control volume (b) ■ Figure E5.8 is getting smaller. To evaluate the first term of Eq. 2, we note that ∫ ρ d V = ρ(ℓA1 + V needle ) (3) cv where ℓ is the changing length of the control volume (see Fig. E5.8) and Vneedle is the volume of the needle. From Eq. 3, we obtain ∂ ∂t ∫ cv ρ d V = ρA1 ∂ℓ ∂t (4) Free ebooks ==> www.Ebook777.com 210 Chapter 5 ■ Finite Control Volume Analysis Solving Eq. 8 for Vp yields Note that −   ∂ℓ = Vp ∂t Vp = (5) where Vp is the speed of the plunger sought in the problem statement. Combining Eqs. 2, 4, and 5 we obtain −ρA1Vp + m ˙ 2 + ρQleak = 0 Vp = However, from Eq. 5.6, we see that (7) −ρA1Vp + ρQ2 + ρQleak = 0 EXAMPLE 5.9 (8) Q2 + 0.1Q2 1.1Q2 = A1 A1 and Vp = and Eq. 6 becomes (9) Since Qleak = 0.1Q2, Eq. 9 becomes (6) ˙ 2 = ρQ2 m Q2 + Qleak A1 (1.1) (300 cm3/min) (500 mm2 ) = 660 mm/min 1000 mm3 ( cm3 ) (Ans) Conservation of Mass—Deforming Control Volume GIVEN Consider Example 5.5. FIND Solve the problem of Example 5.5 using a deforming control volume that includes only the carbon dioxide accumulating in the trench. SOLUTION For this deforming control volume, Eq. 5.17 leads to ∂ ∂t ∫ ρ co2 volume dV + ∫ ρW ∙ nˆ dA = 0 (1) cs The first term of Eq. 1 can be evaluated as ∂ ∂t ∫ co2 volume ∂ [ρh(10 ft) (5 ft)] ∂t ∂h = ρ (50 ft2 ) ∂t ρ dV = ∫ cs ∂h A ∂t ) co2 Vco2 Aco2 Qco2 ∂h = = ∂t (50 ft2 − Aco2 ) (50 ft2 − Aco2 )   (2) or for Aco2 ⪡ 50 ft2 The second term of Eq. 1 can be evaluated as ρ W ∙ nˆ dA = − ρ(Vco2 + Eq. 3 since they are in opposite directions and therefore, the relative velocity would be the two velocities combined (i.e., like the relative velocity of two cars passing each other going in opposite directions). The overall mass flow term is negative since the carbon dioxide is entering the control volume. Thus, from Eqs. 1, 2, and 3 we obtain (3) where Aco2 and Vco2 are the cross-sectional area and velocity of the carbon dioxide flowing from the street down into the trench. Note that the velocity of the carbon dioxide entering the control surface and the velocity of the control surface are summed in ∂h 10 ft3/min = = 0.2 ft /min ∂t (50 ft2 ) (Ans) COMMENT Note that these results using a deforming control volume are the same as those obtained in Example 5.5 with a fixed control volume. The conservation of mass principle is easily applied to the contents of a control volume. The appropriate selection of a specific kind of control volume (for example, fixed and nondeforming, moving and nondeforming, or deforming) can make the solution of a particular problem less complicated. In general, where fluid flows through the control surface, it is advisable to make the control surface perpendicular to the flow. In the sections ahead we learn that the conservation of mass principle is primarily used in combination with other important laws to solve problems. www.Ebook777.com 5.2 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 211 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 5.2.1 Derivation of the Linear Momentum Equation VIDEO V5.4 Smokestack plume momentum Newton’s second law of motion for a system is Time rate of change of the = sum of external forces linear momentum of the system acting on the system Since momentum is mass times velocity, the momentum of a small particle of mass ρd V is V ρd V. Thus, the momentum of the entire system is ʃsysVρd V and Newton’s law becomes D Dt Forces acting on a flowing fluid can change its velocity magnitude and/or direction. ∫ Vρ d V = ∑ Fsys (5.19) sys Any reference or coordinate system for which this statement is true is called inertial. A fixed coordinate system is inertial. A coordinate system that moves in a straight line with constant velocity and is thus without acceleration is also inertial. We proceed to develop the control volume formula for this important law. When a control volume is coincident with a system at an instant of time, the forces acting on the system and the forces acting on the contents of the coincident control volume (see Fig. 5.2) are instantaneously identical, that is, ∑ Fsys = ∑ Fcontents of the (5.20) coincident control volume Furthermore, for a system and the contents of a coincident control volume that is fixed and nondeforming, the Reynolds transport theorem [Eq. 4.19 with b set equal to the velocity (i.e., momentum per unit mass), and Bsys being the system momentum] allows us to conclude that VIDEO D Dt V5.5 Marine propulsion ∫ Vρ d V = sys ∂ ∂t ∫ Vρ d V + cv ∫ VρV ∙ nˆ dA (5.21) cs or time rate of change Time rate of change net rate of flow of the linear of the linear of linear momentum = momentum of the + momentum of the through the contents of the system control surface control volume Equation 5.21 states that the time rate of change of system linear momentum is expressed as the sum of the two control volume quantities: the time rate of change of the linear momentum of the contents of the control volume, and the net rate of linear momentum flow through the control surface. As particles of mass move into or out of a control volume through the control surface, they carry linear momentum in or out. Thus, linear momentum flow should seem no more unusual than mass flow. For a control volume that is fixed (and thus inertial) and nondeforming, Eqs. 5.19, 5.20, and 5.21 provide an appropriate mathematical statement of Newton’s second law of motion as ∂ ∂t ∫ Vρ d V + cv ∫ VρV ∙ nˆ dA = ∑ F cs contents of the control volume We call Eq. 5.22 the linear momentum equation. Coincident control volume FB FC FD FE FA System ■ Figure 5.2 External forces acting on system and coincident control volume. (5.22) 212 Chapter 5 ■ Finite Control Volume Analysis Flow out Flow in Control volume Ffluid out 𝒲 Fwall Ffluid in VIDEO In our application of the linear momentum equation, we initially confine ourselves to fixed, nondeforming control volumes for simplicity. Subsequently, we discuss the use of a moving but inertial, nondeforming control volume. We do not consider deforming control volumes and accelerating (noninertial) control volumes. If a control volume is noninertial, the acceleration components involved (for example, translation acceleration, Coriolis acceleration, and centrifugal acceleration) require consideration. The forces involved in Eq. 5.22 are body and surface forces that act on what is contained in the control volume as shown in the sketch in the margin. The only body force we consider in this chapter is the one associated with the action of gravity. We experience this body force as weight, 𝒲. The surface forces are basically exerted on the contents of the control volume by material just outside the control volume in contact with material just inside the control volume. For example, a wall in contact with fluid can exert a reaction surface force on the fluid it bounds. Similarly, fluid just outside the control volume can push on fluid just inside the control volume at a common interface, usually an opening in the control surface through which fluid flow occurs. An immersed object can resist fluid motion with surface forces. The linear momentum terms in the momentum equation deserve careful explanation. We clarify their physical significance in the following sections. 5.2.2 Application of the Linear Momentum Equation V5.6 Force due to a water jet The linear momentum equation for an inertial control volume is a vector equation (Eq. 5.22). In engineering applications, components of this vector equation resolved along orthogonal coordinates, for example, x, y, and z (rectangular coordinate system) or r, θ, and x (cylindrical coordinate system), will normally be used. A simple example involving steady, incompressible flow is considered first. EXAMPLE 5.10 Linear Momentum—Change in Flow Direction GIVEN As shown in Fig. E5.10a, a horizontal jet of water FIND Determine the anchoring force needed to hold the vane exits a nozzle with a uniform speed of V1 = 10 ft/s, strikes a vane, and is turned through an angle θ. A similar situation is also shown in Video V5.6. stationary if gravity and viscous effects are negligible. SOLUTION We select a control volume that includes the vane and a portion of the water (see Figs. E5.10b, c) and apply the linear momentum equation to this fixed control volume. The only portions of the control surface across which fluid flows are section (1) (the entrance) and section (2) (the exit). Hence, the x and z components of Eq. 5.22 become 0 (flow is steady) ∂ ∂t ∫ u ρ dV + cv ∫ u ρ V ∙ nˆ dA = ∑ Fx cs and 0 (flow is steady) ∂ ∂t ∫ cv w ρ dV + ∫ w ρ V ∙ nˆ dA = ∑ Fz cs or u2 ρA2 V2 − u1 ρA1V1 = ∑ Fx (1) w2 ρA2V2 − w1ρA1V1 = ∑ Fz (2) and where V = u ˆi + w kˆ , and ΣFx and ΣFz are the net x and z components of force acting on the contents of the control volume. Depending on the particular flow situation being considered and the coordinate system chosen, the x and z components of velocity, u and w, can be positive, negative, or zero. In this example the flow is in the positive direction at both the inlet and the outlet. The water enters and leaves the control volume as a free jet at atmospheric pressure. Hence, there is atmospheric pressure surrounding the entire control volume, and the net pressure force on the control volume surface is zero. If we neglect the weight of the water and vane, the only forces applied to the control volume contents are the horizontal and vertical components of the anchoring force, FAx and FAz, respectively. With negligible gravity and viscous effects, and since p1 = p 2, the speed of the fluid remains constant so that V1 = V 2 = 10 ft/s (see the Bernoulli equation, Eq. 3.7). Hence, at section (1), u 1 = V1, w 1 = 0, and at section (2), u2 = V 1 cos θ, w 2 = V1 sin θ. By using this information, Eqs. 1 and 2 can be written as V1cos θ ρ A2V1 − V1 ρ A1V1 = FAx (3) 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations and θ V1sin θ ρ A2V1 − 0 ρ A1V1 = FAz Vane A1 = 0.06 ft2 213 Nozzle V1 (4) Equations 3 and 4 can be simplified by using conservation of mass, which states that for this incompressible flow A1V1 = A2V2, or A1 = A2 since V1 = V2. Thus, FAx = −ρA1V 21 + ρA1V 21 cos θ = −ρA1V 21 (1 − cos θ) (5) and (a) Control volume FAz = ρA1V 21 sin θ (6) With the given data we obtain FAx = −(1.94 slugs/ft3 ) (0.06 ft2 ) (10 ft/s) 2 (1 − cos θ) Nozzle = −11.64(1 − cos θ) slugs ∙ ft/s2 V1 = −11.64(1 − cos θ) lb (Ans) and (b) FAz = (1.94 slugs/ft3 ) (0.06 ft2 ) (10 ft/s) 2 sin θ V2 θ = 11.64 sin θ lb (Ans) (2) z V1 (1) x FAx FAz (c) 15 10 FAz FAx or FAz, lb 5 0 0 30 60 90 120 150 180 –5 COMMENTS The values of FAx and FAz as a function of θ are shown in Fig. E5.10d. Note that if θ = 0 (i.e., the vane does not turn the water), the anchoring force is zero. The inviscid fluid merely slides along the vane without putting any force on it. If θ = 90°, then FAx = −11.64 lb and FAz = 11.64 lb. It is necessary to push on the vane (and, hence, for the vane to push on the water) to the left (FAx is negative) and up in order to change the direction of flow of the water from horizontal to vertical. This momentum change requires a force. If θ = 180°, the water jet is turned back on itself. This requires no vertical force (FAz = 0), but the horizontal force (FAx = −23.3 lb) is two times that required if θ = 90°. This horizontal fluid momentum change requires a horizontal force only. Note that the anchoring force (Eqs. 5, 6) can be written in terms of the mass flowrate, m ˙ = ρA1V1, as ˙ V1 (1 − cos θ) FAx = −m –10 –15 and FAx –20 –25 ■ Figure E5.10 ˙ V1 sin θ FAz = m In this example, exerting a force on a fluid flow resulted in a change in its direction only (i.e., change in its linear momentum). θ, deg (d) THE WIDE WORLD OF FLUIDS Where the plume goes Commercial airliners have wheel brakes forward. As predicted by the momentum equation, the exhaust very similar to those on highway vehicles. In fact, antilock brakes passing through this system produces a substantial braking now found on most new cars were first developed for use on force—the reverse thrust. Designers must know the flow pattern airplanes. However, when landing, the major braking force comes of the exhaust plumes to eliminate potential problems. For from the engine rather than the wheel brakes. Upon touchdown, example, the plumes of hot exhaust must be kept away from a piece of engine cowling translates aft and blocker doors drop parts of the aircraft where repeated heating and cooling could down, directing the engine airflow into a honeycomb structure cause premature fatigue. Also the plumes must not reenter the called a cascade. The cascade reverses the direction of the engine inlet, blow debris from the runway in front of the engine, high-speed engine exhausts by nearly 180° so that it flows or envelop the vertical tail. (See Problem 5.74.) 214 Chapter 5 ■ Finite Control Volume Analysis EXAMPLE 5.11 Linear Momentum—Weight, Pressure, and Change in Speed GIVEN As shown in Fig. E5.11a, water flows through a FIND Determine the anchoring force required to hold the nozzle attached to the end of a laboratory sink faucet with a flowrate of 0.6 liters/s. The nozzle inlet and exit diameters are 16 mm and 5 mm, respectively, and the nozzle axis is vertical. The mass of the nozzle is 0.1 kg. The pressure at section (1) is 464 kPa. nozzle in place. SOLUTION Note that the positive direction is considered “up” for the forces. We will use this same sign convention for the fluid velocity, w, in Eq. 1. In Eq. 1, the dot product, V ∙ nˆ , is “+” for flow out of the control volume and “−” for flow into the control volume. For this particular example The anchoring force sought is the reaction force between the faucet and nozzle threads. To evaluate this force, we select a control volume that includes the entire nozzle and the water contained in the nozzle at an instant, as is indicated in Figs. E5.11a and E5.11b. All of the vertical forces acting on the contents of this control volume are identified in Fig. E5.11b. The action of atmospheric pressure cancels out in every direction and is not shown. Gage pressure forces do not cancel out in the vertical direction and are shown. Application of the vertical or z direction component of Eq. 5.22 to the contents of this control volume leads to V ∙ nˆ dA = ± w dA with the “+” used for flow out of the control volume and “−” used for flow in. To evaluate the control surface integral in Eq. 1, we need to assume a distribution for fluid velocity, w, and fluid density, ρ. For simplicity, we assume that w is uniformly distributed or constant, with magnitudes of w1 and w2 over crosssectional areas A1 and A2. Also, this flow is incompressible so the fluid density, ρ, is constant throughout. Proceeding further we obtain for Eq. 1 0 (flow is steady) ∂ ∂t ∫ wρ d V + cv ∫ wρV ∙ nˆ dA = FA − 𝒲n − p1A1 cs − 𝒲w + p2A2 (−m ˙ 1 ) (−w1 ) + m˙ 2 (−w2 ) (1) = FA − 𝒲n − p1A1 − 𝒲w + p2A2 where w is the z direction component of fluid velocity, and the various parameters are identified in the figure. ˙ = ρAV is the mass flowrate. where m z FA x g Control volume Control volume 𝒲n D1 = 16 mm w1 p1A1 Section (1) h = 30 mm w1 z 𝒲w FA = anchoring force that holds Section (2) p2 A2 nozzle in place 𝒲n = weight of nozzle 𝒲w = weight of water contained in the nozzle D2 = 5 mm p1 = gage pressure at section (1) A1 = cross section area at section (1) p2 = gage pressure at section (2) A2 = cross section area at w2 section (2) w1 = z direction velocity at control volume entrance w2 ■ Figure E5.11a (2) ■ Figure E5.11b w2 = z direction velocity at control volume exit (3) 5.2 Note that −w1 and −w2 are used because both of these velocities are “down.” Also, −m ˙ 1 is used because it is associated with flow into the control volume. Similarly, +m ˙ 2 is used because it is associated with flow out of the control volume. Solving Eq. 3 for the anchoring force, FA, we obtain ˙ 1w1 − m˙ 2w2 + 𝒲n + p1A1 + 𝒲w − p2A2 FA = m ×[ ˙ 1 = m˙ 2 = m˙ m which when combined with Eq. 4 gives ˙ (w1 − w2 ) + 𝒲n + p1A1 + 𝒲w − p2A2 FA = m (6) It is instructive to note how the anchoring force is affected by the different actions involved. As expected, the nozzle weight, 𝒲n, the water weight, 𝒲w, and gage pressure force at section (1), p1 A1, all increase the anchoring force, while the gage pressure force at section (2), p2 A2, acts to decrease the anchoring force. The change in the vertical momentum flowrate, m ˙ (w1 − w2 ), will, in this instance, decrease the anchoring force because this change is negative (w2 > w1 ). To complete this example, we use quantities given in the problem statement to quantify the terms on the right-hand side of Eq. 6. From Eq. 5.6, m ˙ = ρw1A1 = ρQ = (999 kg/m3 ) (0.6 liter/s) (10−3 m3/liter) = 0.599 kg/s (7) and = Q Q = A1 π(D124) (0.6 liter/s) (10 −3 m3/liter) π(16 mm) 24(10002 mm2/m2 ) = 2.98 m/s (8) Also from Eq. 5.6, w2 = = (10002 mm2/m2 ) = 2.84 × 10 m 𝒲w = (999 kg/m3 ) (2.84 × 10−6 m3 ) (9.81 m/s2 ) = 0.0278 N (11) The gage pressure at section (2), p2, is zero since, as discussed in Section 3.6.1, when a subsonic flow discharges to the atmosphere as in the present situation, the discharge pressure is essentially atmospheric. The anchoring force, FA, can now be determined from Eqs. 6 through 11 with FA = (0.599 kg/s) (2.98 m/s − 30.6 m/s) + 0.981 N + (464 kPa) (1000 Pa/kPa) + 0.0278 N − 0 π(16 mm) 2 4(10002 mm2/m2 ) or FA = −16.5 N + 0.981 N + 93.3 N + 0.0278 N = 77.8 N (Ans) Since the anchoring force, FA, is positive, it acts upward in the z direction. The nozzle would be pushed off the pipe if it were not fastened securely. COMMENT The control volume selected to solve problems such as these is not unique. The following is an alternate solution that involves two other control volumes—one containing only the nozzle and the other containing only the water in the nozzle. These control volumes are shown in Figs. E5.11c and E5.11d along with the vertical forces acting on the contents of each control volume. The new force involved, Rz, represents the interaction between the water and the conical inside surface of the nozzle. It includes the net pressure and viscous forces at this interface. Application of Eq. 5.22 to the contents of the control volume of Fig. E5.11c leads to FA = 𝒲n + Rz − patm (A1 − A2 ) Q Q = A2 π(D224) (0.6 liter/s) (10 −3 m3/liter) π(5 mm) 24(10002 mm2/m2 ) = 30.6 m/s ] 3 Thus, (4) (5) (16 mm) 2 + (5 mm) 2 + (16 mm) (5 mm) −6 From the conservation of mass equation, Eq. 5.12, we obtain w1 = 215 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations FA (12) patm (9) The weight of the nozzle, 𝒲n, can be obtained from the nozzle mass, mn, with 𝒲n = mn g = (0.1 kg) (9.81 m/s2 ) = 0.981 N (10) 𝒲n The weight of the water in the control volume, 𝒲w, can be obtained from the water density, ρ, and the volume of water, Vw, in the truncated cone of height h. That is, 𝒲w = ρ Vwg where Rz Vw = = 1 2 12 πh(D1 + D 22   + D1D2 ) (30 mm) 1 π 12 (1000 mm/m) ■ Figure E5.11c 216 Chapter 5 ■ Finite Control Volume Analysis (p1 + patm)A1 𝒲w the projection of the surface area on a plane perpendicular to the axis of the nozzle. The projection of this area on a plane perpendicular to the z direction is A1 − A2. The effect of the atmospheric pressure on the internal area (between the nozzle and the water) is already included in Rz, which represents the net force on this area. Similarly, for the control volume of Fig. E5.11d we obtain Rz Rz = m ˙ (w1 − w2 ) + 𝒲w + ( p1 + patm )A1 − (p2 − patm )A2 (1) w1 (2) (p2 + patm)A2 w2 ■ Figure E5.11d The term patm (A1 − A2 ) is the resultant force from the atmospheric pressure acting on the exterior surface of the nozzle (i.e., that portion of the surface of the nozzle that is not in contact with the water). Recall that the pressure force on a curved surface (such as the exterior surface of the nozzle) is equal to the pressure times VIDEO Vρ V ⋅ ^ n>0 V ^ ^ n n Control volume y x ^ ^ n n V Vρ V ⋅^ n<0 where p1 and p2 are gage pressures. From Eq. 13 it is clear that the value of Rz depends on the value of the atmospheric pressure, patm, since A1 ≠ A2. That is, we must use absolute pressure, not gage pressure, to obtain the correct value of Rz. From Eq. 13 we can easily identify which forces acting on the flowing fluid change its velocity magnitude and thus linear momentum. By combining Eqs. 12 and 13 we obtain the same result for FA as before (Eq. 6): FA = m ˙ (w1 − w2 ) + 𝒲n + p1A1 + Ww − p2A2 Note that although the force between the fluid and the nozzle wall, Rz, is a function of patm, the anchoring force, FA, is not. That is, we were correct in using gage pressure when solving for FA by means of the original control volume shown in Fig. E5.11b. Several important generalities about the application of the linear momentum equation (Eq. 5.22) are apparent in the example just considered. V5.7 Air jet momentum flux V (13) V 1. When the flow is uniformly distributed over a section of the control surface where flow into or out of the control volume occurs, the integral operations are simplified. Thus, one-dimensional flows are easier to work with than flows involving nonuniform velocity distributions. 2. Linear momentum is directional; it can have components in as many as three orthogonal coordinate directions. Furthermore, along any one coordinate, the linear momentum of a fluid particle can be in the positive or negative direction and thus be considered as a positive or a negative quantity. In Example 5.11, only the linear momentum in the z direction was considered (all of it was in the negative z direction and was hence treated as being negative). 3. The flow of positive or negative linear momentum into a control volume involves a negative V ∙ nˆ product. Momentum flow out of the control volume involves a positive V ∙ nˆ product. The correct algebraic sign (+ or −) to assign to momentum flow (VρV ∙ nˆ dA) will depend on the sense of the velocity (+ in positive coordinate direction, − in negative coordinate direction) and the V ∙ nˆ product (+ for flow out of the control volume, − for flow into the control volume). This is shown in the figure in the margin. In Example 5.11, the momentum flow into the control volume past section (1) was a positive (+) quantity while the momentum flow out of the control volume at section (2) was a negative (−) quantity. 4. The time rate of change of the linear momentum of the contents of a nondeforming control volume (i.e., ∂∂t ʃcv Vρ d V) is zero for steady flow. Almost all of the momentum problems considered in this text involve steady flow. 5. If the control surface is selected so that it is perpendicular to the flow where fluid enters or leaves the control volume, the surface force exerted at these locations by fluid outside the control volume on fluid inside will be due to pressure, as shown by the figure in the margin on the next page. Furthermore, when subsonic flow exits from a control volume into the atmosphere, atmospheric pressure prevails at the exit cross section. In Example 5.11, the flow was subsonic and so we set the exit flow pressure at the atmospheric level. The continuity equation (Eq. 5.12) allowed us to evaluate the fluid flow velocities w1 and w2 at sections (1) and (2). 5.2 pout Aout Vout (out) Control volume (in) Vin pin Ain A control volume diagram is similar to a free-body diagram. Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 217 6. The forces due to atmospheric pressure acting on the control surface may need consideration as indicated by Eq. 13 in Example 5.11 for the reaction force between the nozzle and the fluid. When calculating the anchoring force, FA, the forces due to atmospheric pressure on the control surface cancel each other (for example, after combining Eqs. 12 and 13 the atmospheric pressure forces are no longer involved) and gage pressures may be used. 7. The external forces have an algebraic sign—positive if the force is in the assigned positive coordinate direction and negative otherwise. 8. Only external forces acting on the contents of the control volume are considered in the linear momentum equation (Eq. 5.22). If the fluid alone is included in a control volume, reaction forces between the fluid and the surface or surfaces in contact with the fluid [wetted surface(s)] will need to be in Eq. 5.22. If the fluid and the wetted surface or surfaces are within the control volume, the reaction forces between fluid and wetted surface(s) do not appear in the linear momentum equation (Eq. 5.22) because they are internal, not external forces. The anchoring force that holds the wetted surface(s) in place is an external force, however, and must therefore be in Eq. 5.22. 9. The force required to anchor an object will generally exist in response to surface pressure and /or shear forces acting on the control surface, to a change in linear momentum flow through the control volume containing the object, and to the weight of the object and the fluid contained in the control volume. In Example 5.11 the nozzle anchoring force was required mainly because of pressure forces and partly because of a change in linear momentum flow associated with accelerating the fluid in the nozzle. The weight of the water and the nozzle contained in the control volume influenced the size of the anchoring force only slightly. THE WIDE WORLD OF FLUIDS Motorized surfboard When Bob Montgomery, a former profes- response to the change in linear momentum of the water stream sional surfer, started to design his motorized surfboard (called a jet as it enters through the inlet passage and exits through an appro- board), he discovered that there were many engineering challenges. priately designed nozzle. Some of the fluid dynamic problems The idea is to provide surfing to anyone, no matter where they live, associated with designing the craft included one-way valves so near or far from the ocean. The rider stands on the device like a that water does not get into the engine (at both the intake or surfboard and steers it like a surfboard by shifting his or her body exhaust ports), buoyancy, hydrodynamic lift, drag, thrust, and hull weight. A new, sleek, compact 45-hp engine and pump was stability. (See Problem 5.76.) designed to fit within the surfboard hull. Thrust is produced in To further demonstrate the use of the linear momentum equation (Eq. 5.22), we consider another one-dimensional flow example before moving on to other facets of this important equation. E X AMPLE 5.12 Linear Momentum—Pressure and Change in Flow Direction GIVEN Water flows through a horizontal, 180° pipe bend as shown in Fig. E5.12a and illustrated in Fig. E5.12b. The flow cross-sectional area is constant at a value of 0.1 ft2 through the bend. The magnitude of the flow velocity everywhere in the bend is axial and 50 ft/s. The absolute pressures at the entrance and exit of the bend are 30 psia and 24 psia, respectively. FIND Calculate the horizontal (x and y) components of the anchoring force required to hold the bend in place. (a) ■ Figure E5.12 218 Chapter 5 ■ Finite Control Volume Analysis SOLUTION From the given data we can calculate the mass flowrate, m ˙ , from Eq. 5.6 as Since we want to evaluate components of the anchoring force to hold the pipe bend in place, an appropriate control volume (see dashed line in Fig. E5.12b) contains the bend and the water in the bend at an instant. The horizontal forces acting on the contents of this control volume are identified in Fig. E5.12c. Note that the weight of the water is vertical (in the negative z direction) and does not contribute to the x and y components of the anchoring force. All of the horizontal normal and tangential forces exerted on the fluid and the pipe bend are resolved and combined into the two resultant components, FAx and FAy. These two forces act on the control volume contents, and thus for the x direction, Eq. 5.22 leads to ∫ uρV ∙ nˆ dA = FAx m ˙ = ρ1A1υ1 = (1.94 slugs/ft3 ) (0.1 ft2 ) (50 ft/s) = 9.70 slugs/s For determining the anchoring force, FAy, the effects of atmospheric pressure cancel, and thus gage pressures for p1 and p2 are appropriate. By substituting numerical values of variables into Eq. 6 and using the fact that 1 lb = 1 slug ∙ ft/s2 , we get FAy = −(9.70 slugs/s) (50 ft/s + 50 ft/s) −(30 psia − 14.7 psia) (144 in.2/ft2 ) (0.1 ft2 ) −(24 psia − 14.7 psia) (144 in.2/ft2 ) (0.1 ft2 ) (1) FAy = −970 lb − 220 lb − 134 lb = −1324 lb cs (Ans) At sections (1) and (2), the flow is in the y direction and therefore u = 0 at both cross sections. There is no x direction momentum flow into or out of the control volume and we conclude from Eq. 1 that The negative sign for FAy is interpreted as meaning that the y component of the anchoring force is actually in the negative y direction, not the positive y direction as originally indicated in Fig. E5.12c. FAx = 0 the pipe bend is independent of the atmospheric pressure. However, the force that the bend exerts on the fluid inside of it, Ry, depends on the atmospheric pressure. We can see this by using a control volume which surrounds only the fluid within the bend as shown in Fig. E5.12d. Application of the momentum equation to this situation gives COMMENT As with Example 5.11, the anchoring force for (Ans) For the y direction, we get from Eq. 5.22 ∫ υρV ∙ nˆ dA = FAy + p1A1 + p2A2 (2) cs For one-dimensional flow, the surface integral in Eq. 2 is easy to evaluate and Eq. 2 becomes Ry = −m ˙ (υ1 + υ2 ) − p1A1 − p2A2 Note that the y component of velocity is positive at section (1) but is negative at section (2). Also, the mass flowrate term is negative at section (1) (flow in) and is positive at section (2) (flow out). From the continuity equation (Eq. 5.12), we get where p1 and p2 must be in terms of absolute pressure because the force between the fluid and the pipe wall, Ry, is the complete pressure effect (i.e., absolute pressure). We see that forces exerted on the flowing fluid result in a change in its velocity direction (a change in linear momentum). Thus, we obtain m ˙ = m˙ 1 = m˙ 2 Ry = −(9.70 slugs/s) (50 ft/s + 50 ft/s) (+υ1 ) (−m ˙ 1 ) + (−υ2 ) (+m˙ 2 ) = FAy + p1A1 + p2A2 (3) (4) − (30 psia) (144 in.2/ft2 ) (0.1 ft2 ) and thus Eq. 3 can be written as ˙ (υ1 + υ2 ) = FAy + p1A1 + p2A2 −m − (24 psia) (144 in.2/ft2 ) (0.1 ft2 ) (5) = −1748 lb Solving Eq. 5 for FAy we obtain FAy = −m ˙ (υ1 + υ2 ) − p1A1 − p2A2 V = 50 ft/s 50 ft/s Pipe bend and water z w z x V= We can use the control volume that includes just the pipe bend (without the fluid inside it) as shown in Fig. E5.12e to determine (6) A = 0.1 ft2 Section (1) x u v p1A1 y v1 FAy p2 A2 v2 180° pipe bend Section (2) (b) ■ Figure E5.12 cont. FAz y Control volume (7) Control volume 𝒲 FAx (c) 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 219 FAy, the anchoring force component in the y direction necessary to hold the bend stationary. The y component of the momentum equation applied to this control volume gives accounted for by Ry. By combining Eqs. 7 and 8 and using the fact that patm = 14.7 lbin.2 (144 in.2/ft2 ) = 2117 lb/ft2 , we obtain FAy = Ry + patm (A1 + A2 ) FAy = −1748 lb + 2117 lb/ft2 (0.1 ft2 + 0.1 ft2 ) (8) where Ry is given by Eq. 7. The patm (A1 + A2 ) term represents the net pressure force on the outside portion of the control volume. Recall that the pressure force on the inside of the bend is Water in 180° bend z w u x v = −1324 lb in agreement with the original answer obtained using the control volume of Fig. E5.12c. Rz Pipe bend only p1A1 y v1 FAy Ry Ry patm(A1 + A2) p2A2 v2 Control volume 𝒲 Rx Control volume (d) (e) ■ Figure E5.12 cont. V5.8 Fire hose VIDEO In Examples 5.10 and 5.12 the force exerted on a flowing fluid resulted in a change in flow direction only. This force was associated with constraining the flow, with a vane in Example 5.10, and with a pipe bend in Example 5.12. In Example 5.11 the force exerted on a flowing fluid resulted in a change in velocity magnitude only. This force was associated with a converging nozzle. Anchoring forces are required to hold a vane or conduit stationary. They are most easily estimated with a control volume that contains the vane or conduit and the flowing fluid involved. Alternately, two separate control volumes can be used, one containing the vane or conduit only and one containing the flowing fluid only. E X AMPLE 5.13 Linear Momentum—Pressure, Change in Speed, and Friction GIVEN Air flows steadily between two cross sections in a FIND Determine the frictional force exerted by the pipe wall long, straight portion of 4-in. inside diameter pipe as indicated in Fig. E5.13, where the uniformly distributed temperature and pressure at each cross section are given. If the average air velocity at section (2) is 1000 ft/s, we found in Example 5.2 that the average air velocity at section (1) must be 219 ft/s. Assume uniform velocity distributions at sections (1) and (2). on the airflow between sections (1) and (2). p1A1 V1 Control volume Rx Section (1) p1 = 100 psia T1 = 540 °R SOLUTION The control volume of Example 5.2 is appropriate for this problem. The forces acting on the air between sections (1) and (2) are identified in Fig. E5.13. The weight of air is considered negligibly small. The reaction force between the wetted wall of the pipe and the flowing air, Rx, is the frictional force sought. Application of the axial component of Eq. 5.22 to this control volume yields ∫ cs uρV ∙ nˆ dA = −Rx + p1A1 − p2 A2 (1) Pipe V2 = 1000 ft/s p2 A2 Flow y x Section (2) p2 = 18.4 psia T2 = 453 °R ■ Figure E5.13 The positive x direction is set as being to the right. Furthermore, for uniform velocity distributions (one-dimensional flow), Eq. 1 becomes (+u1 ) (−m ˙ 1 ) + (+u2 ) (+m˙ 2 ) = −Rx + p1A1 − p2A2 (2) 220 Chapter 5 ■ Finite Control Volume Analysis From conservation of mass (Eq. 5.12) we get ˙ = m˙ 1 = m˙ 2 m ˙= Hence, m (3) × so that Eq. 2 becomes ˙ (u2 − u1 ) = −Rx + A2 ( p1 − p2 ) m (4) Rx = (5) The equation of state gives (1000 ft/s) = 9.57 lbm/s (8) π(4 in.) 2 (100 psia − 18.4 psia) 4 − (9.57 lbm) (1000 ft/s − 219 ft/s) = 1025 lb − 232 lb (6) or and the equation for area A2 is Rx = 793 lb πD22 4 (Ans) (7) COMMENT For this compressible flow, the pressure differ- Thus, from Eqs. 3, 6, and 7 p πD2 m ˙ = ( 2 )( 2 ) u2 RT2 4 English Engineering (EE) units are often used for this kind of flow. The gas constant, R, for air in EE units is R= 4(144 in.2/ft2 ) 32.174(lbm ∙ ft)(lb ∙ s2 ) p2 ρ2 = RT2 A2 = π(4 in.) 2 Thus, from Eqs. 5 and 8 Solving Eq. 4 for Rx, we get ˙ (u2 − u1 ) R x = A2 ( p 1 − p 2 ) − m (18.4 psia) (144 in.2/ft2 ) [53.3(ft ∙ lb)(lbm ∙ °R)] (453 °R) ence drives the motion, which results in a frictional force, Rx, and because the pressure decrease causes a density decrease, an acceleration of the fluid (i.e., a velocity magnitude increase). For a similar incompressible flow, a pressure difference results in fluid motion with a frictional force only (i.e., no change in velocity magnitude). 1716(ft ∙ lb)(slug ∙ °R) = 53.3(ft ∙ lb)(lbm ∙ °R) 32.174(lbm/slug) EXAMPLE 5.14 Linear Momentum—Weight, Pressure, Friction, and Nonuniform Velocity Profile GIVEN Consider the flow of Example 5.4 to be vertically upward as shown by Fig. E5.14. FIND Develop an expression for the fluid pressure drop that occurs between sections (1) and (2). SOLUTION A control volume (see dashed lines in Fig. E5.14) that includes only fluid from section (1) to section (2) is selected. The forces acting on the fluid in this control volume are identified in Fig. E5.14. The application of the axial component of Eq. 5.22 to the fluid in this control volume results in ∫ wρV ∙ nˆ dA = p1A1 − Rz − 𝒲 − p2A2 (1) cs where Rz is the resultant force of the wetted pipe wall on the fluid. Further, for uniform flow at section (1), and because the flow at section (2) is out of the control volume, Eq. 1 becomes ∫ w2 ρw2 dA2 = ρ A2 ∫ R w22 2πr dr 0 = 2πρ ∫ 0 R r 2 2 (2w1 ) 2 [1 − ( ) ] r dr R or ∫ w2 ρw2 dA2 = 4πρw21 A2 R2 3 Combining Eqs. 2 and 3 we obtain 4 (+w1 ) (−m ˙ 1) + ∫ −w21ρπR2 + 3w21ρπR2 = p1A1 − Rz − 𝒲 − p2A2 (+w2 )ρ(+w2 dA2 ) = p1A1 − Rz A2 − 𝒲 − p2A2 (2) The positive direction is considered up. The surface integral over the cross-sectional area at section (2), A2, is evaluated by using the parabolic velocity profile obtained in Example 5.4, w2 = 2w1[1 − (r/R) 2], as (3) (4) Solving Eq. 4 for the pressure drop from section (1) to section (2), p1 − p2, we obtain p 1 − p2 = Rz ρw21 𝒲 + + 3 A1 A1 (Ans) 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations Flow COMMENT We see that the drop in pressure from section (1) to section (2) occurs because of the following: 1. The change in momentum flow between the two sections associated with going from a uniform velocity profile to a parabolic velocity profile, ρw213 2. Pipe wall friction, Rz 3. The weight of the water column, 𝒲; a hydrostatic pressure effect If the velocity profiles had been identically parabolic at sections (1) and (2), the momentum flowrate at each section would have been identical, a condition we call “fully developed” flow. Then the pressure drop, p1 − p2, would be due only to pipe wall friction and the weight of the water column. If in addition to being fully developed, the flow involved negligible weight effects (for example, horizontal flow of liquids or the flow of gases in any direction), the drop in pressure between any two sections, p1 − p2, would be a result of pipe wall friction only. 221 p2A2 [ ()] –r w2 = 2w1 1 – R 2 Section (2) r Control volume R 𝒲 Fluid only Rz w1 Section (1) z p1A1 w v u y x Momentum flux in nonuniform flow can be evaluated using average velocity and momentum coefficient. ■ Figure E5.14 In Example 5.14, the average velocity is the same at section (1) as it is at section (2) (V1 = V2 = w1 ), but the momentum flux across section (1) is not the same as it is across section (2). If it were, the left-hand side of Eq. (4) would be zero. For a nonuniform flow the momentum flux can be written in terms of the average velocity, V, and the momentum coefficient, β, β= ∫ wρV ∙ nˆ dA ρV 2A Hence the momentum flux can be written as ∫ wρV ∙ nˆ dA = −β1w21ρπR2 + β2w21ρπR2 cs For laminar flow in a pipe, β = 4/3 and for turbulent flow in a pipe β ≈ 1.02. For uniform flow, β = 1. E X AMPLE 5.15 Linear Momentum—Thrust Control volume GIVEN A static thrust stand as sketched in Fig. E5.15 is to be designed for testing a jet engine. The following conditions are known for a typical test: Intake air velocity = 200 m/s; exhaust gas velocity = 500 m/s; intake cross-sectional area = 1 m2; intake static pressure = −22.5 kPa = 78.5 kPa (abs); intake static temperature = 268 K; exhaust static pressure = 0 kPa = 101 kPa (abs). FIND design. Estimate the nominal anchoring force for which to patm(A1 – A2) p2A2 u2 p1A1 u1 Section (2) Section (1) Fth z x ■ Figure E5.15 u 222 Chapter 5 ■ Finite Control Volume Analysis SOLUTION The cylindrical control volume outlined with a dashed line in Fig. E5.15 is selected. The external forces acting in the axial direction are also shown. Application of the momentum equation (Eq. 5.22) to the contents of this control volume yields ∫ We need to determine the mass flowrate, m ˙ , to calculate Fth, and to calculate m ˙ = ρ1A1u1, we need ρ1. From the ideal gas equation of state ρ1 = uρV ∙ nˆ dA = p1A1 + Fth − p2A2 cs − patm (A1 − A2 ) (1) where the pressures are absolute. Thus, for one-dimensional flow, Eq. 1 becomes = 1.02 kg/m3 Thus, m ˙ = ρ1A1u1 = (1.02 kg/m3 ) (1 m2 ) (200 m/s) (2) The positive direction is to the right. The conservation of mass equation (Eq. 5.12) leads to ˙ = m˙ 1 = ρ1A1u1 = m˙ 2 = ρ2A2u2 m Fth = −(1 m2 ) (−22.5 kPa) (1000 Pa/kPa)[1(N/m2 )Pa] + (204 kg/s) (500 m/s − 200 m/s)[1 N/(kg ∙ m/s2 )] or Fth = 22,500 N + 61,200 N = 83,700 N (4) (Ans) COMMENT The force of the thrust stand on the engine is Solving Eq. 4 for the thrust force, Fth, we obtain ˙ (u2 − u1 ) Fth = −p1A1 + p2A2 + m Finally, combining Eqs. 5 and 6 and substituting given data with p2 = 0, we obtain (3) Combining Eqs. 2 and 3 and using gage pressure we obtain ˙ (u2 − u1 ) = p1A1 − p2A2 + Fth m (6) = 204 kg/s ˙ 1 ) + (+u2 ) (+m˙ 2 ) = ( p1 − patm )A1 (+u1 ) (−m − ( p2 − patm )A2 + Fth p1 (78.5 kPa) (1000 Pa/kPa)[1(N/m2 )Pa] = RT1 (286.9 J/kg ∙ K) (268 K) (1 N ∙ m/J) (5) directed toward the right. Conversely, the engine pushes to the left on the thrust stand (or aircraft). THE WIDE WORLD OF FLUIDS Bow thrusters In the past, large ships required the use of maneuvering. Sometimes a second unit is installed in the tugboats for precise maneuvering, especially when docking. stern. Initially used in the bows of ferries, these versatile con- Nowadays, most large ships (and many moderate to small trol devices have become popular in offshore oil servicing ones as well) are equipped with bow thrusters to help steer in boats, fishing vessels, and larger oceangoing craft. They per- close quarters. The units consist of a mechanism (usually a mit unassisted maneuvering alongside of oil rigs, vessels, ducted propeller mounted at right angles to the fore/aft axis loading platforms, fishing nets, and docks. They also provide of the ship) that takes water from one side of the bow and precise control at slow speeds through locks, narrow chan- ejects it as a water jet on the other side. The momentum flux nels, and bridges, where the rudder becomes very ineffective. of this jet produces a starboard or port force on the ship for (See Problem 5.77.) EXAMPLE 5.16 Linear Momentum—Nonuniform Pressure GIVEN A sluice gate across a channel (Fig. E5.16a) of width b is shown in the closed and open positions in Figs. E5.16b and E5.16c. FIND Is the anchoring force required to hold the gate in place larger when the gate is closed or when it is open? ■ Figure E5.16a (Photograph courtesy SOLUTION We will answer this question by comparing expressions for the horizontal reaction force, Rx, between the gate and the water when the gate is closed and when the gate is open. The control volume used in each case is indicated with dashed lines in Figs. E5.16b and E5.16c. When the gate is closed, the horizontal forces acting on the contents of the control volume are identified in Fig. E5.16d. (a) of Pend Oreille Public Utility District.) Application of Eq. 5.22 to the contents of this control volume yields 0 (no flow) ∫ cs 1 uρV ∙ nˆ dA = 2 γH2b − Rx (1) 5.2 Note that the hydrostatic pressure force, γH2b2, is used. From Eq. 1, the force exerted on the water by the gate (which is equal to the force necessary to hold the gate stationary) is 1 Rx = 2 γH2b Open sluice gate Closed sluice gate Control volume Control volume z (2) (b) (c) Water only Control volume Control volume (3) cs Note that because the water at sections (1) and (2) is flowing along straight, horizontal streamlines, the pressure distribution at those locations is hydrostatic, varying from zero at the free surface to γ times the water depth at the bottom of the channel (see Chapter 3, Section 3.4). Thus, the pressure forces at sections (1) and (2) (given by the pressure at the centroid times the area) are γH2b2 and γh2b2, respectively. Also the frictional force between the channel bottom and the water is specified as Ff. The surface integral in Eq. 3 is nonzero only where there is flow across the control surface. With the assumption of uniform velocity distributions, ∫ Thus, Eqs. 3 and 4 combine to form 1 −ρu21Hb + ρu22hb = 2 γH2b − Rx − 2 γh2b − Ff (5) If H ⪢ h, the upstream velocity, u1, is much less than u2 so that the contribution of the incoming momentum flow to the control surface integral can be neglected and from Eq. 5 we obtain 1 1 Rx = 2 γH2b − 2 γh2b − Ff − ρu22hb The linear momentum equation can be written for a moving control volume. V5.9 Jellyfish (12_ γ H ) Hb ( (6) Rx ) 1 _ γ H Hb 2 Section (2) u2 Section (1) (d) Ff (12_ γ h ) hb (e) ■ Figure E5.16 By using the continuity equation, m ˙ = ρbHu1 = ρbhu2, Eq. 6 can be rewritten as 1 Rx = 2 γH2b − 2 γh2b − Ff − m ˙ (u2 − u1 ) (4) cs 1 u1 Rx 1 uρV ∙ nˆ dA = (u1 )ρ(−u1 )Hb + (+u2 )ρ(+u2 )hb h u Water only 1 1 uρV ∙ nˆ dA = 2 γH2b − Rx − 2 γh2b − Ff x H H which is equal in magnitude to the hydrostatic force exerted on the gate by the water. When the gate is open, the horizontal forces acting on the contents of the control volume are shown in Fig. E5.16e. Application of Eq. 5.22 to the contents of this control volume leads to ∫ 223 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations (7) Hence, since u2 > u1 , by comparing the expressions for Rx (Eqs. 2 and 7), we conclude that the reaction force between the gate and the water (and therefore the anchoring force required to hold the gate in place) is smaller when the gate is open than when it is closed. (Ans) COMMENT That the anchoring force on the gate is smaller can be deduced from physical reasoning. When the gate is closed, all of the hydrostatic pressure pushes on the gate. When the gate is open, some of the hydrostatic pressure is used to push the water out. All of the linear momentum examples considered thus far have involved stationary and nondeforming control volumes, which are thus inertial because there is no acceleration. A nondeforming control volume translating in a straight line at constant speed is also inertial because there is no acceleration. For a system and an inertial, moving, nondeforming control volume that are both coincident at an instant of time, the Reynolds transport theorem (Eq. 4.23) is D Dt ∫ ∂ ∂t Vρ d V = sys ∫ cv Vρ d V + ∫ VρW ∙ nˆ dA (5.23) cs where W is the relative velocity. When we combine Eq. 5.23 with Eqs. 5.19 and 5.20, we get VIDEO ∂ ∂t ∫ Vρ d V + cv ∫ cs VρW ∙ nˆ dA = ∑ Fcontents of the (5.24) control volume When the equation relating absolute, relative, and control volume velocities (Eq. 5.14) is used with Eq. 5.24, the result is ∂ ∂t ∫ cv (W + Vcv )ρ d V + ∫ cs (W + Vcv )ρW ∙ nˆ dA = ∑ Fcontents of the control volume (5.25) 224 VIDEO Chapter 5 ■ Finite Control Volume Analysis For a constant control volume velocity, Vcv, and steady flow in the control volume reference frame, V5.10 Running on water ∂ ∂t ∫ (W + Vcv )ρ d V = 0 (5.26) cv Also, for this inertial, nondeforming control volume ∫ (W + Vcv )ρW ∙ nˆ dA = cs ∫ WρW ∙ nˆ dA + V ∫ ρW ∙ nˆ dA cv cs (5.27) cs For steady flow (on an instantaneous or time-average basis), Eq. 5.15 gives ∫ ρW ∙ nˆ dA = 0 (5.28) cs The linear momentum equation for a moving control volume involves the relative velocity. Combining Eqs. 5.25, 5.26, 5.27, and 5.28, we conclude that the linear momentum equation for an inertial, moving, nondeforming control volume that involves steady (instantaneous or time-average) flow is ∫ WρW ∙ nˆ dA = ∑ F (5.29) contents of the control volume cs Example 5.17 illustrates the use of Eq. 5.29. EXAMPLE 5.17 Linear Momentum—Moving Control Volume GIVEN A vane on wheels moves with constant velocity V0 when a stream of water having a nozzle exit velocity of V1 is turned 45° by the vane as indicated in Fig. E5.17a. Note that this is the same moving vane considered in Section 4.4.6 earlier. The speed of the water jet leaving the nozzle is 100 ft /s, 45° A1 = 0.006 ft2 and the vane is moving to the right with a constant speed of 20 ft /s. FIND Determine the magnitude and direction of the force, R, exerted by the stream of water on the vane surface. Moving vane Nozzle V1 V0 ■ Figure E5.17 (a) VCV = V0 Moving control volume Moving vane 1 ft Nozzle V1 V0 Rz x Rx (1) 𝒲w (c) (b) (2) z 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 225 SOLUTION To determine the magnitude and direction of the force, R, exerted by the water on the vane, we apply Eq. 5.29 to the contents of the moving control volume shown in Fig. E5.17b. The forces acting on the contents of this control volume are indicated in Fig. E5.17c. Note that since the ambient pressure is atmospheric, all pressure forces cancel each other out. Equation 5.29 is applied to the contents of the moving control volume in component directions. For the x direction (positive to the right), we get ∫W x ρ W ∙ nˆ dA = −Rx Combining results we get Rx = ρW 21 A1 (1 − cos 45°) or Rx = (1.94 slugs/ft3 ) (80 ft/s) 2 (0.006 ft2 ) (1 − cos 45°) = 21.8 lb Also, Rz = ρW 21 (sin 45°)A1 + 𝒲w where cs 𝒲w = ρgA1ℓ or ˙ 1 ) + (+W2 cos 45°) (+m˙ 2 ) = −Rx (+W1 ) (−m (1) where ˙ 1 = ρ1W1A1  and  m˙ 2 = ρ2W2A2. m For the vertical or z direction (positive up) we get z R = √R2x + R2z = [(21.8 lb) 2 + (53 lb) 2]12 = 57.3 lb w cs The angle of R from the x direction, α, is or ˙ 2 ) = Rz − 𝒲w (+W2 sin 45°) (+m (2) We assume for simplicity that the water flow is frictionless and that the change in water elevation across the vane is negligible. Thus, from the Bernoulli equation (Eq. 3.7) we conclude that the speed of the water relative to the moving control volume, W, is constant or W1 = W2 The relative speed of the stream of water entering the control volume, W1, is W1 = V1 − V0 = 100 ft/s − 20 ft/s = 80 ft/s = W2 The water density is constant so that ρ1 = ρ2 = 1.94 slugs/ft3 Application of the conservation of mass principle to the contents of the moving control volume (Eq. 5.16) leads to m ˙ 1 = ρ1W1A1 = ρ2W2A2 = m˙ 2 VIDEO V5.11 Thrust from water nozzle Rz = (1.94 slugs/ft3 ) (80 ft/s) 2 (sin 45°) (0.006 ft2 ) + (62.4 lb/ft3 ) (0.006 ft2 ) (1 ft) = 52.6 lb + 0.37 lb = 53 lb Combining the components we get ∫ W ρW ∙ nˆ dA = R − 𝒲 z Thus, α = tan−1 Rz Rx = tan−1 (53 lb/21.8 lb) = 67.6° The force of the water on the vane is equal in magnitude but opposite in direction from R; thus it points to the right and down at an angle of 67.6° from the x direction and is equal in magnitude to 57.3 lb. (Ans) COMMENT The force of the fluid on the vane in the x direc- tion, Rx = 21.8 lb, is associated with x-direction motion of the vane at a constant speed of 20 ft/s. Since the vane is not accelerating, this x-direction force is opposed by wheel friction or some other resisting force of the same magnitude. From basic physics we recall that the power generated by the water jet is the product of force and speed. Thus, Ẇ = R V x 0 (21.8 lb) (20 ft/s) 550(ft ∙ lb)(hp ∙ s) = 0.79 hp = It is clear from the preceding examples that a flowing fluid can be forced to 1. 2. 3. 4. 5. change direction speed up or slow down have a velocity profile change do only some or all of the above do none of the above A net force on the fluid is required for achieving any or all of the first four above, which all represent a change in fluid momentum. The forces on a flowing fluid balance out with no net force for the fifth. Typical forces considered in this book include (a) pressure (b) friction (c) weight 226 Chapter 5 ■ Finite Control Volume Analysis and involve some type of constraint such as a vane, channel, or conduit to guide the flowing fluid. A flowing fluid can cause a vane, channel, or conduit to move. When this happens, power is produced. The selection of a control volume is an important matter. For determining anchoring forces, consider including fluid and its constraint in the control volume. For determining force between a fluid and its constraint, consider including only the fluid in the control volume. 5.2.3 Derivation of the Moment-of-Momentum Equation1 The momentof-momentum equation is derived from Newton’s second law. In many engineering problems, the moment of a force with respect to an axis, namely, torque, is important. Newton’s second law of motion has already led to a useful relationship between forces and linear momentum flow. The linear momentum equation can also be used to solve problems involving torques. However, by forming the moment of the linear momentum and the moment of the resultant force associated with each particle of fluid with respect to a point in an inertial coordinate system, we will develop a moment-of-momentum equation that relates torques and angular momentum flow for the contents of a control volume. When torques are important, the moment-ofmomentum equation is often more convenient to use than the linear momentum equation. Application of Newton’s second law of motion to a particle of fluid yields D (Vρ δ V) = δ Fparticle Dt (5.30) where V is the particle velocity measured in an inertial reference system, ρ is the particle density, δ V is the infinitesimally small particle volume, and δ Fparticle is the infinitesimal resultant external force acting on the particle. If we form the moment of each side of Eq. 5.30 with respect to the origin of an inertial coordinate system, we obtain r× D (Vρ δ V) = r × δ Fparticle Dt (5.31) where r is the position vector from the origin of the inertial coordinate system to the fluid particle (Fig. 5.3). We note that D(Vρ δ V) D Dr [(r × V)ρ δ V ] = × Vρ δ V + r × Dt Dt Dt (5.32) Dr =V Dt (5.33) V×V=0 (5.34) and Thus, since by combining Eqs. 5.31, 5.32, 5.33, and 5.34, we obtain the expression D [(r × V)ρ δ V] = r × δ Fparticle Dt (5.35) z V r δ Fparticle y x 1 ■ Figure 5.3 Inertial coordinate system. This section may be omitted, along with Sections 5.2.4 and 5.3.5, without loss of continuity in the text material. However, these sections are recommended for those interested in Chapter 12. 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 227 Equation 5.35 is valid for every particle of a system. For a system (collection of fluid particles), we need to use the sum of both sides of Eq. 5.35 to obtain D [(r × V)ρ d V] = ∑ (r × F) sys sys Dt (5.36) ∑ r × δFparticle = ∑ (r × F) sys (5.37) ∫ where We note that D Dt ∫ (r × V)ρ d V = sys D [(r × V)ρ d V] Dt sys ∫ (5.38) since the sequential order of differentiation and integration can be reversed without consequence. (Recall that the material derivative, D( )Dt, denotes the time derivative following a given system; see Section 4.2.1.) Thus, from Eqs. 5.36 and 5.38 we get D Dt z T=r×F T ∫ (r × V)ρ d V = ∑ (r × F) sys (5.39) sys or r Time rate of change of the = sum of external torques moment-of-momentum of the system acting on the system y F The sketch in the margin illustrates what torque, T = r × F, is. For a control volume that is instantaneously coincident with the system, the torques acting on the system and on the control volume contents will be identical: x ∑ (r × F) sys = ∑ (r × F) cv (5.40) Further, for the system and the contents of the coincident control volume that is fixed and nondeforming, the Reynolds transport theorem (Eq. 4.19) leads to D Dt ∫ (r × V)ρ d V = sys ∂ ∂t ∫ (r × V)ρ d V + cv ∫ (r × V)ρV ∙ nˆ dA (5.41) cs or Time rate of change time rate of change net rate of flow of the moment-ofof the moment-ofof the moment-ofmomentum of the = momentum of the + momentum through system contents of the the control surface control volume For a control volume that is fixed (and therefore inertial) and nondeforming, we combine Eqs. 5.39, 5.40, and 5.41 to obtain the moment-of-momentum equation: For a system, the rate of change of moment-ofmomentum equals the net torque. ∂ ∂t ∫ cv (r × V)ρ d V + ∫ (r × V)ρV ∙ nˆ dA = ∑ (r × F) contents of the cs (5.42) control volume An important category of fluid mechanical problems that is readily solved with the help of the moment-of-momentum equation (Eq. 5.42) involves machines that rotate around a single axis. Examples of these machines include rotary lawn sprinklers, ceiling fans, lawn mower blades, wind turbines, turbochargers, and gas turbine engines. As a class, these devices are often called turbomachines. 5.2.4 Application of the Moment-of-Momentum Equation2 We simplify our use of Eq. 5.42 in several ways: 1. We assume that flows considered are one-dimensional (uniform distributions of average velocity at any section). 2 This section may be omitted, along with Sections 5.2.3 and 5.3.5, without loss of continuity in the text material. However, these sections are recommended for those interested in Chapter 12. 228 Chapter 5 ■ Finite Control Volume Analysis z ^ e z ^ e VIDEO θ Control volume V5.12 Rotating lawn sprinkler r r ^ e θ Control volume W2 U2 ω Flow out Tshaft Flow out V2 Section (2) ω Section (1) r2 Section (2) Vθ 2 Tshaft Section (1) U2 = r2ω W2 (a) (b) Flow in Control volume ■ Figure 5.4 (a) Rotary water sprinkler. Section (1) Flow (b) Rotary water sprinkler, plane view. (c) Rotary water sprinkler, side view. (c) 2. We confine ourselves to steady or steady-in-the-mean cyclical flows. Thus, ∂ ∂t ∫ (r × V)ρ d V = 0 cv at any instant of time for steady flows or on a time-average basis for cyclical unsteady flows. 3. We work only with the component of Eq. 5.42 resolved along the axis of rotation. Change in moment of fluid momentum around an axis can result in torque and rotation around that same axis. Consider the rotating sprinkler sketched in Fig. 5.4. Because the direction and magnitude of the flow through the sprinkler from the inlet [section (1)] to the outlet [section (2)] of the arm changes, the water exerts a torque on the sprinkler head, causing it to tend to rotate or to actually rotate in the direction shown, much like a turbine rotor. In applying the moment-of-momentum equation (Eq. 5.42) to this flow situation, we elect to use the fixed and nondeforming control volume shown in Fig. 5.4. This disk-shaped control volume contains within its boundaries the spinning or stationary sprinkler head and the portion of the water flowing through the sprinkler contained in the control volume at an instant. The control surface cuts through the sprinkler head’s solid material so that the shaft torque that resists motion can be clearly identified. When the sprinkler is rotating, the flow field in the stationary control volume is cyclical and unsteady but steady in the mean. We proceed to use the axial component of the moment-of-momentum equation (Eq. 5.42) to analyze this flow. The integrand of the moment-of-momentum flow term in Eq. 5.42, ∫ (r × V)ρV ∙ nˆ dA cs z V (1) r r×V can be nonzero only where fluid is crossing the control surface. Everywhere else on the control surface this term will be zero because V ∙ nˆ = 0. Water enters the control volume axially through the hollow stem of the sprinkler at section (1). At this portion of the control surface, the component of r × V resolved along the axis of rotation is zero because as illustrated by the figure in the margin, r × V lies in the plane of section (1), perpendicular to the axis of rotation. Thus, there is no axial moment-of-momentum flow in at section (1). Water leaves the control volume through each of the two nozzle openings at section (2). For the exiting flow, the magnitude of the axial component of r × V is r2Vθ2, where r2 is the radius from the axis of rotation to the nozzle centerline and Vθ2 is the value of the tangential component of the velocity of the flow exiting each nozzle as observed from a frame of reference attached to the fixed and nondeforming control volume. The fluid velocity measured relative to a fixed control surface is an absolute velocity, V. The velocity 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 229 of the nozzle exit flow as viewed from the nozzle is called the relative velocity, W. The absolute and relative velocities, V and W, are related by the vector relationship V=W+U (5.43) where U is the velocity of the moving nozzle as measured relative to the fixed control surface. The cross product and the dot product involved in the moment-of-momentum flow term of Eq. 5.42, ∫ (r × V)ρV ∙ nˆ dA cs The algebraic sign of r × V is obtained by the right-hand rule. can each result in a positive or negative value. For flow into the control volume, V ∙ nˆ is negative. For flow out, V ∙ nˆ is positive. The relative directions of r, V and r × V as well as correct algebraic sign to assign the axis component of r × V can be ascertained by using the right-hand rule. The positive direction along the axis of rotation is the direction the thumb of the right-hand points when it is extended, and the remaining fingers are curled around the rotation axis in the positive direction of rotation as illustrated in Fig. 5.5. The direction of the axial component of r × V is similarly ascertained by noting the direction of the cross product of the radius from the axis of rotation, reˆr, and the tangential component of absolute velocity, Vθ eˆ θ. Thus, for the sprinkler of Fig. 5.4, we can state that [ ∫ cs (r × V)ρV ∙ nˆ dA] = (−r2Vθ2 )(+m ˙) (5.44) axial where, because of mass conservation, m ˙ is the total mass flowrate through both nozzles. As was demonstrated in Example 5.7, the mass flowrate is the same whether the sprinkler rotates or not. The correct algebraic sign of the axial component of r × V can be easily remembered in the following way: if Vθ and U are in the same direction, use +; if Vθ and U are in opposite directions, use −. The torque term [ ∑ (r × F) contents of the control volume] of the moment-of-momentum equation (Eq. 5.42) is analyzed next. Confining ourselves to torques acting with respect to the axis of rotation only, we conclude that the shaft torque is important. The net torque with respect to the axis of rotation associated with normal forces exerted on the contents of the control volume will be very small, if not zero. The net axial torque due to fluid tangential forces is also negligibly small for the control volume of Fig. 5.4. Thus, for the sprinkler of Fig. 5.4 ∑ [ (r × F) contents of the ] control volume axial = Tshaft (5.45) Note that we have entered Tshaft as a positive quantity in Eq. 5.45. This is equivalent to assuming that Tshaft is in the same direction as rotation. For the sprinkler of Fig. 5.4, the axial component of the moment-of-momentum equation (Eq. 5.42) is, from Eqs. 5.44 and 5.45 −r2Vθ2m ˙ = Tshaft (5.46) We interpret Tshaft being a negative quantity from Eq. 5.46 to mean that the shaft torque actually opposes the rotation of the sprinkler arms as shown in Fig. 5.4. The shaft torque, Tshaft, opposes rotation in all turbine devices. r×V + r V ■ Figure 5.5 Right-hand rule (a) (b) convention. Free ebooks ==> www.Ebook777.com 230 Chapter 5 ■ Finite Control Volume Analysis Power is equal to angular velocity times torque. We evaluate the shaft power, Ẇ shaft, associated with shaft torque, Tshaft, by forming the product of Tshaft and the rotational speed of the shaft, ω. [We use the notation that W = work, ( ˙ ) = d( )dt, and thus Ẇ = power.] Thus, from Eq. 5.46 we get Ẇ shaft = Tshaft ω = −r2Vθ2m ˙ω (5.47) Since r2ω is the speed of each sprinkler nozzle, U, we can also state Eq. 5.47 in the form Ẇ shaft = −U2Vθ2m ˙ (5.48) Shaft work per unit mass, wshaft, is equal to Ẇ shaft m ˙ . Dividing Eq. 5.48 by the mass flowrate, m˙ , we obtain wshaft = −U2Vθ2 VIDEO V5.13 Impulsetype lawn sprinkler (5.49) Negative shaft work as in Eqs. 5.47, 5.48, and 5.49 is work out of the control volume, that is, work done by the fluid on the rotor and thus its shaft. The principles associated with this sprinkler example can be extended to handle most simplified turbomachine flows. The fundamental technique is not difficult. However, the geometry of some turbomachine flows is quite complicated. Example 5.18 further illustrates how the axial component of the moment-of-momentum equation (Eq. 5.46) can be used. EXAM PLE 5.18 Moment-of-Momentum—Torque GIVEN Water enters a rotating lawn sprinkler through its base at the steady rate of 1000 ml/s as sketched in Fig. E5.18a. The exit area of each of the two nozzles is 30 mm2 and the flow leaving each nozzle is in the tangential direction. The radius from the axis of rotation to the centerline of each nozzle is 200 mm. FIND (a) Determine the resisting torque required to hold the sprinkler head stationary. r2 = 200 mm Nozzle exit area = 30 mm2 Control volume Flow out Flow out Tshaft (b) Determine the resisting torque associated with the sprinkler rotating with a constant speed of 500 rev/min. (c) Determine the speed of the sprinkler if no resisting torque is applied. Q = 1000 ml/s (a) SOLUTION To solve parts (a), (b), and (c) of this example we can use the same fixed and nondeforming, disk-shaped control volume illustrated in Fig. 5.4. As indicated in Fig. E5.18a, the only axial torque considered is the one resisting motion, Tshaft. V2 = Vθ 2 V2 V1 (a) When the sprinkler head is held stationary as specified in part (a) of this example problem, the velocities of the fluid entering and leaving the control volume are shown in Fig. E5.18b. Equation 5.46 applies to the contents of this control volume. Thus, Tshaft = −r2Vθ2m ˙ (b) V2 = Vθ 2 (1) W2 Since the control volume is fixed and nondeforming and the flow exiting from each nozzle is tangential, U2 Vθ2 = V2 ω W2 V2 (2) V1 (c) Equations 1 and 2 give ˙ Tshaft = −r2V2m (3) U2 = r2ω ■ Figure E5.18 www.Ebook777.com 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations In Example 5.7, we ascertained that V2 = 16.7 m/s. Thus, from Eq. 3 with m ˙ = Qρ = (1000 ml/s) (10−3 m3/liter) (999 kg/m3 ) (1000 ml/liter) (c) When no resisting torque is applied to the rotating sprinkler head, a maximum constant speed of rotation will occur as demonstrated below. Application of Eqs. 3, 4, and 5 to the contents of the control volume results in Tshaft = −r2 (W2 − r2ω)m ˙ = 0.999 kg/s (6) For no resisting torque, Eq. 6 yields we obtain Tshaft = − 231 (200 mm) (16.7 m/s) (0.999 kg/s)[1 (N/kg)(m/s2 )] (1000 mm/m) ˙ 0 = −r2 (W2 − r2ω)m Thus, or Tshaft = −3.34 N ∙ m ω= (Ans) (b) When the sprinkler is rotating at a constant speed of 500 rpm, the flow field in the control volume is unsteady but cyclical. Thus, the flow field is steady in the mean. The velocities of the flow entering and leaving the control volume are as indicated in Fig. E5.18c. The absolute velocity of the fluid leaving each nozzle, V2, is from Eq. 5.43, V2 = W2 − U2 (4) ω= U2 = r2ω (5) Application of the axial component of the moment-of-momentum equation (Eq. 5.46) leads again to Eq. 3. From Eqs. 4 and 5, V2 = 16.7 m/s − r2ω = 16.7 m/s − W2 (16.7 m/s) (1000 mm/m) = = 83.5 rad/s r2 (200 mm) or ω= as determined in Example 5.7. The speed of the nozzle, U2, is obtained from (200 mm) (500 rev/min) (2π rad/rev) (1000 mm/m) (60 s/min) (83.5 rad/s) (60 s/min) = 797 rpm 2 π rad/rev COMMENT Note that forcing a change in direction of a flowing fluid, in this case with a sprinkler, resulted in rotary motion and a useful “sprinkling” of water over an area. By repeating the calculations for various values of the angular velocity, ω, the results shown in Fig. E5.18d are obtained. It is seen that the magnitude of the resisting torque associated with rotation is less than the torque required to hold the rotor stationary. Even in the absence of a resisting torque, the rotor maximum speed is finite. 0 V2 = 16.7 m/s − 10.5 m/s = 6.2 m/s Tshaft = −1.24 N ∙ m (Ans) COMMENT Note that the resisting torque associated with sprinkler head rotation is much less than the resisting torque that is required to hold the sprinkler stationary. 0 200 400 600 800 –1 TShaft , N . m or ω, rpm –0.5 ˙ = 0.999 kg/s (as calculated previThus, using Eq. 3, with m ously), we get (200 mm) (6.2 m/s) 0.999 kg/s [1 (N/kg)(m/s2 )] (1000 mm/m) (Ans) For this condition (Tshaft = 0), the water both enters and leaves the control volume with zero angular momentum. or Tshaft = − (7) In Example 5.4, we learned that the relative velocity of the fluid leaving each nozzle, W2, is the same regardless of the speed of rotation of the sprinkler head, ω, as long as the mass flowrate of the fluid, m ˙ , remains constant. Thus, by using Eq. 7 we obtain where W2 = 16.7 m/s W2 r2 –1.5 –2 –2.5 –3 –3.5 –4 ■ Figure E5.18d When the moment-of-momentum equation (Eq. 5.42) is applied to a more general, onedimensional flow through a rotating machine, we obtain Tshaft = (−m ˙ in )(±rinVθ in ) + m˙ out (±routVθout ) (5.50) 232 Chapter 5 ■ Finite Control Volume Analysis rVθ > 0 θ r Vθ W Vr V U by applying the same kind of analysis used with the sprinkler of Fig. 5.4. The “−” is used with mass flowrate into the control volume, m ˙ in, and the “+” is used with mass flowrate out of the control volume, m ˙ out, to account for the sign of the dot product, V ∙ nˆ , involved. Whether “+” or “−” is used with the rVθ product depends on the direction of (r × V) axial. A simple way to determine the sign of the rVθ product is to compare the direction of Vθ and the blade speed, U. As shown in the margin, if Vθ and U are in the same direction, then the rVθ product is positive. If Vθ and U are in opposite directions, the rVθ product is negative. The sign of the shaft torque is “+” if Tshaft is in the same direction along the axis of rotation as ω, and “−” otherwise. The shaft power, Ẇ shaft, is related to shaft torque, Tshaft, by Ẇ shaft = Tshaft ω (5.51) Thus, using Eqs. 5.50 and 5.51 with a “+” sign for Tshaft in Eq. 5.50, we obtain Ẇ shaft = (−m ˙ in )(±rinωVθin ) + m˙ out (±rout ωVθout ) (5.52) or since rω = U r rVθ < 0 Vθ Ẇ shaft = (−m ˙ in )(±UinVθin ) + m˙ out (±UoutVθout ) Vr V W U When shaft torque and shaft rotation are in the same (opposite) direction, power is into (out of) the fluid. (5.53) The “+” is used for the UVθ product when U and Vθ are in the same direction; the “−” is used when U and Vθ are in opposite directions. Also, since +Tshaft was used to obtain Eq. 5.53, when Ẇ shaft is positive, power is into the fluid (for example, a pump), and when Ẇ shaft is negative, power is out of the fluid (for example, a turbine). The shaft work per unit mass, wshaft, can be obtained from the shaft power, Ẇ shaft, by dividing Eq. 5.53 by the mass flowrate, m ˙ . By conservation of mass, m ˙ = m˙ in = m˙ out From Eq. 5.53, we obtain wshaft = −(±UinVθin ) + ( ±UoutVθout ) (5.54) The application of Eqs. 5.50, 5.53, and 5.54 is demonstrated in Example 5.19. More examples of the application of Eqs. 5.50, 5.53, and 5.54 are included in Chapter 12. EXAMPLE 5.19 Moment-of-Momentum—Power GIVEN An air fan has a bladed rotor of 12-in. outside diameter and 10-in. inside diameter as illustrated in Fig. E5.19a. The height of each rotor blade is constant at 1 in. from blade inlet to outlet. The flowrate is steady, on a time-average basis, at 230 ft3/min and the absolute velocity of the air at blade inlet, V1, is radial. The blade discharge angle is 30° from the tangential direction. The rotor rotates at a constant speed of 1725 rpm. FIND Estimate the power required to run the fan. SOLUTION We select a fixed and nondeforming control volume that includes the rotating blades and the fluid within the blade row at an instant, as shown with a dashed line in Fig. E5.19a. The flow within this control volume is cyclical, but steady in the mean. The only torque we consider is the driving shaft torque, Tshaft. This torque is provided by a motor. We assume that the entering and leaving flows are each represented by uniformly distributed velocities and flow properties. Since shaft power is sought, Eq. 5.53 is appropriate. Application of Eq. 5.53 to the contents of the control volume in Fig. E5.19 gives 0 (V1 is radial) ˙ shaft = −m˙ 1 ( ±U1Vθ1 ) + m˙ 2 ( ±U2Vθ2 ) W (1) From Eq. 1 we see that to calculate fan power, we need mass flowrate, m ˙ , rotor exit blade velocity, U2, and fluid tangential 5.2 Newton’s Second Law—The Linear Momentum and Moment-of-Momentum Equations 233 Section (1) 30° Fixed control volume Tshaft W2 V1 ω Section (2) D2 = 2r2 = 12 in. D1 = 2r1 = 10 in. h= 1 in. ω U2 Fixed control volume Wr2 W2 Tshaft Vr2 V2 Vθ 2 30° (a) (b) ■ Figure E5.19 velocity at blade exit, Vθ2. The mass flowrate, m ˙ , is easily obtained from Eq. 5.6 as m ˙ = ρQ = (2.38 × 10−3 slug/ft3 ) (230 ft3/min) (60 s/min) = 0.00912 slug/s W2 sin 30° = Vr 2 (2) Often, problems involving fans are solved using English Engineering units. Since 1 slug = 32.174 lbm, we could have used as the density of air ρair = (2.38 × 10−3 slug/ft3 ) (32.174 lbm/slug) = 0.0766 lbm/ft3. Then 3 m ˙= 3 (0.0766 lbm/ft ) (230 ft /min) = 0.294 lbm/s (60 s/min) m ˙ = ρA2Vr 2 (7) A2 = 2 π r2h (8) or since (9) Taking Eqs. 6 and 9 together we get W2 = (3) To determine the fluid tangential speed at the fan rotor exit, Vθ2, we use Eq. 5.43 to get (4) The vector addition of Eq. 4 is shown in the form of a “velocity triangle” in Fig. E5.19b. From Fig. E5.19b, we can see that Vθ2 = U2 − W2 cos 30° where Vr2 is the radial component of either W2 or V2. Also, using Eq. 5.6, we obtain ˙ = ρ2πr2hVr 2 m (6 in.) (1725 rpm) (2π rad/rev) U2 = r2ω = (12 in./ft) (60 s/min) V2 = W2 + U2 (6) where h is the blade height, Eqs. 7 and 8 combine to form The rotor exit blade speed, U2, is = 90.3 ft/s To solve Eq. 5 for Vθ2 we need a value of W2, in addition to the value of U2 already determined (Eq. 3). To get W2, we recognize that (5) = ρQ m ˙ = ρ2πr2h sin 30° ρ2πr2h sin 30° Q 2πr2h sin 30° Substituting known values into Eq. 10, we obtain W2 = (230 ft3/min) (12 in./ft) (12 in./ft) (60 s/min)2π(6 in.) (1 in.) sin 30° = 29.3 ft/s (10) 234 Chapter 5 ■ Finite Control Volume Analysis COMMENT Note that the “+” was used with the U2Vθ 2 product because U2 and Vθ 2 are in the same direction. This result, 0.097 hp, is the power that needs to be delivered through the fan shaft for the given conditions. Ideally, all of this power would go into the flowing air. However, because of fluid friction, only some of this power will produce useful effects (e.g., movement and pressure rise) on the air. How much useful effect depends on the efficiency of the energy transfer between the fan blades and the fluid. Also, extra power would be needed from the motor to overcome friction in shaft bearings and other mechanical resistance. By using this value of W2 in Eq. 5 we get Vθ2 = U2 − W2 cos 30° = 90.3 ft/s − (29.3 ft/s) (0.866) = 64.9 ft/s Equation 1 can now be used to obtain ˙ U2Vθ2 = Ẇ shaft = m (0.00912 slug/s) (90.3 ft/s) (64.9 ft/s) [1 (slug ∙ ft/s2 )lb][550 (ft ∙ lb)(hp ∙ s)] with BG units. With EE units Ẇ shaft = (0.294 lbm/s) (90.3 ft/s) (64.9 ft/s) [32.174 (lbm ∙ ft)(lb/s2 )][550 (ft ∙ lb)(hp ∙ s)] In either case Ẇ shaft = 0.097 hp 5.3 (Ans) First Law of Thermodynamics—The Energy Equation 5.3.1 Derivation of the Energy Equation The first law of thermodynamics for a system is, in words, Time rate of net time rate of net time rate of increase of the energy addition by energy addition by = + total stored energy heat transfer into work transfer into of the system the system the system In symbolic form, this statement is D Dt The first law of thermodynamics is a statement of conservation of energy. ∫ sys eρ d V = (∑ Q̇ in − ∑ Q̇ out) sys + (∑ Ẇ in − ∑ Ẇ out) sys or D Dt ∫ eρ d V = (Q̇ net + Ẇ net ) sys in sys (5.55) in Some of these variables deserve a brief explanation before proceeding further. The total stored energy per unit mass for each particle in the system, e, is related to the internal energy per unit mass, uˇ , the kinetic energy per unit mass, V 2/2, and the potential energy per unit mass, gz, by the equation e = uˇ + V2 + gz 2 (5.56) In this equation we do not include chemical energy stored in molecular structure and nuclear energy stored in atomic structure because we will not be considering changes of these types of energy. The net rate of heat transfer into the system is denoted with Q̇ net in, and the net rate of work transfer into the system is labeled Ẇ net in. Heat transfer and work transfer are considered “+” going into the system and “−” coming out. [It should be noted that the opposite sign convention is often used for work (see Ref. 3 for example).] Equation 5.55 is valid for inertial and noninertial reference systems. We proceed to develop the control volume statement of the first law of thermodynamics. For the control volume that is coincident with the system at an instant of time (Q̇ net + Ẇ net ) sys = (Q̇ net + Ẇ net ) coincident (5.57) in in in in control volume 5.3 First Law of Thermodynamics—The Energy Equation 235 Furthermore, for the system and the contents of the coincident control volume that is fixed and nondeforming, the Reynolds transport theorem (Eq. 4.19 with the parameter b set equal to e) allows us to conclude that D Dt ∫ ∂ ∂t eρ d V = sys ∫ eρ d V + cv ∫ eρV ∙ nˆ dA (5.58) cs or in words, Time rate net rate of flow time rate of increase of increase of the total stored energy of the total stored of the total = + out of the control energy of the contents stored energy volume through the of the control volume of the system control surface Combining Eqs. 5.55, 5.57, and 5.58, we get the control volume formula for the first law of thermodynamics: The energy equation relates stored energy and heat and work transfer. • Q3 ∂ ∂t • Q2 • Q4 • Q1 • Control Volume • • • • Qnet = Q1 + Q2 – Q3 – Q4 in ∫ eρ d V + cv ∫ eρV ∙ nˆ dA = (Q̇ net + Ẇ net ) cv in cs (5.59) in The total stored energy per unit mass, e, in Eq. 5.59 is for fluid particles entering, leaving, and within the control volume. Further explanation of the heat transfer and work transfer involved in this equation follows. The heat transfer rate, Q̇ , represents all of the ways in which energy is exchanged between the control volume contents and surroundings because of a temperature difference. Thus, radiation, conduction, and/or convection are possible. As shown by the figure in the margin, heat transfer into the control volume is considered positive; heat transfer out is negative. In many engineering applications, the process is adiabatic; the heat transfer rate, Q̇ , is zero. The net heat transfer rate, Q̇ net in, can also be zero when ∑ Q̇ in − ∑ Q̇ out = 0. The work transfer rate, Ẇ , also called power, is positive when work is done on the contents of the control volume by the surroundings. Otherwise, it is considered negative. Work can be transferred across the control surface in several ways. In the following paragraphs, we consider some important forms of work transfer. In many instances, work is transferred across the control surface by a moving shaft. In rotary devices such as turbines, fans, and propellers, a rotating shaft transfers work across that portion of the control surface that slices through the shaft. Even in reciprocating machines like positive displacement internal combustion engines and compressors that utilize piston-in-cylinder arrangements, a rotating crankshaft is used. Since work is the dot product of force and related displacement, rate of work (or power) is the dot product of force and related displacement per unit time. For a rotating shaft, the power transfer, Ẇ shaft, is related to the shaft torque that causes the rotation, Tshaft, and the angular velocity of the shaft, ω, by the relationship Ẇ shaft = Tshaft ω When the control surface cuts through the shaft material, the shaft torque is exerted by shaft material at the control surface. To allow for consideration of problems involving more than one shaft we use the notation Ẇ shaft = ∑ Ẇ shaft − ∑ Ẇ shaft net in in (5.60) out Work transfer can also occur at the control surface when a force associated with fluid normal stress acts over a distance. Consider the simple pipe flow illustrated in Fig. 5.6 and the control volume shown. For this situation, the fluid normal stress, σ, is simply equal to the negative of fluid pressure, p, in all directions; that is, σ = −p (5.61) 236 Chapter 5 ■ Finite Control Volume Analysis Section (1) R Control volume umax r [ ()] u1 = umax 1 - _r R V F Section (2) 2 Pipe umax ■ Figure 5.6 Simple, fully developed [ ()] u2 = umax 1 - _r R 2 pipe flow. This relationship can be used with varying amounts of approximation for many engineering problems (see Chapter 6). ̇ , associated with a force F acting on an object moving with velocity V The power transfer, W is given by the dot product F ∙ V. This is illustrated by the figure in the margin. Hence, the power transfer associated with normal stresses acting on a single fluid particle, δẆ normal stress, can be evaluated as the dot product of the normal stress force, δ Fnormal stress, and the fluid particle velocity, V, as δẆ normal stress = δ Fnormal stress ∙ V θ • W = F•V = FV cos θ If the normal stress force is expressed as the product of local normal stress, σ = −p, and fluid particle surface area, nˆ δA, the result is δ Ẇ normal stress = σ nˆ δA ∙ V = −pnˆ δA ∙ V = −pV ∙ nˆ δA For all fluid particles on the control surface of Fig. 5.6 at the instant considered, power transfer due to fluid normal stress, Ẇ normal stress, is Ẇ normal = stress τ V ^ n • δ Wtangential stress = 0 σV ∙ nˆ dA = cs ∫ −pV ∙ nˆ dA (5.62) cs The work represented by the pressure integral in Eq. 5.62 is often called flow work. Physically, it represents the sum of work done on the system by fluid particles pushing into the control volume (+) and work done on the surrounding fluid by particles pushing out from the control volume (−). Note that the value of Ẇ normal stress for particles on the wetted inside surface of the pipe is zero because V ∙ nˆ is zero there. Thus, Ẇ normal stress can be nonzero only where fluid enters and leaves the control volume. Although only a simple pipe flow was considered, Eq. 5.62 is quite general, and the control volume used in this example can serve as a general model for other cases. Work transfer can also occur at the control surface because of tangential stress forces. Rotating shaft work is transferred by tangential stresses in the shaft material. For a fluid particle, shear stress force power, δẆ tangential stress, can be evaluated as the dot product of tangential stress force, δFtangential stress, and the fluid particle velocity, V. That is, δẆ tangential stress = δFtangential stress ∙ V Work is transferred by rotating shafts, normal stresses, and tangential stresses. p ∫ For the control volume of Fig. 5.6, the fluid particle velocity is zero everywhere on the wetted inside surface of the pipe. Thus, no tangential stress work is transferred across that portion of the control surface. Furthermore, if we select the control surface so that it is perpendicular to the fluid particle velocity, then the tangential stress force is also perpendicular to the velocity. Therefore, the tangential stress work transfer is zero on that part of the control surface. This is illustrated in the figure in the margin. Thus, in general, we select control volumes like the one of Fig. 5.6 and consider fluid tangential stress power transfer to be negligibly small. Using the information we have developed about power, we can express the first law of thermodynamics for the contents of a control volume by combining Eqs. 5.59, 5.60, and 5.62 to obtain ∂ ∂t ∫ cv eρ d V + ∫ cs eρV ∙ nˆ dA = Q̇ net + Ẇ shaft − in net in ∫ pV ∙ nˆ dA cs (5.63) 5.3 First Law of Thermodynamics—The Energy Equation 237 Finally, we multiply and divide by ρ inside the pressure integral, substitute the stored energy from Eq. 5.56, and group terms to obtain the energy equation: ∂ ∂t ∫ eρ d V + cv 2 ∫ (uˇ + ρp + V2 + gz) ρV ∙ nˆ dA = Q̇ net in cs + Ẇ shaft (5.64) net in 5.3.2 Application of the Energy Equation The term p/ρ in the energy equation represents flow work, not stored energy. In Eq. 5.64, the term ∂∂t ʃcv eρ d V represents the time rate of change of the total stored energy, e, of the contents of the control volume. This term is zero when the flow is steady. This term is also zero in the mean when the flow is steady in the mean (cyclical). In Eq. 5.64, the integrand of 2 ∫ (uˇ + ρp + V2 cs + gz) ρV ∙ nˆ dA can be nonzero only where fluid crosses the control surface (V ∙ nˆ ≠ 0). Otherwise, V ∙ nˆ is zero and the integrand is zero for that portion of the control surface. If the properties within parentheses, uˇ , pρ, V 2/2, and gz, are all assumed to be uniformly distributed over the flow cross-sectional areas involved, the integration becomes simple and gives 2 2 ∫ (uˇ + ρp + V2 + gz)ρV ∙ nˆ dA = ∑ (uˇ + pρ + V2 flow out cs + gz) m ˙ p V2 − ∑ (uˇ + + + gz) m ˙ ρ 2 flow (5.65) in Furthermore, if there is only one stream entering and leaving the control volume, then 2 ∫ (uˇ + ρp + V2 cs + gz) ρV ∙ nˆ dA = p p V2 V2 ˇ+ + ˇ+ + u + gz m ˙ − u + gz) m ˙ in out ( )out ( ρ ρ 2 2 in (5.66) Uniform flow as described above will occur in an infinitesimally small diameter streamtube as illustrated in Fig. 5.7. This kind of streamtube flow is representative of the steady flow of a particle of fluid along a pathline. We can also idealize actual conditions by disregarding nonuniformities in a finite cross section of flow. We call this one-dimensional flow, and although such uniform flow rarely occurs in reality, the simplicity achieved with the one-dimensional approximation often justifies its use. More details about the effects of nonuniform distributions of velocities and other fluid flow variables are considered in Section 5.3.4 and in Chapters 8, 9, and 10. If shaft work is involved, the flow must be unsteady, at least locally (see Refs. 1 and 2). The flow in any fluid machine that involves shaft work is unsteady within that machine. For example, the velocity and pressure at a fixed location near the rotating blades of a fan are unsteady owing to blade passing. However, upstream and downstream of the machine, the flow may be steady. Most often shaft work is associated with flow that is unsteady in a recurring or cyclical way. On a timeaverage basis for flow that is one-dimensional, cyclical, and involves only one stream of fluid entering m• out dA Streamtube • min V ■ Figure 5.7 Streamtube flow. 238 Chapter 5 ■ Finite Control Volume Analysis and leaving the control volume, such as the hair dryer shown in the figure in the margin, Eq. 5.64 can be simplified with the help of Eqs. 5.9 and 5.66 to form m• out p p V 2 − V 2in m ˙ [uˇ out − uˇ in + ( ) − ( ) + out + g(zout − zin ) ] = Q̇ net + Ẇ shaft ρ out ρ in 2 in net in m• in m• in = m• out = m• (5.67) We call Eq. 5.67 the one-dimensional energy equation for steady-in-the-mean flow. Note that Eq. 5.67 is valid for incompressible and compressible flows. Often, the fluid property called enthalpy, hˇ , where p hˇ = uˇ + ρ The energy equation is sometimes written in terms of enthalpy. (5.68) is used in Eq. 5.67. With enthalpy, the one-dimensional energy equation for steady-in-the-mean flow (Eq. 5.67) is V 2 − V 2in m ˙ [hˇ out − hˇ in + out + g(zout − zin ) ] = Q̇ net + Ẇ shaft 2 in net in (5.69) Equation 5.69 is often used for solving compressible flow problems. Examples 5.20 and 5.21 illustrate how Eqs. 5.67 and 5.69 can be used. EXAM PLE 5.20 Energy—Pump Power GIVEN A pump delivers water at a steady rate of 300 gal/min as shown in Fig. E5.20. Just upstream of the pump [section (1)] where the pipe diameter is 3.5 in., the pressure is 18 psi. Just downstream of the pump [section (2)] where the pipe diameter is 1 in., the pressure is 60 psi. The change in water elevation across the pump is zero. The rise in internal energy of water, uˇ 2 − uˇ 1 , associated with a temperature rise across the pump is 93 ft ∙ lb/lbm. The pumping process is considered to be adiabatic. • Wshaft = ? Control volume D2 = 1 in. Q= D1 = Section (2) Section (1) p1 = 18 psi FIND Determine the power (hp) required by the pump. 300 gal/min. Pump 3.5 in. p2 = 60 psi ^ ^ u2 – u1 = 93 ft • lb/lbm ■ Figure E5.20 SOLUTION We include in our control volume the water contained in the pump between its entrance and exit sections. Application of Eq. 5.67 to the contents of this control volume on a time-average basis yields out of the pump, V2. All other quantities in Eq. 1 are given in the problem statement. From Eq. 5.6, we get m ˙ = ρQ = 0 (no elevation change) ˙ [uˇ 2 − uˇ 1 + ( m (1.94 slugs/ft3 ) (300 gal/min) (32.174 lbm/slug) (7.48 gal/ft3 ) (60 s/min) = 41.8 lbm/s p p V 22 − V 21 + g(z2 − z1 ) −( + ] ) ) ρ 2 ρ 1 2 Also from Eq. 5.6, V= 0 (adiabatic flow) = Q˙ net + Ẇ shaft in (2) (1) net in We can solve directly for the power required by the pump, Ẇ shaft net in, from Eq. 1, after we first determine the mass flowrate, m ˙ , the speed of flow into the pump, V1, and the speed of the flow Q Q = A πD24 so V1 = (300 gal/min)4 (12 in./ft) 2 Q = A1 (7.48 gal/ft3 ) (60 s/min)π (3.5 in.) 2 = 10.0 ft/s (3) 5.3 First Law of Thermodynamics—The Energy Equation and (300 gal/min)4 (12 in./ft) 2 Q = A2 (7.48 gal/ft3 ) (60 s/min)π (1 in.) 2 V2 = = 123 ft/s Substituting the values of Eqs. 2, 3, and 4 and values from the problem statement into Eq. 1 we obtain Ẇ shaft = (41.8 lbm/s) (93 ft ∙ lb/lbm) [ net in + − (60 psi) (144 in.2/ft2 ) (1.94 slugs/ft3 ) (32.174 lbm/slug) (123 ft/s) 2 − (10.0 ft/s) 2 + 2[32.174 (lbm∙ft)(lb∙s2 )] ] × 1 = 32.2 hp [550(ft∙lb/s)/hp] (4) 239 (Ans) COMMENT Of the total 32.2 hp, internal energy change accounts for 7.09 hp, the pressure rise accounts for 7.37 hp, and the kinetic energy increase accounts for 17.8 hp. An actual pump would require slightly more than 32.2 hp owing to mechanical friction loss in bearings and shaft seals. The sum of the pressure rise, the kinetic energy increase, and any increase in elevation is sometimes called water horsepower. (18 psi) (144 in.2/ft2 ) (1.94 slugs/ft3 ) (32.174 lbm/slug) E X AMPLE 5.21 Energy—Turbine Power per Unit Mass of Flow GIVEN A steam turbine generator unit used to produce elec- tricity is shown in Fig. E5.21a. Assume the steam enters a turbine with a velocity of 30 m/s and enthalpy, hˇ 1 , of 3348 kJ/kg (see Fig. E5.21b). The steam leaves the turbine as a mixture of vapor and liquid having a velocity of 60 m/s and an enthalpy of 2550 kJ/kg. The flow through the turbine is adiabatic, and changes in elevation are negligible. FIND Determine the work output involved per unit mass of steam through-flow. SOLUTION We use a control volume that includes the steam in the turbine from the entrance to the exit as shown in Fig. E5.21b. Applying Eq. 5.69 to the steam in this control volume we get 0 (elevation change is negligible) 0 (adiabatic flow) V2 − V21 ˙ [hˇ 2 − hˇ 1 + 2 + g(z2 − z1 ) ] = Q̇ net + Ẇ shaft m 2 in net in ■ Figure E5.21a Control volume (1) Steam turbine The work output per unit mass of steam through-flow, wshaft net in, can be obtained by dividing Eq. 1 by the mass flowrate, m ˙ , to obtain Ẇ shaft wshaft = net in net in m ˙ V22 − V21 = hˇ 2 − hˇ 1 + 2 Since wshaft net out = −wshaft net in, we obtain V21 − V22 wshaft = hˇ 1 − hˇ 2 + 2 net out or wshaft = 3348 kJ/kg − 2550 kJ/kg net out + [(30 m/s) 2 − (60 m/s) 2][1 J/(N∙m)] 2[1 (kg∙m)(N∙s2 )](1000 J/kJ) Section (1) V ^1 = 30 m/s h1 = 3348 kJ/kg wshaft = ? (2) ■ Figure E5.21b Section (2) V ^2 = 60 m/s h2 = 2550 kJ/kg Thus, wshaft = 3348 kJ/kg − 2550 kJ/kg − 1.35 kJ/kg net out = 797 kJ/kg (Ans) COMMENT Note that in this particular example, the change in kinetic energy is small in comparison to the difference in enthalpy involved. This is often true in applications involving steam flows. ˙ shaft , we must know the mass To determine the power output, W flowrate, m ˙ . If it is not specified, steam enthalpy can be found from steam tables or appropriate software (see Ref. 3). Free ebooks ==> www.Ebook777.com 240 Chapter 5 ■ Finite Control Volume Analysis If the flow is truly steady throughout, so that no work is done, one-dimensional, and only one fluid stream is involved, then setting the shaft work to zero, the energy equation becomes p p V 2 − V 2in m ˙ [uˇ out − uˇ in + ( ) − ( ) + out + g(zout − zin ) ] = Q̇ net ρ out ρ in 2 in VIDEO V5.14 Pelton wheel turbine (5.70) We call Eq. 5.70 the one-dimensional, steady-flow energy equation. This equation is valid for incompressible and compressible flows. For compressible flows, enthalpy is most often used in the one-dimensional, steady-flow energy equation and, thus, we have V 2 − V 2in m ˙ [hˇ out − hˇ in + out + g(zout − zin ) ] = Q̇ net 2 in (5.71) An example of the application of Eq. 5.70 follows. EXAMPLE 5.22 Energy—Temperature Change GIVEN The 420-ft waterfall shown in Fig. E5.22a involves steady flow from one large body of water to another. FIND Determine the temperature change associated with this flow. SOLUTION To solve this problem, we consider a control volume consisting of a small cross-sectional streamtube from the nearly motionless surface of the upper body of water to the nearly motionless surface of the lower body of water as is sketched in Fig. E5.22b. We need to determine T2 − T1. This temperature change is related to the change of internal energy of the water, uˇ 2 − uˇ 1, by the relationship T2 − T1 = uˇ 2 − uˇ 1 cˇ Section (1) (1) ■ Figure E5.22a [Photograph of Akaka Falls (Hawaii) courtesy of Scott and Margaret Jones.] Control volume where cˇ = 1 Btu(lbm ∙ °R) is the specific heat of water. The application of Eq. 5.70 to the contents of this control volume leads to 420 ft Section (2) p p V 2 − V 21 m ˙ [uˇ 2 − uˇ 1 + ( ) − ( ) + 2 + g(z2 − z1 ) ] ρ 2 ρ 1 2 = Q̇ net (2) in We assume that the flow is adiabatic. Thus Q̇ net in = 0. Also, because the flow is incompressible (ρ1 = ρ2) and atmospheric pressure prevails at sections (1) and (2) (p1 = p2). p p = ( ρ )1 ( ρ )2 ■ Figure E5.22b www.Ebook777.com (3) 5.3 Furthermore, V 1 = V2 = 0 (4) First Law of Thermodynamics—The Energy Equation T2 − T1 = (32.2 ft/s2 ) (420 ft) [778 ft ∙ lb/(lbm ∙ °R)][32.2 (lbm ∙ ft)(lb ∙ s2 )] = 0.540 °R because the surface of each large body of water is considered motionless. Thus, Eqs. 1 through 4 combine to yield 241 (Ans) COMMENT Note that it takes a considerable change of potential energy to produce even a small increase in temperature. g(z1 − z2 ) T2 − T1 = cˇ so that with cˇ = [1 Btu/(lbm ∙ °R)] (778 ft ∙ lb/ Btu) = [778 ft ∙ lb/(lbm ∙ °R)] A form of the energy equation that is most often used to solve incompressible flow problems is developed in the next section. 5.3.3 The Mechanical Energy Equation and the Bernoulli Equation When the one-dimensional energy equation for steady-in-the-mean flow, Eq. 5.67, is applied to a flow that is truly steady, Eq. 5.67 becomes the one-dimensional, steady-flow energy equation, Eq. 5.70. The only difference between Eq. 5.67 and Eq. 5.70 is that shaft power, Ẇ shaft net in, is zero if the flow is steady throughout the control volume (fluid machines involve locally unsteady flow). If in addition to being steady, the flow is incompressible (ρout = ρin = ρ), we get from Eq. 5.70 pin pout V 2out − V 2in − + + g(zout − zin ) ] = Q̇ net ρ ρ 2 in . Dividing Eq. 5.72 by the mass flowrate, m, and rearranging terms we obtain m ˙ [uˇ out − uˇ in + pout pin V 2out V 2in + + + gzout = + gzin − (uˇ out − uˇ in − qnet ) ρ ρ 2 2 in (5.72) (5.73) where qnet = in Q̇ net in m ˙ is the heat transfer rate per mass flowrate, or heat transfer per unit mass. Note that Eq. 5.73 involves energy per unit mass and is applicable to one-dimensional, steady, incompressible flow of a single stream of fluid between two sections or flow along a streamline between two sections. Now assume that the steady, incompressible flow that we are considering is also frictionless. In this case, the Bernoulli equation, Eq. 3.17, can also be used to describe what happens along a streamline between two sections in the flow pout + ρV 2out ρV 2in + γzout = pin + + γzin 2 2 (5.74) Now we divide Eq. 5.74 by density, ρ, and obtain pout pin V 2out V 2in + + + gzout = + gzin ρ ρ 2 2 (5.75) Equation 5.75 is the Bernoulli equation in energy form (see Sec. 3.4). It is derived from Newton’s second law of motion, which is a different physical principle from conservation of energy, so Eqs. 5.73 and 5.75 are independently true. A comparison of Eqs. 5.73 and 5.75 prompts us to conclude that uˇ out − uˇ in − qnet = 0 in (5.76) 242 Chapter 5 ■ Finite Control Volume Analysis when the steady incompressible flow is frictionless. For steady incompressible flow with friction, we learn from experience (the second law of thermodynamics; see Sec. 5.4 for details) that uˇ out − uˇ in − qnet > 0 (5.77) in In Eqs. 5.73 and 5.75, we can consider the combination of variables p V2 + + gz ρ 2 as equal to useful energy. This energy is useful because it can be completely used to produce work. Useful energy is sometimes called available energy. Thus, from inspection of Eqs. 5.73 and 5.75, we can conclude that uˇ out − uˇ in − qnet in represents the loss of useful or available energy that occurs in an incompressible fluid flow because of friction. In equation form we have uˇ out − uˇ in − qnet = loss (5.78) in For a frictionless flow, Eqs. 5.73 and 5.75 tell us that loss equals zero. It is often convenient to express Eq. 5.73 in terms of loss as pout pin V 2out V 2in + + + gzout = + gzin − loss ρ ρ 2 2 (5.79) An example of the application of Eq. 5.79 follows. EXAMPLE 5.23 Energy—Effect of Loss of Available Energy GIVEN As shown in Fig. E5.23a, air flows from a room through two different vent configurations: a cylindrical hole in the wall having a diameter of 120 mm and the same diameter cylindrical hole in the wall but with a well-rounded entrance. The room pressure is held constant at 1.0 kPa above atmospheric pressure. Both vents exhaust into the atmosphere. As discussed in Section 8.4.2, the loss in available energy associated with flow through the cylindrical vent from the room to the vent exit is 0.5V 22 2 where V2 is the uniformly distributed exit velocity of air. The loss in available energy associated with flow through the rounded entrance vent from the room to the vent exit is 0.05V 22 2, where V2 is the uniformly distributed exit velocity of air. FIND Compare the volume flowrates associated with the two different vent configurations. Control volume SOLUTION We use the control volume for each vent sketched in Fig. E5.23a. What is sought is the flowrate, Q = A2V2, where A2 is the vent exit cross-sectional area, and V2 is the uniformly distributed exit velocity. For both vents, application of Eq. 5.79 leads to 0 (no elevation change) V 22 p2 p1 V 21 + gz2 = + gz1 − 1loss2 + + ρ ρ 2 2 0 (V1 ≈ 0) D2 = 120 mm Section (2) Section (1) for both vents is in the room and involves V1 = 0 p1 = 1.0 kPa (1) D2 = 120 mm where 1loss2 is the loss between sections (1) and (2). Solving Eq. 1 for V2 we get V2 = 2 B [( p1 − p 2 − loss2] ρ ) 1 (2) V 22 2 (3) V2 V2 Section (2) Control volume Since 1loss2 = KL where KL is the loss coefficient (KL = 0.5 and 0.05 for the two vent configurations involved), we can combine Eqs. 2 and 3 to get V2 = 2 B [( p1 − p 2 V 22 − KL ] ) ρ 2 (4) ■ Figure E5.23a Solving Eq. 4 for V2 we obtain V2 = p1 − p 2 B ρ[(1 + KL )2] (5) 5.3 Therefore, for flowrate, Q, we obtain Q = A2V2 = p1 − p 2 πD 22 4 B ρ[(1 + KL )2] (6) For the rounded entrance cylindrical vent, Eq. 6 gives Q= × π(120 mm) 2 4(1000 mm/m) 2 243 First Law of Thermodynamics—The Energy Equation COMMENT By repeating the calculations for various values of the loss coefficient, KL, the results shown in Fig. E5.23b are obtained. Note that the rounded entrance vent allows the passage of more air than does the cylindrical vent because the loss associated with the rounded entrance vent is less than that for the cylindrical one. For this flow the pressure drop, p1 − p2, has two purposes: (1) overcome the loss associated with the flow, and (2) produce the kinetic energy at the exit. Even if there were no loss (i.e., KL = 0), a pressure drop would be needed to accelerate the fluid through the vent. (1.0 kPa) (1000 Pa /kPa)[1(N/m2 )(Pa)] B (1.23 kg /m3 )[(1 + 0.05)2][1(N∙s2 )(kg∙m)] 0.5 or Q = 0.445 m3/s × (0.5, 0.372 m3/s) Q, m3/s For the cylindrical vent, Eq. 6 gives us Q= 0.4 (0.05, 0.445 m3/s) (Ans) π(120 mm) 2 4(1000 mm /m) 2 0.3 0.2 0.1 (1.0 kPa) (1000 Pa/kPa)[1(N/m2 )(Pa)] B (1.23 kg/m3 )[(1 + 0.5)2][1(N∙s2 )(kg∙m)] or 0 0 0.1 0.2 0.3 0.4 0.5 KL Q = 0.372 m3/s (Ans) ■ Figure E5.23b An important group of fluid mechanics problems involves one-dimensional, incompressible, steady-in-the-mean flow with friction and shaft work. Included in this category are flows through pumps, blowers, fans, and turbines. For this kind of flow, Eq. 5.67 becomes pin pout V 2out − V 2in − + + g(zout − zin ) ] = Q̇ net + Ẇ shaft (5.80) ρ ρ 2 in net in Dividing Eq. 5.80 by mass flowrate and using the work per unit mass, w shaft = Ẇ shaft m ˙ , we obtain m ˙ [uˇ out − uˇ in + net in net in pout pin V 2out V 2in + + + gzout = + gzin + wshaft − (uˇ out − uˇ in − qnet ) ρ ρ 2 2 net in in (5.81) Since the flow is incompressible, Eq. 5.78 shows that uˇ out − uˇ in − qnet in equals the loss of useful energy and Eq. 5.81 can be expressed as pout pin V 2out V 2in + + + gzout = + gzin + w shaft − loss ρ ρ 2 2 net in VIDEO V5.15 Energy transfer Minimizing loss is a central goal of fluid mechanical design. (5.82) This is a form of the energy equation for steady-in-the-mean flow that is often used for incompressible flow problems. It is sometimes called the mechanical energy equation or the extended Bernoulli equation. Note that Eq. 5.82 involves energy per unit mass (ft ∙ lb/slug = ft2/s2 or N ∙ m = m2/s2 ). According to Eq. 5.82, when the shaft work is into the control volume, as for example with a pump, a larger amount of loss will result in more shaft work being required for the same rise in available energy. Similarly, when the shaft work is out of the control volume (for example, a turbine), a larger loss will result in less shaft work out for the same drop in available energy. Designers spend a great deal of effort on minimizing losses in fluid flow components. The following examples demonstrate why losses should be kept as small as possible in fluid systems. 244 Chapter 5 ■ Finite Control Volume Analysis EXAMPLE 5.24 Energy—Fan Work and Efficiency GIVEN An axial-flow ventilating fan driven by a motor that FIND Determine how much of the work to the air actually delivers 0.4 kW of power to the fan blades produces a 0.6-mdiameter axial stream of air having a speed of 12 m/s. The flow far upstream of the fan has negligible speed. produces useful effects, that is, fluid motion and a rise in available energy. Estimate the aerodynamic efficiency of this fan. SOLUTION We select a fixed and nondeforming control volume as is illustrated in Fig. E5.24. The application of Eq. 5.82 to the contents of this control volume leads to 0 (atmospheric pressures cancel) Section (1) Stream surface V1 = 0 0 (V1 ≈ 0) Control volume p2 p1 V 22 V 21 + gz2) − ( + + gz1) (1) w shaft − loss = ( + ρ ρ 2 2 net in 0 (no elevation change) Fan motor Fan D2 = 0.6 m Section (2) V2 = 12 m/s where wshaft net in − loss is the amount of work added to the air that produces a useful effect. Equation 1 leads to w shaft − loss = net in V 22 (12 m/s) 2 = 2 2[1(kg∙m)(N∙s2 )] = 72.0 N∙m/kg (2) (Ans) ■ Figure E5.24 A reasonable estimate of efficiency, η, would be the ratio of amount of work that produces a useful effect, Eq. 2, to the amount of work delivered to the fan blades. That is, wshaft η= net in − loss wshaft wshaft = net in = (3) net in net in (ρπD224)V2 (0.4 kW)[1000 (Nm)(skW)] (1.23 kg/m3 )[(π) (0.6 m) 24](12 m/s) or To calculate the efficiency, we need a value of wshaft net in, which is related to the power delivered to the blades,Ẇ shaft net in. We note that Ẇ shaft net in wshaft = (4) m ˙ net in ˙ , is (from Eq. 5.6) where the mass flowrate, m ˙ = ρAV = ρ m Ẇ shaft πD 22 V 4 2 (5) For fluid density, ρ, we use 1.23 kg/m3 (standard air) and, thus, from Eqs. 4 and 5 we obtain wshaft = 95.8 N∙m/kg (6) net in From Eqs. 2, 3, and 6 we obtain η= 72.0 N∙m/kg = 0.752 95.8 N∙m/kg (Ans) COMMENT Note that only 75% of the power that was delivered to the air resulted in useful effects and, thus, 25% of the shaft power is lost to air friction. The power input to the motor would be more than 0.4 kW because of electrical losses in the motor and mechanical friction in the bearings. THE WIDE WORLD OF FLUIDS Curtain of air An air curtain is produced by blowing air through is the added power requirements to operate the air curtain, a long rectangular nozzle to produce a high-velocity sheet of air, although the advantages can outweigh the disadvantage for or a “curtain of air.” This air curtain is typically directed over a various industrial applications. New applications for current air doorway or opening as a replacement for a conventional door. curtain designs continue to be developed. For example, the use The air curtain can be used for such things as keeping warm air of air curtains as a means of road tunnel fire security is currently from infiltrating dedicated cold spaces, preventing dust and being investigated. In such an application, the air curtain would other contaminants from entering a clean environment, and act to isolate a portion of the tunnel where fire has broken out even just keeping insects out of the workplace, still allowing and not allow smoke and fumes to infiltrate the entire tunnel people to enter or exit. A disadvantage over conventional doors system. (See Problem 5.120.) 5.3 First Law of Thermodynamics—The Energy Equation 245 If Eq. 5.82, which involves energy per unit mass, is multiplied by fluid density, ρ, we obtain pout + VIDEO V5.16 Water plant aerator ρV 2out ρV 2in + γzout = pin + + γzin + ρwshaft − ρ(loss) 2 2 net in where γ = ρg is the specific weight of the fluid. Equation 5.83 involves energy per unit volume and the units involved are identical with those used for pressure (ft ∙ lb/ft3 = lb/ft2 or N ∙ m/m3 = N/m2 ). If Eq. 5.82 is divided by the acceleration of gravity, g, we get pout pin V 2out V 2in + zout = + zin + hs − hL + + γ γ 2g 2g where Ẇ shaft hs = wshaft net in g = net in m ˙g (5.84) Ẇ shaft = net in γQ (5.85) is the shaft work head and hL = loss/g is called the head loss. Equation 5.84 involves energy per unit weight (ft ∙ lb/lb = ft or N ∙ m/N = m). In Section 3.7, we introduced the notion of “head,” which is energy per unit weight. Units of length (for example, ft, m) are used to quantify the amount of head involved. If a turbine is in the control volume, hs is negative because it is associated with shaft work done by the fluid. For a pump in the control volume, hs is positive because it is associated with shaft work done on the fluid. We can define a total head, H, as follows: H= p V2 +z + γ 2g Then Eq. 5.84 can be expressed as Hout = Hin + hs − hL Recall that hs is pump work input minus turbine work output. Some important possible values of Hout in comparison to Hin are shown in Fig. 5.8. Note that hL (head loss) always reduces the value of Hout , except in the ideal case when it is zero. Note also that hL lessens the effect of shaft work that can be extracted from a fluid. When hL = 0 (ideal condition) the shaft work head, hs, and the change in total head are the same. This head change is sometimes called ideal head change. The corresponding ideal shaft work head is the minimum required to achieve a desired effect. For work out, it is the maximum possible. Designers usually strive to minimize loss. In Chapter 12 we learn of one instance when minimum loss is sacrificed for survivability of fish coursing through a turbine rotor. Hout hs hs – hL Hin hs + hL hL > 0, hs < 0 hL = 0, hs < 0 hL > 0, hs > 0 hs hL = 0, hs > 0 hL > 0, hs = 0 hL hL = 0, hs = 0 Head is energy per unit weight. (5.83) ■ Figure 5.8 Total-head change in fluid flows. 246 Chapter 5 ■ Finite Control Volume Analysis EXAMPLE 5.25 Energy—Head Loss and Power Loss GIVEN The pump shown in Fig. E5.25a adds 10 horsepower Section (2) to the water as it pumps water from the lower lake to the upper lake. The elevation difference between the lake surfaces is 30 ft and the head loss is 15 ft. Control volume FIND Determine (a) the flowrate and 30 ft Flow (b) the power loss associated with this flow. SOLUTION (a) Section (1) Pump The energy equation (Eq. 5.84) for this flow is p1 p2 V22 V12 + z2 = + z1 + hs − hL + + γ γ 2g 2g Flow (1) where points 2 and 1 (corresponding to “out” and “in” in Eq. 5.84) are located on the lake surfaces. Thus, p2 = p1 = 0 and V2 = V1 = 0 so that Eq. 1 becomes hs = hL + z2 − z1 ■ Figure E5.25a (b) The power lost due to friction can be obtained from Eq. 5.85 as Ẇ loss = γ QhL = (62.4 lb/ft3 ) (1.96 ft3/s) (15 ft) = 1830 ft ∙ lb/s (1 hp/550 ft ∙ lb/s) (2) where z 2 = 30 ft, z1 = 0, and hL = 15 ft. The pump head is obtained from Eq. 5.85 as hs = Ẇ shaft net in γ Q = (10 hp) (550 ft ∙ lb /s/ hp)(62.4 lb /ft3 ) Q = 88.1/Q where hs is in ft when Q is in ft3/s. Hence, from Eq. 2, 88.1Q = 15 ft + 30 ft or Q = 1.96 ft3/s (Ans) = 3.33 hp (Ans) The remaining 10 hp − 3.33 hp = 6.67 hp that the pump adds to the water is used to lift the water from the lower to the upper lake. This energy is not “lost,” but it is stored as potential energy. By repeating the calculations for various head losses, hL, the results shown in Fig. E5.25b are obtained. Note that as the head loss increases, the flowrate decreases because an increasing portion of the 10 hp supplied by the pump is lost and, therefore, not available to lift the fluid to the higher elevation. Note that in this example the purpose of the pump is to lift the water (a 30-ft head) and overcome the head loss (a 15-ft head); it does not, overall, alter the water’s pressure or velocity. COMMENTS 3.5 3 Q, ft3/s 2.5 (15 ft, 1.96 ft3/s) 2 1.5 1 0.5 0 0 5 10 15 hL, ft ■ Figure E5.25b 20 25 5.3 First Law of Thermodynamics—The Energy Equation 247 A comparison of the energy equation and the Bernoulli equation has led to the concept of loss of available energy in incompressible fluid flows with friction. In Chapter 8, we discuss in detail some methods for estimating loss in incompressible flows with friction. In Section 5.4, we demonstrate that loss of available energy is also an important factor to consider in compressible flows with friction. THE WIDE WORLD OF FLUIDS Smart shocks Vehicle shock absorbers are dampers used to electric coil on the shock piston in a manner that changes the provide a smooth, controllable ride. When going over a bump, fluid viscosity, going anywhere from essentially no damping to the relative motion between the tires and the vehicle body a solid almost instantly. A computer adjusts the current to the displaces a piston in the shock and forces a viscous fluid coil to select the proper viscosity for the given conditions through a small orifice or channel. The viscosity of the fluid (i.e., wheel speed, vehicle speed, steering-wheel angle, lateral produces a head loss that dissipates energy to dampen the acceleration, brake application, and temperature). The goal of vertical motion. Current shocks use a fluid with fixed viscosity. these adjustments is an optimally tuned shock that keeps the However, recent technology has been developed that uses a vehicle on a smooth, even keel while maximizing the contact of synthetic oil with millions of tiny iron balls suspended in it. the tires with the pavement for any road conditions. (See These tiny balls react to a magnetic field generated by an Problem 5.104.) 5.3.4 Application of the Energy Equation to Nonuniform Flows The forms of the energy equation discussed in Sections 5.3.2 and 5.3.3 are applicable to onedimensional flows, flows that are approximated with uniform velocity distributions where fluid crosses the control surface. If the velocity profile at any section where flow crosses the control surface is not uniform, inspection of the energy equation for a control volume, Eq. 5.64, suggests that the integral The kinetic energy coefficient is used to account for nonuniform flows. V2 ρV ∙ nˆ dA cs 2 ∫ will require special attention. The other terms of Eq. 5.64 can be accounted for as already discussed in Sections 5.3.2 and 5.3.3. For one stream of fluid entering and leaving the control volume, we can define the relationship α V2 α V2 V2 ρV ∙ nˆ dA = m ˙ ( out out − in in ) 2 2 cs 2 ∫ where α is the kinetic energy coefficient and V is the average velocity defined earlier in Eq. 5.7. From the above we can conclude that m ˙ αV 2 = 2 ∫ A V2 ρV ∙ nˆ dA 2 for flow through surface area A of the control surface. Thus, α=2 α= Parabolic (laminar) α~ ~ 1.08 Turbulent ∫ (V 22)ρV ∙ nˆ dA A m ˙ V 22 (5.86) It can be shown that for any velocity profile, α ≥ 1, with α = 1 only for uniform flow. Some typical velocity profile examples for flow in a conventional pipe are shown in the sketch in the margin. Therefore, for nonuniform velocity profiles, the energy equation on an energy per unit mass basis for the incompressible flow of one stream of fluid through a control volume that is steady in the mean is α=1 Uniform pin pout αoutV 2out αinV 2in + + + gzout = + gzin + w shaft − loss ρ ρ 2 2 net in (5.87) 248 Chapter 5 ■ Finite Control Volume Analysis On an energy per unit volume basis we have pout + ραoutV 2out ραinV 2in + γzout = pin + + γzin + ρw shaft − ρ(loss) 2 2 net in (5.88) and on an energy per unit weight or head basis we have pout pin αoutV 2out αinV 2in + zout = + zin + + + γ γ 2g 2g w shaft net in g − hL (5.89) The following examples illustrate the use of the kinetic energy coefficient. EXAMPLE 5.26 Energy—Effect of Nonuniform Velocity Profile D2 = 30 mm GIVEN The small fan shown in Fig. E5.26 moves air at a mass flowrate of 0.1 kg/min. Upstream of the fan, the pipe diameter is 60 mm, the flow is laminar, the velocity distribution is parabolic, and the kinetic energy coefficient, α1, is equal to 2.0. Downstream of the fan, the pipe diameter is 30 mm, the flow is turbulent, the velocity profile is quite uniform, and the kinetic energy coefficient, α2, is equal to 1.08. The rise in static pressure across the fan is 0.1 kPa, and the fan motor draws 0.14 W. Turbulent flow Section (2) α 2 = 1.08 FIND Compare the value of loss calculated: (a) assuming uniform Control volume velocity distributions, (b) considering actual velocity distributions. SOLUTION D1 = 60 mm Application of Eq. 5.87 to the contents of the control volume shown in Fig. E5.26 leads to Section (1) α 1 = 2.0 0 (change in gz is negligible) • p2 + ρ α2V 22 2 + gz2 = p1 + ρ α1V 21 2 Laminar flow m = 0.1 kg/min + gz1 − loss + w shaft ■ Figure E5.26 (1) net in = or solving Eq. 1 for loss we get loss = w shaft net in p2 − p 1 α2V 22 α1V 21 −( − + ) ρ 2 2 (2) To proceed further, we need values of wshaft net in, V1, and V2. These quantities can be obtained as follows. For shaft work wshaft = net in net in (3) For the average velocity at section (1), V1, from Eq. 5.11 we obtain V1 = = m ˙ ρA1 m ˙ ρ(πD214) = 0.479 m/s For the average velocity at section (2), V2, V2 = (0.1 kg/min) (1 min/60 s) (1000 mm/m) 2 (5) (a) For the assumed uniform velocity profiles (α1 = α2 = 1.0), Eq. 2 yields (0.14 W)[(1 N ∙ m/s)W] (60 s/min) = 0.1 kg/min = 84.0 N ∙ m/kg (1.23 kg/m3 )[π(60 mm) 24] (1.23 kg/m3 )[π(30 mm) 24] = 1.92 m/s power to fan motor m ˙ or wshaft (0.1 kg/min) (1 min/60 s) (1000 mm/m) 2 loss = wshaft − ( net in (6) Using Eqs. 3, 4, and 5 and the pressure rise given in the problem statement, Eq. 6 gives loss = 84.0 (4) p2 − p 1 V 22 V 21 − + ρ ) 2 2 (0.1 kPa) (1000 Pa/kPa) (1 N/m2/Pa) N∙m − kg 1.23 kg/m3 5.3 + (0.479 m /s) 2 2[1 (kg ∙ m)(N ∙ s2 )] − First Law of Thermodynamics—The Energy Equation (1.92 m /s) 2 2[1 (kg ∙ m)(N + 2 ˙ s )] or 2(0.479 m/s) 2 2[1 (kg ∙ m)(N ∙ s2 )] − 249 1.08(1.92 m/s) 2 2[1 (kg ∙ m)(Ν ∙ s2 )] or loss = 84.0 N ∙ m/kg − 81.3 N ∙ m/kg loss = 84.0 N ∙ m/kg − 81.3 N ∙ m/kg + 0.115 N ∙ m/kg − 1.84 N ∙ m/kg = 0.975 N ∙ m/kg (Ans) (b) For the actual velocity profiles (α1 = 2, α2 = 1.08), Eq. 1 gives loss = wshaft − ( net in + 0.230 N ∙ m /kg − 1.99 N ∙ m /kg (Ans) = 0.940 N ∙ m/kg p2 − p 1 − α2 + α1 ρ ) 2 2 V 21 V 22 (7) COMMENT The difference in loss calculated assuming uni- form velocity profiles and actual velocity profiles is not large compared to wshaft net in for this fluid flow situation. If we use Eqs. 3, 4, and 5 and the given pressure rise, Eq. 7 yields loss = 84 N ∙ m /kg − (0.1 kPa) (1000 Pa /kPa) (1 N/m2/Pa) 1.23 kg/m3 EX AMPLE 5.27 Energy—Effect of Nonuniform Velocity Profile GIVEN Consider the flow situation of Example 5.14. FIND Apply Eq. 5.87 to develop an expression for the fluid pressure drop that occurs between sections (1) and (2). By com- paring the equation for pressure drop obtained presently with the result of Example 5.14, obtain an expression for loss between sections (1) and (2). SOLUTION Application of Eq. 5.87 to the flow of Example 5.14 (see Fig. E5.14) leads to Then, substituting Eq. 4 into Eq. 3, we obtain ρ8w 232π 0 (no shaft work) p1 p2 α2w 22 α1w 21 + gz2 = + gz1 − loss + wshaft + + ρ ρ 2 2 net in Solving Eq. 1 for the pressure drop, p1 − p2, we obtain α2w 22 α1w 21 − + g(z2 − z1 ) + loss] p1 − p 2 = ρ [ 2 2 Since the fluid velocity at section (1), w1, is uniformly distributed over cross-sectional area A1, the corresponding kinetic energy coefficient, α1, is equal to 1.0. The kinetic energy coefficient at section (2), α2, needs to be determined from the velocity profile distribution given in Example 5.14. Using Eq. 5.86 we get ∫ α2 = ρw 32 dA2 A2 mw 22 ˙ (3) Substituting the parabolic velocity profile equation into Eq. 3 we obtain ρ α2 = ∫ [1 − (rR) 2]3r dr ρπR2 w 32 or α2 = (2) R 0 α2 = (1) ∫ 16 R2 ∫ R [1 − 3(r/R) 2 + 3(r/R) 4 − (r/R) 6]r dr 0 =2 (5) Now we combine Eqs. 2 and 5 to get p1 − p2 = ρ [ 2.0w 22 1.0w 21 − + g(z2 − z1 ) + loss] 2 2 However, from conservation of mass w2 = w1 = w so that Eq. 6 becomes p1 − p2 = ρw 2 + ρg(z2 − z1 ) + ρ(loss) 2 (2w1 ) 3[1 − (rR) 2]3 2πr dr ρg(z2 − z1 ) = (ρA2w2 )w 22 𝒲 A (8) Thus, combining Eqs. 7 and 8 we get From conservation of mass, since A1 = A2 w 1 = w2 (7) The term associated with change in elevation, ρg(z2 − z1 ), is equal to the weight per unit cross-sectional area, 𝒲/A, of the water contained between sections (1) and (2) at any instant, R 0 (6) (4) p 1 − p2 = ρw 2 𝒲 + + ρ(loss) 2 A (9) 250 Chapter 5 ■ Finite Control Volume Analysis Rz ρw 2 ρw 2 𝒲 𝒲 + + ρ(loss) = + + 2 A 3 A A The pressure drop between sections (1) and (2) is due to: 1. The change in kinetic energy between sections (1) and (2) associated with going from a uniform velocity profile to a parabolic velocity profile. 2. The weight of the water column, that is, hydrostatic pressure effect. 3. Viscous loss. Comparing Eq. 9 for pressure drop with the one obtained in Example 5.14 (i.e., the answer of Example 5.14), we obtain (10) or loss = Rz ρA − w2 6 (Ans) COMMENT We conclude that while some of the pipe wall friction force, Rz, resulted in loss of available energy, a portion of this friction, ρAw 2/6, led to the velocity profile change. 5.3.5 Combination of the Energy Equation and the Moment-of-Momentum Equation3 If Eq. 5.82 is used for one-dimensional incompressible flow through a turbomachine, we can use Eq. 5.54, developed in Section 5.2.4 from the moment-of-momentum equation (Eq. 5.42), to evaluate shaft work. This application of both Eqs. 5.54 and 5.82 allows us to ascertain the amount of loss that occurs in incompressible turbomachine flows as is demonstrated in Example 5.28. EXAM PLE 5.28 Energy—Fan Performance GIVEN Consider the fan of Example 5.19. FIND Show that only some of the shaft power into the air is converted into useful effects. Develop a meaningful efficiency equation and a practical means for estimating lost shaft energy. SOLUTION We use the same control volume used in Example 5.19. Application of Eq. 5.82 to the contents of this control volume yields p1 p2 V 22 V 21 + gz2 = + gz1 + wshaft − loss + + ρ ρ 2 2 net in useful effect = wshaft − loss net in =( p1 p2 V 21 + gz2) − ( + + gz1) + ρ ρ 2 2 (2) (Ans) In other words, only a portion of the shaft work delivered to the air by the fan blades is used to increase the available energy of the air; the rest is lost because of fluid friction. A meaningful efficiency equation involves the ratio of shaft work converted into a useful effect (Eq. 2) to shaft work into the air, wshaft net in. Thus, we can express efficiency, η, as wshaft − loss η= net in wshaft (4) net in Combining Eqs. 2, 3, and 4, we obtain (1) As in Example 5.26, we can see with Eq. 1 that a “useful effect” in this fan can be defined as V 22 wshaft = +U2Vθ2 (3) η = {[(p2ρ) + (V 22 2) + gz2] − [( p1ρ) + (V 212) + gz1]} U2Vθ2 (5) (Ans) Equation 5 provides us with a practical means to evaluate the efficiency of the fan of Example 5.19. Combining Eqs. 2 and 4, we obtain loss = U2Vθ2 − [( −( p2 V 22 + gz2) + ρ 2 p1 V 21 + gz1)] + ρ 2 (6) (Ans) COMMENT Equation 6 provides us with a useful method of evaluating the loss due to fluid friction in the fan of Example 5.19 in terms of fluid mechanical variables that can be measured. The efficiency calculated by Eq. 3 is called aerodynamic efficiency if the machine is a fan and hydraulic efficiency if the machine is a pump or water turbine. net in However, when Eq. 5.54, which was developed from the moment-of-momentum equation (Eq. 5.42), is applied to the contents of the control volume of Fig. E5.19, we obtain 3 This section may be omitted without loss of continuity in the text material. This section should not be considered without prior study of Sections 5.2.3 and 5.2.4. All of these sections are recommended for those interested in Chapter 12. 5.4 5.4 Second Law of Thermodynamics—Irreversible Flow 251 Second Law of Thermodynamics—Irreversible Flow4 The second law of thermodynamics affords us with a means to verify the inequality The second law of thermodynamics formalizes the notion of loss. uˇ 2 − uˇ 1 − qnet ≥ 0 (5.90) in for steady, incompressible, one-dimensional flow with friction (see Eqs. 5.73 and 5.77). In this section we continue to develop the notion of loss of useful energy for flow with friction. 5.4.1 Semi-infinitesimal Control Volume Statement of the Energy Equation If we apply the one-dimensional, steady flow energy equation, Eq. 5.70, to the contents of a control volume that is infinitesimally thin as illustrated in Fig. 5.8, the result is p V2 m ˙ [duˇ + d ( ) + d( ) + g (dz) ] = δQ̇ net ρ 2 in (5.91) For all pure substances including common engineering fluids, such as air, water, and oil, the following relationship among fluid properties, called the first Tds equation, is valid (see, for example, Ref. 3). 1 T ds = duˇ + pd( ) ρ (5.92) where T is the absolute temperature and s is the entropy per unit mass. (See Chapter 11 and Ref. 3 for more information about entropy.) Combining Eqs. 5.91 and 5.92 we get p 1 V2 m ˙ [T ds − pd( ) + d( ) + d( ) + g dz] = δQ̇ net ρ ρ 2 in or, dividing through by m ˙ and letting δqnet = δQ̇ net m˙ , we obtain in in dp V2 +d ) ( 2 ) + g dz = −(T ds − δqnet ρ in (5.93) 5.4.2 Semi-infinitesimal Control Volume Statement of the Second Law of Thermodynamics A general statement of the second law of thermodynamics is (see Ref. 3) δQ̇ innet D sρ dV ≥ ∑ ( Dt sys T )sys ∫ Semi-infinitesimal control volume z θ x g θ d Flow ■ Figure 5.9 Semi-infinitesimal control volume. 4 This entire section may be omitted without loss of continuity in the text material. (5.94) 252 Chapter 5 ■ Finite Control Volume Analysis or in words, the time rate of increase of the sum of the ratio of net heat entropy of a system ≥ transfer rate into system from the surroundings to absolute temperature for each particle of mass in the system The second law of thermodynamics involves entropy, heat transfer, and temperature. The right-hand side of Eq. 5.94 is identical for the system and control volume at the instant when system and control volume are coincident; thus, δQ̇ innet δQ̇ innet = ∑ ( T )cv T )sys ∑( (5.95) With the help of the Reynolds transport theorem (Eq. 4.19) the system time derivative can be expressed for the contents of the coincident fixed and nondeforming control volume. Using Eq. 4.19, we obtain D Dt ∫ sρ d V = sys ∂ ∂t ∫ sρ d V + ∫ sρV ∙ nˆ dA (5.96) δQ̇ net sρV ∙ nˆ dA ≥ ∑( in T )cv cs (5.97) ∫ (5.98) cv cs Eqs. 5.94, 5.95, and 5.96 combine to give ∂ ∂t ∫ cv sρ d V + ∫ At any instant for steady flow ∂ ∂t sρ dV −=0 cv If the flow consists of only one stream through the control volume and if the properties are uniformly distributed (one-dimensional flow), Eqs. 5.97 and 5.98 lead to δQ̇ net ˙ (sout − sin ) ≥ ∑ in (5.99) m T For the infinitesimally thin control volume of Fig. 5.8, Eq. 5.99 yields δQ̇ net ˙ ds ≥ ∑ m The relationship between entropy and heat transfer is different for reversible and irreversible processes. in T (5.100) If all of the fluid in the infinitesimally thin control volume is considered as being at a uniform temperature, T, then from Eq. 5.100 we get T ds ≥ δqnet in or T ds − δqnet ≥ 0 (5.101) in The equality is for any reversible (frictionless) process; the inequality is for all irreversible (friction) processes. 5.4.3 Combination of the Equations of the First and Second Laws of Thermodynamics Combining Eqs. 5.93 and 5.101, we conclude that dp V2 −[ + d( ) + g dz] ≥ 0 ρ 2 (5.102) 5.5 Chapter Summary and Study Guide 253 The equality is for any steady, reversible (frictionless) flow, an important example being flow for which the Bernoulli equation (Eq. 3.7) applies. The inequality is for all steady, irreversible (friction) flows. The degree of the inequality has physical significance. It represents the loss of useful energy that occurs because of irreversible flow phenomena including viscous friction. Thus, Eq. 5.102 can be expressed as dp V2 −[ + d( ) + g dz] = δ(loss) = (T ds − δqnet ) ρ 2 in (5.103) The loss is zero for a frictionless flow and greater than zero for a flow with frictional effects. Note that when the flow is frictionless, Eq. 5.103 multiplied by density, ρ, is identical to Eq. 3.5. Thus, for steady frictionless flow, Newton’s second law of motion (see Section 3.1) and the first and second laws of thermodynamics lead to the same differential equation, dp V2 +d ( 2 ) + g dz = 0 ρ (5.104) If shaft work is involved, then the flow must be at least locally unsteady and the appropriate form of the energy equation for the contents of an infinitesimally thin control volume can be developed starting with Eq. 5.67. The resulting equation is dp V2 −[ + d( ) + g dz] = δ(loss) − δwshaft ρ 2 net in (5.105) Equations 5.103 and 5.105 are valid for incompressible and compressible flows. If we combine Eqs. 5.92 and 5.103, we obtain 1 duˇ + pd( ) − δqnet = δ(loss) ρ in (5.106) For incompressible flow, d(1ρ) = 0 and, thus, from Eq. 5.106, duˇ − δqnet = δ(loss) (5.107) in Integrating for a finite control volume, we obtain uˇ out − uˇ in − qnet = loss in which is the same conclusion we reached earlier (see Eq. 5.78) for incompressible flows. For compressible flow, d(1/ρ) ≠ 0, and thus when we apply Eq. 5.106 to a finite control volume we obtain uˇ out − uˇ in + ∫ out in 1 pd( ) − qnet = loss ρ in (5.108) indicating that uˇ out − uˇ in − qnet in is not equal to loss. 5.5 Chapter Summary and Study Guide In this chapter the flow of a fluid is analyzed by using important principles including conservation of mass, Newton’s second law of motion, and the first law of thermodynamics as applied to control volumes. The Reynolds transport theorem is used to convert basic system-oriented laws into corresponding control volume formulations. The continuity equation, a statement of the fact that mass is conserved, is obtained in a form that can be applied to any flow—steady or unsteady, incompressible or compressible. Simplified forms of the continuity equation help us to keep track of fluid in a control volume, where it enters, where it leaves, and within. Mass or volume flowrates of fluid entering or leaving a control volume and rate of accumulation or depletion of fluid within a control volume can be evaluated. 254 Chapter 5 ■ Finite Control Volume Analysis conservation of mass continuity equation mass flowrate linear momentum equation moment-ofmomentum equation shaft power shaft torque first law of thermodynamics heat transfer rate flow work energy equation loss shaft work head head loss kinetic energy coefficient The linear momentum equation, a form of Newton’s second law of motion applicable to flow of fluid through a control volume, is obtained and used to solve flow problems. Net force results from or causes changes in linear momentum (velocity magnitude and/or direction) of fluid flowing through a control volume. The moment-of-momentum equation, which involves the relationship between torque and changes in angular momentum, is obtained and used to solve flow problems dealing with turbines (energy extracted from a fluid) and pumps (energy supplied to a fluid). The steady-state energy equation, obtained from the first law of thermodynamics (conservation of energy), is written in several forms. The first (Eq. 5.69) involves power terms. The second form (Eq. 5.82 or 5.84) is termed the mechanical energy equation or the extended Bernoulli equation. This equation is used for different applications than the Bernoulli equation because it contains an extra term that accounts for energy losses due to friction in the flow, as well as terms accounting for the work of pumps or turbines. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ ■ ■ ■ ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. select an appropriate control volume for a given problem and draw an accurately labeled control volume diagram. use the continuity equation and a control volume to solve problems involving mass or volume flowrate. use the linear momentum equation and a control volume, in conjunction with the continuity equation as necessary, to solve problems involving forces related to linear momentum change. use the moment-of-momentum equation to solve problems involving torque and related work and power due to angular momentum change. use the energy equation, in one of its appropriate forms, to solve problems involving losses due to friction (head loss) and energy input by pumps or extraction by turbines. use the momentum coefficient in the linear momentum equation and the kinetic energy coefficient in the energy equation to account for nonuniform flows. Some of the important equations in this chapter are: ∂ ∂t Conservation of mass ∫ ρ dV + cv ρV ∙ nˆ dA = 0 m ˙ = ρQ = ρAV V= Average velocity ∫ (5.6) ρV ∙ nˆ dA A (5.7) ρA ∑ m˙ out − ∑ m˙ in = 0 Steady-flow mass conservation ∂ ∂t Moving control volume mass conservation DMsys Deforming control volume mass conservation Dt ∂ ∂t ∫ cv ∫ = cv (5.5) cs Mass flowrate Force related to change in linear momentum ∫ ρ dV + ∂ ∂t Vρ d V + ∫ (5.9) ρW ∙ nˆ dA = 0 (5.16) cs ∫ ρ dV + cv ∫ ρW ∙ nˆ dA = 0 ∫ VρV ∙ nˆ dA = cs (5.17) cs ∑ Fcontents of the control volume (5.22) References Moving control volume force related to change in linear momentum ∫ WρW ∙ nˆ dA = ∑ F contents of the control volume cs Vector addition of absolute and relative velocities V=W+U ∑ [ (r × F) contents of the ] Shaft torque from force control volume axial 255 (5.29) (5.43) = Tshaft (5.45) Shaft torque related to change in moment-of-momentum (angular momentum) Tshaft = (−m ˙ in )(±rinVθin ) + m˙ out ( ±routVθout ) (5.50) Shaft power related to change in moment-of-momentum (angular momentum) Ẇ shaft = (−m ˙ in )(±UinVθin ) + m˙ out ( ±UoutVθout ) (5.53) First law of thermodynamics (Conservation of energy) First law of thermodynamics (simplified form) ∂ ∂t ∫ eρ d V + cv ∫ V 2 − V 2in m ˙ [hˇ out − hˇ in + out + g(zout − zin ) ] = Q̇ net + Ẇ shaft 2 in net in Mechanical energy equation (incompressible flow) Mechanical energy equation (general) p V2 ˇ+ + u + gz) ρV ∙ nˆ dA = Q̇ net + Ẇ shaft (5.64) ( ρ 2 in net in cs pout pin V 2out V 2in + + + gzout = + gzin + wshaft − loss ρ ρ 2 2 net in −[dp/ρ + d(V2/2) + gdz] = δ(loss) − δwshaft (5.69) (5.82) (5.105) net in References 1. Eck, B., Technische Stromungslehre, Springer-Verlag, Berlin, Germany, 1957. 2. Dean, R. C., “On the Necessity of Unsteady Flow in Fluid Machines,” ASME Journal of Basic Engineering 81D; 24–28, March 1959. 3. Moran, M. J., and Shapiro, H. N., Fundamentals of Engineering Thermodynamics, 8th Ed., Wiley, New York, 2014. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are no unique answers to these problems. 256 Chapter 5 ■ Finite Control Volume Analysis Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. Inlet 0.08-ft diameter Section 5.1.1 Derivation of the Continuity Equation 5.1 Use the Reynolds transport theorem (Eq. 4.19) with B = volume and, therefore, b = volume/mass = 1/density to obtain the continuity equation for steady or unsteady incompressible flow through a fixed control volume: 0.1 ft Blades ∫ V ∙ nˆ dA = 0. cv 5.2 An incompressible fluid flows horizontally in the x–y plane with a velocity given by 0.6 ft y u = 30(1 − e−4 h ) m/s, v = 0 60° V = 10 ft/s where y and h are in meters and h is a constant. Determine the average velocity for the portion of the flow between y = 0 and y = h. ■ Figure P5.4 Section 5.1.2 Fixed, Nondeforming Control Volume— Uniform Velocity Profile or Average Velocity 5.3 Water flows steadily through the horizontal piping system †5.5 Estimate the rate (in gal/hr) that your car uses gasoline when shown in Fig. P5.3. The velocity is uniform at section (1), the mass it is being driven on an interstate highway. Determine how long it flowrate is 10 slugs/s at section (2), and the velocity is nonuniform would take to empty a 12-oz soft-drink container at this flowrate. D at section (3). (a) Determine the value of the quantity ρd − V , List all assumptions and show calculations. Dt sys 5.6 The pump shown in Fig. P5.6 produces a steady flow of 10 gal/s where the system is the water contained in the pipe bounded by through the nozzle. Determine the nozzle exit diameter, D2, if the sections (1), (2), and (3). (b) Determine the mean velocity at exit velocity is to be V2 = 100 ft/s. section (2). (c) Determine, if possible, the value of the integral ∫ ∫ ρV ∙ nˆ dA over section (3). If it is not possible, explain what (3) Section (1) additional information is needed to do so. Pump 15 ft/s (1) Area = 0.7 ft2 10 slugs/s Area = 0.3 ft2 (2) V2 Section (2) D2 (3) Area = 0.7 ft2 ■ Figure P5.3 ■ Figure P5.6 5.4 Water flows out through a set of thin, closely spaced blades as shown in Fig. P5.4 with a speed of V = 10 ft/s around the entire circumference of the outlet. Determine the mass flowrate through the inlet pipe. 5.7 The fluid axial velocities shown in Fig. P5.7 are the average velocities measured in ft/s in each annular area of a duct. Find the volume flowrate for the flowing fluid. Problems 5.11 A woman is emptying her aquarium at a steady rate with a small pump. The water pumped to a 12-in.-diameter cylindrical bucket, and its depth is increasing at the rate of 4.0 in. per minute. Find the rate at which the aquarium water level is dropping if the aquarium measures 24 in. (wide) × 36 in. (long) × 18 in. (high). 3.2 4.4 4.7 5.0 1'' + 2'' 2½'' 5.12 An evaporative cooling tower (see Fig. P5.12) is used to cool water from 110 to 80 °F. Water enters the tower at a rate of 250,000 lbm/ hr. Dry air (no water vapor) flows into the tower at a rate of 151,000 lbm/ hr. If the rate of wet airflow out of the tower is 156,900 lbm/ hr, determine the rate of water evaporation in lbm/ hr and the rate of cooled water flow in lbm/ hr. 3'' ■ Figure P5.7 5.8 The human circulatory system consists of a complex branching pipe network ranging in diameter from the aorta (largest) to the capillaries (smallest). The average radii and the number of these vessels are shown in the table. Does the average blood velocity increase, decrease, or remain constant as it travels from the aorta to the capillaries? Vessel Aorta Arteries Arterioles Capillaries 257 Average Radius, mm Number 12.5 2.0 0.03 0.006 1 159 1.4 × 107 3.9 × 109 Wet air out Water in Air in Air in Water out ■ Figure P5.12 5.9 Air flows steadily between two cross sections in a long, straight section of 0.1-m-inside-diameter pipe. The static temperature and pressure at each section are indicated in Fig. P5.9. If the average air velocity at section (1) is 205 m/s, determine the average air velocity at section (2). D = 0.1 m Section (1) Section (2) p1 = 77 kPa (abs) T1 = 268 K V1 = 205 m/s 5.13 At cruise conditions, air flows into a jet engine at a steady rate of 65 lbm/s. Fuel enters the engine at a steady rate of 0.60 lbm/s. The average velocity of the exhaust gases is 1500 ft/s relative to the engine. If the engine exhaust effective cross-sectional area is 3.5 ft2, estimate the density of the exhaust gases in lbm/ft3. 5.14 Water at 0.1 m3/s and alcohol (SG = 0.8) at 0.3 m3/s are mixed in a y-duct as shown in Fig. 5.14. What is the average density of the mixture of alcohol and water? p2 = 45 kPa (abs) T2 = 240 K Water and alcohol mixture ■ Figure P5.9 5.10 A hydraulic jump (see Video V10.11) is in place downstream from a spillway as indicated in Fig. P5.10. Upstream of the jump, the depth of the stream is 0.6 ft and the average stream velocity is 18 ft/s. Just downstream of the jump, the average stream velocity is 3.4 ft/s. Calculate the depth of the stream, h, just downstream of the jump. Water Q = 0.1 m3/s Alcohol (SG = 0.8) Q = 0.3 m3/s ■ Figure P5.14 0.6 ft h 18 ft/s ■ Figure P5.10 3.4 ft/s 5.15 In the vortex tube shown in Fig. P5.15, air enters at 202 kPa absolute and 300 K. Hot air leaves at 150 kPa absolute and 350 K, whereas cold air leaves at 101 kPa absolute and 250 K. The hot air mass flow rate, m ˙ H , equals the cold air mass flow rate, m˙ C . Find the ratio of the hot air exit area to cold air exit area for equal exit velocities. 258 Chapter 5 ■ Finite Control Volume Analysis r (in.) Axial Velocity (ft/s) 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 2.9 2.95 2.98 3.00 30 29.71 29.39 29.06 28.70 28.31 27.89 27.42 26.90 26.32 25.64 24.84 23.84 22.50 20.38 18.45 16.71 14.66 0 TC = 250 K pC = 101 kPa m∙ TH = 350 K pH = 150 kPa m∙ H C T = 300 K p = 202 kPa ■ Figure P5.15 5.16 Molten plastic at a temperature of 510 °F is augered through 3 an extruder barrel by a screw occupying 5 of the barrel’s volume (Fig. P5.16). The extruder is 16 ft long and has an inner diameter of 8 in. The barrel is connected to an adapter having a volume of 0.48 ft3. The adapter is then connected to a die of equal volume. The plastic exiting the die is immediately rolled into sheets. The line is producing 4-ft widths of material at a rate of 30 ft/min and a gauge thickness of 187 mil. What is the axial velocity, V1, of the plastic in the barrel? Assume that the plastic density is constant as it solidifies from a liquid (in the extruder) into a solid sheet. Rolls Die Adapter V2 = 30 ft/min Extruder barrel Screw V1 t 5.19 Two rivers merge to form a larger river as shown in Fig. P5.19. At a location downstream from the junction (before the two streams completely merge), the nonuniform velocity profile is as shown and the depth is 6 ft. Determine the value of V. Thin sheet ■ Figure P5.16 Section 5.1.2 Fixed, Nondeforming Control Volume— Nonuniform Velocity Profile 5.17 A water jet pump (see Fig. P5.17) involves a jet cross-sectional area of 0.01 m2, and a jet velocity of 30 m/s. The jet is surrounded by entrained water. The total cross-sectional area associated with the jet and entrained streams is 0.075 m2. These two fluid streams leave the pump thoroughly mixed with an average velocity of 6 m/s through a cross-sectional area of 0.075 m2. Determine the pumping rate (i.e., the entrained fluid flowrate) involved in liters/s. 6 m/s 0.8 V 30 ft 3 ft/s 70 ft V 50 ft Depth = 3 ft 80 ft Depth = 5 ft Entrained water 4 ft/s 30 m/s jet ■ Figure P5.19 Entrained water ■ Figure P5.17 *5.18 To measure the mass flowrate of air through a 6-in.-insidediameter pipe, local velocity data are collected at different radii from the pipe axis (see Table). Determine the mass flowrate corresponding to the data listed in the following table. Plot the velocity profile and comment. 5.20 Various types of attachments can be used with the shop vac shown in Video V5.2. Two such attachments are shown in Fig. P5.20—a nozzle and a brush. The flowrate is 1 ft3/s. (a) Determine the average velocity through the nozzle entrance, Vn. (b) Assume the air enters the brush attachment in a radial direction all around the brush with a velocity profile that varies linearly from 0 to Vb along the length of the bristles as shown in the figure. Determine the value of Vb. Problems where R = 1.5 mm. The oil has a specific gravity S = 0.86 and flows in the inlet pipe at a rate of 0.006 kg/s. Compute the average velocity V1 of the oil in the inlet pipe and the average velocity V2 at the outlet (plane 2) and the maximum velocity in the oil inlet pipe (U0). Assume radial flow. Q = 1 ft3/s Q = 1 ft3/s D2 = 10 cm Outlet plane 2 1.5 in. 2-in. dia. Vb Vn 259 Circular disc D1 = 3.0 mm = 2R Inlet pipe 3-in. dia. u = U0 1 – r ■ Figure P5.20 h = 2.0 mm (Rr ( 2 = oil velocity profile ■ Figure P5.24 5.21 An appropriate turbulent pipe flow velocity profile is V = uc( R − r 1n ˆ i R ) where uc = centerline velocity, r = local radius, R = pipe radius, and ˆi = unit vector along pipe centerline. Determine the ratio of average velocity, u, to centerline velocity, uc, for (a) n = 4, (b) n = 6, (c) n = 8, (d) n = 10. Compare the different velocity profiles. 5.22 As shown in Fig. P5.22, at the entrance to a 3-ft-wide channel the velocity distribution is uniform with a velocity V. Further downstream the velocity profile is given by u = 4y − 2y2, where u is in ft /s and y is in ft. Determine the value of V. V y 0.75 ft 1 ft 5.25 Flow of a viscous fluid over a flat plate surface results in the development of a region of reduced velocity adjacent to the wetted surface as depicted in Fig. P5.25. This region of reduced flow is called a boundary layer. At the leading edge of the plate, the velocity profile may be considered uniformly distributed with a value U. All along the outer edge of the boundary layer, the fluid velocity component parallel to the plate surface is also U. If the x-direction velocity profile at section (2) is u 3 y 1 y 3 = ( )− ( ) U 2 δ 2 δ develop an expression for the volume flowrate through the edge of the boundary layer from the leading edge to a location downstream at x where the boundary layer thickness is δ. u = 4y – 2y2 x Section (2) Outer edge of boundary layer U Section (1) U ■ Figure P5.22 δ 5.23 The cross-sectional area of a rectangular duct is divided into 16 equal rectangular areas, as shown in Fig. P5.23. The axial fluid velocity measured in feet per second in each smaller area is given in the figure. Estimate the volume flowrate and average axial velocity. 20.0 in. 16.0 in. 3.0 3.4 3.6 3.1 3.7 4.0 3.9 3.8 3.9 4.6 4.5 4.2 3.7 4.4 4.3 3.9 Velocities in ft/s x ■ Figure P5.25 Section 5.1.2 Fixed, Nondeforming Control Volume— Unsteady Flow 5.26 Air at standard conditions enters the compressor shown in Fig. P5.26 at a rate of 10 ft3/s. It leaves the tank through a 1.2-in.-diameter pipe with a density of 0.0035 slugs/ft3 and a uniform speed of 700 ft/s. (a) Determine the rate (slugs/s) at which the mass of air in the tank is increasing or decreasing. (b) Determine the average time rate of change of air density within the tank. ■ Figure P5.23 Compressor 5.24 Oil for lubricating the thrust bearing shown in Fig. P5.24 flows into the space between the bearing surfaces through a circular inlet pipe with velocity r 2 u = U0[1 − ( ) ], R 10 ft3/s 0.00238 slugs/ft3 ■ Figure P5.26 Tank volume = 20 ft3 1.2 in. 700 ft/s 0.0035 slugs/ft3 Free ebooks ==> www.Ebook777.com 260 Chapter 5 ■ Finite Control Volume Analysis A1 = 5 acres 5.27 Estimate the time required to fill with water a cone-shaped container (see Fig. P5.27) 5 ft high and 5 ft across at the top if the filling rate is 20 gal/min. A2 = 4 acres 5 ft Reservoir 2 h1 Reservoir 1 h2 5 ft ■ Figure P5.27 d3 = 0.50 m Section 5.1.3 Moving, Nondeforming Control Volume †5.28 For an automobile moving along a highway, describe the control volume you would use to estimate the flowrate of air across the radiator. Explain how you would estimate the velocity of that air. 5.29 A water jet leaves a fixed nozzle with a velocity of 10 m/s. The jet diameter is 10 cm. A 30o cone is pushed into the water jet at a speed of 5 m/s. The water impinges on the cone with the jet axis and the cone axis in perfect alignment so that the water is divided evenly by the cone. Bernoulli’s equation suggests that because the pressure on the jet boundary is constant, the water velocity relative to the cone surface is constant. Determine the thickness of the water stream when it reaches the base of the cone. 10 cm 30° 5 m/s 10 m/s ■ Figure P5.31 5.32 The Hoover Dam (see Video V2.4) backs up the Colorado River and creates Lake Mead, which is approximately 115 miles long and has a surface area of approximately 225 square miles. If during flood conditions the Colorado River flows into the lake at a rate of 45,000 cfs and the outflow from the dam is 8000 cfs, how many feet per 24-hour day will the lake level rise? 5.33 Storm sewer backup causes your basement to flood at the steady rate of 1 in. of depth per hour. The basement floor area is 1500 ft2. What capacity (gal/min) pump would you rent to (a) keep the water accumulated in your basement at a constant level until the storm sewer is blocked off, and (b) reduce the water accumulation in your basement at a rate of 3 in./hr even while the backup problem exists? 5.34 (See The Wide World of Fluids article “‘Green’ 1.6-gpf standards,” Section 5.1.2.) When a toilet is flushed, the water depth, h, in the tank as a function of time, t, is as given in the table. The size of the rectangular tank is 19 in. by 7.5 in. (a) Determine the volume of water used per flush, gpf. (b) Plot the flowrate for 0 ≤ t ≤ 6 s. 20 cm ■ Figure P5.29 Section 5.1.4 Deforming Control Volume 5.30 A hypodermic syringe (see Fig. P5.30) is used to apply a vaccine. If the plunger is moved forward at the steady rate of 20 mm/s and if vaccine leaks past the plunger at 0.1 of the volume flowrate out the needle opening, calculate the average velocity of the needle exit flow. The inside diameters of the syringe and the needle are 20 mm and 0.7 mm. Qleak V3 Qout ■ Figure P5.30 5.31 Figure P5.31 shows a two-reservoir water supply system. The water level in reservoir 1 drops at the rate of 0.01 m/min, and the water level in reservoir 2 drops at the rate of 0.015 m/ min. Calculate the average velocity V3 in the 0.50-m-diameter pipe. t (s) h (in.) 0 0.5 1.0 2.0 3.0 4.0 5.0 6.0 5.70 5.33 4.80 3.45 2.40 1.50 0.75 0 Section 5.2.1 Derivation of the Linear Momentum Equation 5.35 Find the force components Fx and Fy required to hold the box shown in Fig. P5.35 stationary. The fluid is oil and has specific gravity of 0.85. Neglect gravity effects. Atmospheric pressure acts around the entire box. Steady flow conditions prevail. www.Ebook777.com Problems the horizontal plane required to hold the bend in place is shown on the figure. Determine the flowrate through the bend and the y-direction anchoring force, FAy, required to hold the bend in place. The flow is not frictionless. Q3 = 3.0 m3/s V3 = 1.5 m/s V2 = 2.0 m/s 261 θ 3 = 60° θ2 = 10 psi 45° Bend Q=? Q1 = 5 m3/s V1 = 2.0 m/s Fx Water Area = 0.2 ft2 FAx = 1440 lb FAy = ? Area = 0.1 ft2 45° Fy ■ Figure P5.35 ■ Figure P5.38 Section 5.2.2 Application of the Linear Momentum Equation (also see Lab Problems 5.1LP, 5.2LP, 5.3LP, and 5.4LP) †5.36 When a baseball player catches a ball, the force of the ball on her glove is as shown as a function of time in Fig. P5.36. Describe how this situation is similar to the force generated by the deflection of a jet of water by a vane. Note: Consider many baseballs being caught in quick succession. 5.39 Find the magnitude of the force F required to hold the plate in Fig. P5.39 stationary. V2 V1 = 30 m/s m∙ = 5 kg/s F Water 45° F V3 ■ Figure P5.39 t ■ Figure P5.36 5.37 Find the horizontal and vertical forces to hold stationary the nozzle shown in Fig. P5.37. The fluid flowing through it is 10 °C liquid water; A1 = 1.0 m2, A2 = 0.25 m2, V1 = 20 m/s, p2 = patm, and p1 = patm + 30 kPa. Neglect gravity. 5.40 Water enters the horizontal, circular cross-sectional, sudden contraction nozzle sketched in Fig. P5.40 at section (1)  with a uniformly distributed velocity of 25 ft/s and a pressure of 75 psi. The water exits from the nozzle into the atmosphere at section (2) where the uniformly distributed velocity is 100 ft/s. Determine the axial component of the anchoring force required to hold the contraction in place. 45° A2, p2 = patm y p2 = p1 = 75 psi V1 = 25 ft/s 0 psi V2 = 100 ft/s Fx V1 = 20 m/s x Section (2) D1 = 3 in. Section (1) A1 Fy patm ■ Figure P5.40 p1 †5.41 A truck carrying chickens is too heavy for a bridge that it needs to cross. The empty truck is within the weight limits; with the chickens it is overweight. It is suggested that if one 5.38 Water flows through a horizontal bend and discharges into could get the chickens to fly around the truck (i.e., by banging the atmosphere as shown in Fig. P5.38. When the pressure gage on the truck side) it would be safe to cross the bridge. Do you reads 10 psi, the resultant x-direction anchoring force, FAx, in agree? Explain. ■ Figure P5.37 262 Chapter 5 ■ Finite Control Volume Analysis 5.42 Exhaust (assumed to have the properties of standard air) leaves the 4-ft-diameter chimney shown in Video V5.4 and Fig. P5.42 with a speed of 6 ft/s. Because of the wind, after a few diameters downstream the exhaust flows in a horizontal direction with the speed of the wind, 15 ft/s. Determine the horizontal component of the force that the blowing wind exerts on the exhaust gases. 5.45 Determine the magnitude and direction of the anchoring force needed to hold the horizontal elbow and nozzle combination shown in Fig. P5.45 in place. Atmospheric pressure is 100 kPa(abs). The gage pressure at section (1) is 100 kPa. At section (2), the water exits to the atmosphere. 160 mm V2 15 ft/s 15 ft/s y Section (2) x 300 mm 6 ft/s V1 4 ft Water Section (1) p1 = 100 kPa V1 = 2 m/s ■ Figure P5.42 ■ Figure P5.45 5.43 Air at T1 = 300 K, p1 = 303 kPa, and V1 = 0.5 m/s enters the Venturi shown in Fig. P5.43. The air leaves at T2 = 220 K and p2 = 101 kPa; A1 = 0.6 m2 and A2 = 1.0 m2. Calculate the horizontal force required to hold the Venturi stationary. 5.46 Figure P5.46 shows a lateral pipe fitting. This particular fitting has a mainline diameter of 4.0 in. The diameter of the lateral is 3.0 in., and the lateral angle is 45°; 60 °F water is flowing in the lateral. Measurements show that the pressure at point 1 is 34.0 psig, the pressure at point 2 is 35.0 psig, the pressure at point 3 is 33.5 psig, and the flow rate at point 2 is 1.0 ft3/s. Determine the horizontal and vertical force components (Fx and Fy) required to hold the lateral fitting stationary. Neglect gravity. Q1 = 1.63 ft3/s. V1 = 0.5 m/s V2 T1, p1 T2, p2 Q3 D3 = 3 in. F Fx patm = 101 kPa 3 ■ Figure P5.43 D1 = 4 in. 5.44 Water flows steadily from a tank mounted on a cart as shown in Fig. P5.44. After the water jet leaves the nozzle of the tank, it falls and strikes a vane attached to another cart. The cart’s wheels are frictionless, and the fluid is inviscid. (a) Determine the speed of the water leaving the tank, V1, and the water speed leaving the cart, V2. (b) Determine the tension in rope A. (c) Determine the tension in rope B. θ = 45° Q1 Q2 1 2 y x Fy ■ Figure P5.46 5.47 Water flows steadily between fixed vanes, as shown in Fig. P5.47. Find the x and y components of the water’s force on the vanes. The total volume flow rate is 100 m3/s, pressure p1 = 150 kPa, and pressure p2 = 101 kPa. Nozzle area = 0.01 m2 2m 4m Horizontal free jets V1 = 10 m/s V1 30° 45° V2 A2 = A1 y Rope A ■ Figure P5.44 Rope B A1 x ■ Figure P5.47 Problems 5.48 The hydraulic dredge shown in Fig. P5.48 is used to dredge sand from a river bottom. Estimate the thrust needed from the propeller to hold the boat stationary. Assume the specific gravity of the sand/water mixture is SG = 1.4. 2-ft diameter 9 ft 7 ft Vp D0 263 Dp Vs 30 ft/s D0 = 1.0 m Dp = 0.25 m Vs = 1.07 m/s Vp = 4.0 m/s 30° Prop 1 2 ■ Figure P5.52 ■ Figure P5.48 5.49 A static thrust stand is to be designed for testing a specific jet engine, knowing the following conditions for a typical test. intake air velocity = 700 ft/s exhaust gas velocity = 1640 ft/s intake cross section area = 10 ft2 intake static pressure = 11.4 psia intake static temperature = 480 °R exhaust gas pressure = 0 psi 5.53 Air flows into the atmosphere from a nozzle and strikes a vertical plate as shown in Fig. P5.53. A horizontal force of 12 N is required to hold the plate in place. Determine the reading on the pressure gage. Assume the flow to be incompressible and frictionless. p=? 12N Estimate a nominal thrust to design for. 5.50 A vertical jet of water leaves a nozzle at a speed of 10 m/s and a diameter of 20 mm. It suspends a plate having a mass of 1.5 kg as indicated in Fig. P5.50. What is the vertical distance h? Area = 0.003 m2 Area = 0.01 m2 ■ Figure P5.53 h 5.54 Water flows from a large tank into a dish as shown in Fig. P5.54. (a) If at the instant shown the tank and the water in it weigh W1 lb, what is the tension, T1, in the cable supporting the tank? (b) If at the instant shown the dish and the water in it weigh W2 lb, what is the force, F2, needed to support the dish? T1 ■ Figure P5.50 5.51 A horizontal, circular cross-sectional jet of air having a diameter of 6 in. strikes a conical deflector as shown in Fig. P5.51. A horizontal anchoring force of 5 lb is required to hold the cone in place. Estimate the nozzle flowrate in ft3/s. The magnitude of the velocity of the air remains constant. Tank 10 ft 0.1-ft diameter 60° 6 in. 12 ft FA = 5 lb 2 ft Dish F2 ■ Figure P5.51 ■ Figure P5.54 5.52 Calculate the pressure change (p2 − p1) for the jet pump shown in Fig. P5.52. The fluid is 20 °C water. Assume negligible friction at the walls and uniform pressure over each flow area. 5.55 Figure P5.55 shows the configuration of the center (tailmounted) jet engine on an airliner. The airliner is cruising at altitude, and the velocities shown are relative to an observer 264 Chapter 5 ■ Finite Control Volume Analysis on board. Calculate the thrust force that the engine exerts on the airplane. D1 = 1.0 m pulley diameter is 2.0 ft. Assume that there is no friction between the belt and the other rollers. Coal hopper m• fuel = 3 kg/s D2 = 0.75 m 15.0′ Engine V1 = 250 m/s ρ1 = 0.4 kg/m3 p1 = pamb F = thrust force of engine on plane ρ2 = 0.44 kg/m3 p2 = 63 kPa = 40 kPa A = 12 ft2 4.0′ (nozzle exit only) ■ Figure P5.55 5.56 The plate shown in Fig. P5.56 is 0.5 m wide perpendicular to the paper. Calculate the velocity of the water jet required to hold the plate upright. ■ Figure P5.58 5.59 Determine the magnitude of the horizontal component of the anchoring force required to hold in place the sluice gate shown in Fig. 5.59. Compare this result with the size of the horizontal component of the anchoring force required to hold in place the sluice gate when it is closed and the depth of water upstream is 10 ft. d = 10.0 cm V 20°C water jet H = 0.5 m 10 ft 20°C water 1.5 ft 4 ft/s Pivot ■ Figure P5.56 ■ Figure P5.59 5.57 Two water jets of equal size and speed strike each other as shown in Fig. P5.57. Determine the speed, V, and direction, θ, of the resulting combined jet. Gravity is negligible. 5.60 Water flows steadily into and out of a tank that sits on frictionless wheels as shown in Fig. P5.60. Determine the diameter D so that the tank remains motionless if F = 0. V θ d V2 = 10 ft /s 0.1 ft F 90° 0.1 ft V1 = 10 ft /s D d ■ Figure P5.60 ■ Figure P5.57 5.58 Figure P5.58 shows coal being dropped from a hopper onto a conveyor belt at a constant rate of 25 ft3/s. The coal has a specific gravity ranging from 1.12 to 1.50. The belt has a speed of 5.0 ft/s and a loaded length of 15.0 ft. Estimate the torque required to turn the drive pulley of the conveyor belt. The drive 5.61 The rocket shown in Fig. P5.61 is held stationary by the horizontal force, Fx, and the vertical force, Fz. The velocity and pressure of the exhaust gas are 5000 ft/s and 20 psia at the nozzle exit, which has a cross section area of 60 in.2. The exhaust mass flowrate is constant at 21 lbm/s. Determine the value of the restraining force Fx. Assume the exhaust flow is essentially horizontal. Problems 265 Q Fx Fz x h = 0.01 m ■ Figure P5.61 0.2 m 5.62 Air discharges from a 2-in.-diameter nozzle and strikes a curved vane, which is in a vertical plane as shown in Fig. P5.62. A stagnation tube connected to a water U-tube manometer is located in the free air jet. Determine the horizontal component of the force that the air jet exerts on the vane. Neglect the weight of the air and all friction. 10 m/s y 0 m/s ■ Figure P5.64 Stagnation tube Air Fixed vane 2-in. dia. Open Water Free air jet where u is the velocity in ft/s and y is the distance on either side of the centerline in feet (see Fig. P5.65). Assume that the body shape does not change in the direction normal to the paper. Calculate the drag force (reaction force in x direction) exerted on the air by the body per unit length normal to the plane of the sketch. V2 = 100 ft/s 7 in. V1 = 100 ft/s 30° Body u 3 ft ■ Figure P5.62 5.63 Water is sprayed radially outward over 180° as indicated in Fig. P5.63. The jet sheet is in the horizontal plane. If the jet velocity at the nozzle exit is 20 ft/s, determine the direction and magnitude of the resultant horizontal anchoring force required to hold the nozzle in place. 8 in. 0.5 in. V= 20 ft/s Section (1) 3 ft Section (2) ■ Figure P5.65 5.66 A variable mesh screen produces a linear and axisymmetric velocity profile as indicated in Fig. P5.66 in the airflow through a 2-ft-diameter circular cross-sectional duct. The static pressures upstream and downstream of the screen are 0.2 and 0.15 psi and are uniformly distributed over the flow cross-sectional area. Neglecting the force exerted by the duct wall on the flowing air, calculate the screen drag force. Variable mesh screen ■ Figure P5.63 5.64 A sheet of water of uniform thickness (h = 0.01 m) flows from the device shown in Fig. P5.64. The water enters vertically through the inlet pipe and exits horizontally with a speed that varies linearly from 0 to 10 m/s along the 0.2-m length of the slit. Determine the y component of anchoring force necessary to hold this device stationary. 5.65 The results of a wind tunnel test to determine the drag on a body (see Fig. P5.65) are summarized below. The upstream [section (1)] velocity is uniform at 100 ft/s. The static pressures are given by p1 = p2 = 14.7 psia. The downstream velocity distribution, which is symmetrical about the centerline, is given by ∣ y∣    y ≤ 3 ft u = 100 − 30(1 − ∣∣ 3) u = 100   ∣ y ∣ > 3 ft D = 2 ft Section (1) Section (2) p1 = 0.2 psi V1 = 100 ft/s p2 = 0.15 psi ■ Figure P5.66 5.67 Consider unsteady flow in the constant diameter, horizontal pipe shown in Fig. P5.67. The velocity is uniform throughout the entire pipe, but it is a function of time: V = u(t) ˆi . Use the x component of the unsteady momentum equation to determine the pressure difference p1 − p2. Discuss how this result is related to Fx = max. 266 Chapter 5 ■ Finite Control Volume Analysis u (t) ℓ D x 3.5-in.-diameter outlet jet ρ = density (1) (2) 30° ■ Figure P5.67 25-in.2 inlet area 5.68 In a turbulent pipe flow that is fully developed, the axial velocity profile is, r 1/7 u = uc 1 − [ ( R )] as is illustrated in Fig. P5.68. Compare the axial direction momentum flowrate calculated with the average velocity, u, with the axial direction momentum flowrate calculated with the nonuniform velocity distribution taken into account. r R ■ Figure P5.71 about the aircraft’s mass center (cg) for the conditions indicated in the figure. (b) By how much is the thrust (force along the centerline of the aircraft) reduced for the case indicated compared to normal flight when the exhaust is parallel to the centerline? pout = 0 Vout = 1500 ft/s cg 20 ft Vin = 300 ft/s pin = 0 Vane u θ = 8 deg uc mt in = mt out = 16 slugs/s ■ Figure P5.72 ■ Figure P5.68 5.69 Water from a garden hose is sprayed against your car to rinse dirt from it. Estimate the force that the water exerts on the car. List all assumptions and show calculations. 5.70 A Pelton wheel vane directs a horizontal, circular crosssectional jet of water symmetrically as indicated in Fig. P5.70 and Video V5.6. The jet leaves the nozzle with a velocity of 100 ft/s. Determine the x-direction component of anchoring force required to (a) hold the vane stationary, (b) confine the speed of the vane to a value of 10 ft/s to the right. The fluid speed magnitude remains constant along the vane surface. 45° 5.73 The exhaust gas from the rocket shown in Fig. P5.73a leaves the nozzle with a uniform velocity parallel to the x axis. The gas is assumed to be discharged from the nozzle as a free jet. (a) Show that the thrust that is developed is equal to ρAV2, where A = πD2/4. (b) The exhaust gas from the rocket nozzle shown in Fig. P5.73b is also uniform, but rather than being directed along the x axis, it is directed along rays from point 0 as indicated. Determine the thrust for this rocket. mt D 45° 100 ft/s D = 1 in. y V 100 ft/s x (a) 10 ft/s D = 1 in. 45° O 45° (a) θ mt D (b) V (b) ■ Figure P5.70 ■ Figure P5.73 5.71 The thrust developed to propel the jet ski shown in Video V9.18 and Fig. P5.71 is a result of water pumped through the vehicle and exiting as a high-speed water jet. For the conditions shown in the figure, what flowrate is needed to produce a 300-lb thrust? Assume the inlet and outlet jets of water are free jets. 5.72 Thrust vector control is a technique that can be used to greatly improve the maneuverability of military fighter aircraft. It consists of using a set of vanes in the exit of a jet engine to deflect the exhaust gases as shown in Fig. P5.72. (a) Determine the pitching moment (the moment tending to rotate the nose of the aircraft up) 5.74 (See The Wide World of Fluids article titled “Where the Plume goes,” Section 5.2.2.) Air flows into the jet engine shown in Fig. P5.74 at a rate of 9 slugs/s and a speed of 300 ft/s. Upon landing, the engine exhaust exits through the reverse thrust mechanism with a speed of 900 ft/s in the direction indicated. Determine the reverse thrust applied by the engine to the airplane. Assume the inlet and exit pressures are atmospheric and that the mass flowrate of fuel is negligible compared to the air flowrate through the engine. Problems V3 = 900 ft/s 5.77 (See The Wide World of Fluids article titled “Bow Thrusters,” Section 5.2.2.) The bow thruster on the boat shown in Fig. P5.77 is used to turn the boat. The thruster produces a 1-m-diameter jet of water with a velocity of 10 m/s. Determine the force produced by the thruster. Assume that the inlet and outlet pressures are zero and that the momentum of the water entering the thruster is negligible. (3) 4-ft diameter (1) V1 = 300 ft/s 267 30° (2) V = 10 m/s V2 = 900 ft/s D=1m ■ Figure P5.74 5.75 Figure P5.75 shows a sharp-edged splitter plate used to control the flow of a liquid jet W units wide by H1 units high. Write expressions for the deflection angle θ and the force F of the jet on the splitter plate as a function of the fluid density ρ, H1, W, V, and plate insertion h. This force F has no components parallel to the plate. Assume that the jet flow is inviscid, that the jet width W remains constant, H2 H3 = H′2 H′3 = h/(H1 − h), and constant fluid density. V ■ Figure P5.77 5.78 Water flows from a two-dimensional open channel and is diverted by an inclined plate as illustrated in Fig. P5.78. When the velocity at section (1) is 10 ft/s, what horizontal force (per unit width) is required to hold the plate in position? At section (1) the pressure distribution is hydrostatic, and the fluid acts as a free jet at section (2). Neglect friction. patm A1 = WH1 V A3 = WH3 θ H3′ H2′ h 20° Section (1) A2 = WH2 patm V 10 ft/s Plate 4 ft patm Section (2) ■ Figure P5.75 5.76 (See The Wide World of Fluids article titled “Motorized Surfboard,” Section 5.2.2.) The thrust to propel the powered surfboard shown in Fig. P5.76 is a result of water pumped through the board that exits as a high-speed 2.75-in.-diameter jet. Determine the flowrate and the velocity of the exiting jet if the thrust is to be 300 lb. Neglect the momentum of the water entering the pump. 1.0 ft ■ Figure P5.78 5.79 If a valve in a pipe is suddenly closed, a large pressure surge may develop. For example, when the electrically operated shutoff valve in a dishwasher closes quickly, the pipes supplying the dishwasher may rattle or “bang” because of this large pressure pulse. Explain the physical mechanism for this “water hammer” phenomenon. How could this phenomenon be analyzed? 5.80 A snowplow mounted on a truck clears a path 12 ft through heavy wet snow, as shown in Figure P5.80. The snow is 8 in. deep and its density is 10 lbm/ft3. The truck travels at 30 mph. The snow is discharged from the plow at an angle of 45° from the direction of travel and 45° above the horizontal, as shown in Figure P5.80. Estimate the force required to push the plow. U = 30 mph θ = 45° (in plane of blade) d = 8 in. ■ Figure P5.76 ■ Figure P5.80 268 Chapter 5 ■ Finite Control Volume Analysis Section 5.2.3 Derivation of the Moment-of-Momentum Equation 5.81 An incompressible fluid flows outward through a blower as indicated in Fig. P5.81. The shaft torque involved, Tshaft, is estimated with the following relationship: Nozzle exit area normal to relative velocity = 18 mm2 r = 0.5m θ Tshaft = m ˙ r2Vθ2 where m ˙ = mass flowrate through the blower, r2 = outer radius of blower, and Vθ 2 = tangential component of absolute fluid velocity leaving the blower. State the flow conditions that make this formula valid. Q = 5 liters/s ■ Figure P5.83 V2 5.84 Figure P5.84 shows a simplified sketch of a dish-washer water supply manifold. Find the resisting torque for a water temperature of 140 °F. Q = 0.25 gal/min and ω = 3 rpm. Vθ2 r2 ¼-in.-diameter holes r1 ω Q ω Q Q Q Q Q Q Q 45° 3 in. 2 in.2 in. 2 in. ■ Figure P5.84 ■ Figure P5.81 Section 5.2.4 Application of the Moment-of-Momentum Equation 5.82 Water at 60 °F is flowing through the 2-in. steel pipe shown in Fig. P5.82 at the rate of 90 gal/min. Determine the torque developed at the base where the pipe is supported. Neglect the pipe and water weights. Steady-state conditions apply. 5.85 The hydraulic turbine shown in Fig. P5.85 has a 10 °C water flow rate of 36.4 m3/s, an inlet radius R1 = 1.0 m, an outlet radius R2 = 0.50 m, a blade depth (perpendicular to paper) h = 0.50 m, and a rotational speed N = 360 rpm; V1 = 50 m/s, V2 = 30 m/s, α1 = 13.4°, and α2 = 40°. Calculate the power transferred by the fluid to the rotor, the inlet relative velocity W1, and the direction W1 makes with the radius at the inlet. V1 β 1 α1 U1 V2 α2 Horizontal section of pipe is 24 in. long. 13 in. Pipe supported by wall 9.5 in. Q = 90 gal/min ■ Figure P5.82 5.83 Five liters/s of water enter the rotor shown in Video V5.10 and Fig. P5.83 along the axis of rotation. The crosssectional area of each of the three nozzle exits normal to the relative velocity is 18 mm2. How large is the resisting torque required to hold the rotor stationary? How fast will the rotor spin steadily if the resisting torque is reduced to zero and (a) θ = 0°, (b) θ = 30°, (c) θ = 60°? W1 W2 Blade U2 ω R2 R1 ■ Figure P5.85 5.86 A fan (see Fig. P5.86) has a bladed rotor of 12-in. outside diameter and 5-in. inside diameter and runs at 1725 rpm. The width of each rotor blade is 1 in. from blade inlet to outlet. The volume flowrate is steady at 230 ft3/min, and the absolute velocity of the air at blade inlet, V1, is purely radial. The blade discharge angle is 30° measured with respect to the tangential direction at the outside diameter of the rotor. (a) What would be a reasonable blade inlet angle (measured with respect to the tangential direction at the inside diameter of the rotor)? (b) Find the power required to run the fan. Problems 269 Q = 230 ft3/min 30° Section (1) α1 r1 V1 r2 12 in. 5 in. V2 = 12 m/s 1725 rpm Section (2) 1 in. ■ Figure P5.89 ■ Figure P5.86 5.87 Calculate the torque required to drive the pump shown in Fig. P5.87 at 30 Hz and to deliver 20 °C water at 3.0 m3/min. Absolute velocity, V2 5.90 A sketch of the arithmetic mean radius blade sections of an axial-flow water turbine stage is shown in Fig. P5.90. The rotor speed is 1000 rpm. (a) Sketch and label velocity triangles for the flow entering and leaving the rotor row. Use V for absolute velocity, W for relative velocity, and U for blade velocity. Assume flow enters and leaves each blade row at the blade angles shown. (b) Calculate the work per unit mass delivered at the shaft. b2 = 1.0 cm θ2 = 45° ω, T Absolute velocity, V1 R2 = 0.13 m rm = 6 in. 30° = θ1 1000 rpm b1 = 3.0 cm R1 = 0.065 m Rotor Stator 70° U Impeller ■ Figure P5.87 45° 5.88 An axial flow turbomachine rotor involves the upstream (1) and downstream (2) velocity triangles shown in Fig. P5.88. Is this turbomachine a turbine or a fan? Sketch an appropriate blade section and determine energy transferred per unit mass of fluid. 1 W2 U1 = 30 ft/s V1 = 20 ft/s Blade sections at the arithmetic mean radius 45° ■ Figure P5.90 Section 5.3.1 Derivation of the Energy Equation 5.91 Distinguish between shaft work and other kinds of work associated with a flowing fluid. W1 = W2 W1 U U2 = 30 ft/s 60° 5.92 An incompressible fluid flows along a 0.20-m-diameter pipe with a uniform velocity of 3 m/s. If the pressure drop between the upstream and downstream sections of the pipe is 20 kPa, determine the power transferred to the fluid due to fluid normal stresses. ■ Figure P5.88 Section 5.3.2 Application of the Energy Equation—No Shaft Work and Section 5.3.3 The Mechanical Energy Equation and the Bernoulli Equation 5.89 An inward flow radial turbine (see Fig. P5.89) involves a nozzle angle, α1, of 60° and an inlet rotor tip speed, U1, of 6 m/s. The ratio of rotor inlet to outlet diameters is 1.8. The absolute velocity leaving the rotor at section (2) is radial with a magnitude of 12 m/s. Determine the energy transfer per unit mass of fluid flowing through this turbine if the fluid is (a) air, (b) water. 5.93 A horizontal Venturi flow meter consists of a converging– diverging conduit as indicated in Fig. P5.93. The diameters of cross sections (1) and (2) are 6 and 4 in. The velocity and static pressure are uniformly distributed at cross sections (1) and (2). Determine the volume flowrate (ft3/s) through the meter if p1 − p2 = 3 psi, the flowing fluid is oil (p = 56 lbm/ft3), and the loss per unit mass from (1) to (2) is negligibly small. 270 Chapter 5 ■ Finite Control Volume Analysis D1 = 6 in. 5.97 Oil (SG = 0.9) flows downward through a vertical pipe contraction as shown in Fig. P5.97. If the mercury manometer reading, h, is 100 mm, determine the volume flowrate for frictionless flow. Is the actual flowrate more or less than the frictionless value? Explain. Section (1) D2 = 4 in. 300 mm Section (2) ■ Figure P5.93 5.94 Figure P5.94 shows the mixing of two streams. The shear stress between each fluid and its adjacent walls is negligible. Why can’t Bernoulli’s equation be applied between points in stream 1 and the mixed stream or between points in stream 2 and the mixed stream? p1 = 50 psia 0.6 m p3 h 5 ft/s 2 ft2 p1 = 62.4 lbm/ft3 100 mm ■ Figure P5.97 2 ft2 p2 = 62.4 lbm/ft3 5.98 An incompressible liquid flows steadily along the pipe shown in Fig. P5.98. Determine the direction of flow and the head loss over the 6-m length of pipe. 10 ft/s 0.75 m p2 = 50 psia 1.0 m ■ Figure P5.94 3m 5.95 Liquid water at 40 °F flows down a vertical, thermally insulated, 2-in. steel pipe. The temperature change of the water is related to its internal energy change by ~ u = 32.2 Btu/slug ∙ °F(T − T ). u −~ 2 1 2 1 What is the temperature change of the water per 100 ft of drop if the pressure drop (p1 − p2) per 100 ft is 12.0 psi? *5.96 A simplified schematic drawing of the carburetor of a gasoline (S = 0.75) engine is shown in Fig. P5.96. The throat area is 0.5 in.2. The running engine draws air downward through the carburetor Venturi and maintains a throat pressure of 14.3 psia. The low throat pressure draws fuel from the float chamber and into the airstream. The energy losses in the 0.07-in.-diameter fuel metering line and valve are given by ghL = KV2 , 2g 6m 1.5 m ■ Figure P5.98 5.99 A siphon is used to draw water at 70 °F from a large container as indicated in Fig. P5.99. The inside diameter of the siphon line is 1 in. and the pipe centerline rises 3 ft above the essentially constant water level in the tank. Show that by varying the length of the siphon below the water level, h, the rate of flow through the siphon can be changed. Assuming frictionless flow, determine the maximum flowrate possible through the siphon. The limiting condition is the occurrence of cavitation in the siphon. Will the actual maximum flow be more or less than the frictionless value? Explain. where K = 6.0 and V is the fuel velocity in the metering line. Assume that the air is an ideal fluid having a constant density ρA = 0.075 lbm/ft3. The atmospheric pressure is 14.7 psia. Calculate the air-to-fuel ratio (m ˙ am˙ f ). Air 3 ft Air vent h Fuel 2 0.5 in. ■ Figure P5.96 1 in. Float chamber ■ Figure P5.99 Problems 5.100 A water siphon having a constant inside diameter of 3 in. is arranged as shown in Fig. P5.100. If the friction loss between A and B is 0.8V2/2, where V is the velocity of flow in the siphon, determine the flowrate involved. 271 5.103 An automobile engine will work best when the back pressure at the interface of the exhaust manifold and the engine block is minimized. Show how reduction of losses in the exhaust manifold, piping, and muffler will also reduce the back pressure. How could losses in the exhaust system be reduced? What primarily limits the minimization of exhaust system losses? 5.104 (See The Wide World of Fluids article titled “Smart Shocks,” Section 5.3.3.) A 200-lb force applied to the end of the piston of the shock absorber shown in Fig. P5.104 causes the two ends of the shock absorber to move toward each other with a speed of 5 ft/s. Determine the head loss associated with the flow of the oil through the channel. Neglect gravity and any friction force between the piston and cylinder walls. 4 ft A 12 ft 4 ft 3 in. B Gas ■ Figure P5.100 p=0 Oil Piston 5.101 Figure P5.101 shows a test rig for evaluating the loss coefficient, K, for a valve. Mechanical energy losses in valves are modeled by the equation: ghL = K( V2 , 2) where ghL is the mechanical energy loss and V is the flow velocity entering the valve. In a particular test, the pressure gage reads 40 kPa, gage, and the 1.5-m3 catch tank fills in 2 min 55 s. Calculate the loss coefficient for a water temperature of 20 °C. D1 = 6 cm D2 = 4 cm 1 Channel 1-in. diameter 200 lb ■ Figure P5.104 Section 5.3.2 Application of the Energy Equation— With Shaft Work 5.105 Based on flowrate and pressure rise information, estimate the power output of a human heart. 2 Valve 5.106 Oil (SG = 0.88) flows in an inclined pipe at a rate of 5 ft3/s as shown in Fig. P5.106. If the differential reading in the mercury manometer is 3 ft, calculate the power that the pump supplies to the oil if head losses are negligible. Pressure gage Catch tank ■ Figure P5.101 6 in. 5.102 For the 180° elbow and nozzle flow shown in Fig. P5.102, determine the loss in available energy from section (1) to section (2). How much additional available energy is lost from section (2) to where the water comes to rest? h Oil P H 6 in. 12 in. y Section (2) x 12 in. p1 = 15 psi V1 = 5 ft/s ■ Figure P5.106 Section (1) ■ Figure P5.102 3 ft 5.107 The pumper truck shown in Fig. P5.107 is to deliver 1.5 ft3/s to a maximum elevation of 60 ft above the hydrant. The pressure 272 Chapter 5 ■ Finite Control Volume Analysis at the 4-in.-diameter outlet of the hydrant is 10 psi. If head losses are negligibly small, determine the power that the pump must add to the water. in the pipe. According to the pump manufacturer, the relationship between the pump head and the flowrate is as shown in Fig. P5.110b: hp = 20 − 2000 Q2, where hp is in meters and Q is in m3/s. Determine the flowrate, Q. hp, m 20 6m 10 hp = 20–2000Q2 60 ft 0 0 10 psi 4-in. diameter Pump 0.05 Q, m3/s 0.07 m (a) 0.10 (b) ■ Figure P5.110 Hydrant ■ Figure P5.107 5.108 The hydroelectric turbine shown in Fig. P5.108 passes 8 million gal/min across a head of 600 ft. What is the maximum amount of power output possible? Why will the actual amount be less? 5.111 Water is pumped steadily through the apparatus shown in Fig. P5.111. The pipe area and gage pressure are shown for both outlet sections 1 and 2. Assume that the 40 °F water is frictionless and incompressible. Compute the horsepower input to the pump. The total volume flow rate QT = 1.0 ft3/s. Da = 4.0 in. A1 = 0.04 ft2 p1 = 10 psig patm = 14.7 psi Q1 1 600 ft Q2 25 ft 2 15 ft A2 = 0.03 ft2 p2 = 15 psig a 7 ft Pump 0 Turbine ■ Figure P5.108 5.109 A pump is to move water from a lake into a large, pressurized tank as shown in Fig. P5.109 at a rate of 1000 gal in 10 min or less. Will a pump that adds 3 hp to the water work for this purpose? Support your answer with appropriate calculations. Repeat the problem if the tank were pressurized to 3, rather than 2, atmospheres. ■ Figure P5.111 5.112 Water is pumped from the large tank shown in Fig. P5.112. The head loss is known to be equal to 4V2/2g and the pump head is hp = 20 − 4Q2, where hp is in ft when Q is in ft3/s. Determine the flowrate. p = 2 atm Air 13 ft 20 ft Q Pump V Pump Pipe area = 0.10 ft2 ■ Figure P5.112 ■ Figure P5.109 5.110 Water is pumped from the tank shown in Fig. P5.110a. The head loss is known to be 1.2 V2/2g, where V is the average velocity 5.113 Water flows by gravity from one lake to another as sketched in Fig. P5.113 at the steady rate of 80 gpm. What is the loss in available energy associated with this flow? If this same amount of Problems 273 Section (2) loss is associated with pumping the fluid from the lower lake to the higher one at the same flowrate, estimate the amount of pumping power required. 8-in. insidediameter pipe 50 ft Section (1) 50 ft Pump ■ Figure P5.116 ■ Figure P5.113 5.114 The turbine shown in Fig. P5.114 develops 100 hp when the flowrate of water is 20 ft3/s. If all losses are negligible, determine (a) the elevation h, (b) the pressure difference across the turbine, and (c) the flowrate expected if the turbine were removed. 5.117 Determine the volume flow rate and minimum power input to the water pump in Fig. P5.117. Determine the actual power if the hydraulic efficiency is 75% and losses in the motor and bearings are negligible. patm = 14.7 psia p2 p2 = 49.5 psia d2 = 4.0 in. p3 p4 d1 = 6.0 in. p1 p1 = h 13.0 psia Q 12 in. T 16.0 ft 8.0 in. 12 in. Free jet ■ Figure P5.114 ■ Figure P5.117 5.115 Figure P5.115 shows a pump testing setup. Water is drawn from a sump and pumped through a pipe containing a valve. The water is discharged into a catch tank sitting on a scale. During a test run, 800 lb of water is collected in the catch tank in 15 s. The pump power input to the fluid during this period is 700 ft ∙ lb/s. Calculate the water velocity in the pipe and the mechanical energy loss (ft ∙ lb/lbm) in the pipe and valve. 5.118 A pump moves water horizontally at a rate of 0.02 m3/s. Upstream of the pump where the pipe diameter is 90 mm, the pressure is 120 kPa. Downstream of the pump where the pipe diameter is 30 mm, the pressure is 400 kPa. If the loss in energy across the pump due to fluid friction effects is 170 N ∙ m/kg, determine the hydraulic efficiency of the pump. 5.119 Water is to be pumped from the large tank shown in Fig. P5.119 with an exit velocity of 6 m/s. It was determined that the original pump (pump 1) that supplies 1 kW of power to the water did not produce the desired velocity. Hence, it is proposed that an additional pump (pump 2) be installed as indicated to increase the flowrate to the desired value. How much power must pump 2 add to the water? The head loss for this flow is hL = 250 Q2, where hL is in m when Q is in m3/s. 4-in. I.D. pipe Valve 5 ft Catch tank 60°F H2O Sump ■ Figure P5.115 5.116 Water is to be moved from one large reservoir to another at a higher elevation as indicated in Fig. P5.116. The loss of available energy associated with 2.5 ft3/s being pumped from sections (1) to (2) is loss = 61V 2/2 ft2/s2, where V is the average velocity of water in the 8-in. inside-diameter piping involved. Determine the amount of shaft power required. V = 6 m/s Nozzle area = 0.01 m2 Pipe area = 0.02 m2 Pump #2 Pump #1 2m ■ Figure P5.119 5.120 (See The Wide World of Fluids article titled “Curtain of Air,” Section 5.3.3.) The fan shown in Fig. P5.120 produces an air 274 Chapter 5 ■ Finite Control Volume Analysis curtain to separate a loading dock from a cold storage room. The air curtain is a jet of air 10 ft wide, 0.5 ft thick moving with speed V = 30 ft/s. The loss associated with this flow is loss = KLV2/2, where KL = 5. How much power must the fan supply to the air to produce this flow? principles, derive the following expression for h2, h2 = − h1 h1 2 2V 21h1 + ( ) + g 2 B 2 The loss of available energy across the jump can also be determined if energy conservation is considered. Derive the loss expression jump loss = Fan V = 30 ft/s g(h2 − h1 ) 3 4h1h2 Air curtain (0.5-ft thickness) Open door h1 10 ft h2 V1 ■ Figure P5.120 Section 5.3.3 Application of the Energy Equation and the Bernoulli Equation—Combined with Linear Momentum 5.121 When the pump shown in Fig. P5.121 is stopped, there is no flow through the system and the spring force is zero. With the pump running, a 6-in.-diameter jet leaves the pipe, and the spring force is 420 lb. The water surface elevation in the tank is constant. Determine the water flow rate and the power consumed by the pump. Assume quasi-steady flow. ■ Figure P5.123 5.124 Water flows steadily down the inclined pipe as indicated in Fig P5.124. Determine the following: (a) the difference in pressure p1 − p2, (b) the loss between sections (1) and (2), (c) the net axial force exerted by the pipe wall on the flowing water between sections (1) and (2). Flo w 5 ft Section (1) 3.0' Spring 6 in. 6-in. diameter Pump Jet 30° 1.0' 30° Section (2) Frictionless rollers 6 in. ■ Figure P5.121 5.122 Air flows past an object in a pipe of 2-m diameter and exits as a free jet as shown in Fig. P5.122. The velocity and pressure upstream are uniform at 10 m/s and 50 N/m2, respectively. At the pipe exit the velocity is nonuniform as indicated. The shear stress along the pipe wall is negligible. (a) Determine the head loss associated with a particle as it flows from the uniform velocity upstream of the object to a location in the wake at the exit plane of the pipe. (b) Determine the force that the air exerts on the object. Mercury ■ Figure P5.124 5.125 When fluid flows through an abrupt expansion as indicated in Fig. P5.125, the loss in available energy across the expansion, lossex, is often expressed as lossex = (1 − Air 2-m-dia. Wake 1-m dia. 4 m/s 12 m/s p = 50 N/m2 V = 10 m/s where A1 = cross-sectional area upstream of expansion, A2 = cross-sectional area downstream of expansion, and V1 = velocity of flow upstream of expansion. Derive this relationship. Exit ■ Figure P5.122 5.123 Near the downstream end of a river spillway, a hydraulic jump often forms, as illustrated in Fig. P5.123 and Video V10.11. The velocity of the channel flow is reduced abruptly across the jump. Using the conservation of mass and linear momentum A1 2 V 12 A2 ) 2 Section (1) Section (2) ■ Figure P5.125 Problems 5.126 Water (60 °F) flows through an annular space formed by inserting a 1-in.-radius solid cylinder into a 1.5-in.-radius tube. The following axial velocities were measured in the annular space. (r − ri)(r0 − ri) u (ft/s) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 12 23 28 33 34 31 28 21 10 area is uniform. At section (2), the velocity profile is V = wc ( R − r 1/ 7 ˆ k R ) where V = local velocity vector, wc = centerline velocity in the axial direction, R = pipe inside radius, and r = radius from pipe axis. Develop an expression for the loss in available energy between sections (1) and (2). 5.130 Calculate the kinetic energy correction factor for each of the following velocity profiles for a circular pipe: r ; R) 2 r (b) u = umax 1 − 2 ; ( R) (a) u = umax(1 − r 1/9 (c) u = umax 1 − . ( R) Assume that the no-slip condition (u = 0) exists at the solid boundaries. What are the rates of mass, momentum, and kinetic energy flow through the annular space? 5.127 Find the acceleration of the cart shown in Fig. P5.127 as a function of the water height in the cart, which varies with time. The initial total mass is m0, and the fluid density is ρ0. Assume frictionless bearings, a frictionless surface, constant fluid density, uniform velocity over area AN, all the fluid in the cart has the cart velocity, no air drag, and A >> AN . R is the pipe radius and r is the radial coordinate. 5.131 The cross-sectional area of a rectangular duct is divided into 16 equal rectangular areas, as shown in Fig. P5.131. The axial fluid velocity measured in feet per second in each smaller area is shown. Estimate the kinetic energy correction factor. 20.0 in. 16.0 in. A g AN 275 h0 = Initial fluid height ρ0 3.0 3.4 3.6 3.1 3.7 4.0 3.9 3.8 3.9 4.6 4.5 4.2 3.7 4.4 4.3 3.9 Velocities in ft/s ■ Figure P5.131 ■ Figure P5.127 5.128 Two water jets collide and form one homogeneous jet as shown in Fig. P5.128. (a) Determine the speed, V, and direction, θ, of the combined jet. (b) Determine the loss for a fluid particle flowing from (1) to (3), from (2) to (3). Gravity is negligible. V 0.12 m (3) θ V2 = 6 m/s (2) 90° (1) 0.10 m V1 = 4 m/s 5.132 A small fan moves air at a mass flowrate of 0.004 lbm/s. Upstream of the fan, the pipe diameter is 2.5 in., the flow is laminar, the velocity distribution is parabolic, and the kinetic energy coefficient, α1, is equal to 2.0. Downstream of the fan, the pipe diameter is 1 in., the flow is turbulent, the velocity profile is quite flat, and the kinetic energy coefficient, α2, is equal to 1.08. If the rise in static pressure across the fan is 0.015 psi and the fan shaft draws 0.00024 hp, compare the value of loss calculated: (a) assuming uniform velocity distributions, (b) considering actual velocity distributions. Section 5.3.5 Combination of the Energy Equation and the Moment-of-Momentum Equation 5.133 Air enters a radial blower with zero angular momentum. It leaves with an absolute tangential velocity, Vθ , of 200 ft/s. The rotor blade speed at rotor exit is 170 ft/s. If the stagnation pressure rise across the rotor is 0.4 psi, calculate the loss of available energy across the rotor and the rotor efficiency. Section 5.3.4 Application of the Energy Equation to Nonuniform Flows 5.134 Water enters a pump impeller radially. It leaves the impeller with a tangential component of absolute velocity of 10 m/s. The impeller exit diameter is 60 mm, and the impeller speed is 1800 rpm. If the stagnation pressure rise across the impeller is 45 kPa, determine the loss of available energy across the impeller and the hydraulic efficiency of the pump. 5.129 Water flows vertically upward in a circular cross-sectional pipe. At section (1), the velocity profile over the cross-sectional 5.135 Water enters an axial-flow turbine rotor with an absolute velocity tangential component, Vθ , of 15 ft/s. The corresponding ■ Figure P5.128 276 Chapter 5 ■ Finite Control Volume Analysis blade velocity, U, is 50 ft/s. The water leaves the rotor blade row with no angular momentum. If the stagnation pressure drop across the turbine is 12 psi, determine the hydraulic efficiency of the turbine. 5.136 An inward flow radial turbine (see Fig. P5.136) involves a nozzle angle, α1, of 60° and an inlet rotor tip speed, U1, of 30 ft/s. The ratio of rotor inlet to outlet diameters is 2.0. The radial component of velocity remains constant at 20 ft/s through the rotor, and the flow leaving the rotor at section (2) is without angular momentum. If the flowing fluid is water and the stagnation pressure drop across the rotor is 16 psi, determine the loss of available energy across the rotor and the hydraulic efficiency involved. 60° Vr1 = 20 ft/s U1 = 30 ft/s ■ Figure P5.136 r1 r2 2 1 ■ Lifelong Learning Problems 5.1 LL What are typical efficiencies associated with swimming and how can they be improved? 5.2 LL Magnetohydrodynamics (MHD) is the study of the dynamics of electrically conducting fluids. Discuss the special features of MHD flows and some engineering applications for these flows. 5.3 LL Discuss the main causes of loss of available energy in a turbo-pump and how they can be minimized. What are typical turbo-pump efficiencies? 5.4 LL Most of the loss of useful energy in fluid flows is due to turbulence. Explain the mechanisms by which turbulence and fluid viscosity work together to dissipate kinetic energy into heat/ internal energy. Find a short “poem” that describes this process. 6 Differential Analysis of Fluid Flow Learning Objectives After completing this chapter, you should be able to: ■ determine various kinematic elements of the flow given the velocity field. ■ explain the conditions necessary for a velocity field to satisfy the continuity equation. ■ apply the concepts of stream function and velocity potential. ■ characterize simple potential flow fields. ■ analyze certain types of flows using the Navier–Stokes equations. The previous chapter focused on the use of finite control volumes for solving a variety of fluid mechanics problems. This approach is very practical and useful, since it does not require detailed knowledge of the pressure and velocity variations within the control volume. Typically, we found that only conditions on the surface of the control volume were needed, and thus problems could be solved without detailed knowledge of the flow field. Unfortunately, many situations arise in which the details of the flow are important and the finite control volume approach will not yield the desired information. For example, we may need to know how the velocity varies over the cross section of a pipe, or how the pressure and shear stress vary along the surface of an airplane wing. In these circumstances we need to develop relationships that apply at a point, or at least in a very small region within a given flow field. This approach, which involves an infinitesimal control volume, as distinguished from a finite control volume, is commonly referred to as differential analysis, since (as we will soon discover) the governing equations are differential equations. In this chapter we will provide an introduction to the differential equations that describe (in detail) the motion of fluids. Unfortunately, we will also find that these equations are complicated, nonlinear partial differential equations that cannot be solved exactly except in a few simple cases. Thus, although differential analysis has the potential for supplying very detailed information about flow fields, this information is not easily extracted. Nevertheless, this approach provides a fundamental basis for the study of fluid mechanics. We do not want to be too discouraging at this point, since there are some exact solutions for laminar flow that can be obtained, and these have proved to be very useful. A few of these are included in this chapter. In addition, by making some simplifying assumptions, many other analytical solutions can be obtained. For example, in some circumstances it may be reasonable to assume that the effect of viscosity is small and can be neglected. This rather drastic assumption greatly simplifies the analysis and provides the opportunity to obtain detailed solutions to a variety of complex flow problems. Some examples of these so-called inviscid flow solutions are also described in this chapter. 277 278 VIDEO VIDEO Chapter 6 ■ Differential Analysis of Fluid Flow V6.1 Spinning footballpressure contours V6.2 Spinning football-velocity vectors 6.1 Fluid element motion consists of translation, linear deformation, rotation, and angular deformation. It is known that for certain types of flows the flow field can be conceptually divided into two regions—a very thin region near the boundaries of the system in which viscous effects are important, and a region away from the boundaries in which the flow is essentially inviscid. By making certain assumptions about the behavior of the fluid in the thin layer near the boundaries, and using the assumption of inviscid flow outside this layer, a large class of problems can be solved using differential analysis. These boundary layer problems are discussed in Chapter 9. Finally, with the availability of powerful computers it is feasible to attempt to solve the differential equations using the techniques of numerical analysis. Although it is beyond the scope of this book to delve extensively into this approach, which is generally referred to as computational fluid dynamics (CFD), the reader should be aware of this approach to complex flow problems. CFD has become a common engineering tool, and a brief introduction can be found in Appendix A. To introduce the power of CFD, two animations based on numerical computations are provided as shown in the margin. We begin our introduction to differential analysis by reviewing and extending some of the ideas associated with fluid kinematics that were introduced in Chapter 4. With this background the remainder of the chapter will be devoted to the derivation of the basic differential equations (which will be based on the principle of conservation of mass and Newton’s second law of motion) and to some applications. Fluid Element Kinematics In this section we will be concerned with the mathematical description of the motion of fluid elements moving in a flow field. A small fluid element in the shape of a cube that is initially in one position will move to another position during a short time interval δt as illustrated in Fig. 6.1a. Because of the generally complex velocity variation within the field, we expect the element not only to translate from one position to another but to be deformed as well. Even though they occur simultaneously, we can break the element’s complex motion into four components: translation, rotation, linear deformation, and angular deformation, as shown in Fig. 6.1b. Since element motion and deformation are intimately related to the velocity and variation of velocity throughout the flow field, we will briefly review the manner in which velocity and acceleration fields can be described. y y y Particle at time t + δ t Particle at time t (a) x Translation x Rotation y x y Linear deformation x Angular deformation x (b) ■ Figure 6.1 General fluid element motion and its components: (a) total element motion; (b) components of element motion. 6.1 Fluid Element Kinematics 279 6.1.1 Velocity and Acceleration Fields Revisited As discussed in detail in Section 4.1, the velocity field can be described by specifying the velocity V at all points, and at all times, within the flow field of interest. Thus, in terms of rectangular coordinates, the notation V (x, y, z, t) means that the velocity of a fluid particle depends on where it is located within the flow field (as determined by its coordinates, x, y, and z) and when it occupies the particular point (as determined by the time, t). As is pointed out in Section 4.1.1, this method of describing the fluid motion is called the Eulerian method. It is also convenient to express the velocity in terms of three rectangular components so that V = uiˆ + υjˆ + wkˆ z V w ^ k ^ i x ^ j v u y (6.1) where u, υ, and w are the velocity components in the x, y, and z directions, respectively, and ˆi , jˆ, and kˆ are the corresponding unit vectors, as shown by the figure in the margin. Of course, each of these components will, in general, be a function of x, y, z, and t. One of the goals of differential analysis is to determine how these velocity components specifically depend on x, y, z, and t for a particular problem. With this description of the velocity field it was also shown in Section 4.2.1 that the acceleration of a fluid particle can be expressed as a= ∂V ∂V ∂V ∂V +u +υ +w ∂t ∂x ∂y ∂z (6.2) ax = ∂u ∂u ∂u ∂u +u +υ +w ∂t ∂x ∂y ∂z (6.3a) ay = ∂υ ∂υ ∂υ ∂υ +u +υ +w ∂t ∂x ∂y ∂z (6.3b) az = ∂w ∂w ∂w ∂w +u +υ +w ∂t ∂x ∂y ∂z (6.3c) and in component form: The acceleration of a fluid particle is described using the material derivative. The acceleration is also concisely expressed as a= DV Dt (6.4) where the operator ∂( ) ∂( ) ∂( ) D( ) ∂( ) = +u +υ +w Dt ∂t ∂x ∂y ∂z (6.5) is termed the material derivative, or substantial derivative. In vector notation D( ) ∂( ) = + (V ∙ 𝛁)( ) Dt ∂t (6.6) where the gradient operator, 𝛁( ), is 𝛁( ) = ∂( ) ˆ ∂( ) ˆ ∂( ) ˆ i + j + k ∂x ∂y ∂z (6.7) which was introduced in Chapter 2. As we will see in the following sections, the motion and deformation of a fluid element depend on the velocity field. The relationship between the motion and the forces causing the motion depends on the acceleration field. 6.1.2 Linear Motion and Deformation The simplest type of motion that a fluid element can undergo is translation, as illustrated in Fig. 6.2. In a small time interval δt a particle located at point O will move to point O′ as is illustrated in the figure. If all points in the element have the same velocity (which is only true if there are no velocity gradients), then the element will simply translate from one position to another. However, Free ebooks ==> www.Ebook777.com 280 Chapter 6 ■ Differential Analysis of Fluid Flow v O O' v δt u ■ Figure 6.2 Translation of a fluid element. u δt because of the presence of velocity gradients, the element will generally be deformed and rotated as it moves. For example, consider the effect of a single velocity gradient, ∂u∂x, on a small cube having sides δx, δy, and δz. As is shown in Fig. 6.3a, if the x component of velocity of O and B is u, then at nearby points A and C the x component of the velocity can be expressed as u + ( ∂u∂x) δx. This difference in velocity causes a “stretching” of the volume element by an amount (∂u∂x)(δx)(δt) during the short time interval δt in which line OA stretches to OA′ and BC to BC′ (Fig. 6.3b). The corresponding change in the original volume, δ− V = δx δy δz, would be Change in δ− V =( ∂u δx (δy δz)(δt) ∂x ) and the rate at which the volume δ V is changing per unit volume due to the gradient ∂u∂x is (∂u∂x) δt V) 1 d(δ− ∂u = lim [ ] = ∂x δt→0 δ− V dt δt The rate of volume change per unit volume is related to the velocity gradients. (6.8) If velocity gradients ∂υ∂y and ∂w∂z are also present, then using a similar analysis it follows that, in the general case, V) 1 d(δ− ∂u ∂υ ∂w = + + =𝛁∙V (6.9) δ− V dt ∂x ∂y ∂z This rate of change of the volume per unit volume is called the volumetric dilatation rate. Thus, we see that the volume of a fluid element may change as the element moves from one location to another in the flow field. However, for an incompressible fluid the volumetric dilatation rate must be zero, since the element volume cannot change without a change in fluid density (the element mass must be conserved). Variations in the velocity in the direction of the velocity, as represented by the derivatives ∂u∂x, ∂υ∂y, and ∂w∂z, simply cause a linear deformation of the element in the sense that the shape of the element does not change. Cross derivatives, such as ∂u∂y and ∂υ∂x, will cause the element to rotate and generally to undergo an angular deformation, which changes the shape of the element. 6.1.3 Angular Motion and Deformation For simplicity we will consider motion in the x–y plane, but the results can be readily extended to the more general three-dimensional case. The velocity variation that causes rotation and angular deformation is illustrated in Fig. 6.4a. In a short time interval δt the line segments OA and OB will B u C ∂u u + __ δ x ∂x δy C C' A A' δy ∂u u + __ δ x ∂x u O B A δx O δx ) ∂__∂ux δ x ) δt (a) (b) www.Ebook777.com ■ Figure 6.3 Linear deformation of a fluid element. 6.1 281 Fluid Element Kinematics ) ∂∂__yu δ y) δt ∂u u + ___ δ y ∂y B B B' δβ δy δy v O A' ∂v v + ___ δ x ∂x u δx δα O A (a) δx A ) ∂∂__xv δ x) δ t ■ Figure 6.4 Angular motion and deformation of a fluid element. (b) rotate through the angles δα and δβ to the new positions OA′ and OB′, as is shown in Fig. 6.4b. The angular velocity of line OA, ωOA, is VIDEO V6.3 Shear deformation ωOA = lim δt→0 δα δt For small angles tan δα ≈ δα = (∂υ∂x) δx δt ∂υ = δt δx ∂x (6.10) so that ωOA = lim [ δt→0 ( ∂υ∂x) δt ∂υ ] = ∂x δt Note that if ∂υ∂x is positive, ωOA will be counterclockwise. Similarly, the angular velocity of the line OB is δβ δt→0 δt ωOB = lim and tan δβ ≈ δβ = ( ∂u∂y) δy δt ∂u = δt δy ∂y (6.11) so that ωOB = lim [ δt→0 Rotation of fluid particles is related to certain velocity gradients in the flow field. (∂u∂y) δt ∂u = ] δt ∂y In this instance if ∂u∂y is positive, ωOB will be clockwise. The rotation, ωz, of the element about the z axis is defined as the average of the angular velocities ωOA and ωOB of the two mutually perpendicular lines OA and OB.1 Thus, if counterclockwise rotation is considered to be positive, it follows that 1 ∂υ ∂u ωz = ( − ) 2 ∂x ∂y (6.12) Rotation of the fluid element about the other two coordinate axes can be obtained in a similar manner with the result that for rotation about the x axis 1 ∂w ∂υ ωx = ( − ) 2 ∂y ∂z (6.13) 1 ∂u ∂w ωy = ( − 2 ∂z ∂x ) (6.14) and for rotation about the y axis 1 With this definition ωz can also be interpreted to be the angular velocity of the bisector of the angle between the lines OA and OB. 282 Chapter 6 ■ Differential Analysis of Fluid Flow The three components, ωx, ωy, and ωz can be combined to give the rotation vector, ω, in the form ω = ω ˆi + ω ˆj + ω kˆ (6.15) x y z An examination of this result reveals that ω is equal to one-half the curl of the velocity vector. That is, 1 1 ω = 2 curl V = 2 𝛁 × V (6.16) since in Cartesian coordinates the vector operation 𝛁 × V is ˆi ˆj kˆ 1 1 𝛁×V= 2 2 | | ∂ ∂x u ∂ ∂y υ ∂ ∂z w 1 ∂w ∂υ 1 ∂u ∂w ˆ + 1 ∂υ − ∂u kˆ = ( − )ˆi + ( − j 2( ∂x ∂y ) 2 ∂y ∂z 2 ∂z ∂x ) Vorticity in a flow field is twice the fluid particle rotation. The vorticity, ζ, is defined as a vector that is twice the rotation vector; that is, ζ=2ω=𝛁×V (6.17) The use of the vorticity to describe the rotational characteristics of the fluid simply eliminates the 1 ( 2 ) factor associated with the rotation vector. The figure in the margin shows vorticity contours of the wing tip vortex flow shortly after an aircraft has passed. The lighter colors indicate stronger vorticity. (See also Fig. 4.3.) We observe from Eq. 6.12 that the fluid element will rotate about the z axis as an undeformed block (i.e., ωOA = −ωOB ) only when ∂u∂y = −∂υ∂x. Otherwise the rotation will be associated with an angular deformation. We also note from Eq. 6.12 that when ∂u∂y = ∂υ∂x the rotation around the z axis is zero. More generally if 𝛁 × V = 0, then the rotation (and the vorticity) are zero, and flow fields for which this condition applies are termed irrotational. We will find in Section 6.4 that the condition of irrotationality often greatly simplifies the analysis of complex flow fields. However, it is probably not immediately obvious why some flow fields would be irrotational, and we will need to examine this concept more fully in Section 6.4. Wing Engine EXAMPLE 6 .1 Vorticity GIVEN For a certain two-dimensional flow field the velocity FIND Is this flow irrotational? is given by the equation V = (x2 − y2 )iˆ − 2xyjˆ SOLUTION For an irrotational flow the rotation vector, ω, having the components given by Eqs. 6.12, 6.13, and 6.14 must be zero. For the prescribed velocity field u = x2 − y2  υ = −2xy  w = 0 zero, since by definition of two-dimensional flow u and υ are not functions of z, and w is zero. In this instance the condition for irrotationality simply becomes ωz = 0 or ∂υ∂x = ∂u∂y. The streamlines for the steady, two-dimensional flow of this example are shown in Fig. E6.1. (Information about how to calculate and therefore 1 ∂w ∂υ − )= 0 ωx = ( 2 ∂y ∂z 1 ∂u ∂w =0 ωy = ( − 2 ∂z ∂x ) y ∂u 1 1 ∂υ ωz = ( − ) = [(−2y) − (−2y)] = 0 2 ∂x ∂y 2 Thus, the flow is irrotational. (Ans) x COMMENTS It is to be noted that for a two-dimensional flow field (where the flow is in the x–y plane) ωx and ωy will always be ■ Figure E6.1 6.2 Conservation of Mass 283 lines OA and OB of Fig. 6.4 rotate with the same speed but in opposite directions. As shown by Eq. 6.17, the condition of irrotationality is equivalent to the fact that the vorticity, ζ, is zero or the curl of the velocity is zero. streamlines for a given velocity field is given in Sections 4.1.4 and 6.2.3.) It is noted that all of the streamlines (except for the one through the origin) are curved. However, because the flow is irrotational, there is no rotation of the fluid elements. That is, In addition to the rotation associated with the derivatives ∂u∂y and ∂υ∂x, it is observed from Fig. 6.4b that these derivatives can cause the fluid element to undergo an angular deformation, which results in a change in shape of the element. The change in the original right angle formed by the lines OA and OB is termed the shearing strain, δγ, and from Fig. 6.4b δγ = δα + δβ where δγ is considered to be positive if the original right angle is decreasing. The rate of change of δγ is called the rate of shearing strain or the rate of angular deformation and is commonly denoted with the symbol γ˙. The angles δα and δβ are related to the velocity gradients through Eqs. 6.10 and 6.11 so that γ˙ = lim δγ = lim (∂υ∂x) δt + (∂u∂y) δt ] δt δt→0 δt δt→0 [ and, therefore, γ˙ = ∂υ ∂u + ∂x ∂y (6.18) As we will learn in Section 6.8, the rate of angular deformation is related to a corresponding shearing stress that causes the fluid element to change in shape. From Eq. 6.18 we note that if ∂u∂y = −∂υ∂x, the rate of angular deformation is zero, and this condition corresponds to the case in which the element is simply rotating as an undeformed block (Eq. 6.12). In the remainder of this chapter we will see how the various kinematical relationships developed in this section play an important role in the development and subsequent analysis of the differential equations that govern fluid motion. 6.2 Conservation of mass requires that the mass of a system remain constant. Conservation of Mass As is discussed in Section 5.1, conservation of mass requires that the mass, M, of a system remain constant as the system moves through the flow field. In equation form this principle is expressed as DMsys =0 Dt We found it convenient to use the control volume approach for fluid flow problems, with the control volume representation of the conservation of mass written as ∂ ∂t ∫ ρ dV −+ cv ∫ ρ V ∙ nˆ dA = 0 (6.19) cs where the equation (commonly called the continuity equation) can be applied to a finite control volume (cv), which is bounded by a control surface (cs). The first integral on the left side of Eq. 6.19 represents the rate at which the mass within the control volume is changing, and the second integral represents the net rate at which mass is flowing out through the control surface (rate of mass outflow − rate of mass inflow). To obtain the differential form of the continuity equation, Eq. 6.19 is applied to an infinitesimal control volume. 6.2.1 Differential Form of Continuity Equation We will take as our control volume the small, stationary cubical element shown in Fig. 6.5a. At the center of the element the fluid density is ρ, and the velocity has components u, υ, and w. Since the element is small, the volume integral in Eq. 6.19 can be expressed as ∂ ∂t ∫ cv ρ dV −≈ ∂ρ δx δy δz ∂t (6.20) 284 Chapter 6 ■ Differential Analysis of Fluid Flow v ρ y w δx δ x δ y δz (ρu) __ ρu – ∂_____ ∂x 2 δy u y δz δz δ x δ y δz (ρu) __ ρu + ∂_____ ∂x 2 δx x z δy ρu x z (a) (b) ■ Figure 6.5 A differential element for the development of conservation of mass equation. The rate of mass flow through the surfaces of the element can be obtained by considering the flow in each of the coordinate directions separately. For example, in Fig. 6.5b flow in the x direction is depicted. If we let ρu represent the x component of the mass rate of flow per unit area at the center of the element, then on the right face ρu x+ (δx2) = ρu + ∂ (ρu) δx ∂x 2 (6.21) and on the left face ∂ (ρu) δx (6.22) ∂x 2 Note that we are really using a Taylor series expansion of ρu and neglecting higher order terms such as (δx) 2, (δx) 3, and so on. This is indicated by the figure in the margin. When the right sides of Eqs. 6.21 and 6.22 are multiplied by the area δy δz, the rate at which mass is crossing the right and left sides of the element is obtained as is illustrated in Fig. 6.5b. When these two expressions are combined, the net rate of mass flowing from the element through the two surfaces can be expressed as ρu x− (δx2) = ρu − ∂ (ρ u) δ x ∂x 2 (ρ u) Slope = ∂ ∂x ρu δx 2 δx 2 x ∂ (ρu) δx Net rate of mass = ρu + δy δz outflow in x direction [ ∂x 2 ] ∂ (ρu) δx ∂ (ρu) δy δz = δx δy δz (6.23) ∂x 2 ] ∂x For simplicity, only flow in the x direction has been considered in Fig. 6.5b, but, in general, there will also be flow in the y and z directions. An analysis similar to the one used for flow in the x direction shows that ∂ (ρυ) Net rate of mass = δx δy δz (6.24) outflow in y direction ∂y − [ρu − and ∂ (ρw) Net rate of mass δx δy δz = outflow in z direction ∂z (6.25) ∂ (ρu) ∂ (ρυ) ∂ (ρw) Net rate of = + + δx δy δz mass outflow [ ∂x ∂y ∂z ] (6.26) Thus, We now combine Eqs. 6.19, 6.20, and 6.26, cancel the common term δx δy δz, and take the limit as δx, δy, and δz, approach zero, shrinking the tiny control volume to a point. The result is the differential equation for conservation of mass The continuity equation is one of the fundamental equations of fluid mechanics. ∂ρ ∂ (ρu) ∂ (ρυ) ∂ (ρw) + + + =0 ∂t ∂x ∂y ∂z (6.27) As previously mentioned, this equation is also commonly referred to as the continuity equation. 6.2 Conservation of Mass 285 The continuity equation is one of the fundamental equations of fluid mechanics and, as expressed in Eq. 6.27, is valid for steady or unsteady flow, and compressible or incompressible flow. In vector notation, Eq. 6.27 can be written as ∂ρ + 𝛁 ∙ ρV = 0 ∂t (6.28) Two special cases are of particular interest. For steady compressible flow 𝛁 ∙ ρV = 0 or ∂ (ρu) ∂ (ρυ) ∂ (ρw) + + =0 ∂x ∂y ∂z For incompressible flow the continuity equation reduces to a simple relationship involving certain velocity gradients. (6.29) This follows since by definition ρ is not a function of time for steady flow, but could be a function of position. For incompressible flow the fluid density, ρ, is a constant throughout the flow field so that Eq. 6.28 becomes 𝛁∙V=0 (6.30) ∂u ∂υ ∂w + + =0 ∂x ∂y ∂z (6.31) or Equation 6.31 applies to both steady and unsteady incompressible flow. Note that Eq. 6.31 is the same as that obtained by setting the volumetric dilatation rate (Eq. 6.9) equal to zero. This result should not be surprising since both relationships are based on conservation of mass for incompressible fluids. However, the expression for the volumetric dilatation rate was developed from a system approach, whereas Eq. 6.31 was developed from a control volume approach. In the former case the deformation of a particular differential mass of fluid was studied, and in the latter case mass flow through a fixed differential volume was studied. E X AMPLE 6 .2 Continuity Equation GIVEN The velocity components for a certain incompressible, steady-flow field are FIND Determine the form of the z component, w, required to satisfy the continuity equation. u=x +y +z 2 2 2 υ = xy + yz + z w=? so that the required expression for ∂w∂z is ∂w = −2x − (x + z) = −3x − z ∂z SOLUTION Any physically possible velocity distribution must for an incompressible flow satisfy conservation of mass as expressed by the continuity equation ∂u ∂υ ∂w + + =0 ∂x ∂y ∂z For the given velocity distribution ∂u ∂υ = 2x  and   = x + z ∂x ∂y Integration with respect to z yields w = −3xz − z2 + f (x, y) 2 (Ans) COMMENT The third velocity component cannot be explicitly determined since the function f (x, y) can have any form and conservation of mass will still be satisfied. The specific form of this function will be governed by the flow field described by these velocity components—that is, some additional information is needed to completely determine w. 286 Chapter 6 ■ Differential Analysis of Fluid Flow y ^ eθ P vθ P r vr ^ e r ^ e z vz x θ z ■ Figure 6.6 The representation of velocity components in cylindrical polar coordinates. 6.2.2 Cylindrical Polar Coordinates For some problems, velocity components expressed in cylindrical polar coordinates will be convenient. For some problems it is more convenient to express the various differential relationships in cylindrical polar coordinates rather than Cartesian coordinates. As is shown in Fig. 6.6, with cylindrical coordinates a point is located by specifying the coordinates r, θ, and z. The coordinate r is the radial distance from the z axis, θ is the angle measured from a line parallel to the x axis (with counterclockwise taken as positive), and z is the coordinate along the z axis. The velocity components, as sketched in Fig. 6.6, are the radial velocity, υr, the tangential velocity, υθ, and the axial velocity, υz. Thus, the velocity at some arbitrary point P can be expressed as V = υrêr + υθêθ + υzêz (6.32) where êr, êθ, and êz are the unit vectors in the r, θ, and z directions, respectively, as are illustrated in Fig. 6.6. The use of cylindrical coordinates is particularly convenient when the boundaries of the flow system are cylindrical. Several examples illustrating the use of cylindrical coordinates will be given in succeeding sections in this chapter. The differential form of the continuity equation in cylindrical coordinates is ∂ (ρυz ) ∂ρ 1 ∂ (rρυr ) 1 ∂ (ρυθ ) + + + =0 r r ∂t ∂r ∂θ ∂z (6.33) This equation can be derived by following the same procedure used in the preceding section. For steady, compressible flow ∂ (ρυz ) 1 ∂ (rρυr ) 1 ∂ (ρυθ ) + + =0 r r ∂θ ∂r ∂z (6.34) For incompressible flow (steady or unsteady) ∂υz 1 ∂ (rυr ) 1 ∂υθ + + =0 r ∂r r ∂θ ∂z (6.35) 6.2.3 The Stream Function Steady, incompressible, plane, two-dimensional flow represents one of the simplest types of flow of practical importance. By plane, two-dimensional flow we mean that there are only two velocity components, such as u and υ, when the flow is considered to be in the x–y plane. For this flow the continuity equation, Eq. 6.31, reduces to ∂u ∂υ + =0 ∂x ∂y (6.36) 6.2 287 We have two variables, u and υ, to deal with, but they must be related in a special way as indicated by Eq. 6.36. This equation suggests that if we define a function ψ(x, y), called the stream function, which relates the velocities shown by the figure in the margin as Velocity components in a two-dimensional flow field can be expressed in terms of a stream function. u= ∂ψ ∂ψ   υ = − ∂y ∂x (6.37) then the continuity equation is identically satisfied. This conclusion can be verified by simply substituting the expressions for u and υ into Eq. 6.36 so that y V u= v=– ∂ψ ∂y x V v u Streamlines Conservation of Mass y x ∂ψ ∂ 2ψ ∂ 2ψ ∂ ∂ψ ∂ + − = − =0 ∂x( ∂y ) ∂y( ∂x ) ∂x ∂y ∂y ∂x ∂ψ ∂x Thus, whenever the velocity components are defined in terms of the stream function we know that conservation of mass will be satisfied. Of course, we still do not know what ψ(x, y) is for a particular problem, but at least we have simplified the analysis by having to determine only one unknown function, ψ(x, y), rather than the two functions, u(x, y) and υ(x, y). Another particular advantage of using the stream function is related to the fact that lines along which ψ is constant are streamlines. Recall from Section 4.1.4 that streamlines are lines in the flow field that are everywhere tangent to the velocities, as is illustrated in Fig. 6.7 and the figure in the margin. It follows from the definition of the streamline that the slope at any point along a streamline is given by dy υ = u dx The change in the value of ψ as we move from one point (x, y) to a nearby point (x + dx, y + dy) is given by the relationship: dψ = ∂ψ ∂ψ dx + dy = −υ dx + u dy ∂x ∂y Along a line of constant ψ we have dψ = 0 so that −υ dx + u dy = 0 and, therefore, along a line of constant ψ dy υ = u dx which is the defining equation for a streamline. Thus, if we know the function ψ(x, y), we can plot lines of constant ψ to provide the family of streamlines that are helpful in visualizing the flow pattern. V v Streamlines u y x ■ Figure 6.7 Velocity and velocity components along a streamline. 288 Chapter 6 ■ Differential Analysis of Fluid Flow ψ + dψ C dq ψ ψ2 q u dy A y B ψ1 – v dx x (a) (b) ■ Figure 6.8 The flow between two streamlines.  The change in the value of the stream function is related to the volume rate of flow. An infinite number of streamlines make up a particular flow field, since for each constant value assigned to ψ a streamline can be drawn. The actual numerical value associated with a particular streamline is not of particular significance, but the change in the value of ψ is related to the volume rate of flow. Consider two closely spaced streamlines, shown in Fig. 6.8a. The lower streamline is designated ψ and the upper one ψ + dψ. Let dq represent the volume rate of flow (per unit width perpendicular to the x–y plane) passing between the two streamlines. Note that flow never crosses streamlines, since by definition the velocity is tangent to the streamline. From conservation of mass we know that the inflow, dq, crossing the arbitrary surface AC of Fig. 6.8a must equal the net outflow through surfaces AB and BC. Thus, dq = u dy − υ dx or in terms of the stream function dq = ∂ψ ∂ψ dy + dx ∂y ∂x (6.38) The right-hand side of Eq. 6.38 is equal to dψ so that dq = dψ (6.39) Thus, the volume rate of flow, q, between two streamlines such as ψ1 and ψ2 of Fig. 6.8b can be determined by integrating Eq. 6.39 to yield ψ2 > ψ1 q q= ψ1 ∫ ψ2 dψ = ψ2 − ψ1 (6.40) ψ1 ψ2 < ψ1 q The relative value of ψ2 with respect to ψ1 determines the direction of flow, as shown by the figure in the margin. In cylindrical coordinates the continuity equation (Eq. 6.35) for incompressible, plane, twodimensional flow reduces to 1 ∂ (rυr ) 1 ∂υθ + =0 r ∂r r ∂θ ψ1 (6.41) and the velocity components, υr and υθ, can be related to the stream function, ψ(r, θ), through the equations υr = ∂ψ vr = _1r_ ∂θ vθ = – r V ∂ψ ∂r θ ∂ψ 1 ∂ψ   υθ = − r ∂θ ∂r (6.42) as shown by the figure in the margin. Substitution of these expressions for the velocity components into Eq. 6.41 shows that the continuity equation is identically satisfied. The stream function concept can be extended to axisymmetric flows, such as flow in pipes or flow around bodies of revolution, and to two-dimensional compressible flows. However, the concept is not applicable to general three-dimensional flows using a single stream function. 6.3 E X AMPLE 6 .3 The Linear Momentum Equation 289 Stream Function GIVEN The velocity components in a steady, incompressible, two-dimensional flow field are u = 2y FIND (a) Determine the corresponding stream function and (b) Show on a sketch several streamlines. Indicate the direction of flow along the streamlines. υ = 4x SOLUTION (a) From the definition of the stream function (Eqs. 6.37) u= y ψ=0 ψ=0 ∂ψ = 2y ∂y and υ=− ∂ψ = 4x ∂x x The first of these equations can be integrated to give ψ = y2 + f1 (x) where f1 (x) is an arbitrary function of x. Similarly from the second equation ψ = −2x2 + f2 (y) where f2 (y) is an arbitrary function of y. It now follows that in order to satisfy both expressions for the stream function ψ = −2x2 + y2 + C (Ans) where C is an arbitrary constant. Since the velocities are related to the derivatives of the stream function, an arbitrary constant can always be added to the function, and the value of the constant is actually of no consequence. Usually, for simplicity, we set C = 0 so that for this particular example the simplest form for the stream function is ψ = −2x2 + y2 (1) (Ans) Either answer indicated would be acceptable. (b) Streamlines can now be determined by setting ψ = constant and plotting the resulting curve. With the preceding expression for ψ (with C = 0) the value of ψ at the origin is zero, so that the equation of the streamline passing through the origin (the ψ = 0 streamline) is 0 = −2x2 + y2 6.3 ■ Figure E6.3 or y = ± √2x Other streamlines can be obtained by setting ψ equal to various constants. It follows from Eq. 1 that the equations of these streamlines (for ψ ≠ 0) can be expressed in the form y2 x2 =1 − ψ ψ2 which we recognize as the equation of a hyperbola. Thus, the streamlines are a family of hyperbolas with the ψ = 0 streamlines as asymptotes. Several of the streamlines are plotted in Fig. E6.3. Since the velocities can be calculated at any point, the direction of flow along a given streamline can be easily deduced. For example, υ = −∂ψ∂x = 4x so that υ > 0 if x > 0 and υ < 0 if x < 0. The direction of flow is indicated on the figure. COMMENT One could sketch the streamlines by drawing several velocity vectors and then sketching curves tangent to them. The Linear Momentum Equation To develop the differential momentum equations we can start with the linear momentum equation (Newton’s Second Law) DP F= (6.43) Dt sys | where F is the resultant force acting on a fluid mass, P is the linear momentum defined as P= ∫ sys V dm 290 Chapter 6 ■ Differential Analysis of Fluid Flow and the operator D( )Dt is the material derivative (see Section 4.2.1 and Eqs. 6.5–6.7). In the last chapter it was demonstrated how Eq. 6.43 in the form ∑ Fcontents of the = control volume ∂ ∂t ∫ Vρ dV −+ cv ∫ VρV ∙ nˆ dA (6.44) cs could be applied to a finite control volume to solve a variety of flow problems. To obtain the differential form of the linear momentum equation, we can either apply Eq. 6.43 to a differential system, consisting of a mass, δm, or apply Eq. 6.44 to an infinitesimal control volume, δ− V, which initially bounds the mass δm. It is probably simpler to use the system approach since application of Eq. 6.43 to the differential mass, δm, yields δF = D(V δm) Dt where δ F is the resultant force acting on δm. Using this system approach δm can be treated as a constant so that DV δ F = δm Dt But DVDt is the acceleration, a, of the element. Thus, δ F = δm a (6.45) which is simply Newton’s second law applied to the mass δm. This is the same result that would be obtained by applying Eq. 6.44 to an infinitesimal control volume (see Ref. 1). Before we can proceed, it is necessary to examine how the force δ F can be most conveniently expressed. 6.3.1 Description of Forces Acting on the Differential Element Both surface forces and body forces generally act on fluid particles. In general, two types of forces need to be considered: surface forces, which act on the surface of the differential element, and body forces, which are distributed throughout the element. For our purpose, the only body force, δ Fb, of interest is the weight of the element, which can be expressed as δ Fb = δm g (6.46) where g is the vector representation of the acceleration of gravity. In component form δFbx = δm gx δFby = δm gy δFbz = δm gz (6.47a) (6.47b) (6.47c) where gx, gy, and gz are the components of the acceleration of gravity vector in the x, y, and z directions, respectively. Surface forces act on the element as a result of its interaction with its surroundings. At any arbitrary location within a fluid mass, the force acting on a small area, δA, that lies in an arbitrary surface, can be represented by δ Fs, as is shown in Fig. 6.9. In general, δ Fs will be inclined with respect to the surface. The force δ Fs can be resolved into three components, δFn, δF1, and δF2, where δFn is normal to the area, δA, and δF1 and δF2 are parallel to the area and orthogonal to each other. The normal stress, σn , is defined as δFn σn = lim δA→0 δA δ Fn δ Fs δA δ F2 δ F1 Arbitrary surface ■ Figure 6.9 Components of force acting on an arbitrary differential area. 6.3 The Linear Momentum Equation C C' τ xz τ xz σxx τ xy y τ xy D D' 291 σxx B B' A A' (b) (a) x z ■ Figure 6.10 Double subscript notation for stresses. and the shearing stresses are defined as τ1 = lim δF1 δA τ2 = lim δF2 δA δA→0 and δA→0 Surface forces can be expressed in terms of the shear and normal stresses. We will use σ for normal stresses and τ for shearing stresses. The intensity of the force per unit area at a point in a body can thus be characterized by a normal stress and two shearing stresses, if the orientation of the area is specified. For purposes of analysis it is usually convenient to reference the area to the coordinate system. For example, for the rectangular coordinate system shown in Fig. 6.10 we choose to consider the stresses acting on planes parallel to the coordinate planes. On the plane ABCD of Fig. 6.10a, which is parallel to the y–z plane, the normal stress is denoted σxx and the shearing stresses are denoted as τxy and τxz. To easily identify the particular stress component we use a double subscript notation. The first subscript indicates the direction of the normal to the plane on which the stress acts, and the second subscript indicates the direction of the stress. Thus, normal stresses have repeated subscripts, whereas the subscripts for the shearing stresses are always different. It is also necessary to establish a sign convention for the stresses. We define the positive direction for the stress as the positive coordinate direction on the surfaces for which the outward normal is in the positive coordinate direction. This is the case illustrated in Fig. 6.10a where the outward normal to the area ABCD is in the positive x direction. The positive directions for σxx, τxy, and τxz are as shown in Fig. 6.10a. If the outward normal points in the negative coordinate direction, as in Fig. 6.10b for the area A′B′C′D′, then the stresses are considered positive if directed in the negative coordinate directions. Thus, the stresses shown in Fig. 6.10b are considered to be positive when directed as shown. Note that positive normal stresses are tensile stresses; that is, they tend to “stretch” the material. It should be emphasized that the state of stress at a point in a material is not completely defined by simply three components of a “stress vector.” This follows, since any particular stress vector depends on the orientation of the plane passing through the point. However, it can be shown that the normal and shearing stresses acting on any plane passing through a point can be expressed in terms of the stresses acting on three orthogonal planes passing through the point (Ref. 2). We now can express the surface forces acting on a small cubical element of fluid in terms of the stresses acting on the faces of the element as shown in Fig. 6.11. It is expected that in general the stresses will vary from point to point within the flow field. Thus, through the use of Taylor series expansions we will express the stresses on the various faces in terms of the corresponding stresses at the center of the element of Fig. 6.11 and their gradients in the coordinate directions. For simplicity only the forces in the x direction are shown. Note that the stresses must be multiplied by the area on which they act to obtain the force. Summing all these forces in the x direction yields ∂τyx ∂τzx ∂σxx δFsx = + + δx δy δz ( ∂x ∂y ∂z ) (6.48a) 292 Chapter 6 ■ Differential Analysis of Fluid Flow (τ yx ∂τ yx __ δy δ x δ z + ____ ∂y 2 ( δy (σ xx (τ (τ zx (σ ∂σ xx δ__x δ yδ z – ____ ∂x 2 ( zx + ∂τ z x __ δ z δx δy – ____ ∂z 2 ∂τz x __ δz δx δy ____ ∂z 2 ( ( xx + ∂σxx δ__x ____ δ y δz ∂x 2 ( δz δx y (τ yx ∂τ yx __ δy δ x δ z – ____ ∂y 2 ( ■ Figure 6.11 Surface x forces in the x direction acting on a fluid element. z for the resultant surface force in the x direction. In a similar manner the resultant surface forces in the y and z directions can be expressed as ∂τxy + δFsy = ( ∂x ∂τxz δFsz = ( + ∂x ∂σyy + ∂τzy δx δy δz ∂z ) ∂σzz + δx δy δz ∂y ∂z ) ∂y ∂τyz The resultant surface force can now be expressed as δ Fs = δFsxˆi + δFsy jˆ + δFszkˆ (6.48b) (6.48c) (6.49) and this force combined with the body force, δ Fb, yields the resultant force, δ F, acting on the differential mass, δm. That is, δ F = δ Fs + δ Fb. 6.3.2 Equations of Motion The expressions for the body and surface forces can now be used in conjunction with Eq. 6.45 to develop the equations of motion. In component form Eq. 6.45 can be written as δFx = δm ax δFy = δm ay δFz = δm az where δm = ρ δx δy δz, and the acceleration components are given by Eq. 6.3. It now follows (using Eqs. 6.47 and 6.48 for the forces on the element) that The motion of a fluid is governed by a set of nonlinear differential equations. ρgx + ρgy + ρgz + ∂τyx ∂τzx ∂σxx ∂u ∂u ∂u ∂u + + = ρ( + u +υ +w ) ∂x ∂y ∂z ∂t ∂x ∂y ∂z ∂τxy ∂x ∂τxz ∂x + + ∂σyy ∂y ∂τyz ∂y + + ∂τzy ∂z ∂σzz ∂z (6.50a) = ρ( ∂υ ∂υ ∂υ ∂υ +u +υ +w ) ∂t ∂x ∂y ∂z (6.50b) = ρ( ∂w ∂w ∂w ∂w +u +υ +w ) ∂t ∂x ∂y ∂z (6.50c) where the element volume δx δy δz cancels out before taking the limit, shrinking the tiny particle to a point. Equations 6.50, sometimes called the Cauchy Equations, are the general differential equations of motion for a fluid. In fact, they are applicable to any continuum (solid or fluid) in motion or at rest. However, before we can use the equations to solve specific problems, some additional information about the stresses must be obtained. Otherwise, we will have more unknowns (thirteen, all of the stresses and velocities and the density) than equations (including continuity, four). It should not be 6.4 Inviscid Flow 293 too surprising that the differential analysis of fluid motion is complicated. We are attempting to describe, in detail, complex fluid motion. 6.4 Inviscid Flow z p = –σzz y p = – σxx p = – σyy As is discussed in Section 1.6, shearing stresses develop in a moving fluid because of the viscosity and the rate of strain of the fluid. We know that for some common fluids, such as air and water, the viscosity is small, and therefore it seems reasonable to assume that under some circumstances we may be able to simply neglect the effect of viscosity (and thus shearing stresses), at least in regions where the strain rate is small. Flow fields in which the shearing stresses are assumed to be negligible are said to be inviscid, nonviscous, or frictionless. These terms are used interchangeably. As is discussed in Section 2.1, for fluids in which there are no shearing stresses the normal stress at a point is independent of direction—that is, σxx = σyy = σzz. In this instance we define the pressure, p, as the negative of the normal stress so that, as indicated by the figure in the margin, x −p = σxx = σyy = σzz The negative sign is used so that a compressive normal stress (which is what we expect in a fluid) will give a positive value for p. In Chapter 3 the inviscid flow concept was used in the development of the Bernoulli equation, and numerous applications of this important equation were considered. In this section we will again consider the Bernoulli equation and will show how it can be derived from the general equations of motion for inviscid flow. 6.4.1 Euler’s Equations of Motion For an inviscid flow in which all the shearing stresses are zero, and the normal stresses are replaced by −p, the general equations of motion (Eqs. 6.50) reduce to ρgx − ∂p ∂u ∂u ∂u ∂u = ρ( + u +υ +w ) ∂x ∂t ∂x ∂y ∂z (6.51a) ρgy − ∂p ∂υ ∂υ ∂υ ∂υ = ρ( + u +υ +w ) ∂y ∂t ∂x ∂y ∂z (6.51b) ρgz − ∂p ∂w ∂w ∂w ∂w = ρ( +u +υ +w ) ∂z ∂t ∂x ∂y ∂z (6.51c) These equations are commonly referred to as Euler’s equations of motion, named in honor of Leonhard Euler (1707–1783), a famous Swiss mathematician who pioneered work on the relationship between pressure and flow. In vector notation Euler’s equations can be expressed as Euler’s equations of motion apply to an inviscid flow field. ρg − 𝛁p = ρ [ ∂V + (V ∙ 𝛁)V] ∂t (6.52) Although Eqs. 6.51 are considerably simpler than the general equations of motion, Eqs. 6.50, they are still not amenable to a general analytical solution that would allow us to determine the pressure and velocity at all points within an inviscid flow field. (At least, when the continuity equation is included, the number of equations equals the number of unknowns.) The main difficulty in solving the equations arises from the nonlinear velocity terms (u ∂u∂x, υ ∂u∂y, etc.), which appear in the convective acceleration, or the (V ∙ 𝛁)V term. Because of these terms, Euler’s equations are nonlinear partial differential equations for which we do not have a general method of solving. However, under some circumstances we can use them to obtain useful information about inviscid flow fields. For example, as shown in the following section we can integrate Eq. 6.52 to obtain a relationship (the Bernoulli equation) between elevation, pressure, and velocity along a streamline. 6.4.2 The Bernoulli Equation In Section 3.2 the Bernoulli equation was derived by a direct application of Newton’s second law to a fluid particle moving along a streamline. In this section we will again derive this important 294 Chapter 6 ■ Differential Analysis of Fluid Flow Streamline ds Streamline z z y ds y x x ■ Figure 6.12 The notation for differential length along a streamline. equation, starting from Euler’s equations. Of course, we should obtain the same result since Euler’s equations simply represent a statement of Newton’s second law expressed in a general form that maintains the restriction of zero viscosity. We will restrict our attention to steady flow so Euler’s equation in vector form becomes ρg − 𝛁p = ρ(V ∙ 𝛁)V (6.53) z g y We wish to integrate this differential equation along some arbitrary streamline (Fig. 6.12) and select the coordinate system with the z axis vertical (with “up” being positive) so that, as shown by the figure in the margin, the acceleration of gravity vector can be expressed as g = −g𝛁z x g = −g∇z ^ ^ ^ = 0i + 0j − gk where g is the magnitude of the acceleration of gravity vector. Also, it will be convenient to use the vector identity 1 (V ∙ 𝛁)V = 2𝛁(V ∙ V) − V × (𝛁 × V) Equation 6.53 can now be written in the form ρ −ρg𝛁z − 𝛁p = 𝛁(V ∙ V) − ρV × (𝛁 × V) 2 and this equation can be rearranged to yield 𝛁p 1 + 𝛁(V 2 ) + g𝛁z = V × (𝛁 × V) ρ 2 We next take the dot product of each term with a differential length ds along a streamline (Fig. 6.12). Thus, 𝛁p 1 ∙ ds + 𝛁(V 2 ) ∙ ds + g𝛁z ∙ ds = [V × (𝛁 × V)] ∙ ds (6.54) ρ 2 V ×( ∇ × V) Since ds has a direction along the streamline, the vectors ds and V are parallel. However, as shown by the figure in the margin, the vector V × (𝛁 × V) is perpendicular to V (why?), so it follows that [V × (𝛁 × V)] ∙ ds = 0 ∇×V V Recall also that the dot product of the gradient of a scalar and a differential length gives the differential change in the scalar in the direction of the differential length. That is, with ds = dx ˆi + dy jˆ + dz kˆ we can write 𝛁p ∙ ds = (∂p∂x) dx + (∂p∂y)dy + ( ∂p∂z)dz = dp. Thus, Eq. 6.54 becomes dp 1 + d(V 2 ) + g dz = 0 (6.55) ρ 2 where the change in p, V, and z is along the streamline. Equation 6.55 can now be integrated to give Euler’s equations can be integrated to give the relationship among pressure, velocity, and elevation for inviscid flow. ∫ dp V2 + + gz = constant ρ 2 (6.56) which indicates that the sum of the three terms on the left side of the equation must remain a constant along a given streamline. Equation 6.56 is valid for both compressible and incompressible inviscid flows, but for compressible fluids the variation in ρ with p must be specified before the first term in Eq. 6.56 can be evaluated. 6.4 Inviscid Flow 295 For inviscid, incompressible fluids (commonly called ideal fluids) Eq. 6.56 can be written as p V2 + + gz = constant along a streamline ρ 2 (6.57) and this equation is the Bernoulli equation used extensively in Chapter 3. It is often convenient to write Eq. 6.57 between two points (1) and (2) along a streamline and to express the equation in the “head” form by dividing each term by g so that p1 p2 V 12 V 22 + z1 = + z2 + + γ γ 2g 2g (6.58) It should be again emphasized that the Bernoulli equation, as expressed by Eqs. 6.57 and 6.58, is restricted to the following: ■ ■ inviscid flow steady flow ■ ■ incompressible flow flow along a streamline Note that these are precisely the same restrictions that appeared in Chapter 3. You may want to go back and review some of the examples in Chapter 3 that illustrate the use of the Bernoulli equation. 6.4.3 Irrotational Flow The vorticity is zero in an irrotational flow field. If we make one additional assumption—that the flow is irrotational—the analysis of inviscid flow problems is further simplified. Recall from Section 6.1.3 that the rotation of a fluid element is equal 1 to 2 (𝛁 × V), and an irrotational flow field is one for which 𝛁 × V = 0 (i.e., the curl of velocity is zero). Since the vorticity, ζ, is defined as 𝛁 × V, it also follows that in an irrotational flow field the vorticity is zero. The concept of irrotationality may seem to be a rather strange condition for a flow field. Why would a flow field be irrotational? To answer this question we note that if 1 2 (𝛁 × V) = 0, then each of the components of this vector, as are given by Eqs. 6.12, 6.13, and 6.14, must be equal to zero. Since these components include the various velocity gradients in the flow field, the condition of irrotationality imposes specific relationships among these velocity gradients. For example, for rotation about the z axis to be zero, it follows from Eq. 6.12 that 1 ∂υ ∂u ωz = ( − ) = 0 2 ∂x ∂y and, therefore, ∂υ ∂u = ∂x ∂y (6.59) Similarly from Eqs. 6.13 and 6.14 ∂w ∂υ = (6.60) ∂y ∂z ∂u ∂w = (6.61) ∂z ∂x A general flow field would not satisfy these three equations. However, a uniform flow as is illustrated in Fig. 6.13 does. Since u = U (a constant), υ = 0, and w = 0, it follows that Eqs. 6.59, 6.60, and 6.61 are all satisfied. Therefore, a uniform flow field (in which there are no velocity gradients) is certainly an example of an irrotational flow. u = U (constant) v=0 y w=0 x z ■ Figure 6.13 Uniform flow in the x direction. 296 Chapter 6 ■ Differential Analysis of Fluid Flow Flow fields involving real fluids often include both regions of negligible shearing stresses and regions of significant shearing stresses. Uniform flows by themselves are not very interesting. However, many interesting and important flow problems include uniform flow in some part of the flow field. Two examples are shown in Fig. 6.14. In Fig. 6.14a a solid body is placed in a uniform stream of fluid. Far away from the body the flow remains uniform, and in this far region the flow is irrotational. In Fig. 6.14b, flow from a large reservoir enters a pipe through a streamlined entrance where the velocity distribution is essentially uniform. Thus, at the entrance the flow is irrotational. For an inviscid fluid there are no shearing stresses—the only forces acting on a fluid element are its weight and pressure forces. Since the weight acts through the element center of gravity, and the pressure acts in a direction normal to the element surface, neither of these forces can cause the element to rotate. Therefore, for an inviscid fluid, if some part of the flow field is irrotational, the fluid elements emanating from this region will not take on any rotation as they progress through the flow field. This phenomenon is illustrated in Fig. 6.14a in which fluid elements flowing far away from the body have irrotational motion, and as they flow around the body the motion remains irrotational except very near the boundary. Near the boundary the velocity changes rapidly from zero at the boundary (no-slip condition) to some relatively large value in a short distance from the boundary. This rapid change in velocity gives rise to a large velocity gradient normal to the boundary and produces significant shearing stresses, even though the viscosity is small. Of course if we had a truly inviscid fluid, the fluid would simply Boundary layer Uniform approach velocity Separation Wake Inviscid irrotational flow region (a) (b) Entrance region Uniform entrance velocity Fully developed region Inviscid, irrotational core Boundary layer (c) ■ Figure 6.14 Various regions of flow: (a) around bodies; (b) actual flow over a circular cylinder showing wake behind the cylinder; (c) through channels. 6.4 Inviscid Flow 297 “slide” past the boundary and the flow would be irrotational everywhere. But this is not the case for real fluids, so we will typically have a layer (usually relatively thin) near any fixed surface in a moving stream in which shearing stresses are not negligible. This layer is called the boundary layer. Outside the boundary layer the flow can be treated as an irrotational flow. Another possible consequence of the boundary layer is that the main stream may “separate” from the surface and form a wake downstream from the body (Fig. 6.14(c)). The wake would include a region of slow, perhaps randomly moving fluid. To completely analyze this type of flow it is necessary to consider both the inviscid, irrotational flow outside the boundary layer, and the viscous, rotational flow within the boundary layer and to somehow “match” these two regions. This type of analysis is considered in Chapter 9. As is illustrated in Fig. 6.14c, the flow in the entrance to a pipe may be uniform (if the entrance is streamlined) and thus will be irrotational. In the central core of the pipe the flow remains irrotational for some distance. However, a boundary layer will develop along the wall and grow in thickness until it fills the pipe. Thus, for this type of internal flow there will be an entrance region in which there is a central irrotational core, followed by a so-called fully developed region in which viscous forces are dominant. The flow is rotational in the fully developed region. This type of internal flow is considered in detail in Chapter 8. The two preceding examples are intended to illustrate the possible applicability of irrotational flow to some “real” flow problems and to indicate some limitations of the irrotationality assumption. We proceed to develop some useful equations based on the assumptions of inviscid, incompressible, irrotational flow, with the admonition to use caution when applying the equations. 6.4.4 The Bernoulli Equation for Irrotational Flow In the development of the Bernoulli equation in Section 6.4.2, Eq. 6.54 was integrated along a streamline. This restriction was imposed so the right side of the equation could be set equal to zero; that is, [V × (𝛁 × V)] ∙ ds = 0 (since ds is parallel to V). However, for irrotational flow, 𝛁 × V = 0 , so the right side of Eq. 6.54 is zero regardless of the direction of ds. We can now follow the same procedure used to obtain Eq. 6.55, where the differential changes dp, d(V 2 ), and dz can be taken in any direction. Integration of Eq. 6.55 again yields ∫ dp V2 + + gz = constant ρ 2 (6.62) where for irrotational flow the constant is the same throughout the flow field. Thus, for incompressible, irrotational flow the Bernoulli equation can be written as p1 p2 V 21 V 22 + z1 = + z2 + + γ γ 2g 2g The Bernoulli equation can be applied between any two points in an irrotational flow field. (6.63) between any two points in the flow field. Equation 6.63 is exactly the same form as Eq. 6.58 but is not limited to application along a streamline. However, Eq. 6.63 is restricted to ■ ■ inviscid flow steady flow ■ ■ incompressible flow irrotational flow It may be worthwhile to review the use and misuse of the Bernoulli equation for rotational flow as is illustrated in Example 3.17. 6.4.5 The Velocity Potential For an irrotational flow the velocity gradients are related through Eqs. 6.59, 6.60, and 6.61. In this case the velocity components can be expressed in terms of a scalar function ϕ(x, y, z, t) as u= ∂ϕ ∂ϕ ∂ϕ   υ =   w = ∂x ∂y ∂z (6.64) 298 Chapter 6 ■ Differential Analysis of Fluid Flow where ϕ is called the velocity potential. Direct substitution of these expressions for the velocity components into Eqs. 6.59, 6.60, and 6.61 will verify that a velocity field defined by Eqs. 6.64 is indeed irrotational. In vector form, Eqs. 6.64 can be written as V = 𝛁ϕ (6.65) so that for an irrotational flow the velocity is expressible as the gradient of a scalar function ϕ. The velocity potential is a consequence of the irrotationality of the flow field, whereas the stream function is a consequence of conservation of mass (see Section 6.2.3). It is to be noted, however, that the velocity potential can be defined for a general three-dimensional flow, whereas a single stream function is restricted to two-dimensional flows. For an incompressible fluid we know from conservation of mass that 𝛁∙V=0 Inviscid, incompressible, irrotational flow fields are governed by Laplace’s equation and are called potential flows. ∇2 ϕ = 0 Streamlines ∇2 ϕ ≠ 0 and therefore for incompressible, irrotational flow (with V = 𝛁ϕ) it follows that 𝛁2 ϕ = 0 (6.66) where 𝛁 ( ) = 𝛁 ∙ 𝛁 ( ) is the Laplacian operator. In Cartesian coordinates 2 ∂ 2ϕ ∂ 2ϕ ∂ 2ϕ + + =0 ∂x2 ∂y2 ∂z2 This differential equation arises in many different areas of engineering and physics and is called Laplace’s equation. Thus, inviscid, incompressible, irrotational flow fields are governed by Laplace’s equation. This type of flow is commonly called a potential flow. To complete the mathematical formulation of a given problem, boundary conditions have to be specified. These are usually velocities specified on the boundaries of the flow field of interest. It follows that if the potential function can be determined, then the velocity at all points in the flow field can be determined from Eq. 6.64, and the pressure at all points can be determined from the Bernoulli equation (Eq. 6.63). Although the concept of the velocity potential is applicable to both steady and unsteady flow, we will confine our attention to steady flow. Potential flows, governed by Eqs. 6.64 and 6.66, are irrotational flows. That is, the vorticity is zero throughout. If vorticity is present (e.g., boundary layer, wake), then the flow cannot be described by Laplace’s equation. The figure in the margin illustrates a flow in which the vorticity is not zero in two regions—the separated region behind the bump and the boundary layer next to the solid surface. This is discussed in detail in Chapter 9. For some problems it will be convenient to use cylindrical coordinates, r, θ, and z. In this coordinate system the gradient operator is 𝛁( ) = ∂( ) ∂( ) 1 ∂( ) êr + êθ + êz r ∂θ ∂r ∂z (6.67) so that 𝛁ϕ = Vorticity contours ∂ϕ ∂ϕ 1 ∂ϕ êr + êθ + êz r ∂θ ∂r ∂z (6.68) where ϕ = ϕ(r, θ, z). Since V = υrêr + υθêθ + υzêz (6.69) it follows for an irrotational flow (with V = 𝛁ϕ) υr = ∂ϕ ∂ϕ 1 ∂ϕ   υθ =   υz = r ∂θ ∂r ∂z (6.70) Also, Laplace’s equation in cylindrical coordinates is ∂ϕ ∂ 2ϕ 1 ∂ 1 ∂ 2ϕ r + + =0 r ∂r ( ∂r ) r2 ∂θ2 ∂z2 (6.71) 6.4 E X AMPLE 6 .4 Inviscid Flow 299 Velocity Potential and Inviscid Flow Pressure GIVEN The two-dimensional flow of a nonviscous, incom- pressible fluid in the vicinity of the 90° corner of Fig. E6.4a is described by the stream function that is, there is no difference in elevation between points (1) and (2). FIND ψ = 2r 2 sin 2θ (a) where ψ has units of m2/s when r is in meters. Assume the fluid density is 103 kgm3 and the x–y plane is horizontal— Determine, if possible, the corresponding velocity potential. (b) If the pressure at point (1) on the wall is 30 kPa, what is the pressure at point (2)? y y Streamline ( ψ = constant) Equipotential line ( φ = constant) (2) r 0.5 m θ (1) x 1m x (a) (b) α α (c) ■ Figure E6.4 SOLUTION (a) The radial and tangential velocity components can be obtained from the stream function as (see Eq. 6.42) υr = 1 ∂ψ = 4r cos 2θ r ∂θ and therefore by integration ϕ = 2r2 cos 2θ + f1 (θ) where f1 (θ) is an arbitrary function of θ. Similarly υθ = and υθ = − ∂ψ = −4r sin 2θ ∂r (1) 1 ∂ϕ = −4r sin 2θ r ∂θ and integration yields ϕ = 2r 2 cos 2θ + f2 (r) (2) where f2 (r) is an arbitrary function of r. To satisfy both Eqs. 1 and 2, the velocity potential must have the form Since υr = ∂ϕ ∂r it follows that ∂ϕ = 4r cos 2θ ∂r ϕ = 2r2 cos 2θ + C (Ans) where C is an arbitrary constant. As is the case for stream functions, the specific value of C is not important, and it is customary to let C = 0 so that the velocity potential for this corner flow is ϕ = 2r2 cos 2θ (Ans) 300 Chapter 6 ■ Differential Analysis of Fluid Flow COMMENT In the statement of this problem, it was implied by the wording “if possible” that we might not be able to find a corresponding velocity potential. The reason for this concern is that we can always define a stream function for two-dimensional flow, but the flow must be irrotational if there is a corresponding velocity potential. Thus, the fact that we were able to determine a velocity potential means that the flow is irrotational. Several streamlines and lines of constant ϕ are plotted in Fig. E6.4b. These two sets of lines are orthogonal. The reason why streamlines and lines of constant ϕ are always orthogonal is explained in Section 6.5. (b) Since we have an irrotational flow of a nonviscous, incompressible fluid, the Bernoulli equation can be applied between any two points. Thus, between points (1) and (2) with no elevation change p1 p2 V 12 V 22 = + + γ γ 2g 2g This result indicates that the square of the velocity at any point depends only on the radial distance, r, to the point. Note that the constant, 16, has units of s−2. Thus, V 12 = (16 s −2 ) (1 m) 2 = 16 m2s2 and V 22 = (16 s −2 ) (0.5 m) 2 = 4 m2s2 Substitution of these velocities into Eq. 3 gives p2 = 30 × 103 Nm2 + = 36 kPa 103 kgm3 (16 m2s2 − 4 m2s2 ) 2 (Ans) COMMENT The stream function used in this example could also be expressed in Cartesian coordinates as ψ = 2r 2 sin 2θ = 4r 2 sin θ cos θ     or ψ = 4xy or p2 = p 1 + ρ 2 (V 1 − V 22 ) 2 (3) since x = r cos θ and y = r sin θ. However, in the cylindrical polar form the results can be generalized to describe flow in the vicinity of a corner of angle α (see Fig. E6.4c) with the equations Since V 2 = υ 2r + υ 2θ V = (4r cos 2θ) + (−4r sin 2θ) 2 2 = 16r (cos 2θ + sin 2θ) 2 2 πθ α ϕ = Ar πα cos πθ α and it follows that for any point within the flow field 2 ψ = Ar πα sin 2   = 16r 2 6.5 where A is a constant. Some Basic, Plane Potential Flows A major advantage of Laplace’s equation is that it is a linear partial differential equation. Since it is linear, various solutions can be added to obtain other solutions—that is, if ϕ1 (x, y, z) and ϕ2 (x, y, z) are two solutions to Laplace’s equation, then ϕ3 = ϕ1 + ϕ2 is also a solution. The practical implication of this result is that if we have certain basic solutions we can combine them to obtain more complicated and interesting solutions. In this section several basic velocity potentials, which describe some relatively simple flows, will be determined. In the next section these basic potentials will be combined to represent more complicated flows. For simplicity, only plane (two-dimensional) flows will be considered. In this case, by using Cartesian coordinates For potential flow, basic solutions can be simply added to obtain more complicated solutions. y V u= v= ∂φ ∂x x vr = V θ ∂ϕ ∂ϕ   υ = ∂x ∂y (6.72) ∂ϕ 1 ∂ϕ   υθ = r ∂θ ∂r (6.73) u= or by using cylindrical coordinates υr = ∂φ ∂r ∂φ vθ = _1r_ ∂θ r ∂φ ∂y as shown by the figure in the margin. Since we can define a stream function for plane flow, we can also let ∂ψ ∂ψ u=   υ = − (6.74) ∂y ∂x 6.5 Some Basic, Plane Potential Flows 301 or υr = ∂ψ 1 ∂ψ   υθ = − r ∂θ ∂r (6.75) where the stream function was previously defined in Eqs. 6.37 and 6.42. We know that by defining the velocities in terms of the stream function, conservation of mass is identically satisfied. If we now impose the condition of irrotationality, it follows from Eq. 6.59 that ∂u ∂υ = ∂y ∂x and in terms of the stream function ∂ψ ∂ ∂ψ ∂ = − ∂y( ∂y ) ∂x( ∂x ) or ∂ 2ψ ∂ 2ψ + =0 ∂x2 ∂y2 Thus, for a plane irrotational flow we can use either the velocity potential or the stream function— both must satisfy Laplace’s equation in two dimensions. It is apparent from these results that the velocity potential and the stream function are somehow related. We have previously shown that lines of constant ψ are streamlines; that is, dy dx | = along ψ=constant υ u (6.76) The change in ϕ as we move from one point (x , y) to a nearby point (x + dx, y + dy) is given by the relationship y a b a dϕ = b a_ × – b_ = –1 b a ( ) Along a line of constant ϕ we have dϕ = 0 so that dy dx x Δψ Streamwise acceleration ∂ϕ ∂ϕ dx + dy = u dx + υ dy ∂x ∂y | =− along ϕ=constant u υ (6.77) A comparison of Eqs. 6.76 and 6.77 shows that lines of constant ϕ (called equipotential lines) are orthogonal to lines of constant ψ (streamlines) at all points where they intersect. (Recall that two lines are orthogonal if the product of their slopes is −1, as illustrated by the figure in the margin.) For any potential flow field a “flow net” can be drawn that consists of a family of streamlines and equipotential lines. The flow net is useful in visualizing flow patterns and can be used to obtain graphical solutions by sketching in streamlines and equipotential lines and adjusting the lines until the lines are approximately orthogonal at all points where they intersect. An example of a flow net is shown in Fig. 6.15. Velocities can be estimated from the flow net, since the velocity is inversely proportional to the streamline spacing, as shown by the figure in the margin. Thus, for example, from Fig. 6.15 we can see that the velocity near the inside corner will be higher than the velocity along the outer part of the bend. 6.5.1 Uniform Flow The simplest plane flow is one for which the streamlines are all straight and parallel, and the magnitude of the velocity is constant. This type of flow is called a uniform flow. For example, consider a uniform flow in the positive x direction as is illustrated in Fig. 6.16a. In this instance, u = U and υ = 0, and in terms of the velocity potential Streamwise deceleration ∂ϕ ∂ϕ = U   =0 ∂x ∂y 302 Chapter 6 ■ Differential Analysis of Fluid Flow Equipotential line ( φ = constant) d d1 > d V V1 < V d V1 V V2 d2 < d V2 > V Streamline (ψ = constant) ■ Figure 6.15 Flow net for a 90° bend. (Adapted from Ref. 3.) These two equations can be integrated to yield ϕ = Ux + C where C is an arbitrary constant, which can be set equal to zero. Thus, for a uniform flow in the positive x direction ϕ = Ux (6.78) The corresponding stream function can be obtained in a similar manner, since ∂ψ ∂ψ = U   =0 ∂y ∂x and, therefore, ψ = Uy (6.79) These results can be generalized to provide the velocity potential and stream function for a uniform flow at an angle α with the x axis, as in Fig. 6.16b. For this case ϕ = U(x cos α + y sin α) (6.80) ψ = U(y cos α − x sin α) (6.81) and 6.5.2 Source and Sink Consider a fluid flowing radially outward from a line through the origin perpendicular to the x–y plane as is shown in Fig. 6.17. Let m be the volume rate of flow emanating from the line (per unit length), and therefore to satisfy conservation of mass (2πr)υr = m or υr = y = y U = 1 = 2 = 3 = 4 m 2πr 1 = U ■ Figure 6.16 Uniform flow: x (b) 4 φ = φ2 φ = φ1 φ = φ2 x (a) 3 = α φ = φ1 2 = (a) in the x direction; (b) in an arbitrary direction, α. 6.5 Some Basic, Plane Potential Flows 303 y ψ = constant φ = constant vr r θ x ■ Figure 6.17 The streamline pattern for a source. A source or sink represents a purely radial flow. Also, since the flow is a purely radial flow, υθ = 0, the corresponding velocity potential can be obtained by integrating the equations ∂ϕ m 1 ∂ϕ =    =0 r ∂θ ∂r 2πr It follows that ϕ= vr ~ _1r_ vr r m ln r 2π (6.82) If m is positive, the flow is radially outward, and the flow is considered to be a source flow. If m is negative, the flow is toward the origin, and the flow is considered to be a sink flow. The flowrate, m, is the strength of the source or sink. As shown by the figure in the margin, at the origin where r = 0 the velocity becomes infinite, which is of course physically impossible. Thus, sources and sinks do not really exist in real flow fields, and the line representing the source or sink is a mathematical singularity in the flow field. However, some real flows can be approximated at points away from the origin by using sources or sinks. Also, the velocity potential representing this hypothetical flow can be combined with other basic velocity potentials to approximately describe some real flow fields. This idea is further discussed in Section 6.6. The stream function for the source can be obtained by integrating the relationships υr = ∂ψ 1 ∂ψ m =   υθ = − =0 r ∂θ 2πr ∂r to yield ψ= m θ 2π (6.83) It is apparent from Eq. 6.83 that the streamlines (lines of ψ = constant) are radial lines, and from Eq. 6.82 that the equipotential lines (lines of ϕ = constant) are concentric circles centered at the origin. E X AMPLE 6 .5 Potential Flow—Sink GIVEN A nonviscous, incompressible fluid flows between y wedge-shaped walls into a small opening as shown in Fig. E6.5. The velocity potential (in ft2s) that approximately describes this flow is ϕ = −2 ln r vr R FIND Determine the volume rate of flow (per unit length) into the opening. π _ 6 r θ x ■ Figure E6.5 304 Chapter 6 ■ Differential Analysis of Fluid Flow SOLUTION COMMENT Note that the radius R is arbitrary since the flowrate crossing any curve between the two walls must be the same. The negative sign indicates that the flow is toward the opening, that is, in the negative radial direction. The units are ft2/s because q is the flowrate per unit width. The components of velocity are υr = ∂ϕ 2 1 ∂ϕ = −   υθ = =0 r r ∂θ ∂r which indicates we have a purely radial flow. The flowrate per unit width, q, crossing the arc of length Rπ6 can thus be obtained by integrating the expression q= ∫ 0 π6 υr R dθ = − ∫ 0 π6 2 ( R ) R dθ π = − = −1.05 ft2s 3 (Ans) 6.5.3 Vortex We next consider a flow field in which the streamlines are concentric circles—that is, we interchange the velocity potential and stream function for the source. Thus, let A vortex represents a flow in which the streamlines are concentric circles. vθ r (6.84) ψ = −K ln r (6.85) and where K is a constant. In this case the streamlines are concentric circles as are illustrated in Fig. 6.18, with υr = 0 and υθ = vθ ~ _1_ r ϕ = Kθ ∂ψ 1 ∂ϕ K =− = r ∂θ r ∂r (6.86) This result indicates that the tangential velocity varies inversely with the distance from the origin, as shown by the figure in the margin, with a singularity occurring at r = 0 (where the velocity becomes infinite). It may seem strange that this vortex motion is irrotational (and it is since the flow field is described by a velocity potential). However, it must be recalled that rotation refers to the orientation of a fluid element and not the path followed by the element. Thus, for an irrotational vortex, if a pair of small sticks were placed in the flow field at location A, as indicated in Fig. 6.19a, the sticks would rotate as they move to location B. One of the sticks, the one that is aligned along the streamline, would follow a circular path and rotate in a counterclockwise direction. The other stick y ψ = constant vθ r θ x φ = constant ■ Figure 6.18 The streamline pattern for a vortex. 6.5 B _ vθ ~ 1 r A Some Basic, Plane Potential Flows A 305 vθ ~ r B r r (a) (b) ■ Figure 6.19 Motion of fluid element from A to B: (a) for irrotational (free) vortex; (b) for rotational (forced) vortex. Vortex motion can be either rotational or irrotational. would rotate in a clockwise direction due to the nature of the flow field—that is, the part of the stick nearest the origin moves faster than the opposite end. Although both sticks are rotating, the average angular velocity of the two sticks is zero and the flow is irrotational. If the fluid were rotating as a rigid body, such that υθ = K1r where K1 is a constant, then sticks similarly placed in the flow field would rotate as is illustrated in Fig. 6.19b. This type of vortex motion is rotational and cannot be described with a velocity potential. The rotational vortex is commonly called a forced vortex, whereas the irrotational vortex is usually called a free vortex. The swirling motion of the water as it drains from a bathtub is similar to that of a free vortex, whereas the motion of a liquid contained in a tank that is rotated about its axis with angular velocity ω corresponds to a forced vortex. A combined vortex is one with a forced vortex as a central core and a velocity distribution corresponding to that of a free vortex outside the core. Thus, for a combined vortex υθ = ωr  r ≤ r0 (6.87) K   r > r0 r (6.88) and υθ = where K and ω are constants and r0 corresponds to the radius of the central core. The pressure distribution in both the free and forced vortex was previously considered in Example 3.3. (See Fig. E6.6a for an approximation of this type of flow.) THE WIDE WORLD OF FLUIDS Some hurricane facts One of the most interesting, yet poten- pressure is at a minimum and may be 10% less than standard tially devastating, naturally occurring fluid flow phenomena is a atmospheric pressure. This low pressure creates strong down- hurricane. Broadly speaking, a hurricane is a rotating mass of air drafts of dry air from above. Hurricanes are classified into five circulating around a low-pressure central core. In some respects categories based on their wind speeds: the motion is similar to that of a free vortex. The Caribbean and Gulf of Mexico experience the most hurricanes, with the official Category one—74–95 mph hurricane season being from June 1 to November 30. Hurricanes Category two—96–110 mph are usually 300 to 400 miles wide and are structured around a central eye in which the air is relatively calm. The eye is sur- Category three—111–130 mph rounded by an eye wall, which is the region of strongest winds Category four—131–155 mph and precipitation. As one goes from the eye wall to the eye, Category five—greater than 155 mph the wind speeds decrease sharply and within the eye the air is relatively calm and clear of clouds. However, in the eye the (See Problem 6.46.) 306 Chapter 6 ■ Differential Analysis of Fluid Flow ds Arbitrary curve C V ■ Figure 6.20 The notation for determining circulation around closed curve C. A mathematical concept commonly associated with vortex motion is that of circulation. The circulation, Γ, is defined as the line integral of the tangential component of the velocity taken around a closed curve in the flow field. In equation form, Γ can be expressed as ∫ Γ = ˇ○ V ∙ ds (6.89) C where the integral sign means that the integration is taken around a closed curve, C, in the counterclockwise direction, and ds is a differential length along the curve as is illustrated in Fig. 6.20. For an irrotational flow, V = 𝛁ϕ so that V ∙ ds = 𝛁ϕ ∙ ds = dϕ and, therefore, ∫ Γ = ˇ○ dϕ = 0 C This result indicates that for an irrotational flow the circulation will generally be zero. (Chapter 9 has further discussion of circulation in real flows.) However, if there are singularities enclosed within the curve the circulation may not be zero. For example, for the free vortex with υθ = Kr the circulation around the circular path of radius r shown in Fig. 6.21 is Γ= The numerical value of the circulation may depend on the particular closed path considered. ∫ 0 2π K (r dθ) = 2πK r which shows that the circulation is nonzero and the constant K = Γ2π. However, for irrotational flows the circulation around any path that does not include a singular point will be zero. This can be easily confirmed for the closed path ABCD of Fig. 6.21 by evaluating the circulation around that path. The velocity potential and stream function for the free vortex are commonly expressed in terms of the circulation as ϕ= VIDEO V6.4 Vortex in a beaker Γ θ 2π (6.90) Γ ln r 2π (6.91) and ψ=− The concept of circulation is often useful when evaluating the forces developed on bodies immersed in moving fluids. This application will be considered in Section 6.6.3. vθ ds B C D r A dθ θ ■ Figure 6.21 Circulation around various paths in a free vortex. 6.5 E X AMPLE 6 .6 Some Basic, Plane Potential Flows 307 Potential Flow—Free Vortex GIVEN A liquid drains from a large tank through a small opening as illustrated in Fig. E6.6a. A vortex forms whose velocity distribution away from the tank opening can be approximated as that of a free vortex having a velocity potential ϕ= Γ θ 2π FIND Determine an expression relating the surface shape to the strength of the vortex as specified by the circulation Γ. SOLUTION Since the free vortex represents an irrotational flow field, the Bernoulli equation ■ Figure E6.6a p2 p1 V 12 V 22 + z1 = + z2 + + γ γ 2g 2g z V 22 V 12 = zs + 2g 2g r p = patm can be written between any two points. If the points are selected at the free surface, p1 = p2 = 0, so that zs x (1) (1) y (2) where the free surface elevation, zs, is measured relative to a datum passing through point (1) as shown in Fig. E6.6b. The velocity is given by the equation 1 ∂ϕ Γ = r ∂θ 2πr We note that far from the origin at point (1), V1 = υθ ≈ 0 so that Eq. 1 becomes υθ = Γ2 zs = − 2 2 8π r g (Ans) ■ Figure E6.6b COMMENT The negative sign indicates that the surface falls as the origin is approached as shown in Fig. E6.6. This solution is not valid very near the origin since the predicted velocity becomes excessively large as the origin is approached. which is the desired equation for the surface profile. 6.5.4 Doublet A doublet is formed by an appropriate source–sink pair. The final, basic potential flow to be considered is one that is formed by combining a source and sink in a special way. Consider the equal strength, source–sink pair of Fig. 6.22. The combined stream function for the pair is ψ=− y P r2 θ2 θ Source r r1 θ1 Sink a m (θ − θ2 ) 2π 1 a x ■ Figure 6.22 The combination of a source and sink of equal strength located along the x axis. 308 Chapter 6 ■ Differential Analysis of Fluid Flow which can be rewritten as 2πψ tan θ1 − tan θ2 = tan(θ tan(− − θ ) = 1 2 m ) 1 + tan θ1 tan θ2 (6.92) From Fig. 6.22 it follows that tan θ1 = r sin θ r cos θ − a tan θ2 = r sin θ r cos θ + a and These results substituted into Eq. 6.92 give 2πψ 2ar sin θ = 2 tan(− m ) r − a2   y so that ψ=− Source Sink x m 2ar sin θ tan−1( 2 2π r − a2 ) The figure in the margin shows typical streamlines for this flow. For small values of the distance a ψ=− m 2ar sin θ mar sin θ =− 2 2 2 2π r − a π(r − a2 )   A doublet is formed by letting a source and sink approach one another. (6.93) (6.94)   since the tangent of an angle approaches the value of the angle for small angles. The so-called doublet is formed by letting the source and sink approach one another (a → 0) while increasing the strength m (m → ∞) so that the product maπ remains constant. In this case, since r(r2 − a2 ) → 1r, Eq. 6.94 reduces to ψ=− K sin θ r (6.95) where K, a constant equal to maπ, is called the strength of the doublet. The corresponding velocity potential for the doublet is ϕ= K cos θ r (6.96) Plots of lines of constant ψ reveal that the streamlines for a doublet are circles through the origin tangent to the x axis as shown in Fig. 6.23. Just as sources and sinks are not physically realistic entities, neither are doublets. However, the doublet when combined with other basic potential flows provides a useful representation of some flow fields of practical interest. For example, we will y x ■ Figure 6.23 Streamlines for a doublet. 6.6 309 Superposition of Basic, Plane Potential Flows ■ Table 6.1 Summary of Basic, Plane Potential Flows Description of Flow Field Uniform flow at angle α with the x axis (see Fig. 6.16b) Velocity Potential Stream Function ϕ = U(x cos α + y sin α) ψ = U( y cos α − x sin α) Source or sink (see Fig. 6.17) m > 0 source m < 0 sink m ln r 2π ϕ= m θ 2π ψ= Velocity Componentsa u = U cos α υ = U sin α υr = m 2πr υθ = 0 Free vortex (see Fig. 6.18) Γ > 0 counterclockwise motion Γ < 0 clockwise motion ϕ= Γ θ 2π ψ=− Γ ln r 2π υr = 0 υθ = Doublet (see Fig. 6.23) ϕ= K cos θ r ψ=− K sin θ r Γ 2πr K cos θ r2 K sin θ υθ = − r2 υr = − a Velocity components are related to the velocity potential and stream function through the relationships: u= ∂ϕ ∂x = ∂ψ ∂y   υ = ∂ϕ ∂y =− ∂ψ ∂x   υr = ∂ϕ ∂r = ∂ψ 1 ∂ψ 1 ∂ϕ   υθ = =− . r ∂θ r ∂θ ∂r determine in Section 6.6.3 that the combination of a uniform flow and a doublet can be used to represent the flow around a circular cylinder. Table 6.1 provides a summary of the pertinent equations for the basic, plane potential flows considered in the preceding sections. 6.6 Superposition of Basic, Plane Potential Flows As was discussed in the previous section, potential flows are governed by Laplace’s equation, which is a linear partial differential equation. It therefore follows that the various basic velocity potentials and stream functions can be combined to form new potentials and stream functions. (Why is this true?) Whether such combinations yield useful results remains to be seen. It is to be noted that any streamline in an inviscid flow field can be considered as a solid boundary, since the conditions along a solid boundary and a streamline are the same—that is, there is no flow through the boundary or the streamline. Thus, if we can combine some of the basic velocity potentials or stream functions to yield a streamline that corresponds to a particular body shape of interest, that combination can be used to describe in detail the flow around that body. This method of solving some interesting flow problems, commonly called the method of superposition, is illustrated in the following three sections. 6.6.1 Source in a Uniform Stream—Half-Body Flow around a half-body is obtained by the addition of a source to a uniform flow. Consider the superposition of a source and a uniform flow as shown in Fig. 6.24a. The resulting stream function is ψ = ψuniform flow + ψsource = Ur sin θ + m θ 2π (6.97) 310 Chapter 6 ■ Differential Analysis of Fluid Flow y U ψ = π bU Stagnation point Stagnation point r πb θ x πb Source b b (a) (b) ■ Figure 6.24 The flow around a half-body: (a) superposition of a source and a uniform flow; (b) replacement of streamline ψ = πbU with solid boundary to form a half-body. and the corresponding velocity potential is ϕ = Ur cos θ + V6.5 Half-body VIDEO m ln r 2π (6.98) It is clear that at some point along the negative x axis the velocity due to the source will just cancel that due to the uniform flow and a stagnation point will be created. For the source alone m υr = 2πr so that the stagnation point will occur at x = −b where m U= 2πb or m b= (6.99) 2πU The value of the stream function at the stagnation point can be obtained by evaluating ψ at r = b and θ = π, which yields from Eq. 6.97 m 2 Since m2 = πbU (from Eq. 6.99) it follows that the equation of the streamline passing through the stagnation point is ψstagnation = For inviscid flow, a streamline can be replaced by a solid boundary. πbU = Ur sin θ + bUθ or r= Increasing m U = constant b(π − θ) sin θ (6.100) where θ can vary between 0 and 2π. A plot of this streamline is shown in Fig. 6.24b. If we replace this streamline with a solid boundary, as indicated in the figure, then it is clear that this combination of a uniform flow and a source can be used to describe the flow around the nose of a streamlined body placed in a uniform stream. The body is open at the downstream end and thus is called a half-body. Other streamlines in the flow field can be obtained by setting ψ = constant in Eq. 6.97 and plotting the resulting equation. A number of these streamlines are shown in Fig. 6.24b. Although the streamlines inside the body are shown, they are actually of no interest in this case, since we are concerned with the flow field outside the body. It should be noted that the singularity in the flow field (the source) occurs inside the body, and there are no singularities in the flow field of interest (outside the body). The width of the half-body asymptotically approaches 2πb = m/U. Thus, as shown by the figure in the margin, for a given free stream velocity, U, the width of the half-body increases as the source strength. This follows from Eq. 6.100, which can be written as y = b(π − θ) 6.6 Superposition of Basic, Plane Potential Flows 311 so that as θ → 0 or θ → 2π the half-width approaches ±bπ. With the stream function (or velocity potential) known, the velocity components at any point can be obtained. For the half-body, using the stream function given by Eq. 6.97, υr = 1 ∂ψ m = U cos θ + r ∂θ 2πr and υθ = − ∂ψ = −U sin θ ∂r Thus, the square of the magnitude of the velocity, V, at any point is V 2 = υ2r + υ2θ = U 2 + m 2 Um cos θ +( πr 2πr ) and since b = m2πU b b2 V 2 = U 2(1 + 2 cos θ + 2 ) r r (6.101)   With the velocity known, the pressure at any point can be determined from the Bernoulli equation, which can be written between any two points in the flow field since the flow is irrotational. Thus, applying the Bernoulli equation between a point far from the body, where the pressure is p0 and the velocity is U, and some arbitrary point with pressure p and velocity V, it follows that 1 1 p0 + 2 ρU 2 = p + 2 ρV 2 For a potential flow the fluid is allowed to slip past a fixed solid boundary. (6.102) where elevation changes have been neglected. Equation 6.101 can now be substituted into Eq. 6.102 to obtain the pressure at any point in terms of the reference pressure, p0, and the upstream velocity, U. This relatively simple potential flow provides some useful information about the flow around the front part of a streamlined body, such as a bridge pier or strut placed in a uniform stream. An important point to be noted is that the velocity tangent to the surface of the body is not zero; that is, the fluid “slips” by the boundary. This result is a consequence of neglecting viscosity, the fluid property that causes real fluids to stick to the boundary, thus creating a “no-slip” condition. All potential flows differ from the flow of real fluids in this respect and do not accurately represent the velocity very near the boundary. However, outside this very thin boundary layer the velocity distribution will generally correspond to that predicted by potential flow theory if flow separation does not occur. (See Section 9.2.6.) Also, the pressure distribution along the surface will closely approximate that predicted from the potential flow theory, since the boundary layer is thin and there is little opportunity for the pressure to vary through the thin layer. In fact, as discussed in more detail in Chapter 9, the pressure distribution obtained from potential flow theory is used in conjunction with viscous flow theory to determine the nature of flow within the boundary layer. E X AMPLE 6 .7 Potential Flow—Half-body GIVEN A 40-mi/hr wind blows toward a hill arising from a 40 mi/hr plain that can be approximated with the top section of a half-body as illustrated in Fig. E6.7a. The height of the hill approaches 200 ft as shown. Assume an air density of 0.00238 slugs/ft3. FIND (2) (1) (a) What is the magnitude of the air velocity at a point on the hill directly above the origin [point (2)]? (b) What is the elevation of point (2) above the plain and what is the difference in pressure between point (1) on the plain far from the hill and point (2)? 200 ft y r (3) b ■ Figure E6.7a θ x 312 Chapter 6 ■ Differential Analysis of Fluid Flow SOLUTION (a) and The velocity is given by Eq. 6.101 as b b2 V 2 = U 2(1 + 2 cos θ + 2 ) r r V2 = (47.4 mihr) (   At point (2), θ = π2, and since this point is on the surface (Eq. 6.100), r= it follows that p1 − p2 = b(π − θ) πb = sin θ 2 5280 ftmi = 69.5 fts 3600 shr ) (1) Thus, (0.00238 slugsft3 ) [(69.5 fts) 2 − (58.7 fts) 2] 2 + (0.00238 slugsft3 ) (32.2 fts2 ) (100 ft − 0 ft) = 9.31 lbft2 = 0.0647 psi b2 V 22 = U 2 1 + [ (πb2) 2 ] COMMENTS This result indicates that the pressure on the 4 π2 ) and the magnitude of the velocity at (2) for a 40 mihr approaching wind is = U 2(1 + 4 12 V2 = 1 + 2 (40 mihr) = 47.4 mihr ( π) (Ans) (b) The elevation at (2) above the plain is given by Eq. 1 as πb 2 Since the height of the hill approaches 200 ft and this height is equal to πb, it follows that y2 = y2 = 200 ft = 100 ft 2 hill at point (2) is slightly lower than the pressure on the plain at some distance from the base of the hill with a 0.0533-psi difference due to the elevation increase and a 0.0114-psi difference due to the velocity increase. By repeating the calculations for various values of the upstream wind speed, U, the results shown in Fig. E6.7b are obtained. Note that as the wind speed increases, the pressure difference increases from the calm conditions of p1 − p2 = 0.0533 psi. The maximum velocity along the hill surface does not occur at point (2) but farther up the hill at θ = 63°. At this point Vsurface = 1.26U. The minimum velocity (V = 0) and maximum pressure occur at point (3), the stagnation point. 0.14 (Ans) 0.12 From the Bernoulli equation (with the y axis the vertical axis) so that p1 − p 2 = ρ 2 (V − V 12 ) + γ(y2 − y1 ) 2 2 and with 0.10 p1 – p2, psi p2 p1 V 12 V 22 + y1 = + y2 + + γ γ 2g 2g 0.08 0.06 (40 mph, 0.0647 psi) 0.04 0.02 0 5280 ftmi = 58.7 fts V1 = (40 mihr) ( 3600 shr ) (Ans) 0 20 60 40 80 100 U, mph ■ Figure E6.7b 6.6.2 Rankine Ovals The half-body described in the previous section is a body that is “open” at one end. To simulate the flow around a closed body, a source and a sink of equal strength can be combined with a uniform flow as shown in Fig. 6.25a. The stream function for this combination is m (θ − θ2 ) 2π 1 (6.103) m (ln r1 − ln r2 ) 2π (6.104) ψ = Ur sin θ − and the velocity potential is ϕ = Ur cos θ − 6.6 y U r2 θ2 θ r Stagnation point r1 θ1 +m –m a Sink a Stagnation point ψ=0 x Source 313 Superposition of Basic, Plane Potential Flows h h a a ℓ ( a) ℓ (b) ■ Figure 6.25 The flow around a Rankine oval: (a) superposition of source–sink pair and a uniform flow; (b) replacement of streamline ψ = 0 with solid boundary to form Rankine oval. As discussed in Section 6.5.4, the stream function for the source–sink pair can be expressed as in Eq. 6.93, and, therefore, Eq. 6.103 can also be written as m 2ar sin θ ψ = Ur sin θ − tan−1( 2 2π r − a2 ) or   ψ = Uy − Rankine ovals are formed by combining a source and sink with a uniform flow. 2ay m tan−1( 2 2π x + y2 − a2 ) (6.105) The corresponding streamlines for this flow field are obtained by setting ψ = constant. If several of these streamlines are plotted, it will be discovered that the streamline ψ = 0 forms a closed body as is illustrated in Fig. 6.25b. We can think of this streamline as forming the surface of a body of length 2ℓ and width 2h placed in a uniform stream. The streamlines inside the body are of no practical interest and are not shown. Note that since the body is closed, all of the flow emanating from the source flows into the sink. These bodies have an oval shape and are termed Rankine ovals. Note that the Rankine oval is not an ellipse; it is more blunt. Stagnation points occur at the upstream and downstream ends of the body, as are indicated in Fig. 6.25b. These points can be located by determining where along the x axis the velocity is zero. The stagnation points correspond to the points where the uniform velocity, the source velocity, and the sink velocity all combine to give a zero velocity. The locations of the stagnation points depend on the value of a, m, and U. The body half-length, ℓ (the value of  x that gives V = 0 when y = 0), can be expressed as 12 ma ℓ =( + a2) (6.106) πU or Large Ua/m 12 ℓ m + 1) =( a πUa (6.107) The body half-width, h, can be obtained by determining the value of y where the y axis intersects the ψ = 0 streamline. Thus, from Eq. 6.105 with ψ = 0, x = 0, and y = h, it follows that h= h2 − a2 2πUh tan m 2a (6.108) or h πUa h 1 h 2 = [( ) − 1] tan [2( a m ) a] 2 a Small Ua/m (6.109) Equations 6.107 and 6.109 show that both ℓa and ha are functions of the dimensionless parameter, πUam. Although for a given value of Uam the corresponding value of ℓa can be determined directly from Eq. 6.107, ha must be determined by a trial-and-error solution of Eq. 6.109. A large variety of body shapes with different length to width ratios can be obtained by using different values of Uam, as shown by the figure in the margin. As this parameter becomes large, flow around a long slender body is described, whereas for small values of the parameter, flow 314 Chapter 6 ■ Differential Analysis of Fluid Flow around a more blunt shape is obtained. Downstream from the point of maximum body width, the surface pressure increases with distance along the surface. This condition (called an adverse pressure gradient) typically leads to separation of the flow from the surface, resulting in a large lowpressure wake on the downstream side of the body. Separation is not predicted by potential theory (which simply indicates a symmetrical flow). This is illustrated by the figure in the margin for an extreme blunt shape, showing the differences between potential and viscous flow. Therefore, the potential solution for the Rankine ovals will give a reasonable approximation of the velocity outside the thin, viscous boundary layer and the pressure distribution on the front part of the body only. Potential Flow © S. T. Thoroddsen & Stanford University 6.6.3 Flow around a Circular Cylinder Viscous Flow © Stanford University, with permission A doublet combined with a uniform flow can be used to represent flow around a circular cylinder. As was noted in the previous section, when the distance between the source–sink pair approaches zero, the shape of the Rankine oval becomes more blunt and in fact approaches a circular shape. At the same time, the size of the body gets smaller and smaller. Since the doublet described in Section 6.5.4 was developed by letting a source–sink pair approach one another, it might be expected that a uniform flow in the positive x direction combined with a doublet could be used to represent flow around a circular cylinder. This combination gives for the stream function K sin θ (6.110) ψ = Ur sin θ − r and for the velocity potential K cos θ ϕ = Ur cos θ + (6.111) r In order for the stream function to represent flow around a circular cylinder, it is necessary that ψ = constant for r = a, where a is the radius of the cylinder. Since Eq. 6.110 can be written as V6.6 Circular cylinder VIDEO ψ = (U − it follows that ψ = 0 for r = a if K r sin θ r 2) K =0 a2 which indicates that the doublet strength, K, must be equal to Ua2. Thus, the stream function for flow around a circular cylinder can be expressed as U− ψ = Ur (1 − V6.7 Ellipse a2 sin θ r 2) (6.112) a2 cos θ r 2) (6.113) VIDEO and the corresponding velocity potential is ϕ = Ur (1 + A sketch of the streamlines for this flow field is shown in Fig. 6.26. The velocity components can be obtained from either Eq. 6.112 or 6.113 as ∂ϕ 1 ∂ψ a2 = = U (1 − 2 ) cos θ r ∂θ ∂r r (6.114) ∂ψ 1 ∂ϕ a2 =− = −U(1 + 2 ) sin θ r ∂θ ∂r r (6.115) υr = and 2 υθ = υ θs U On the surface of the cylinder (r = a) it follows from Eq. 6.114 and 6.115 that υr = 0 and 1 υθs = −2U sin θ 0 0 ±π 2 θ ±π As shown by the figure in the margin, the maximum velocity occurs at the top and bottom of the cylinder (θ = ±π2) and has a magnitude of twice the upstream velocity, U. As we move away from the cylinder along the ray θ = π2 the velocity varies, as is illustrated in Fig. 6.26. 6.6 Superposition of Basic, Plane Potential Flows 315 U 2U Ψ=0 r θ a ■ Figure 6.26 The flow around a circular cylinder. The pressure distribution on the cylinder surface is obtained from the Bernoulli equation. The pressure distribution on the cylinder surface is obtained from the Bernoulli equation written from a point far from the cylinder where the pressure is p0 and the velocity is U so that 1 1 p0 + 2 ρU 2 = ps + 2 ρυ2θs where ps is the surface pressure. Elevation changes are neglected. Since υθs = −2U sin θ, the surface pressure can be expressed as 1 ps = p0 + 2 ρU 2 (1 − 4 sin2 θ) VIDEO V6.8 Circular cylinder with separation (6.116) A comparison of this theoretical, symmetrical pressure distribution expressed in dimensionless form with a typical measured distribution is shown in Fig. 6.27. This figure clearly reveals that only on the upstream part of the cylinder is there approximate agreement between the potential flow and the experimental results. Because of the viscous boundary layer that develops on the cylinder, the main flow separates from the surface of the cylinder, leading to the large difference between the theoretical, frictionless fluid solution and the experimental results on the downstream side of the cylinder (see Chapter 9). The resultant force (per unit length) developed on the cylinder can be determined by integrating the pressure over the surface. From Fig. 6.28 it can be seen that Fx = − ∫ 2π ps cos θ a dθ (6.117) 0 VIDEO V6.9 Forces on suspended ball 3 U β 2 1 ps – p0 _______ 1 _ ρU2 2 0 Experimental –1 –2 Theoretical (inviscid) –3 0 30 60 90 β (deg) 120 150 180 ■ Figure 6.27 A comparison of theoretical (inviscid) pressure distribution on the surface of a circular cylinder with typical experimental distribution. 316 Chapter 6 ■ Differential Analysis of Fluid Flow y ps Fy a dθ θ x Fx ■ Figure 6.28 The notation for determining lift and drag on a circular cylinder. VIDEO V6.10 Potential and viscous flow Potential theory incorrectly predicts that the drag on a cylinder is zero. and Fy = − ∫ 2π 0 (6.118) ps sin θ a dθ where Fx is the drag (force parallel to direction of the uniform flow) and Fy is the lift (force perpendicular to the direction of the uniform flow). Substitution for ps from Eq. 6.116 into these two equations, and subsequent integration, reveals that Fx = 0 and Fy = 0. These results indicate that both the drag and lift as predicted by potential theory for a fixed cylinder in a uniform stream are zero. Since the pressure distribution is symmetrical around the cylinder, this is not really a surprising result. However, we know from experience that there is a significant drag developed on a cylinder when it is placed in a moving fluid. This discrepancy is known as d’Alembert’s paradox. The paradox is named after Jean le Rond d’Alembert (1717–1783), a French mathematician and philosopher, who first showed that the drag on bodies immersed in inviscid fluids is zero. It was not until the latter part of the nineteenth century and the early part of the twentieth century that the role viscosity plays in the steady fluid motion was understood and d’Alembert’s paradox explained (see Section 9.1). EXAMPLE 6 .8 Potential Flow—Cylinder GIVEN When a circular cylinder is placed in a uniform FIND stream, a stagnation point is created on the cylinder as is shown in Fig. E6.8a. If a small hole is located at this point, the stagnation pressure, pstag, can be measured and used to determine the approach velocity, U. (a) U Show how pstag and U are related. (b) If the cylinder is misaligned by an angle α (Fig. E6.8b), but the measured pressure is still interpreted as the stagnation pressure, determine an expression for the ratio of the true velocity, U, to the predicted velocity, U′. Plot this ratio as a function of α for the range −20° ≤ α ≤ 20°. U U y p0 a r θ x α β β a Stagnation point (a) (b) (d) 1.5 1.4 U 1.3 __ U' 1.2 1.1 1.0 –20° –10° 0° α (c) 10° 20° ■ Figure E6.8 6.6 Superposition of Basic, Plane Potential Flows 317 SOLUTION (a) The velocity at the stagnation point is zero, so the Bernoulli equation written between a point on the stagnation streamline upstream from the cylinder and the stagnation point gives If we now write the Bernoulli equation between a point upstream of the cylinder and the point on the cylinder where r = a, θ = α, it follows that pstag p0 U2 = + γ γ 2g p0 + 2 ρU 2 = pα + 2 ρ(−2U sin α) 2 1 and, therefore, Thus, 12 2 U = [ ( pstag − p0 ) ] ρ 1 pα − p0 = 2 ρU 2 (1 − 4 sin2 α) (Ans) (2) 1 Since pstag − p0 = 2 ρU 2 it follows from Eqs. 1 and 2 that COMMENT A measurement of the difference between the pressure at the stagnation point and the upstream pressure can be used to measure the approach velocity. This is, of course, the same result that was obtained in Section 3.5 for Pitot-static tubes. (b) If the direction of the fluid approaching the cylinder is not known precisely, it is possible that the cylinder is misaligned by some angle, α. In this instance the pressure actually measured, pα, will be different from the stagnation pressure, but if the misalignment is not recognized the predicted approach velocity, U′, would still be calculated as 12 2 U′ = [ ( pα − p0 ) ] ρ Thus, pstag − p0 12 U(true) =( pα − p0 ) U′ (predicted) 1 (1) The velocity on the surface of the cylinder, υθ, where r = a, is obtained from Eq. 6.115 as U(true) = (1 − 4 sin2 α)−12 U′ (predicted) (Ans) This velocity ratio is plotted as a function of the misalignment angle α in Fig. E6.8c. COMMENT It is clear from these results that significant errors can arise if the stagnation pressure tap is not aligned with the stagnation streamline. As is discussed in Section 3.5, if two additional, symmetrically located holes are drilled on the cylinder, as are illustrated in Fig. E6.8d, the correct orientation of the cylinder can be determined. The cylinder is rotated until the pressures in the two symmetrically placed holes are equal, thus indicating that the center hole coincides with the stagnation streamline. For β = 30° the pressure at the two holes theoretically corresponds to the upstream pressure, p0. With this orientation a measurement of the difference in pressure between the center hole and the side holes can be used to determine U. υθ = −2U sin θ An additional, interesting potential flow can be developed by adding a free vortex to the stream function or velocity potential for the flow around a cylinder. In this case ψ = Ur(1 − a2 Γ sin θ − ln r 2) 2π r (6.119) a2 Γ cos θ + θ 2π r 2) (6.120) and ϕ = Ur(1 + where Γ is the circulation. We note that the circle r = a will still be a streamline (and thus can be replaced with a solid cylinder), since the streamlines for the added free vortex are all circular. However, the tangential velocity, υθ, on the surface of the cylinder (r = a) now becomes υθs = − Flow around a rotating cylinder is approximated by the addition of a free vortex. ∂ψ ∂r | = −2U sin θ + r=a Γ 2πa (6.121) This type of flow field could be approximately created by placing a rotating cylinder in a uniform stream. Because of the presence of viscosity in any real fluid, the fluid in contact with the rotating cylinder would rotate with the same velocity as the cylinder, and the resulting flow field would resemble that developed by the combination of a uniform flow past a cylinder and a free vortex. 318 Chapter 6 ■ Differential Analysis of Fluid Flow Γ=0 Γ <1 4π Ua (a) (b) Stagnation point Γ =1 4π Ua Γ >1 4π Ua (c) (d) ■ Figure 6.29 The location of stagnation points on a circular cylinder: (a) without circulation; (b, c, d ) with circulation. A variety of streamline patterns can be developed, depending on the vortex strength, Γ. For example, from Eq. 6.121 we can determine the location of stagnation points on the surface of the cylinder. These points will occur at θ = θstag where υθ = 0 and therefore from Eq. 6.121 sin θstag = Γ 4πUa (6.122) If Γ = 0, then θstag = 0 or π—that is, the stagnation points occur at the front and rear of the cylinder as are shown in Fig. 6.29a. However, for −1 ≤ Γ4πUa ≤ 1, the stagnation points will occur at some other location on the surface as illustrated in Figs. 6.29b,c. If the absolute value of the parameter Γ4πUa exceeds 1, Eq. 6.122 cannot be satisfied, and the stagnation point is located away from the cylinder as shown in Fig. 6.29d. The force per unit length developed on the cylinder can again be obtained by integrating the differential pressure forces around the circumference as in Eqs. 6.117 and 6.118. For the cylinder with circulation, the surface pressure, ps, is obtained from the Bernoulli equation (with the surface velocity given by Eq. 6.121) p0 + 1 1 Γ 2 ρU 2 = ps + ρ(−2U sin θ + 2 2 2πa ) or ps = p0 + 1 2Γ sin θ Γ2 ρU 2 1 − 4 sin2 θ + − 2 2 2 ( 2 πaU 4π a U ) (6.123) Equation 6.123 substituted into Eq. 6.117 for the drag, and integrated, again yields Fx = 0 Potential flow past a cylinder with circulation gives zero drag but nonzero lift. That is, even for the rotating cylinder no force in the direction of the uniform flow is developed. However, use of Eq. 6.123 with the equation for the lift, Fy (Eq. 6.118), yields Fy = −ρUΓ (6.124) Thus, for the cylinder with circulation, lift is developed equal to the product of the fluid density, the upstream velocity, and the circulation. The negative sign means that if U is positive (in the positive x direction) and Γ is positive (a free vortex with counterclockwise rotation), the direction of the Fy is downward. Of course, if the cylinder is rotated in the clockwise direction (Γ < 0), the direction of Fy would be upward. This can be seen by studying the surface pressure distribution (Eq. 6.123), which is plotted in Fig. 6.30 for two situations. One has Γ/4pUa = 0, which corresponds to no rotation of the cylinder. The other has Γ/4pUa = −0.25, which corresponds to clockwise rotation of the cylinder. With no rotation, the flow is symmetrical both top to bottom and front to back on 6.7 319 Other Aspects of Potential Flow Analysis 1 Γ/4πUa = 0 Γ/4πUa = –0.25 0 –1 ps – p0 _______ 1 2 –2 2 ρU –3 U –4 θ –5 Top half –6 0 90 Bottom half 180 θ, deg 270 360 ■ Figure 6.30 Pressure distribution on a circular cylinder with and without rotation. the cylinder. With rotation, the flow is symmetrical front to back but not top to bottom. In this case the two stagnation points [i.e., ( ps − p0)/(rU2 /2) = 1] are located on the bottom of the cylinder, and the average pressure on the top half of the cylinder is less than that on the bottom half. The result is an upward lift force. It is this force acting in a direction perpendicular to the direction of the approach velocity that causes baseballs and golf balls to curve when they spin as they are propelled through the air. The development of this lift on rotating bodies is called the Magnus effect. (See Section 9.4 for further comments.) Although Eq. 6.124 was developed for a cylinder with circulation, it gives the lift per unit length for any two-dimensional object of any cross-sectional shape placed in a uniform, inviscid stream. The circulation is determined around any closed curve containing the body. The generalized equation relating lift to fluid density, velocity, and circulation is called the Kutta–Joukowski law and is commonly used to determine the lift on airfoils (see Section 9.4.2 and Refs. 2–6). THE WIDE WORLD OF FLUIDS A sailing ship without sails A sphere or cylinder spinning rigging with two vertical cylinders that were 50 ft high and 9 ft in about its axis when placed in an airstream develops a force at diameter. The cylinders looked like smokestacks on the ship. right angles to the direction of the airstream. This phenomenon Their spinning motion was developed by 45-hp motors. The com- is commonly referred to as the Magnus effect and is responsible bination of a wind and the rotating cylinders created a force for the curved paths of baseballs and golf balls. Another lesser- (Magnus effect) to push the ship forward. The ship, named the known application of the Magnus effect was proposed by a Baden Baden, made a successful voyage across the Atlantic, German physicist and engineer, Anton Flettner, in the 1920s. arriving in New York Harbor on May 9, 1926. Although the feasibil- Flettner’s idea was to use the Magnus effect to make a ship ity of the rotor-ship was clearly demonstrated, it proved to be move. To demonstrate the practicality of the “rotor-ship” he pur- less efficient and practical than more conventional vessels and chased a sailing schooner and replaced the ship’s masts and the idea was not pursued. (See Problem 6.62.) 6.7 Other Aspects of Potential Flow Analysis In the preceding section the method of superposition of basic potentials and stream functions has been used to obtain detailed descriptions of irrotational flow around certain body shapes immersed in a uniform stream. For the cases considered, two or more of the basic potentials were combined, and the question is asked: What kind of flow does this combination represent? This approach is relatively simple and does not require the use of advanced mathematical techniques. It is, however, restrictive in its general applicability. It does not allow us to specify a priori the body shape and then determine the velocity potential or stream function that describes the flow around the particular body. 320 Chapter 6 ■ Differential Analysis of Fluid Flow Potential flow solutions are always approximate because the fluid is assumed to be frictionless. VIDEO 6.8 V6.11 Potential flow Determining the velocity potential or stream function for a given body shape is a much more complicated problem. It is possible to extend the idea of superposition by considering a distribution of sources and sinks, vortexes, or doublets, which when combined with a uniform flow can describe the flow around bodies of arbitrary shape. Singularities may be distributed along an axis or over a surface. Techniques are available to determine the required distribution to give a prescribed body shape. Also, for plane potential flow problems it can be shown that complex variable theory (the use of real and imaginary numbers) can be effectively used to obtain solutions to a great variety of important flow problems. There are, of course, numerical techniques that can be used to solve not only plane two-dimensional problems but also the more general three-dimensional problems. Since potential flow is governed by Laplace’s equation, any procedure that is available for solving this equation can be applied to the analysis of irrotational flow of frictionless fluids. Potential flow theory is an old and well-established discipline within the general field of fluid mechanics. The interested reader can find many detailed references on this subject, including Refs. 2–7 given at the end of this chapter. An important point to remember is that regardless of the particular technique used to obtain a solution to a potential flow problem, the solution remains approximate because of the fundamental assumption of a frictionless fluid. Thus, “exact” solutions based on potential flow theory represent, at best, only approximate solutions to real fluid problems. The applicability of potential flow theory to real fluid problems has been alluded to in a number of examples considered in the previous section. As a rule of thumb, potential flow theory will usually provide a reasonable approximation in those circumstances when we are dealing with a low viscosity fluid moving at a relatively high velocity, in regions of the flow field in which the flow is accelerating. Under these circumstances we generally find that the effect of viscosity is confined to the thin boundary layer that develops at a solid boundary. Outside the boundary layer, the velocity distribution and the pressure distribution are closely approximated by the potential flow solution. However, in those regions of the flow field in which the flow is decelerating (for example, in the rearward portion of a bluff body or in the expanding region of a conduit), the pressure near a solid boundary will increase in the direction of flow. This so-called adverse pressure gradient can lead to flow separation, a phenomenon that causes dramatic changes in the flow field which are generally not accounted for by potential theory. However, as discussed in Chapter 9, in which boundary layer theory is developed, it is found that potential flow theory is used to obtain the appropriate pressure distribution that can then be combined with the boundary layer approximation to the viscous flow equations to obtain solutions near the boundary (and also to predict separation). The general differential equations that describe viscous fluid behavior and some simple solutions to these equations are considered in the remaining sections of this chapter. Viscous Flow To incorporate viscous effects into the differential analysis of fluid motion we must return to the previously derived general equations of motion, Eqs. 6.50. Since these equations include both stresses and velocities, there are more unknowns than equations and, therefore, before proceeding it is necessary to establish a relationship between the stresses and velocities. 6.8.1 Stress–Deformation Relationships For incompressible Newtonian fluids it is known that the stresses are linearly related to the rates of deformation and can be expressed in Cartesian coordinates as (for normal stresses) σxx = −p + 2μ ∂u ∂x (6.125a) σyy = −p + 2μ ∂υ ∂y (6.125b) σzz = −p + 2μ ∂w ∂z (6.125c) 6.8 Viscous Flow 321 (for shearing stresses) z σzz σyy σxx x p=– 1 (σ + σyy + σzz) 3 xx For Newtonian fluids, stresses are linearly related to the rate of strain. y ∂u ∂υ τxy = τyx = μ( + ) (6.125d) ∂y ∂x ∂υ ∂w τyz = τzy = μ( + (6.125e) ∂z ∂y ) ∂w ∂u τzx = τxz = μ( + ) (6.125f) ∂x ∂z where p is the pressure, the negative of the average of the three normal stresses; that is, as indicated 1 by the figure in the margin, −p = ( 3 )(σxx + σyy + σzz ). For viscous fluids in motion the normal stresses are not necessarily the same in different directions, thus, the need to define the pressure as the average of the three normal stresses. For fluids at rest, or frictionless fluids, the normal stresses are equal in all directions. (We have made use of this fact in the chapter on fluid statics and in developing the equations for inviscid flow.) Detailed discussions of the development of these stress– velocity gradient relationships can be found in Refs. 3, 8, and 9. An important point to note is that whereas for elastic solids the stresses are linearly related to the deformation (or strain), for Newtonian fluids the stresses are linearly related to the rate of deformation (or rate of strain). In cylindrical polar coordinates the stresses for incompressible Newtonian fluids are expressed as (for normal stresses) ∂υr σrr = −p + 2μ (6.126a) ∂r υr 1 ∂υθ (6.126b) σθθ = −p + 2μ( + ) r ∂θ r σzz = −p + 2μ ∂υz ∂z (6.126c) (for shearing stresses) 1 ∂υr ∂ υθ + ) ( r r ∂θ ] ∂r ∂υθ 1 ∂υz τθz = τzθ = μ( + r ∂θ ) ∂z ∂υz ∂υr τzr = τrz = μ( + ∂z ∂r ) τrθ = τθr = μ[r (6.126d) (6.126e) (6.126f) The double subscript has a meaning similar to that of stresses expressed in Cartesian coordinates— that is, the first subscript indicates the plane on which the stress acts and the second subscript the direction. Thus, for example, σrr refers to a stress acting on a plane perpendicular to the radial direction and in the radial direction (thus a normal stress). Similarly, τrθ refers to a stress acting on a plane perpendicular to the radial direction but in the tangential (θ direction) and is therefore a shearing stress. 6.8.2 The Navier–Stokes Equations The stresses as defined in the preceding section can be substituted into the differential equations of motion (Eqs. 6.50) and simplified by using the continuity equation for incompressible flow (Eq. 6.31). For rectangular coordinates (see the figure in the margin) the results are (x direction) w z u v y x ∂p ∂u ∂u ∂u ∂u ∂ 2u ∂ 2u ∂ 2u ρ( + u +υ + w ) = − + ρgx + μ( 2 + 2 + 2 ) ∂t ∂x ∂y ∂z ∂x ∂x ∂y ∂z (6.127a) ∂p ∂υ ∂υ ∂υ ∂υ ∂ 2υ ∂ 2υ ∂ 2υ ρ( + u +υ + w ) = − + ρgy + μ( 2 + 2 + 2 ) ∂t ∂x ∂y ∂z ∂y ∂x ∂y ∂z (6.127b) (y direction) 322 Chapter 6 ■ Differential Analysis of Fluid Flow (z direction) ∂p ∂w ∂w ∂w ∂w ∂ 2w ∂ 2w ∂ 2w ρ( +u +υ + w ) = − + ρgz + μ( 2 + 2 + 2 ) ∂t ∂x ∂y ∂z ∂z ∂x ∂y ∂z The Navier–Stokes equations are the basic differential equations describing the flow of Newtonian fluids. (6.127c) where u, υ, and w are the x, y, and z components of velocity as shown in the figure in the margin of the previous page. These equations appear quite formidable, and indeed, they are; however, it is important to recall that they are just Newton’s Second Law, F = ma. We have rearranged the equations so that the acceleration terms are on the left side and the force terms are on the right. These equations are commonly called the Navier–Stokes equations, named in honor of the French mathematician L. M. H. Navier (1785–1836) and the English mechanician Sir G. G. Stokes (1819–1903), who were responsible for their formulation. These three equations of motion, when combined with the conservation of mass equation (Eq. 6.31), provide a complete mathematical description of the flow of incompressible Newtonian fluids. We have four equations (three momentum equations plus continuity) and four unknowns (u, υ, w, and p), and therefore when a complete set of boundary conditions and initial conditions are provided the problem is “well-posed” in mathematical terms. Unfortunately, because of the general complexity of the Navier–Stokes equations (they are nonlinear, second-order, partial differential equations), they are not amenable to exact mathematical solutions except in a few instances. However, in those few instances in which solutions have been obtained and compared with experimental results, the results have been in close agreement. Thus, the Navier–Stokes equations are considered to be the governing differential equations of motion for incompressible Newtonian fluids. In terms of cylindrical polar coordinates (see the figure in the margin), the Navier–Stokes equations can be written as (r direction) υθ ∂υr υ2θ ∂υr ∂υr ∂υr + υz ρ + υ + − r ( ∂t r ∂θ r ∂r ∂z ) =− z vz ∂p ∂υr υr ∂ 2 υr 1 ∂ 1 ∂ 2 υr 2 ∂υθ + ρgr + μ [ r − + − + r ∂r ( ∂r ) r 2 ∂r r 2 ∂θ2 r 2 ∂θ ∂z2 ] (6.128a) (θ direction) r ∂υθ υθ ∂υθ υrυθ ∂υθ ∂υθ + υz ρ( + υr + + r r ∂t ∂r ∂θ ∂z ) vθ vr =− y υθ ∂ 2 υθ ∂υθ 1 ∂p 1 ∂ 1 ∂ 2 υθ 2 ∂υr + ρgθ + μ[ r − + + + r ∂θ r ∂r ( ∂r ) r 2 r 2 ∂θ2 r 2 ∂θ ∂z2 ] (6.128b) θ x (z direction) ∂υz ∂υz ∂υz υθ ∂υz ρ( + υr + + υz r ∂θ ∂t ∂r ∂z ) =− 2 ∂ 2 υz ∂υz ∂p 1 ∂ 1 ∂ υz + ρgz + μ[ r + + r ∂r ( ∂r ) r 2 ∂θ2 ∂z ∂z2 ] (6.128c) To provide a brief introduction to the use of the Navier–Stokes equations, a few of the simplest exact solutions are developed in the next section. Although these solutions will prove to be relatively simple, this is not the case in general. In fact, only a few other exact solutions have been obtained. 6.9 Some Simple Solutions for Laminar, Viscous, Incompressible Flows A principal difficulty in solving the Navier–Stokes equations is their nonlinearity arising from the convective acceleration terms (i.e., u ∂u∂x, w ∂υ∂z, etc.). There are no general analytical schemes for solving nonlinear partial differential equations (e.g., superposition of solutions cannot be 6.9 Some Simple Solutions for Laminar, Viscous, Incompressible Flows 323 used), and each problem must be considered individually. For most practical flow problems, fluid particles do have accelerated motion as they move from one location to another in the flow field. Thus, the convective acceleration terms are usually important. However, there are a few special cases for which the convective acceleration vanishes because of the nature of the geometry of the flow system. In these cases exact solutions are often possible. The Navier–Stokes equations apply to both laminar and turbulent flow, but for turbulent flow each velocity component fluctuates in an apparently random fashion, with a very short time scale, and this added complication makes an analytical solution intractable. Thus, the exact solutions referred to are for laminar flows in which the velocity is either independent of time (steady flow) or dependent on time (unsteady flow) in a well-defined manner. 6.9.1 Steady, Laminar Flow between Fixed Parallel Plates An exact solution can be obtained for steady laminar flow between fixed parallel plates. We first consider flow between the two horizontal, infinite parallel plates of Fig. 6.31a. For this geometry the fluid particles move in the x direction parallel to the plates, and there is no velocity in the y or z direction—that is, υ = 0 and w = 0. In this case it follows from the continuity equation (Eq. 6.31) that ∂u∂x = 0. Furthermore, there would be no variation of u in the z direction for infinite plates, and for steady flow ∂u∂t = 0 so that u = u(y). If these conditions are used in the Navier–Stokes equations (Eqs. 6.127), they reduce to ∂p ∂ 2u + μ( 2 ) ∂x ∂y ∂p 0 = − − ρg ∂y ∂p 0=− ∂z 0=− (6.129) (6.130) (6.131) where we have set gx = 0, gy = −g, and gz = 0. That is, the y axis points up. We see that for this particular problem the Navier–Stokes equations reduce to some rather simple equations. Equations 6.130 and 6.131 can be integrated to yield p = −ρgy + f1 (x) (6.132) which shows that the pressure varies hydrostatically in the y direction. Equation 6.129, rewritten as d 2u 1 ∂p = 2 μ ∂x dy can be integrated to give du 1 ∂p = ( ) y + c1 μ ∂x dy VIDEO V6.12 No-slip boundary condition and integrated again to yield u= 1 ∂p 2 y + c1y + c2 2μ( ∂x ) (6.133) Note that for this simple flow the pressure gradient, ∂p∂x, is treated as constant as far as the integration is concerned, since (as shown in Eq. 6.132) it is not a function of y. The two constants c1 and c2 must be determined from the boundary conditions. For example, if the two plates are y h x h u u g umax z (a) (b) ■ Figure 6.31 The viscous flow between parallel plates: (a) coordinate system and notation used in analysis; (b) parabolic velocity distribution for flow between parallel fixed plates. 324 Chapter 6 ■ Differential Analysis of Fluid Flow fixed, then u = 0 for y = ±h (because of the no-slip condition for viscous fluids). To satisfy this condition c1 = 0 and 1 ∂p c2 = − ( ) h2 2μ ∂x Thus, the velocity distribution becomes u= 1 ∂p (y2 − h2 ) 2μ( ∂x ) (6.134) Equation 6.134 shows that the velocity profile between the two fixed plates is parabolic as illustrated in Fig. 6.31b. The volume rate of flow, q, passing between the plates (for a unit width in the z direction) is obtained from the relationship VIDEO V6.13 Liquid– liquid no-slip q= ∫ h −h u dy = ∫ h −h 1 ∂p (y2 − h2 ) dy 2μ( ∂x ) or q=− 2h3 ∂p 3μ ( ∂x ) (6.135) The pressure gradient ∂p∂x is negative, since the pressure decreases in the direction of flow. If we let Δp represent the pressure drop between two points a distance ℓ apart, then Δp ∂p =− ℓ ∂x and Eq. 6.135 can be expressed as q= 2h3 Δp 3μℓ (6.136) The flow is proportional to the pressure gradient, inversely proportional to the viscosity, and strongly dependent ( ~h3 ) on the gap width. In terms of the mean velocity, V, where V = q2h, Eq. 6.136 becomes V= h2 Δp 3μℓ (6.137) Equations 6.136 and 6.137 provide convenient relationships for relating the pressure drop along a parallel-plate channel and the rate of flow or mean velocity. The maximum velocity, umax, occurs midway (y = 0) between the two plates, as shown in Fig. 6.31b, so that from Eq. 6.134 umax = − h2 ∂p 2μ( ∂x ) or 3 umax = 2V The Navier–Stokes equations provide detailed flow characteristics for laminar flow between fixed parallel plates. (6.138) The details of the steady laminar flow between infinite parallel plates are completely predicted by this solution to the Navier–Stokes equations. For example, if the pressure gradient, viscosity, and plate spacing are specified, then from Eq. 6.134 the velocity profile can be determined, and from Eqs. 6.136 and 6.137 the corresponding flowrate and mean velocity determined. In addition, since the pressure gradient in the x direction is constant, from Eq. 6.132 it follows that ∂p f1 (x) = ( ) x + p0 ∂x where p0 is a reference pressure at x = y = 0, and the pressure variation throughout the fluid can be obtained from ∂p p = −ρgy + ( ) x + p0 (6.139) ∂x 6.9 325 Some Simple Solutions for Laminar, Viscous, Incompressible Flows For a given fluid and reference pressure, p0, the pressure at any point can be predicted. This relatively simple example of an exact solution illustrates the detailed information about the flow field which can be obtained. The flow will be laminar if the Reynolds number, Re = ρV(2h)μ, remains below about 1400. For flow with larger Reynolds numbers the flow becomes turbulent and the preceding analysis is not valid since the flow field is complex, three-dimensional, and unsteady. THE WIDE WORLD OF FLUIDS 10 tons on 8 psi Place a golf ball on the end of a garden hose stone to an accuracy of a few thousandths of an inch. This and then slowly turn the water on a small amount until the allows the gap between the ball and its pedestal to be very ball just barely lifts off the end of the hose, leaving a small small (on the order of 5/1000 in.) and the water flowrate cor- gap between the ball and the hose. The ball is free to rotate. respondingly small (on the order of 5 gal/min). Due to the small This is the idea behind the new “floating ball water fountains” gap, the flow in the gap is essentially that of flow between developed in Finland. Massive, 10-ton, 6-ft-diameter stone parallel plates. Although the sphere is very heavy, the pres- spheres are supported by the pressure force of the water on sure under the sphere within the pedestal needs to be only the curved surface within a pedestal and rotate so easily that about 8 psi. A similar fountain can be seen in the Inner Harbor even a small child can change their direction of rotation. The area in Baltimore, Maryland. (See Problem 6.71.) key to the fountain design is the ability to grind and polish 6.9.2 Couette Flow For a given flow geometry, the character and details of the flow are strongly dependent on the boundary conditions. Another simple parallel-plate flow can be developed by fixing one plate and letting the other plate move with a constant velocity, U, as is illustrated in Fig. 6.32a. The Navier–Stokes equations reduce to the same form as those in the preceding section, and the general solution for the pressure and velocity distribution are still given by Eqs. 6.132 and 6.133, respectively. However, for the moving plate problem the boundary conditions for the velocity are different. For this case we locate the origin of the coordinate system at the bottom plate and designate the distance between the two plates as b (see Fig. 6.32a). The two constants c1 and c2 in Eq. 6.133 can be determined from the no-slip boundary conditions, u = 0 at y = 0 and u = U at y = b. It follows that u=U y 1 ∂p + (y2 − by) b 2μ ( ∂x ) (6.140) or, in dimensionless form, y y y u b2 ∂p = − 1 − U b 2μU ( ∂x )( b )( b) (6.141) U Moving plate 1.0 0.8 Backflow U P = –3 b _y b y x –2 –1 0 1 2 3 0.6 b 0.4 0.2 0 z Fixed plate -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 u __ U (a) (b) ■ Figure 6.32 The viscous flow between parallel plates with bottom plate fixed and upper plate moving (Couette flow): (a) coordinate system and notation used in analysis; (b) velocity distribution as a function of parameter, P, where P = −(b 22μU) 𝛛p𝛛x. 326 Chapter 6 ■ Differential Analysis of Fluid Flow Lubricating oil ω ro Rotating shaft ri Housing ■ Figure 6.33 Flow in the narrow gap of a journal bearing. The actual velocity profile will depend on the dimensionless parameter P=− Flow between parallel plates with one plate fixed and the other moving is called Couette flow. b2 ∂p 2μU ( ∂x ) Several profiles are shown in Fig. 6.32b. This type of flow is called Couette flow. The simplest type of Couette flow is one for which the pressure gradient is zero; that is, the fluid motion is caused by the fluid being dragged along by the moving boundary. In this case, with ∂p∂x = 0, Eq. 6.140 simply reduces to u=U y b (6.142) which indicates that the velocity varies linearly between the two plates as shown in Fig. 6.31b for P = 0. This situation would be approximated by the flow between closely spaced concentric cylinders in which one cylinder is fixed and the other cylinder rotates with a constant angular velocity, ω. As illustrated in Fig. 6.33, the flow in an unloaded journal bearing might be approximated by this simple Couette flow if the gap width is very small (i.e., ro − ri ⪡ ri ). In this case U = ri ω, b = ro − ri, and the shearing stress resisting the rotation of the shaft can be simply calculated as τ = μri ω(ro − ri ). When the bearing is loaded (i.e., a force applied normal to the axis of rotation), the shaft will no longer remain concentric with the housing and the flow cannot be treated as flow between parallel boundaries. Such problems are dealt with in lubrication theory (see, for example, Ref. 9). EXAMPLE 6 .9 Plane Couette Flow h GIVEN A wide moving belt passes through a container of a viscous liquid. The belt moves vertically upward with a constant velocity, V0, as illustrated in Fig. E6.9a. Because of viscous forces the belt picks up a film of fluid of thickness h. Gravity tends to make the fluid drain down the belt. Assume that the flow is laminar, steady, and fully developed. Fluid layer V0 y FIND Use the Navier–Stokes equations to determine an expres- sion for the average velocity of the fluid film as it is dragged up the belt. g x SOLUTION Since the flow is assumed to be fully developed, the only velocity component is in the y direction (the υ component) so that u = w = 0. It follows from the continuity equation that ∂υ∂y = 0, and for steady flow ∂υ∂t = 0, so that υ = υ(x). Under these conditions the Navier–Stokes equations for the x direction (Eq. 6.127a) and the z direction (perpendicular to the paper) (Eq.6.127c) simply reduce to ∂p ∂p = 0   = 0 ∂x ∂z ■ Figure E6.9a This result indicates that the pressure does not vary over a horizontal plane, and since the pressure on the surface of the film (x = h) is atmospheric, the pressure throughout the film must be 6.9 atmospheric (or zero gage pressure). The equation of motion in the y direction (Eq. 6.127b) thus reduces to 0 = −ρg + μ and thus q = V0 h − 2 d υ dx2 V = V0 − (1) γ dυ = x + c1 μ dx τxy = μ( (2) dυ dx ) Thus, if τxy = 0 at x = h, it follows from Eq. 2 that c1 = − γh μ υ x 2 x = c( − 2c( +1 V0 h) h) where c = γh2/2μV0. This velocity profile is shown in Fig. E6.9b. Note that even though the belt is moving upward, for c > 1 (e.g., for fluids with small enough viscosity or with a small enough belt speed) there are portions of the fluid that flow downward (as indicated by υ/V0 < 0). It is interesting to note from this result that there will be a net upward flow of liquid (positive V ) only if V0 > γh2/3μ. It takes a relatively large belt speed to lift a small viscosity fluid. 0.8 0.6 v/V0 (3) With the velocity distribution known we can determine the flowrate per unit width, q, from the relationship h 0 υ dx = ∫ 0 h γ 2 γh ( 2μ x − μ x + V0) dx 0.2 0 c = 1.0 0 0.2 0.4 0.6 0.8 1 –0.2 and the velocity distribution is therefore γ 2 γh x − x + V0 μ 2μ c = 0.5 0.4 At the belt (x = 0) the fluid velocity must match the belt velocity, V0, so that c2 = V0 c=0 1 A second integration of Eq. 2 gives the velocity distribution in the film as γ 2 γh x − x + c2 υ= μ 2μ ∫ (Ans) form as On the film surface (x = h) we assume the shearing stress is zero—that is, the drag of the air on the film is negligible. The shearing stress at the free surface (or any interior parallel surface) is designated as τxy , where from Eq. 6.125d q= γh2 3μ COMMENT Equation (3) can be written in dimensionless Integration of Eq. 1 yields υ= γh3 3μ The average film velocity, V (where q = Vh), is therefore or γ d 2υ = 2 μ dx 327 Some Simple Solutions for Laminar, Viscous, Incompressible Flows c = 1.5 –0.4 –0.6 –0.8 –1 c = 2.0 x/ h ■ Figure E6.9b 6.9.3 Steady, Laminar Flow in Circular Tubes An exact solution can be obtained for steady, incompressible, laminar flow in circular tubes. Probably the best known exact solution to the Navier–Stokes equations is for steady, incompressible, laminar flow through a straight circular tube of constant cross section. This type of flow is commonly called Hagen–Poiseuille flow, or simply Poiseuille flow. It is named in honor of J. L. Poiseuille (1799–1869), a French physician, and G. H. L. Hagen (1797–1884), a German hydraulic engineer. Poiseuille was interested in blood flow through capillaries and deduced experimentally the resistance laws for laminar flow through circular tubes. Hagen’s investigation of flow in tubes was also experimental. It was actually after the work of Hagen and Poiseuille that the theoretical results presented in this section were determined, but their names are commonly associated with the solution of this problem. Consider the flow through a horizontal circular tube of radius R as is shown in Fig. 6.34a. Because of the cylindrical geometry it is convenient to use cylindrical coordinates. We assume that the flow is parallel to the walls so that υr = 0 and υθ = 0, and from the continuity equation (6.34) ∂υz∂z = 0. Also, for steady, axisymmetric flow, υz is not a function of t or θ, so the velocity, υz, 328 Chapter 6 ■ Differential Analysis of Fluid Flow y g vz r R vz θ r z z dr (a) (b) ■ Figure 6.34 The viscous flow in a horizontal, circular tube: (a) coordinate system and notation used in analysis; (b) flow through differential annular ring. is only a function of the radial position within the tube—that is, υz = υz (r). Under these conditions the Navier–Stokes equations (Eqs. 6.128) reduce to ∂p ∂r 1 ∂p 0 = −ρg cos θ − r ∂θ ∂υz ∂p 1 ∂ 0 = − + μ[ r ( r ∂z ∂r ∂r )] 0 = −ρg sin θ − (6.143) (6.144) (6.145) where we have used the relationships gr = −g sin θ and gθ = −g cos θ (with θ measured from the horizontal plane). Equations 6.143 and 6.144 can be integrated to give p = −ρg(r sin θ) + f1 (z) or p = −ρgy + f1 (z) VIDEO V6.14 Laminar flow (6.146) Equation 6.146 indicates that the pressure is hydrostatically distributed at any particular cross section, and the z component of the pressure gradient, ∂p∂z, is not a function of r or θ. The equation of motion in the z direction (Eq. 6.145) can be written in the form ∂υz 1 ∂ 1 ∂p r = r ∂r( ∂r ) μ ∂z and integrated (using the fact that ∂p∂z = constant) to give r ∂υz ∂r = 1 ∂p 2 r + c1 2μ( ∂z ) Integrating again we obtain υz = 1 ∂p 2 r + c1 ln r + c2 4μ ( ∂z ) (6.147) Since we wish υz to be finite at the center of the tube (r = 0), it follows that c1 = 0 [since ln (0) = −∞]. At the wall (r = R) the velocity must be zero so that c2 = − The velocity distribution is parabolic for steady, laminar flow in circular tubes. 1 ∂p 2 R 4μ ( ∂z ) and the velocity distribution becomes υz = 1 ∂p (r2 − R2 ) 4μ( ∂z ) (6.148) Thus, at any cross section the velocity distribution is parabolic. To obtain a relationship between the volume rate of flow, Q, passing through the tube and the pressure gradient, we consider the flow through the differential, washer-shaped ring of 6.9 Some Simple Solutions for Laminar, Viscous, Incompressible Flows 329 Fig. 6.34b. Since υz is constant on this ring, the volume rate of flow through the differential area dA = (2πr) dr is dQ = υz (2πr) dr and therefore Q = 2π ∫ R υz r dr (6.149) 0 VIDEO V6.15 Complex pipe flow Equation 6.148 for υz can be substituted into Eq. 6.149, and the resulting equation integrated to yield πR 4 ∂p Q=− (6.150) 8μ ( ∂z ) This relationship can be expressed in terms of the pressure drop, Δp, which occurs over a length, ℓ, along the tube, since Δp ∂p =− ℓ ∂z and therefore Q= Poiseuille’s law relates pressure drop and flowrate for steady, laminar flow in circular tubes. πR 4 Δp 8μℓ (6.151) For a given pressure drop per unit length, the volume rate of flow is inversely proportional to the viscosity and proportional to the tube radius to the fourth power. A doubling of the tube radius produces a 16-fold increase in flow! Equation 6.151 is commonly called Poiseuille’s law. In terms of the mean velocity, V, where V = QπR2, Eq. 6.151 becomes V= R2 Δp 8μℓ (6.152) The maximum velocity υmax occurs at the center of the tube, where from Eq. 6.148 υmax = − R2 Δp R2 ∂p = 4μ( ∂z ) 4μℓ (6.153) so that υmax = 2V vz _r vmax = 1 – R ( ) R r vmax 2 The velocity distribution, as shown by the figure in the margin, can be written in terms of υmax as υz r 2 = 1 −( ) υmax R (6.154) As was true for the similar case of flow between parallel plates (sometimes referred to as plane Poiseuille flow), a very detailed description of the pressure and velocity distribution in tube flow results from this solution to the Navier–Stokes equations. Numerous experiments performed to substantiate the theoretical results show that the theory and experiment are in agreement for the laminar flow of Newtonian fluids in circular tubes or pipes. In general, the flow remains laminar for Reynolds numbers, Re = ρV(2R)μ, below about 2100. Turbulent flow in tubes is considered in Chapter 8. THE WIDE WORLD OF FLUIDS Poiseuille’s law revisited Poiseuille’s law governing laminar Another formula was given for the value of K as a function of the flow of fluids in tubes has an unusual history. It was developed water temperature. It was not until the concept of viscosity was in 1842 by a French physician, J. L. M. Poiseuille, who was inter- introduced at a later date that Poiseuille’s law was derived math- ested in the flow of blood in capillaries. Poiseuille, through a ematically and the constant K found to be equal to π8μ, where series of carefully conducted experiments using water flowing μ is the fluid viscosity. The experiments by Poiseuille have long through very small tubes, arrived at the formula, Q = KΔp D4/ℓ. In been admired for their accuracy and completeness considering this formula Q is the flowrate, K an empirical constant, Δp the the laboratory instrumentation available in the mid-nineteenth pressure drop over the length ℓ, and D the tube diameter. century. 330 Chapter 6 ■ Differential Analysis of Fluid Flow vz ro r ri z ■ Figure 6.35 The viscous flow through an annulus. 6.9.4 Steady, Axial, Laminar Flow in an Annulus An exact solution can be obtained for axial flow in the annular space between two fixed, concentric cylinders. The differential equations (Eqs. 6.143, 6.144, 6.145) used in the preceding section for flow in a tube also apply to the axial flow in the annular space between two fixed, concentric cylinders (Fig. 6.35). Equation 6.147 for the velocity distribution still applies, but for the stationary annulus the boundary conditions become υz = 0 at r = ro and υz = 0 for r = ri. With these two conditions the constants c1 and c2 in Eq. 6.147 can be determined and the velocity distribution becomes υz = r 2i − r 2o 1 ∂p r 2 2 r − r + ln o 4μ ( ∂z ) [ ln(ro ri ) ro ]   (6.155) The corresponding volume rate of flow is Q= ∫ ro υz (2πr) dr = − ri (r 2o − r 2i ) 2 π ∂p 4 4 r − r − o i 8μ ( ∂z ) [ ln(ro ri ) ] or in terms of the pressure drop, Δp, in length ℓ of the annulus Q= (r 2o − r 2i ) 2 πΔp 4 4 r − r − o i 8μℓ [ ln(ro ri ) ]   (6.156) The velocity at any radial location within the annular space can be obtained from Eq. 6.155. The maximum velocity occurs at the radius r = rm where ∂υz ∂r = 0. Thus, rm = [ r 2o − r 2i 12 2 ln(ro ri ) ] (6.157) An inspection of this result shows that the maximum velocity does not occur at the midpoint of the annular space, but rather it occurs nearer the inner cylinder. The specific location depends on ro and ri. These results for flow through an annulus are valid only if the flow is laminar. A criterion based on the conventional Reynolds number (which is defined in terms of the tube diameter) cannot be directly applied to the annulus, since there are really “two” diameters involved. For tube cross sections other than simple circular tubes it is common practice to use an “effective” diameter, termed the hydraulic diameter, Dh, which is defined as Dh = 4 × cross-sectional area wetted perimeter The wetted perimeter is the perimeter in contact with the fluid. For an annulus Dh = 4π(ro2 − ri2 ) = 2(ro − ri ) 2π(ro + ri ) In terms of the hydraulic diameter, the Reynolds number is Re = ρDhVμ (where V = Q/ cross-sectional area), and it is commonly assumed that if this Reynolds number remains below about 2100 the flow will be laminar. A further discussion of the concept of the hydraulic diameter as it applies to other noncircular cross sections is given in Section 8.4.3. 6.9 E X AMPLE 6 .10 331 Some Simple Solutions for Laminar, Viscous, Incompressible Flows Laminar Flow in an Annulus GIVEN A viscous liquid (ρ = 1.18 × 103 kgm3; μ = 0.0045 N ˙ sm2 ) flows at a rate of 12 mls through a horizontal, 4-mm-diameter tube. FIND (a) Determine the pressure drop along a l-m length of the tube which is far from the tube entrance so that the only component of velocity is parallel to the tube axis. (b) If a 2-mm-diameter rod is placed in the 4-mm-diameter tube to form a symmetric annulus, what is the pressure drop along a l-m length if the flowrate remains the same as in part (a)? SOLUTION (a) We first calculate the Reynolds number, Re, to determine whether or not the flow is laminar. With the diameter D = 4 mm = 0.004 m, the mean velocity is V= Q (π4)D2 = so that Δp = (12 mls) (10−6 m3ml)  − (1.18 × 10 kgm ) (0.955 ms) (0.004 m) ρVD = μ 0.0045 N sm2 ˙ Since the Reynolds number is well below the critical value of 2100, we can safely assume that the flow is laminar. Thus, we can apply Eq. 6.151, which gives for the pressure drop 8μℓQ πR4 8(0.0045 N 2 × 10−6 m3s) ˙ sm ) (1 m) (12 4 π(0.002 m) = 8.59 kPa (Ans) (b) For flow in the annulus with an outer radius ro = 0.002 m and an inner radius ri = 0.001 m, the mean velocity is Q − = 1.27 ms π(ro2 ri2 ) = 12 × 10−6 m3s COMMENTS The pressure drop in the annulus is much larger than that of the tube. This is not a surprising result, since to maintain the same flow in the annulus as that in the open tube, the average velocity must be larger (the cross-sectional area is smaller) and the pressure difference along the annulus must overcome the shearing stresses that develop along both an inner and an outer wall. By repeating the calculations for various radius ratios, ri /ro, the results shown in Fig. E6.10 are obtained. It is seen that the pressure drop ratio, Δpannulus/Δptube (i.e., the pressure drop in the annulus compared to that in a tube with a radius equal to the outer radius of the annulus, ro), is a strong function of the radius ratio. Even an annulus with a very small inner radius will have a pressure drop significantly larger than that of a tube. For example, if the inner radius is only 1/100 of the outer radius, Δpannulus/Δptube = 1.28. As shown in the figure, for larger inner radii, the pressure drop ratio is much larger [i.e., Δpannulus / Δptube = 7.94 for ri/ro = 0.50 as in part (b) of this example]. (π)[(0.002 m) 2 − (0.001 m) 2] 8 and the Reynolds number [based on the hydraulic diameter, Dh = 2(ro − ri ) = 2(0.002 m − 0.001 m) = 0.002 m] is 0.0045 N = 666 2 ˙ sm 8μℓQ 4 − Δp = r − r4i − π [o ln(ro ri ) ]   1.1 5 1 4 0 0.005 0.01 3 This value is also well below 2100, so the flow in the annulus should also be laminar. From Eq. 6.156, (r 2o 1.2 6 Δpannulus ________ Δptube = ρDhV μ (1.18 × 103 kgm3 ) (0.002 m) (1.27 ms) (0.50, 7.94) (0.01, 1.28) 1.3 7 Re = (Ans) 3 = 1000 V= [(0.002 m) 2 − (0.001 m) 2]2 −1 ln(0.002 m0.001 m) } = 68.2 kPa 3 = π   × { (0.002 m) − (0.001 m) 4 (π4) (0.004 m) 2 and, therefore, Δp = 2 −6 3 ˙ sm ) (1 m) (12 × 10 m s) 4 = 0.955 ms Re = 8(0.0045 N 2 1 0 r 2i ) 2 −1 0.1 0.2 0.3 ri /r0   ■ Figure E6.10 0.4 0.5 332 Chapter 6 ■ Differential Analysis of Fluid Flow 6.10 Other Aspects of Differential Analysis In this chapter the basic differential equations that govern the flow of incompressible fluids have been developed. The Navier–Stokes equations, which can be compactly expressed in vector notation as ∂V ρ( + V ∙ 𝛁V) = −𝛁p + ρg + μ∇ 2V ∂t (6.158) along with the continuity equation ∇ ∙V=0 Very few practical fluid flow problems can be solved using an exact analytical approach. (6.159) are the general equations of motion for incompressible Newtonian fluids. Although we have restricted our attention to incompressible fluids, these equations can be readily extended to include compressible fluids. It is well beyond the scope of this introductory text to consider in depth the variety of analytical and numerical techniques that can be used to obtain both exact and approximate solutions to the Navier–Stokes equations. Students, however, should be aware of the existence of these very general equations, which are frequently used as the basis for many advanced analyses of fluid motion. A few relatively simple solutions have been obtained and discussed in this chapter to indicate the type of detailed flow information that can be obtained by using differential analysis. However, it is hoped that the relative ease with which these solutions were obtained does not give the false impression that solutions to the Navier–Stokes equations are readily available. This is certainly not true, and as previously mentioned there are actually very few practical fluid flow problems that can be solved by using an exact analytical approach. In fact, there are no known analytical solutions to Eq. 6.158 for flow past any object such as a sphere, cube, or airplane. Because of the difficulty in solving the Navier–Stokes equations, much attention has been given to various types of approximate solutions. For example, if the viscosity is set equal to zero, the Navier–Stokes equations reduce to Euler’s equations. Thus, the frictionless fluid solutions discussed previously are actually approximate solutions to the Navier–Stokes equations. At the other extreme, for problems involving slowly moving fluids, viscous effects may be dominant and the nonlinear (convective) acceleration terms can be neglected. This assumption greatly simplifies the analysis, since the equations now become linear. There are numerous analytical solutions to these “slow flow” or “creeping flow” problems. Another broad class of approximate solutions is concerned with flow in the very thin boundary layer. L. Prandtl showed in 1904 how the Navier–Stokes equations could be simplified to study flow in boundary layers. Such “boundary layer solutions” play a very important role in the study of fluid mechanics. By “patching together” an inviscid flow solution and a boundary layer solution, a relatively complete solution can be found for many important flows. A further discussion of boundary layers is given in Chapter 9. 6.10.1 Numerical Methods VIDEO V6.16 CFD example Numerical methods using digital computers are, of course, commonly utilized to solve a wide variety of flow problems. As discussed previously, although the differential equations that govern the flow of Newtonian fluids [the Navier–Stokes equations (6.158)] were derived many years ago, there are few known analytical solutions to them. With the advent of high-speed digital computers it has become possible to obtain numerical solutions to these (and other fluid mechanics) equations for many different types of problems, including both inviscid flows and boundary layer flows. A brief introduction to computational fluid dynamics (CFD) is given in Appendix A. THE WIDE WORLD OF FLUIDS Fluids in the Academy Awards A computer science professor or CFD) and simulate complex liquid flows. The realism of the at Stanford University and his colleagues were awarded a simulations has found application in the entertainment industry. Scientific and Technical Academy Award for applying the Navier– Movie producers have used the power of these numerical tools to Stokes equations for use in Hollywood movies. These researchers simulate flows from ocean waves in Pirates of the Caribbean to make use of computational algorithms to numerically solve the lava flows in the final duel in Star Wars: Revenge of the Sith. Navier–Stokes equations (also termed computational fluid dynamics, Therefore, even Hollywood has recognized the usefulness of CFD. 6.11 6.11 volumetric dilatation rate vorticity irrotational flow continuity equation stream function Euler’s equations of motion ideal fluid Bernoulli equation velocity potential potential flow equipotential lines flow net uniform flow source and sink vortex circulation doublet method of superposition half-body Rankine oval Navier–Stokes equations Couette flow Poiseuille’s law Chapter Summary and Study Guide 333 Chapter Summary and Study Guide Differential analysis of fluid flow is concerned with the development of concepts and techniques that can be used to provide a detailed, point-by-point, description of a flow field. Concepts related to the motion and deformation of a fluid element are introduced, including the Eulerian method for describing the velocity and acceleration of fluid particles. Linear deformation and angular deformation of a fluid element are described through the use of flow characteristics such as the volumetric dilatation rate, rate of angular deformation, and vorticity. The differential form of the conservation of mass equation (continuity equation) is derived in both rectangular and cylindrical polar coordinates. The stream function for steady, incompressible, plane, two-dimensional flow is introduced. The general equations of motion are developed, and for inviscid flow these equations are reduced to the simpler Euler equations. The Euler equations are integrated to give the Bernoulli equation, and the concept of irrotational flow is introduced. Use of the velocity potential for describing irrotational flow is considered in detail, and several basic velocity potentials are described, including those for a uniform flow, source or sink, vortex, and doublet. The technique of using various combinations of these basic velocity potentials, by superposition, to form new potentials is described. Flows around a half-body, a Rankine oval, and a circular cylinder are obtained using this superposition technique. Basic differential equations describing incompressible, viscous flow (the Navier–Stokes equations) are introduced. Several relatively simple solutions for steady, viscous, laminar flow between parallel plates and through circular tubes are included. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ ■ ■ ■ ■ ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. determine the acceleration of a fluid particle, given the equation for the velocity field. determine the volumetric dilatation rate, vorticity, and rate of angular deformation for a fluid element, given the equation for the velocity field. show that a given velocity field satisfies the continuity equation. use the stream function to describe a flow field. use the velocity potential to describe a flow field. use superposition of basic velocity potentials to describe simple potential flow fields. use the Navier–Stokes equations to determine the detailed flow characteristics of incompressible, steady, laminar, viscous flow between parallel plates and through circular tubes. Some of the important equations in this chapter are: Acceleration of fluid particle a= Vorticity Conservation of mass Stream function Euler’s equations of motion ∂V ∂V ∂V ∂V +u +υ +w ∂t ∂x ∂y ∂z ζ = 2ω = ∇ × V ∂ρ ∂ (ρu) ∂ (ρυ) ∂ (ρw) + + + =0 ∂t ∂x ∂y ∂z ∂ψ ∂ψ u=   υ = − ∂y ∂x ∂p ∂u ∂u ∂u ∂u ρgx − = ρ( + u +υ +w ) ∂x ∂t ∂x ∂y ∂z ∂p ∂υ ∂υ ∂υ ∂υ ρgy − = ρ( + u +υ +w ) ∂y ∂t ∂x ∂y ∂z ρgz − ∂p ∂w ∂w ∂w ∂w = ρ( +u +υ +w ) ∂z ∂t ∂x ∂y ∂z (6.2) (6.17) (6.27) (6.37) (6.51a) (6.51b) (6.51c) 334 Chapter 6 ■ Differential Analysis of Fluid Flow Velocity potential V = 𝛁ϕ (6.65) Laplace’s equation 𝛁2 ϕ = 0 (6.66) Uniform potential flow ϕ = U(x cos α + y sin α) Source and sink ϕ= m ln r 2π Vortex ϕ= Γ θ 2π   ψ=  ψ = − ψ = U(y cos α − x sin α) m θ 2π υr = Γ ln r 2π υr = 0 m 2πr υθ = 0 υθ = ϕ= Doublet K cos θ r   ψ=− K sin θ r u = U cos α υ = U cos α Γ 2πr K cos θ r2 K cos θ υθ = − r2 υr = − The Navier–Stokes equations (x direction) ∂p ∂u ∂u ∂u ∂u ∂ 2u ∂ 2u ∂ 2u ρ( + u +υ + w ) = − + ρgx + μ( 2 + 2 + 2 ) ∂t ∂x ∂y ∂z ∂x ∂x ∂y ∂z (6.127a) (y direction) ∂p ∂υ ∂υ ∂υ ∂υ ∂ 2υ ∂ 2υ ∂ 2υ ρ( + u +υ + w ) = − + ρgy + μ( 2 + 2 + 2 ) ∂t ∂x ∂y ∂z ∂y ∂x ∂y ∂z (6.127b) (z direction) ∂p ∂w ∂w ∂w ∂w ∂ 2w ∂ 2w ∂ 2w ρ( +u +υ + w ) = − + ρgz + μ( 2 + 2 + 2 ) ∂t ∂x ∂y ∂z ∂z ∂x ∂y ∂z (6.127c) References White, F. M., Fluid Mechanics, 5th Ed., McGraw-Hill, New York, 2003. Streeter, V. L., Fluid Dynamics, McGraw-Hill, New York, 1948. Rouse, H., Advanced Mechanics of Fluids, Wiley, New York, 1959. Milne-Thomson, L. M., Theoretical Hydrodynamics, 4th Ed., Macmillan, New York, 1960. Robertson, J. M., Hydrodynamics in Theory and Application, Prentice-Hall, Englewood Cliffs, N.J., 1965. Panton, R. L., Incompressible Flow, 3rd Ed., Wiley, New York, 2005. Hess, J. L., “Review of Integral Equation Techniques for Solving Potential-Flow Problems with Emphasis on the Surface-Source Method,” Computer Methods in Applied Mechanics and Engineering, 5, 1975, pp. 145–196. 8. Li, W. H., and Lam, S. H., Principles of Fluid Mechanics, Addison-Wesley, Reading, Mass., 1964. 9. Schlichting, H., Boundary-Layer Theory, 8th Ed., McGraw-Hill, New York, 2000. 10. Fuller, D. D., Theory and Practice of Lubrication for Engineers, Wiley, New York, 1984. 1. 2. 3. 4. 5. 6. 7. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are not unique answers to these problems. Problems 335 Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. (a) Determine the volumetric dilatation rate and interpret the results. (b) Determine an expression for the rotation vector. Is this an irrotational flow field? Section 6.1 Fluid Element Kinematics 6.5 Determine an expression for the vorticity of the flow field described by V = −xy3ˆi + y4jˆ 6.1 The velocity in a certain two-dimensional flow field is given by the equation V = 2xtiˆ − 2yt jˆ   where the velocity is in ft/s when x, y, and t are in feet and seconds, respectively. Determine expressions for the local and convective components of acceleration in the x and y directions. What is the magnitude and direction of the velocity and the acceleration at the point x = y = 2 ft at the time t = 0? 6.2 The velocity in a certain flow field is given by the equation V = xiˆ + x2z jˆ + yzkˆ   Determine the expressions for the three rectangular components of acceleration. Is the flow irrotational? 6.6 According to Eq. 6.134, the x-velocity in fully developed laminar flow between parallel plates is given by u= 1 ∂p (y2 − h2 ) 2μ ( ∂x ) The y-velocity is υ = 0. Determine the volumetric strain rate, the vorticity, and the rate of angular deformation. What is the shear stress at the plate surface? 6.7 For a certain incompressible, two-dimensional flow field the velocity component in the y direction is given by the equation υ = 3xy + x2y 6.3 The flow in the plane two-dimensional channel shown in Figure P6.3 has x- and y-components of velocity given by Determine the velocity component in the x direction so that the volumetric dilatation rate is zero. y 2 x u = u0(1 + )[1 − ( ) ] ℓ Y 6.8 An incompressible viscous fluid is placed between two large parallel plates as shown in Fig. P6.8. The bottom plate is fixed and the upper plate moves with a constant velocity, U. For these conditions the velocity distribution between the plates is linear and can be expressed as υ = u0[ y3 y x 2 1+ ) − ] ℓ ℓ ℓY 20 ( u=U Calculate the linear acceleration, rotation, vorticity, rate of volumentric strain, and rate of shear deformation for the flow. y b Determine: (a) the volumetric dilatation rate, (b) the rotation vector, (c) the vorticity, and (d) the rate of angular deformation. y U Y0 = 1.0 m + Y (x) u (x, y ) Yℓ = 0.5 m x ℓ = 5.0 m ■ Figure P6.3 6.4 The three components of velocity in a flow field are given by u = x2 + y2 + z2 υ = xy + yz + z2 w = −3xz − z22 + 4 u Moving plate b y Fixed plate ■ Figure P6.8 6.9 A viscous fluid is contained in the space between concentric cylinders. The inner wall is fixed, and the outer wall rotates with an angular velocity ω. (See Fig. P6.9a and Video V6.3.) Assume that the velocity distribution in the gap is linear as illustrated in Fig. P6.9b. For the small rectangular element shown in Fig. P6.9b, determine the rate of change of the right angle γ due to the fluid motion. Express your answer in terms of ro, ri, and ω. 336 Chapter 6 ■ Differential Analysis of Fluid Flow ω 6.16 The stream function for an incompressible, two-dimensional flow field is ro ω ψ = ay − by3 u ri ro γ y ro – ri where a and b are constants. Is this an irrotational flow? Explain. 6.17 For a certain two-dimensional flow field u=0 υ=V x (a) (a) What are the corresponding radial and tangential velocity components? (b) Determine the corresponding stream function expressed in Cartesian coordinates and in cylindrical polar coordinates. (b) ■ Figure P6.9 6.10 Air is delivered through a constant-diameter duct by a fan. The air is inviscid, so the fluid velocity profile is “flat” across each cross section. During the fan start-up, the following velocities were measured at the time t and axial positions x: x=0 t=0s x = 10 m V = 0 ms V = 0 ms x = 20 m V = 0 ms t = 1.0 s V = 1.00 ms V = 1.20 ms V = 1.40 ms t = 2.0 s V = 1.70 ms V = 1.80 ms V = 1.90 ms t = 3.0 s V = 2.10 ms V = 2.15 ms V = 2.20 ms 6.18 In a certain steady, two-dimensional flow field the fluid density varies linearly with respect to the coordinate x: that is, ρ = Ax where A is a constant. If the x component of velocity u is given by the equation u = y, determine an expression for υ. 6.19 In a two-dimensional, incompressible flow field, the x component of velocity is given by the equation u = 2x. (a) Determine the corresponding equation for the y component of velocity if υ = 0 along the x axis. (b) For this flow field, what is the magnitude of the average velocity of the fluid crossing the surface OA of Fig. P6.19? Assume that the velocities are in feet per second when x and y are in feet. y, ft Estimate the local acceleration, the convective acceleration, and the total acceleration at t = 1.0 s and x = 10 m. What is the local acceleration after the fan has reached a steady air flow rate? A 1.0 Section 6.2 Conservation of Mass 6.11 For a certain incompressible flow field it is suggested that the velocity components are given by the equations u = 2xy υ = −x2y w = 0 Is this a physically possible flow field? Explain. 6.12 The velocity components of an incompressible, two-dimensional velocity field are given by the equations u = y − x(1 + x) 2 υ = y(2x + 1) Show that the flow is irrotational and satisfies conservation of mass. 6.13 Consider the two-dimensional channel flow of Problem 6.3. Show that the given velocities satisfy conservation of mass in both differential and control volume forms. O 1.0 x, ft ■ Figure P6.19 6.20 The stream function for an incompressible flow field is given by the equation ψ = 3x2y − y3 where the stream function has the units of m2/s with x and y in meters. (a) Sketch the streamline(s) passing through the origin. (b) Determine the rate of flow across the straight path AB shown in Fig. P6.20. y, m 1.0 B 6.14 The x-velocity profile in a certain laminar boundary layer is approximated as follows π y u = U0 sin( 2 0.1√x ) Determine the y-velocity, υ(x, y). A 0 1.0 x, m ■ Figure P6.20 6.15 For each of the following stream functions, with units of m2/s, determine the magnitude and the angle the velocity vector *6.21 The stream function for an incompressible, two-dimensional makes with the x axis at x = 1 m, y = 2 m. Locate any stagnation flow field is points in the flow field. ψ = 3x2y + y (a) ψ = xy 2 (b) ψ = −2x + y For this flow field, plot several streamlines. Problems 6.22 Consider the incompressible, two-dimensional flow of a nonviscous fluid between the boundaries shown in Fig. P6.22. The velocity potential for this flow field is ϕ=x −y 2 2 (a) Determine the corresponding stream function. (b) What is the relationship between the discharge, q (per unit width normal to plane of paper) passing between the walls and the coordinates xi, yi of any point on the curved wall? Neglect body forces. 337 6.27 A certain flow field is described by the stream function ψ = A θ + B r sin θ where A and B are positive constants. Determine the corresponding velocity potential and locate any stagnation points in this flow field. 6.28 Integrate Bernoulli’s equation for compressible flow, Eq. (6.56), for an ideal gas undergoing an isothermal (constant temperature) process along a streamline. 6.29 Integrate Bernoulli’s equation for compressible flow, Eq. (6.56), for a fluid whose pressure p and density ρ obey the expression p = Cρn along a streamline, where C and n are constants. y 6.30 The velocity potential for a certain inviscid flow field is ϕ = −(3x2y − y3 ) q where ϕ has the units of ft2/s when x and y are in feet. Determine the pressure difference (in psi) between the points (1, 2) and (4, 4), where the coordinates are in feet, if the fluid is water and elevation changes are negligible. (xi, yi) B q ψ =0 A x ■ Figure P6.22 Section 6.3 The Linear Momentum Equation 6.23 A fluid with a density of 2000 kg/m3 flows steadily between two flat plates as shown in Fig. P6.23. The bottom plate is fixed and the top one moves at a constant speed in the x direction. The velocity is V = 0.20 y ˆi m/s where y is in meters. The acceleration of gravity is g = −9.8 jˆ m/s2. The only nonzero shear stresses, τyx = τxy, are constant throughout the flow with a value of 5 N/m2. The normal stress at the origin (x = y = 0) is σxx = −100 kPa. Use the x and y components of the equations of motion (Eqs. 6.50a and b) to determine the normal stress throughout the fluid. Assume that σxx = σyy. Moving plate U 6.31 The stream function for a two-dimensional, nonviscous, incompressible flow field is given by the expression ψ = −2(x − y) where the stream function has the units of ft2/s with x and y in feet. (a) Is the continuity equation satisfied? (b) Is the flow field irrotational? If, so, determine the corresponding velocity potential. (c) Determine the pressure gradient in the horizontal x direction at the point x = 2 ft, y = 2 ft. 6.32 The velocity potential for a certain inviscid, incompressible flow field is given by the equation 2 ϕ = 2x2y − ( 3 )y3 where ϕ has the units of m2/s when x and y are in meters. Determine the pressure at the point x = 2 m, y = 2 m if the pressure at x = 1 m, y = 1 m is 200 kPa. Elevation changes can be neglected, and the fluid is water. 6.33 Consider the two-dimensional air flow around the corner, as shown in Fig. P6.33. The x- and y-direction velocities are u= b and g y υ=− x z υ0 y x sin h ( ) cos h ( ) L L L υ0 y x cos h ( ) sin h ( ), L L L respectively. Assume constant density, steady flow, negligible gravity, and inviscid flow. Find p(x, y). Fixed plate y ■ Figure P6.23 6.24 In Section 6.3, we derived the differential equation(s) of linear momentum by considering the motion of a fluid element. Derive the linear momentum equation(s) by considering a small control volume, like we did for the continuity equation in Section 6.2. v (x, L) L u (L, y ) 6.25 By considering the rotational equilibrium of a fluid mass element, show that τxy = τyx . Section 6.4 Inviscid Flow 6.26 The stream function for a given two-dimensional flow field is ψ = 5x2y − (53)y3 Determine the corresponding velocity potential. 0 0 ■ Figure P6.33 L x 338 Chapter 6 ■ Differential Analysis of Fluid Flow 6.34 The streamlines for an incompressible, inviscid, two-dimensional flow field are all concentric circles, and the velocity varies directly with the distance from the common center of the streamlines; that is A to the center of the tornado. Why can the free vortex model not be used to approximate the tornado throughout the flow field (r ≥ 0)? υθ = Kr y where K is a constant. (a) For this rotational flow, determine, if possible, the stream function. (b) Can the pressure difference between the origin and any other point be determined from the Bernoulli equation? Explain. r Rc 6.35 The velocity potential A   ϕ = −k(x2 − y2 )  (k = constant) may be used to represent the flow against an infinite plane boundary, as illustrated in Fig. P6.35. For flow in the vicinity of a stagnation point, it is frequently assumed that the pressure gradient along the surface is of the form ∂p = Ax ∂x where A is a constant. Use the given velocity potential to show that this is true. y B x 100 ft ■ Figure P6.38 6.39 The motion of a liquid in an open tank is that of a combined vortex consisting of a forced vortex for 0 ≤ r ≤ 2 ft and a free vortex for r > 2 ft. The velocity profile and the corresponding shape of the free surface are shown in Fig. P6.39. The free surface at the center of the tank is a depth h below the free surface at r = ∞. Determine the value of h. Note that h = hforced + hfree, where hforced and hfree are the corresponding depths for the forced vortex and the free vortex, respectively. (See Section 2.12.2 for further discussion of the forced vortex.) 10 vθ , ft/s x ■ Figure P6.35 0 6.36 Water is flowing between wedge-shaped walls into a small opening as shown in Fig. P6.36. The velocity potential with units m2/s for this flow is ϕ = −2 ln r with r in meters. Determine the pressure differential between points A and B. r, ft 2 z 2 r, ft h _π_ 6 r ■ Figure P6.39 A θ 0.5 m B 1.0 m ■ Figure P6.36 6.37 The velocity potential for a given two-dimensional flow field is 6.40 Water flows through a two-dimensional diffuser having a 20° expansion angle as shown in Fig. P6.40. Assume that the flow in the diffuser can be treated as a radial flow emanating from a source at the origin O. (a) If the velocity at the entrance is 20 m/s, determine an expression for the pressure gradient along the diffuser walls. (b) What is the pressure rise between the entrance and exit? Diffuser wall 5 ϕ = ( 3 )x3 − 5xy2 Show that the continuity equation is satisfied and determine the corresponding stream function. Section 6.5 Some Basic, Plane Potential Flows 6.38 As illustrated in Fig. P6.38, a tornado can be approximated by a free vortex of strength Γ for r > Rc, where Rc is the radius of the core. Velocity measurements at points A and B indicate that VA = 125 ft/s and VB = 60 ft/s. Determine the distance from point 2m r O Flow Entrance ■ Figure P6.40 7m 20° Exit Problems 6.41 When water discharges from a tank through an opening in its bottom, a vortex may form with a curved surface profile, as shown in Fig. P6.41 and Video V6.4. Assume that the velocity distribution in the vortex is the same as that for a free vortex. At the same time the water is being discharged from the tank at point A, it is desired to discharge a small quantity of water through the pipe B. As the discharge through A is increased, the strength of the vortex, as indicated by its circulation, is increased. Determine the maximum strength that the vortex can have in order that no air is sucked in at B. Express your answer in terms of the circulation. Assume that the fluid level in the tank at a large distance from the opening at A remains constant and viscous effects are negligible. 1 ft B 2 ft 339 y V _ r e x ■ Figure P6.44 6.45 For a free vortex (see Video V6.4) determine an expression for the pressure gradient (a) along a streamline, and (b) normal to a streamline. Assume that the streamline is in a horizontal plane, and express your answer in terms of the circulation. 6.46 (See The Wide World of Fluids article titled “Some Hurricane Facts,” Section 6.5.3.) Consider a category five hurricane that has a maximum wind speed of 160 mph at the eye wall, 10 mi from the center of the hurricane. If the flow in the hurricane outside of the hurricane’s eye is approximated as a free vortex, determine the wind speeds at locations 20 mi, 30 mi, and 40 mi from the center of the storm. A ■ Figure P6.41 6.42 Water flows over a flat surface at 4 ft/s, as shown in Fig. P6.42. A pump draws off water through a narrow slit at a volume rate of 0.1 ft3/s per foot length of the slit. Assume that the fluid is incompressible and inviscid and can be represented by the combination of a uniform flow and a sink. Locate the stagnation point on the wall (point A) and determine the equation for the stagnation streamline. How far above the surface, H, must the fluid be so that it does not get sucked into the slit? 4 ft/s H A 6.47 Consider the flow of a liquid of viscosity μ and density ρ down an inclined plate making an angle θ with the horizontal. The film thickness is t and is constant. The fluid velocity parallel to the plate is given by Vx = where y is the coordinate normal to the plate. Calculate Φ and Ψ for this flow and show that neither satisfies Laplace’s equation. Why not? 6.48 A horizontal oil-hearing stratum is 3 m high and, after hydraulic fracturing (“fracking”), provides a volume flow rate of Q = 1000 m3/hr. The flow moves radially inward and is collected by a vertical porous pipe having an outer radius of 1.0 m. The Laplace equation for the potential flow model of the oil is ∂2Ф 1 ∂2Ф 1 ∂Ф + + 2 ( 2 ) = 0. 2 ) ( r ∂r ∂r r ∂θ 3 0.1 ft /s (per foot of length of slit) ■ Figure P6.42 ρt2g cos θ y 2 1 − [ ( t ) ], 2μ Find Φ. 6.49 Show that the circulation of a free vortex for any closed path that does not enclose the origin is zero. 6.43 Two sources, one of strength m and the other with strength 3m, are located on the x axis as shown in Fig. P6.43. Determine the location of the stagnation point in the flow produced by these sources. 6.50 Show that the circulation of a free vortex for any closed path that encloses the origin is Г. Section 6.6 Superposition of Basic, Plane Potential Flows 2ℓ y 3ℓ x +m +3m ■ Figure P6.43 6.44 The velocity potential for a spiral vortex flow is given by ϕ = (Γ/2π) θ − (m/2π) ln r, where Γ and m are constants. Show that the angle, α, between the velocity vector and the radial direction is constant throughout the flow field (see Fig. P6.44). 6.51 Potential flow against a flat plate (Fig. P6.51a) can be described with the stream function ψ = Axy where A is a constant. This type of flow is commonly called a stagnation point flow since it can be used to describe the flow in the vicinity of the stagnation point at O. By adding a source of strength m at O, stagnation point flow against a flat plate with a “bump” is obtained as illustrated in Fig. P6.51b. Determine the relationship between the bump height, h, the constant, A, and the source strength, m. 340 Chapter 6 ■ Differential Analysis of Fluid Flow y near the end of the pond so that water can be pumped out. When water is pumped at the rate of 0.08 m3/s through a 3-m-long pipe, what will be the velocity at point A? Hint: Consider the flow inside a half-body. (See Video V6.5.) 6.56 A Rankine oval is formed by combining a source–sink pair, each having a strength of 36 ft2/s and separated by a distance of 12 ft along the x axis, with a uniform velocity of 10 ft/s (in the positive x direction). Determine the length and thickness of the oval. O x (a) 6.57 A 15-mph wind flows over a Quonset hut having a radius of 12 ft and a length of 60 ft, as shown in Fig. P6.57. The upstream pressure and temperature are equal to those inside the Quonset hut: 14.696 psia and 70 °F. Estimate the upward force on the Quonset hut. Find the location θ on the roof of the Quonset hut where the pressure is p∞. y V∞ p∞, T∞ p∞, T∞ h θ x Source 12 ft ■ Figure P6.57 (b) ■ Figure P6.51 6.52 The combination of a uniform flow and a source can be used to describe flow around a streamlined body called a halfbody. (See Video V6.5.) Assume that a certain body has the shape of a half-body with a thickness of 0.5 m. If this body is placed in an airstream moving at 15 m/s, what source strength is required to simulate flow around the body? 6.58 An ideal fluid flows past an infinitely long, semicircular “hump” located along a plane boundary, as shown in Fig. P6.58. Far from the hump the velocity field is uniform, and the pressure is p0. (a) Determine expressions for the maximum and minimum values of the pressure along the hump, and indicate where these points are located. Express your answer in terms of ρ, U, and p0. (b) If the solid surface is the ψ = 0 streamline, determine the equation of the streamline passing through the point θ = π/2, r = 2a. U, p0 6.53 Show that if Φ1 and Φ2 are both solutions of Laplace’s equation, the sum (Φ1 + Φ2) is also a solution. 6.54 The flow in the impeller of a centrifugal pump is modeled by the superposition of a source and a free vortex. The impeller has an outer diameter of 0.5 m and an inner diameter of 0.3 m. At the outlet from the impeller, the flowing water has the following velocity components, relative to the impeller: radial component 2 m/s and tangential component 7 m/s. (a) Find the strength of the source and the vortex required to model this flow. (b) Assume that the impeller blades are shaped like the streamlines and plot an impeller blade shape. (c) Find the radial and tangential components of velocity at the inlet to the impeller. 6.55 One end of a pond has a shoreline that resembles a halfbody as shown in Fig. P6.55. A vertical porous pipe is located r a e ■ Figure P6.58 6.59 Assume that the flow around the long circular cylinder of Fig. P6.59 is nonviscous and incompressible. Two pressures, p1 and p2, are measured on the surface of the cylinder, as illustrated. It is proposed that the free-stream velocity, U, can be related to the pressure difference Δp = p1 − p2 by the equation U=C Δp B ρ where ρ is the fluid density. Determine the value of the constant C. Neglect body forces. U y p2 Pipe A p1 60$ a x ■ Figure P6.59 5m 15 m ■ Figure P6.55 *6.60 Consider the steady potential flow around the circular cylinder shown in Fig. 6.26. On a plot show the variation of the magnitude Problems of the dimensionless fluid velocity, V/U, along the positive y axis. At what distance, y/a (along the y axis), is the velocity within 1% of the free-stream velocity? 6.61 A highway sign has the shape of a half-cylinder with radius 1.0 m and length 25.0 m. A rearward 10-km/hr wind is blowing over the sign as shown in Fig. P6.61. y 6.65 Consider the possibility of using two rotating cylinders to replace the conventional wings on an airplane for lift. Consider an airplane flying at 700 km/hr through the Standard Atmosphere at 10,000 m. Each “wing cylinder” has a 3.0-m radius. The surface velocity of each cylinder is 28 km/hr. Find the length ℓ of each wing to develop a lift of 1550 kN. Assume potential flow and neglect end effects. 6.66 Air at 25 °C flows normal to the axis of an infinitely long cylinder of 1.0-m radius. The approach velocity is 100 km/hr. Find the rotational velocity of the cylinder so that a single stagnation point occurs on the lower-most part of the cylinder. Assume potential flow. Wind p R 2m θ V∞ x Section 6.8 Viscous Flow Message here 6.67 Determine the shearing stress for an incompressible Newtonian fluid with a velocity distribution of V = (3xy2 − 4x3)i + (12x2y − y3)j. 6.68 The two-dimensional velocity field for an incompressible Newtonian fluid is described by the relationship Length Perpendicular to paper = 25 m ■ Figure P6.61 V = (12xy2 − 6x3 )iˆ + (18x2y − 4y3 )jˆ The pressure on the flat face is constant and equal to the pressure at the corner. Find the force on the sign. Neglect end effects. 6.62 Air at 25 °C flows normal to the axis of an infinitely long cylinder of 1.0-m radius. The cylinder is rotating at 10 rad/s, and the approach velocity is 100 km/hr. Find the maximum and minimum pressures on the cylinder surface. Assume potential flow. 6.63 The velocity potential for a cylinder (Fig. P6.63) rotating in a uniform stream of fluid is ϕ = Ur (1 + a2 Γ cos θ + θ 2π r2)   where Γ is the circulation. For what value of the circulation will the stagnation point be located at: (a) point A; (b) point B? U 341 y where the velocity has units of m/s when x and y are in meters. Determine the stresses σxx, σyy, and τxy at the point x = 0.5 m, y = 1.0 m if pressure at this point is 6 kPa and the fluid is glycerin at 20 °C. Show these stresses on a sketch. 6.69 The velocity of a fluid particle moving along a horizontal streamline that coincides with the x axis in a plane, two-dimensional, incompressible flow field was experimentally found to be described by the equation u = x2. Along this streamline determine an expression for (a) the rate of change of the υ component of velocity with respect to y, (b) the acceleration of the particle, and (c) the pressure gradient in the x direction. The fluid is Newtonian. 6.70 “Stokes’s first problem” involves the instantaneous acceleration at time t = 0 of a flat plate to a constant velocity U0 while in contact with a “semi-infinite,” static fluid as shown in Fig. P6.70. For a constant fluid density and viscosity, the simplified Navier– Stokes equation is r θ a A ∂u ∂ 2u =ν 2 ∂t ∂y x B ■ Figure P6.63 6.64 (See The Wide World of Fluids article titled “A Sailing Ship without Sails,” Section 6.6.3.) Determine the magnitude of the total force developed by the two rotating cylinders on the Flettner “rotor-ship” due to the Magnus effect. Assume a wind-speed relative to the ship of (a) 10 mph and (b) 30 mph. Each cylinder has a diameter of 9 ft, a length of 50 ft, and rotates at 750 rev/min. Use Eq. 6.124 and calculate the circulation by assuming the air sticks to the rotating cylinders. Note: This calculated force is at right angles to the direction of the wind and it is the component of this force in the direction of motion of the ship that gives the propulsive thrust. Also, due to viscous effects, the actual propulsive thrust will be smaller than that calculated from Eq. 6.124, which is based on inviscid flow theory. where u is the fluid velocity in the x or velocity U0 direction and y is a coordinate normal to the plate. Find the appropriate boundary conditions and initial conditions for this problem and then solve the differential equation to determine the velocity distribution u/U0 = f(y, t). Hint: Assume that f is a function of a single variable η where η = y/2 √νt. “Semi-infinite” fluid t ≤ 0, u = 0 u ( y, t ) y Plate U0 ■ Figure P6.70 x 342 Chapter 6 ■ Differential Analysis of Fluid Flow Section 6.9.1 Steady, Laminar Flow between Fixed Parallel Plates 6.71 Oil (SAE 30) at 15.6 °C flows steadily between fixed, horizontal, parallel plates. The pressure drop per unit length along the channel is 30 kPa/m, and the distance between the plates is 4 mm. The flow is laminar. Determine: (a) the volume rate of flow (per meter of width), (b) the magnitude and direction of the shearing stress acting on the bottom plate, and (c) the velocity along the centerline of the channel. 6.72 Two fixed, horizontal, parallel plates are spaced 0.4 in. apart. A viscous liquid (μ = 8 × 10−3 lb ⋅ s/ft2, SG = 0.9) flows between the plates with a mean velocity of 0.5 ft/s. The flow is laminar. Determine the pressure drop per unit length in the direction of flow. What is the maximum velocity in the channel? 6.73 A liquid of constant density ρ and constant viscosity μ flows down a wide, long inclined flat plate. The plate makes an angle θ with the horizontal. The velocity components do not change in the direction of the plate, and the fluid depth, h, normal to the plate is constant. There is negligible shear stress by the air on the fluid. Find the velocity profile u(y), where u is the velocity parallel to the plate and y is measured perpendicular to the plate. Write an expression for the volume flow rate per unit width of the plate. 6.74 We will see in Chapter 8 that the pressure drop in fully developed pipe flow is sometimes computed with the aid of a friction factor, defined by Δp D f=1 2 ℓ 2 ρV where V is the average velocity and ℓ is the length of pipe over which Δp occurs. For laminar fully developed flow, f can be evaluated from C f= Re where C is a constant and Re is the Reynolds number, given by Re = ρVD/μ. Find the constant C for a parallel plate channel. Take D as 2h. 6.75 (See The Wide World of Fluids article titled “10 Tons on 8 psi,” Section 6.9.1.) A massive, precisely machined, 6-ft-diameter granite sphere rests on a 4-ft-diameter cylindrical pedestal as shown in Fig. P6.75. When the pump is turned on and the water pressure within the pedestal reaches 8 psi, the sphere rises off the pedestal, creating a 0.005-in. gap through which the water flows. The sphere can then be rotated about any axis with minimal friction. (a) Estimate the pump flowrate, Q0, required to accomplish this. Assume the flow in the gap between the sphere and the pedestal is essentially viscous flow between fixed, parallel plates. (b) Describe what would happen if the pump flowrate were increased to 2Q0. 6.76 The bearing shown in Fig. P6.76 consists of two parallel discs of radius R separated from each other by a small distance h(h << R). The lower disc is made of a porous material, and an incompressible, viscous fluid is pumped through it and into the gap, filling the space between the discs completely. The pores of the lower disc are closely spaced, so that the velocity of the fluid as it leaves the surface of the porous disc may be regarded as a uniform value w0, that is, small compared to the mean radial velocity. Find the radial velocity u and the load L that the bearing can support as a function of w0, μ, R, and h. The inertia terms are negligible, the circumferential velocity is zero (Vθ = 0), and angular symmetry exists ( ∂/∂θ = 0). L R h Flow ■ Figure P6.76 6.77 A Bingham plastic is a fluid in which the stress τ is related to the rate of strain du/dy by τ = τ0 + μ( du , dy ) where τ0 and μ are constants. Consider the flow of a Bingham plastic between two fixed, horizontal, infinitely wide, flat plates. For fully developed flow with du/dx = 0 and dp/dx constant, find u( y). Section 6.9.2 Couette Flow 6 ft 0.005 in. 4 in. 4 ft Pump ■ Figure P6.75 p = 8 psi Two horizontal, infinite, parallel plates are spaced a dis6.78 tance b apart. A viscous liquid is contained between the plates. The bottom plate is fixed, and the upper plate moves parallel to the bottom plate with a velocity U. Because of the no-slip boundary condition (see Video V6.12), the liquid motion is caused by the liquid being dragged along by the moving boundary. There is no pressure gradient in the direction of flow. Note that this is a socalled simple Couette flow discussed in Section 6.9.2. (a) Start with the Navier–Stokes equations and determine the velocity distribution between the plates. (b) Determine an expression for the flowrate passing between the plates (for a unit width). Express your answer in terms of b and U. 6.79 An incompressible, viscous fluid is placed between horizontal, infinite, parallel plates as is shown in Fig. P6.79. The two plates move in opposite directions with constant velocities, U1 and U2, as shown. The pressure gradient in the x direction is zero, and the only body force is due to the fluid weight. Use the Navier–Stokes equations to derive an expression for the velocity distribution between the plates. Assume laminar flow. Problems b 343 U1 connected between two points along the bottom indicates a differential reading of 0.1 in. If the upper plate moves with a velocity of 0.02 ft/s, at what distance from the bottom plate does the maximum velocity in the gap between the two plates occur? Assume laminar flow. x 6.83 A flat block is pulled along a horizontal flat surface by a horizontal rope perpendicular to one of the sides. The block measures 1.0 m × 1.0 m, has a mass of 100 kg and a constant velocity of 1.0 m/s, and is separated from the flat surface by a 0.10-cmthick oil layer of 15 °C SAE 20 crankcase oil. Find the coefficient of sliding friction for the block. Does this coefficient of friction change with the velocity? y U2 ■ Figure P6.79 6.80 Two immiscible, incompressible, viscous fluids having the same densities but different viscosities are contained between two infinite, horizontal, parallel plates (Fig. P6.80). The bottom plate is fixed, and the upper plate moves with a constant velocity U. Determine the velocity at the interface. Express your answer in terms of U, μ1, and μ2. The motion of the fluid is caused entirely by the movement of the upper plate; that is, there is no pressure gradient in the x direction. The fluid velocity and shearing stress are continuous across the interface between the two fluids. Assume laminar flow. U 6.84 A viscosity motor/pump is shown in Fig. P6.84. The rotor is concentric within a stationary housing. The clearance h between the housing and the rotor is small compared to the width w and radius R of the rotor. A seal divides the clearance space as shown. When the device is operated as a motor, the viscous liquid is introduced at the high-pressure side of the seal and leaves at the lowpressure side of the seal. This flow causes the rotor to turn and produce power. For R = 1.0 ft, h = 0.50 in., w = 1.0 ft, 𝜔 = 10 rpm, and a flow rate of 1.5 ft3/min, find the power output of the rotor. The fluid is 60 °F, SAE 30 crankcase oil. Neglect gravity. h ρ , μ1 h y h Housing ρ, μ 2 x Rotor + Fixed plate R ω phigh ■ Figure P6.80 6.81 The disc shown in Fig. P6.81 has a diameter D of 1 m and a rotational speed of 1800 rpm. It is positioned 4 mm from a solid boundary. The gap is filled with 15 °C SAE 10W oil. Find the torque required to overcome the frictional resistance of the oil on the disc. Assume that the circumferential velocity (Vθ ) is linear across the gap at the each radius r. ω 4 mm plow Seal (no drag on rotor) ■ Figure P6.84 6.85 A vertical shaft passes through a bearing and is lubricated with an oil having a viscosity of 0.2 N ⋅ s/m2 as shown in Fig. P6.85. Assume that the flow characteristics in the gap between the shaft and bearing are the same as those for laminar flow between infinite parallel plates with zero pressure gradient in the direction of flow. Estimate the torque required to overcome viscous resistance when the shaft is turning at 80 rev/min. Shaft D 75 mm ■ Figure P6.81 Bearing 6.82 A viscous fluid (specific weight = 80 lb/ft ; viscosity = 0.03 lb ⋅ s/ft2) is contained between two infinite, horizontal parallel plates as shown in Fig. P6.82. The fluid moves between the plates under the action of a pressure gradient, and the upper plate moves with a velocity U while the bottom plate is fixed. A U-tube manometer 3 U = 0.02 ft/s 1.0 in. 6 in. Oil 0.25 mm ■ Figure P6.85 y x Fixed plate 0.1 in. γ = 100 lb/ft3 ■ Figure P6.82 160 mm 6.86 A viscous fluid is contained between two long concentric cylinders. The geometry of the system is such that the flow between the cylinders is approximately the same as the laminar flow between two infinite parallel plates. (a) Determine an expression for the torque required to rotate the outer cylinder with an angular velocity ω. The inner cylinder is fixed. Express your answer in terms of the geometry of the system, the viscosity of the fluid, and the angular velocity. (b) For a small, rectangular element located at the fixed wall, determine an expression for the rate of angular deformation of this element. (See Video V6.3 and Fig. P6.10.) 344 Chapter 6 ■ Differential Analysis of Fluid Flow Section 6.9.3 Steady, Laminar Flow in Circular Tubes 6.87 Verify that the momentum correction factor β for fully developed, laminar flow in a circular tube is 4/3. manometer is connected to the tube as shown to measure the pressure drop along the tube. When the differential reading, Δh, is 9 mm, what is the mean velocity in the tube? 6.88 Verify that the kinetic energy correction factor α for fully developed, laminar flow in a circular tube is 2.0. 6.89 A simple flow system to be used for steady-flow tests consists of a constant head tank connected to a length of 4-mm-diameter tubing as shown in Fig. P6.89. The liquid has a viscosity of 0.015 N ⋅ s/m2, a density of 1200 kg/m3, and discharges into the atmosphere with a mean velocity of 2 m/s. (a) Verify that the flow will be laminar. (b) The flow is fully developed in the last 3 m of the tube. What is the pressure at the pressure gage? (c) What is the magnitude of the wall shearing stress, τrz, in the fully developed region? 4 mm 2m Δh Density of gage fluid = 2000 kg/m3 ■ Figure P6.93 6.94 Fluid with kinematic viscosity ν flows down an inclined circular pipe of length ℓ and diameter D with flow rate Q. Find the vertical drop per unit length of the pipe so that the pressure drop (p1 − p2) is zero for laminar flow. Pressure gage 3m Diameter = 4 mm ■ Figure P6.89 6.90 (a) Show that for Poiseuille flow in a tube of radius R the magnitude of the wall shearing stress, τrz, can be obtained from the relationship  (τrz ) wall  = 6.95 Blood flows at volume rate Q in a circular tube of radius R. The blood cells concentrate and flow near the center of the tube, while the cell-free fluid (plasma) flows in the outer region. The center core of radius Rc has a viscosity μc, and the plasma has a viscosity μp. Assume laminar, fully developed flow for both the core and plasma flows and show that an “apparent” viscosity is defined by μapp ≡ 4μQ πR3 for a Newtonian fluid of viscosity μ. The volume rate of flow is Q. (b) Determine the magnitude of the wall shearing stress for a fluid having a viscosity of 0.004 N ⋅ s/m2 flowing with an average velocity of 130 mm/s in a 2-mm-diameter tube. 6.91 An infinitely long, solid, vertical cylinder of radius R is located in an infinite mass of an incompressible fluid. Start with the Navier–Stokes equation in the θ direction and derive an expression for the velocity distribution for the steady-flow case in which the cylinder is rotating about a fixed axis with a constant angular velocity ω. You need not consider body forces. Assume that the flow is axisymmetric and the fluid is at rest at infinity. is given by μapp = f= C Re where C is a constant and Re is the Reynolds number, given by Re = ρVD/μ. Find the constant C for flow in a circular tube. 6.93 A liquid (viscosity = 0.002 N ⋅ s/m2; density = 1000 kg/m3) is forced through the circular tube shown in Fig. P6.93. A differential 1 − (Rc R) 4 (1 − μp μc ) 6.96 An incompressible Newtonian fluid flows steadily between two infinitely long, concentric cylinders as shown in Fig. P6.96. The outer cylinder is fixed, but the inner cylinder moves with a longitudinal velocity V0 as shown. The pressure gradient in the axial direction is −Δp/ℓ. For what value of V0 will the drag on the inner cylinder be zero? Assume that the flow is laminar, axisymmetric, and fully developed. Fixed wall ro Δp D 1 2 ℓ 2 ρV where V is the average velocity, D is the pipe diameter, and ℓ is the length of pipe over which Δp occurs. For laminar fully developed flow, f can be evaluated from μp Section 6.9.4 Steady, Axial, Laminar Flow in an Annulus 6.92 We will see in Chapter 8 that the pressure drop in fully developed pipe flow is sometimes computed with the aid of a friction factor, defined by f= πR4 Δp 8LQ V0 ri ■ Figure P6.96 6.97 A viscous fluid is contained between two infinitely long, vertical, concentric cylinders. The outer cylinder has a radius ro and rotates with an angular velocity ω. The inner cylinder is fixed and has a radius ri. Make use of the Navier–Stokes equations to obtain an exact solution for the velocity distribution in the gap. Assume that the flow in the gap is axisymmetric (neither velocity Problems nor pressure are functions of angular position θ within the gap) and that there are no velocity components other than the tangential component. The only body force is the weight. 6.98 A double-pipe heat exchanger consists of two concentric tubes with one fluid flowing in the central tube and the other flowing in the annulus between the tubes. In a particular exchanger, cold (15 oC) water flows at 300 liters/min through a 10 cm diameter inner tube and hot (50 oC) SAE 30 oil flows at 400 liters/min in the annulus between the inner tube and a 12 cm diameter outer tube. The heat exchanger is 2 m in length. (a) Calculate the pressure drop in the oil stream (b) Can you calculate the pressure drop in the water using the same equations? Why or why not? 6.99 A wire of diameter d is stretched along the centerline of a pipe of diameter D. For a given pressure drop per unit length of pipe, by how much does the presence of the wire reduce the flowrate if (a) dD = 0.1; (b) dD = 0.01? Section 6.10 Other Aspects of Differential Analysis 6.100 Look through several professional society magazines such as Civil Engineering, Mechanical Engineering, or ASEE Prism. Count the advertisements for CFD software and calculate an “ad density” in advertisements per issue. 6.101 Consider the Navier-Stokes equations as written in Eq. 6.158. Write the physical meaning of each term in the equation; for example, “The term ρg is the gravity force (or weight) per unit volume.” 345 6.102 Find out what CFD software packages your college or university has. How many “seats” of each? How much does your institution spend annually on CFD software? ■ Lifelong Learning Problems 6.1 LL What sometimes appear at first glance to be simple fluid flows can contain subtle, complex fluid mechanics. One such example is the stirring of tea leaves in a teacup. Obtain information about “Einstein’s tea leaves” and investigate some of the complex fluid motions interacting with the leaves. Summarize your findings in a brief report. 6.2 LL Computational fluid dynamics (CFD) has moved from a research tool to a design tool for engineering. Initially, much of the work in CFD was focused in the aerospace industry, but now has expanded into other areas. Obtain information on what other industries (e.g., automotive) make use of CFD in their engineering design. Summarize your findings in a brief report. 6.3 LL Find information about the (proposed) Madaras Rotor Powerplant. What is it? How would it work? When was it first proposed? Has there been interest more recently? Do you think it is feasible? Summarize your findings in a short report. 7 Dimensional Analysis, Similitude, and Modeling Learning Objectives After completing this chapter, you should be able to: ■ apply the Buckingham pi theorem. ■ develop a set of dimensionless variables for a given flow situation. ■ recognize and use common dimensionless groups ■ discuss the use of dimensionless variables in data analysis. ■ apply the concepts of modeling and similitude to develop prediction equations. Experimentation and modeling are widely used techniques in fluid mechanics. VIDEO 346 V7.1 Real and model flies Although many practical engineering problems involving fluid mechanics can be solved by using the equations and analytical procedures described in the preceding chapters, there remain a large number of problems that rely on experimental data for their solution. In fact, it is probably fair to say that very few problems involving real flows can be solved by analysis alone. The solution to many problems is achieved through the use of a combination of theoretical and numerical (computer) analysis and experimental data. Thus, engineers working on fluid mechanics problems should be familiar with the experimental approach to these problems so that they can interpret and make use of data obtained by others, such as might appear in handbooks, or be able to plan and execute the necessary experiments in their own laboratories. In this chapter we consider some techniques and ideas that are important in the planning and execution of experiments, as well as in understanding and correlating data that may have been obtained by other experimenters. An obvious goal of any experiment is to make the results as widely applicable as possible. To achieve this end, the concept of similitude is often used so that measurements made on one system (for example, in the laboratory) can be used to describe the behavior of other similar systems (outside the laboratory). The laboratory systems are usually thought of as models and are used to study the phenomenon of interest under carefully controlled conditions. From these model studies, empirical formulations can be developed, or specific predictions of one or more characteristics of some other similar system can be made. To do this, it is necessary to establish the relationship between the laboratory model and the “other” system. In the following sections, we find out how this can be accomplished in a systematic manner. The specific topic of this chapter is dimensional analysis. Dimensional analysis is a packaging or compacting technique used to reduce the complexity of experimental programs and at the same time increase the generality of experimental information. The information packaging produced by dimensional analysis also has other uses. It is equally applicable to theoretical (analytic) and computational investigations and results. Also, dimensional analysis gives us the tools to decide 7.1 The Need for Dimensional Analysis 347 a priori whether a flow is likely to be laminar or turbulent or whether simplifying assumptions, such as neglecting compressibility, are appropriate. Before we discuss the subject in detail, we need to point out the following about dimensional analysis: ■ ■ ■ Dimensional analysis is sometimes very confusing to beginners. The basic principle involved seems obvious, almost trivial. The execution of the technique, however, often seems almost “magical.” Confusion arises because we usually expect theories to be complete, that is, to begin at the beginning and carry through to the end. Dimensional analysis stops far short of providing a final answer to a particular engineering problem; it simply leads to a more efficient organization of the variables. The ultimate answer comes only in the laboratory or from analytic or computational methods. Dimensional analysis tells us that we can organize variables efficiently and suggests possible ways that the organization can be carried out; however, it does not allow us to find unique results. Two analogies help illustrate the utility of dimensional analysis. Dimensional analysis is similar to a trash compactor. Trash thrown into the compactor is crushed into a smaller space so it is easier to handle. The compactor does not solve the trash problem nor does it ultimately dispose of any of the trash put into it; it only reduces it to a more convenient package. In the same way, dimensional analysis reduces the number of variables that we must consider in an analytic or experimental investigation by “compacting” several of the raw variables into convenient packages. Dimensional analysis also is similar to packing a suitcase for a trip. The traveler must select the items to take. Any item that is not selected will not be in the suitcase. The traveler organizes the items in the suitcase in one of several equally convenient ways. Different travelers prefer different arrangements, but that does not make any arrangement superior. After the suitcase is packed, it is up to the traveler to carry it to the ultimate destination. The suitcase does not go on its own, and just packing it does not accomplish the objective of the trip. THE WIDE WORLD OF FLUIDS Model study of New Orleans levee breach caused by than a mile of the adjacent Lake Pontchartrain front. The objec- Hurricane Katrina Much of the devastation to New Orleans tive of the study is to obtain information regarding the effect from Hurricane Katrina in 2005 was a result of flood waters that that waves had on the breaching of the canal and to investi- surged through a breach of the 17th Street Outfall Canal. To gate the surging water currents within the canals. The waves better understand why this occurred and to determine what are generated by computer-controlled wave generators that can be done to prevent future occurrences, the U.S. Army can produce waves of varying heights, periods, and directions Engineer Research and Development Center Coastal and similar to the storm conditions that occurred during the hurri- Hydraulics Laboratory is conducting tests on a large (1:50 cane. Data from the study will be used to calibrate and validate length scale) 15,000-square-foot hydraulic model that repli- information that will be fed into various numerical model studies cates 0.5 mile of the canal surrounding the breach and more of the disaster. 7.1 The Need for Dimensional Analysis To illustrate a typical fluid mechanics problem in which experimentation is required, consider the steady flow of an incompressible Newtonian fluid through a long, smooth-walled, horizontal, circular pipe. An important characteristic of this system, which would be of interest to an engineer designing a pipeline, is the pressure drop per unit length that develops along the pipe as a result of friction. Although this would appear to be a relatively simple flow problem, it cannot generally be solved analytically (even with the aid of large computers) without the use of experimental data. The first step in planning an experiment to study this problem would be to decide on the variables that will have an effect on the pressure drop per unit length, Δpℓ [(lbft2 )ft = lbft3 or Nm3]. 348 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling We expect the list to include the pipe diameter, D, the fluid density, ρ, fluid viscosity, μ, and the mean velocity, V, at which the fluid is flowing through the pipe. Thus, we can express this relationship as Δpℓ = f (D, ρ, μ, V)   It is important to develop a meaningful and systematic way to perform an experiment. (7.1) which simply indicates mathematically that we expect the pressure drop per unit length to be some function of the factors contained within the parentheses. At this point the nature of the function is unknown, and the objective of the experiments to be performed is to determine the nature of this function. To perform the experiments in a meaningful and systematic manner, it would be necessary to change one of the variables, such as the velocity, while holding all others constant, and measure the corresponding pressure drop. This series of tests would yield data that could be represented graphically as is illustrated in Fig. 7.1a. It is to be noted that this plot would only be valid for the specific pipe and for the specific fluid used in the tests; this certainly does not give us the general formulation we are looking for. We could repeat the process by varying each of the other variables in turn, as is illustrated in Figs. 7.1b, 7.1c, and 7.1d. This approach to determining the functional relationship between the pressure drop and the various factors that influence it, although logical in concept, is fraught with difficulties. Some of the experiments would be hard to carry out—for example, to obtain the data illustrated in Fig. 7.1c it would be necessary to vary fluid density while holding viscosity constant. How would you do this? Finally, once we obtained the various curves shown in Figs. 7.1a, 7.1b, 7.1c, and 7.1d, how could we combine these data to obtain the desired general functional relationship between Δpℓ, D, ρ, μ, and V that would be valid for any similar pipe system? Fortunately, there is a much simpler approach to this problem that will minimize such difficulties. In the following sections we will show that rather than working with the original list of variables, as described in Eq. 7.1, we can collect these into two nondimensional combinations of variables (called dimensionless products or dimensionless groups) so that D Δpℓ ρVD = ϕ( 2 μ ) ρV (7.2) Thus, instead of having to work with five variables, we now have only two. The necessary experiment would simply consist of varying the dimensionless product ρVDμ and determining the corresponding value of D Δpℓ ρV 2. The results of the experiment could then be represented by a Δpℓ Δpℓ D, ρ , μ – constant V, ρ, μ– constant V D (a) (b) Δpℓ Δpℓ D, ρ , V– constant D, V, μ – constant μ ρ (c) (d) ■ Figure 7.1 Illustrative plots showing how the pressure drop in a pipe may be affected by several different factors. 7.2 Buckingham Pi Theorem 349 D Δpℓ _____ ρV 2 ρ VD ____ μ Dimensionless products are important and useful in the planning, execution, and interpretation of experiments. ■ Figure 7.2 An illustrative plot of pressure drop data using dimensionless parameters. single, universal curve as is illustrated in Fig. 7.2. This curve would be valid for any combination of smooth-walled pipe and incompressible Newtonian fluid. To obtain this curve we could choose a pipe of convenient size and a fluid that is easy to work with. Note that we wouldn’t have to use different pipe sizes or even different fluids. It is clear that the experiment would be much simpler, easier to do, and less expensive (which would certainly make an impression on your boss). And think how much less paper (or disk space) would be required to present the results. The basis for this simplification lies in a consideration of the dimensions of the variables involved. As was discussed in Chapter 1, a qualitative description of physical quantities can be given in terms of basic dimensions such as mass, M, length, L, and time, T.1 Alternatively, we could use force, F, L, and T as basic dimensions, since from Newton’s second law F ≐ MLT −2 (Recall from Chapter 1 that the notation ≐ is used to indicate dimensional equality.) The dimensions of the variables in the pipe flow example are Δpℓ ≐ FL−3, D ≐ L, ρ ≐ FL−4T 2, μ ≐ FL −2T, and V ≐ LT −1. [Note that the pressure drop per unit length has the dimensions of (FL2 )L = FL−3.] A quick check of the dimensions of the two groups that appear in Eq. 7.2 shows that they are in fact dimensionless products; that is, D Δpℓ L(FL3 ) ≐ ≐ F 0 L0 T 0 ρV 2 (FL−4T 2 )(LT −1 ) 2 and ρVD (FL −4 T 2 )(LT −1 )(L) ≐ ≐ F 0 L0 T 0 μ (FL−2T)   Not only have we reduced the number of variables from five to two, but also the new groups are dimensionless combinations of variables, which means that the results presented in the form of Fig. 7.2 will be independent of the system of units we choose to use. This type of analysis is called dimensional analysis, and the basis for its application to a wide variety of problems is found in the Buckingham pi theorem described in the following section. 7.2 Buckingham Pi Theorem There are two fundamental questions that we must answer. The first is “Is Eq. 7.2 a fluke, or can we always reduce the number of variables by dimensional analysis? “The second is” How many dimensionless products are required to replace the original list of variables?” The answer to these questions is supplied by the basic theorem of dimensional analysis that states the following: If an equation involving k variables is dimensionally homogeneous, it can be reduced to a relationship among k − r independent dimensionless products, where r is the minimum number of reference dimensions required to describe the variables. 1 As noted in Chapter 1, we will use T to represent the basic dimension of time, although T is also used for temperature in thermodynamic relationships (such as the ideal gas law). The dimension of temperature is often represented by θ. 350 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling Dimensional analysis is based on the Buckingham pi theorem. The dimensionless products are frequently referred to as pi terms, and the theorem is called the Buckingham pi theorem.2 Edgar Buckingham used the symbol Π to represent a dimensionless product, and this notation is commonly used. Although the pi theorem is a simple one, its proof is not so simple and we will not include it here. Many entire books have been devoted to the subject of similitude and dimensional analysis, and a number of these are listed at the end of this chapter (Refs. 1–15). Students interested in pursuing the subject in more depth (including the proof of the pi theorem) can refer to one of these books. The pi theorem is based on the idea of dimensional homogeneity which was introduced in Chapter 1. Essentially for any physically meaningful equation involving k variables, such as u1 = f(u2, u3, . . . , uk ) the dimensions of the variable on the left side of the equal sign must be equal to the dimensions of any term that stands by itself on the right side of the equal sign. The pi theorem then guarantees that we can rearrange the equation into a set of dimensionless products (pi terms) so that Π 1 = ϕ(Π 2, Π 3, . . . , Π k−r ) where ϕ(Π 2, Π 3, . . . , Π k−r ) is a function of Π 2 through Π k−r. The required number of pi terms is fewer than the number of original variables by r, where r is determined by the minimum number of reference dimensions required to describe the original list of variables. Usually the reference dimensions required to describe the variables in fluid mechanics will be the basic dimensions M, L, and T or F, L, and T. However, in some instances perhaps only two dimensions, such as L and T, are required, or maybe just one, such as L. Also, in a few rare cases the variables may be described by some combination of basic dimensions, such as MT 2 and L, and in this case r would be equal to two rather than three. Although the use of the pi theorem may appear to be a little mysterious and complicated, we will develop a simple, systematic procedure for developing the pi terms for a given problem. 7.3 A dimensional analysis can be performed using a series of distinct steps. Determination of Pi Terms Several methods can be used to form the dimensionless products, or pi terms, that arise in a dimensional analysis. To begin we will consider a method that will allow us to systematically form the pi terms so that we are sure that they are dimensionless and independent and that we have the right number. The method we will describe in detail in this section is called the method of repeating variables. It will be helpful to break the repeating variable method down into a series of distinct steps that can be followed for any given problem. With a little practice you will be able to readily complete a dimensional analysis for your problem. Step 1 List all the variables that are involved in the problem. This step is the most difficult one, and it is, of course, vitally important that all pertinent variables be included. Otherwise the dimensional analysis will not be correct! We are using the term variable to include any quantity, including dimensional and nondimensional constants, which play a role in the phenomenon under investigation. All such quantities should be included in the list of variables to be considered for the dimensional analysis. The determination of the variables must be accomplished by knowledge of the problem and the physical laws that govern the phenomenon. Typically the variables will include those that are necessary to describe the geometry of the system (such as a pipe diameter), to define any fluid properties (such as a fluid viscosity), and to indicate external effects that influence the system (such as a driving pressure drop per unit length). These general classes of variables are intended as broad categories that should be helpful in identifying variables. It is likely, however, that there 2 Although several early investigators, including Lord Rayleigh (1842–1919) in the nineteenth century, contributed to the development of dimensional analysis, Edgar Buckingham’s (1867–1940) name is usually associated with the basic theorem. He stimulated interest in the subject in the United States through his publications during the early part of the twentieth century. See, for example, E. Buckingham, On Physically Similar Systems: Illustrations of the Use of Dimensional Equations, Phys. Rev., 4 (1914), 345–376. 7.3 Step 2 Step 3 Step 4 Step 5 Step 6 Determination of Pi Terms 351 will be variables that do not fit easily into one of these categories, and each problem needs to be carefully analyzed. Any dimension must appear in at least two parameters, or obtaining dimensionless groups would not be possible. Whenever a dimension is represented in only one parameter, either that parameter is irrelevant or you must include another parameter that also includes the unique dimension. Since we wish to keep the number of variables to a minimum, so that we can minimize the amount of work, it is important that all variables be independent. For example, if in a certain problem the cross-sectional area of a pipe is an important variable, either the area or the pipe diameter could be used, but not both, since they are obviously not independent. Similarly, if both fluid density, ρ, and specific weight, γ, are important variables, we could list ρ and γ, or ρ and g (acceleration of gravity), or γ and g. However, it would be incorrect to use all three since γ = ρg; that is, ρ, γ, and g are not independent. Note that although g would normally be constant in a given experiment, that fact is irrelevant as far as a dimensional analysis is concerned. Express each of the variables in terms of basic dimensions. For the typical fluid mechanics problem the basic dimensions will be either M, L, and T or F, L, and T. Dimensionally these two sets are related through Newton’s second law (F = ma) so that F ≐ MLT −2. For example, ρ ≐ ML−3 or ρ ≐ FL−4T 2. Thus, either set can be used. The basic dimensions for typical variables found in fluid mechanics problems are listed in Table 1.1 in Chapter 1. Determine the required number of pi terms. This can be accomplished by means of the Buckingham pi theorem, which indicates that the number of pi terms is equal to k − r, where k is the number of variables in the problem (which is determined from Step 1) and r is the number of reference dimensions required to describe these variables (which is determined from Step 2). The reference dimensions usually correspond to the basic dimensions and can be determined by an inspection of the dimensions of the variables obtained in Step 2. As previously noted, there may be occasions (usually rare) in which the basic dimensions appear in combinations so that the number of reference dimensions is less than the number of basic dimensions. In this case, the number of dimensions in the MLT and FLT systems will be different. Then r will be the smaller of the two. This possibility is illustrated in Example 7.2. Select a number of repeating variables, where the number required is equal to the number of reference dimensions. Essentially what we are doing here is selecting from the original list of variables several of which can be combined with each of the remaining variables to form a pi term. All of the required reference dimensions must be included within the group of repeating variables, and each repeating variable must be dimensionally independent of the others (i.e., the dimensions of one repeating variable cannot be reproduced by some combination of products of powers of the remaining repeating variables). This means that the repeating variables cannot themselves be combined to form a dimensionless product. For any given problem we usually are interested in determining how one particular variable is influenced by the other variables. We would consider this variable to be the dependent variable, and we would want this to appear in only one pi term. Thus, do not choose the dependent variable as one of the repeating variables, since the repeating variables will generally appear in more than one pi term. Other tips for choosing repeating variables are do not choose a variable whose effect on the problem is questionable or important only part of the time and select variables as close to dimensionally pure as possible (pipe diameter is pure length, fluid density is almost pure mass, etc.) Form a pi term by multiplying one of the nonrepeating variables by the product of the repeating variables, each raised to an exponent that will make the combination dimensionless. Essentially each pi term will be of the form uiua1iub2iuc3i where ui is one of the nonrepeating variables; u1, u2, and u3 are the repeating variables; and the exponents ai, bi, and ci are determined so that the combination is dimensionless. Repeat Step 5 for each of the remaining nonrepeating variables. The resulting set of pi terms will correspond to the required number obtained from Step 3. If not, check your work—you have made a mistake! 352 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling Step 7 Check all the resulting pi terms to make sure they are dimensionless. It is easy to make a mistake in forming the pi terms. However, this can be checked by simply substituting the dimensions of the variables into the pi terms to confirm that they are all dimensionless. One good way to do this is to express the variables in terms of M, L, and T if the basic dimensions F, L, and T were used initially, or vice versa, and then check to make sure the pi terms are dimensionless. Step 8 Rearrange the pi terms to simplify the problem or to correspond to customary usage. Later, we will see that certain dimensionless groups are commonly used in fluid mechanics and you can communicate better with other engineers if you use these groups. Step 9 Express the final form as a relationship among the pi terms, and think about what it means. Typically the final form can be written as Π 1 = ϕ(Π 2, Π 3, . . . , Π k−r ) By using dimensional analysis, the original problem is simplified and defined with pi terms. (1) (2) D V ρ, μ ℓ Δpℓ = (p1 – p2)/ℓ where Π 1 would contain the dependent variable in the numerator. It should be emphasized that if you started out with the correct list of variables (and the other steps were completed correctly), then the relationship in terms of the pi terms can be used to describe the problem. You need only work with the pi terms—not with the individual variables. However, it should be clearly noted that this is as far as we can go with the dimensional analysis; that is, the actual functional relationship among the pi terms must be determined by experiment or computation. To illustrate these steps we will again consider the problem discussed earlier in this chapter: the steady flow of an incompressible Newtonian fluid through a long, smooth-walled, horizontal circular pipe. We are interested in the pressure drop per unit length, Δpℓ, along the pipe as illustrated by the figure in the margin. First (Step 1) we must list all of the pertinent variables that are involved based on the experimenter’s knowledge of the problem. In this problem we assume that Δpℓ = f(D, ρ, μ, V) where D is the pipe diameter, ρ and μ are the fluid density and viscosity, respectively, and V is the mean velocity. Next (Step 2) we express all the variables in terms of basic dimensions. Using F, L, and T as basic dimensions it follows that Δpℓ ≐ FL−3 D≐L ρ ≐ FL−4T 2 μ ≐ FL−2 T   V ≐ LT −1 Special attention should be given to the selection of repeating variables as detailed in Step 4. We could also use M, L, and T as basic dimensions if desired—the final result will be the same. Note that for density, which is a mass per unit volume (ML−3 ), we have used the relationship F ≐ MLT −2 to express the density in terms of F, L, and T. Do not mix the basic dimensions; that is, use either F, L, and T or M, L, and T. We can now apply the pi theorem to determine the required number of pi terms (Step 3). An inspection of the dimensions of the variables from Step 2 reveals that all three basic dimensions are required to describe the variables. Since there are five (k = 5) variables (do not forget to count the dependent variable, Δpℓ ) and three required reference dimensions (r = 3), then according to the pi theorem there will be (5 − 3), or two pi terms required. The repeating variables to be used to form the pi terms (Step 4) need to be selected from the list D, ρ, μ, and V. Remember, we do not want to use the dependent variable as one of the repeating variables. Since three reference dimensions are required, we will need to select three repeating variables. Generally, we would try to select as repeating variables those that are the simplest, dimensionally. For example, if one of the variables has the dimension of a length, choose it as one of the repeating variables. In this example we will use D, V, and ρ as repeating variables. Note that these 7.3 Determination of Pi Terms 353 are dimensionally independent, since D is a length, V involves both length and time, and ρ involves force, length, and time. This means that we cannot form a dimensionless product from this set. We are now ready to form the two pi terms (Step 5). Typically, we would start with the dependent variable and combine it with the repeating variables to form the first pi term; that is, Π 1 = Δpℓ DaV b ρc   Since this combination is to be dimensionless, it follows that (FL−3 )(L) a (LT −1 ) b (FL−4T 2 ) c ≐ F 0L0T 0 The exponents, a, b, and c must be determined such that the resulting exponent for each of the basic dimensions—F, L, and T — must be zero (so that the resulting combination is dimensionless). Thus, we can write 1 + c = 0  (for F) −3 + a + b − 4c = 0  (for L) −b + 2c = 0  (for T) The solution of this system of algebraic equations gives the desired values for a, b, and c. It follows that a = 1, b = −2, c = −1 and, therefore, Π1 = ΔpℓD ρV2 The process is now repeated for the remaining nonrepeating variables (Step 6). In this example there is only one additional variable (μ) so that Π 2 = μDaVbρc or (FL−2T)(L) a (LT−1 ) b (FL−4T2 ) c ≐ F0L0T0 and, therefore, Step 1 1 + c = 0  (for F) Δpℓ = f (D, ρ, μ, V ) −2 + a + b − 4c = 0  (for L) Step 2 1 − b + 2c = 0  (for T) Δpℓ = FL−3, ... Solving these equations simultaneously it follows that a = −1, b = −1, c = −1 so that Step 3 k–r=2 Π2 = Step 4 Note that we end up with the correct number of pi terms as determined from Step 3. At this point stop and check to make sure the pi terms are actually dimensionless (Step 7). We will check using both FLT and MLT dimensions. Thus, D, V, ρ Step 5 a b c Π1 = ΔpℓD V ρ Step 6 Step 7 ΔpℓD 0 0 0 _____ =F L T ρV2 −1 Π2 = Π2 ρVD = ____ μ Step 9 ρ ΔpℓD ∼ DV _____ = ϕ ____ ρV2 Π1 = ΔpℓD (FL−3 )(L) ≐ ≐ F 0 L0 T 0 ρV 2 (FL−4T 2 )(LT −1 ) 2 Π2 = μ (FL−2 T) ≐ ≐ F 0 L0 T 0 DVρ (L)(LT −1 )(FL−4T 2 ) Π1 = ΔpℓD (ML−2T −2 )(L) ≐ ≐ M 0 L0 T 0 ρV 2 (ML−3 )(LT −1 ) 2 Π2 = μ (ML−1T −1 ) ≐ ≐ M 0 L0 T 0 DVρ (L)(LT −1 )(ML−3 ) a b c Π2 = μD V ρ Step 8 ( μ DVρ μ ) or alternatively,   Soon we will discover that a standard dimensionless parameter, called the Reynolds number, is defined by Re = ρVD/μ. To conform to customary usage (Step 8), we replace Π2 by its reciprocal. 354 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling Finally (Step 9), we can express the result of the dimensional analysis as ΔpℓD ̃ ρVD = ϕ( 2 μ ) ρV This result indicates that this problem can be studied in terms of these two pi terms, rather than the original five variables we started with. The nine steps carried out to obtain this result are summarized by the figure in the margin. Dimensional analysis will not provide the form of the function ϕ̃. This can only be obtained from a suitable set of experiments or computations. All that we know about the relationship between Π 1 and Π 2 is DΔpℓ ______ D Δpℓ ρVD = ϕ( 2 μ ) ρV ρV2 ρVD _____ μ The method of repeating variables can be most easily carried out by following a stepby-step procedure. as shown by the figure in the margin. To summarize, the steps to be followed in performing a dimensional analysis using the method of repeating variables are as follows: Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Step 8 Step 9 EXAMPLE 7.1 List all the variables that are involved in the problem. Express each of the variables in terms of basic dimensions. Determine the required number of pi terms. Select a number of repeating variables, where the number required is equal to the number of reference dimensions (usually the same as the number of basic dimensions). Form a pi term by multiplying one of the nonrepeating variables by the product of repeating variables each raised to an exponent that will make the combination dimensionless. Repeat Step 5 for each of the remaining nonrepeating variables. Check all the resulting pi terms to make sure they are dimensionless and independent. Rearrange the pi terms to simplify the problem or to correspond to customary usage. Express the final form as a relationship among the pi terms and think about what it means. Method of Repeating Variables GIVEN A thin rectangular plate having a width w and a height FIND Determine a suitable set of pi terms to study this problem h is located so that it is normal to a moving stream of fluid as shown in Fig. E7.1. Assume the drag, 𝒟, that the fluid exerts on the plate is a function of w and h, the fluid viscosity and density, μ and ρ, respectively, and the velocity V of the fluid approaching the plate. experimentally. w V ρ, μ h VIDEO V7.2 Flow past a flat plate SOLUTION From the statement of the problem we can write 𝒟 = f(w, h, μ, ρ, V) where this equation expresses the general functional relationship between the drag and the several variables that will affect it. The dimensions of the variables (using the MLT system) are 𝒟 ≐ MLT −2 w≐L h≐L μ ≐ ML−1T −1 ρ ≐ ML−3 V ≐ LT−1 ■ Figure E7.1 We see that all three basic dimensions are required to define the six variables so that the Buckingham pi theorem tells us that three pi terms will be needed (six variables minus three reference dimensions, k − r = 6 − 3). We will next select three repeating variables such as w, V, and ρ. A quick inspection of these three reveals that they are dimensionally independent, since each one contains a basic dimension not included in the others. Note that it would be incorrect to use both w and h as repeating variables since they have the same dimensions. 7.4 Starting with the dependent variable, 𝒟, the first pi term can be formed by combining 𝒟 with the repeating variables such that and, therefore, 1 + c = 0  (for M) −1 + a + b − 3c = 0  (for L) −1 − b = 0  (for T) Π 1 = 𝒟waVbρc and in terms of dimensions (MLT −2 ) (L) a (LT −1 ) b (ML−3 ) c ≐ M 0L0T 0     Thus, for Π 1 to be dimensionless it follows that and, therefore, a = −2, b = −2, and c = −1. The pi term then becomes 𝒟 w V 2ρ μ wVρ Now that we have the three required pi terms we should check to make sure they are dimensionless. To make this check we use F, L, and T, which will also verify the correctness of the original dimensions used for the variables. Thus, (F) 𝒟 ≐ ≐ F 0L0T 0 2 2 −1 2 wV ρ (L) (LT ) (FL−4T 2 ) (L) h ≐ F 0L0T 0 Π2 = ≐ w (L) μ (FL−2T) Π3 = ≐ ≐ F 0L0T 0 wVρ (L) (LT −1 ) (FL−4T 2 ) Π1 = 2 2     Next the procedure is repeated with the second nonrepeating variable, h, so that Π 2 = hw aV bρc       It follows that (L) (L) a (LT −1 ) b (ML−3 ) c ≐ M 0L0T 0   and c = 0  (for M) 1 + a + b − 3c = 0  (for L) b = 0  (for T) so that a = −1, b = 0, c = 0, and therefore h Π2 = w The remaining nonrepeating variable is μ so that If these do not check, go back to the original list of variables and make sure you have the correct dimensions for each of the variables and then check the algebra you used to obtain the exponents a, b, and c. Next, we see that Π3 is the reciprocal of the Reynolds number, so we invert the fraction. We also change hw to wh. Finally, we can express the results of the dimensional analysis in the form 𝒟 w ρVw = ϕ( , h μ ) w2ρV2 (Ans) The ratio of the plate width to height, wh, is called the aspect ratio, and ρVwμ is the Reynolds number. Π 3 = μwaV bρc COMMENT Π2 is simply a ratio of two lengths and might (ML−1T −1 ) (L) a (LT −1 ) b (ML−3 ) c ≐ M 0L0T 0 have been intuitively obvious. To proceed, it would be necessary to perform a set of experiments to determine the nature of the function ϕ, as discussed in Section 7.7. with 7.4 Solving for the exponents, we obtain a = −1, b = −1, c = −1 so that Π3 = 1 + c = 0  (for M) 1 + a + b − 3c = 0  (for L) −2 − b = 0  (for T) Π1 = 355 Some Additional Comments about Dimensional Analysis Some Additional Comments about Dimensional Analysis The preceding section provides a systematic approach for determining pi terms. Other methods, such as formation by inspection, as is discussed in Sec. 7.5, could be used, although we think the method of repeating variables is the easiest for the beginning student to use because, once the significant dimensional variables are known, the procedure becomes algorithmic. In this section we will elaborate on some of the more subtle points that often prove to be puzzling to students. 7.4.1 Selection of Variables One of the most important and difficult steps in applying dimensional analysis is the selection of the variables. As noted previously, for convenience we will use the term variable to indicate any quantity involved, including dimensional and nondimensional constants. There is no simple procedure 356 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling It is often helpful to classify variables into three groups— geometry, material properties, and external effects. whereby the variables can be easily identified. Generally, one must rely on a good understanding of the phenomenon involved and the governing physical laws. If extraneous variables are included, then too many pi terms appear in the final solution, and it may be difficult, time consuming, and expensive to eliminate these experimentally. If important variables are omitted, then an incomplete result will be obtained; and again, this may prove to be costly and difficult to ascertain. It is, therefore, imperative that sufficient time and attention be given to this first step in which the variables are determined. Most engineering problems involve certain simplifying assumptions that have an influence on the variables to be considered. Usually we wish to keep the problem as simple as possible, perhaps even if some accuracy is sacrificed. A suitable balance between simplicity and accuracy is a desirable goal. How “accurate” the solution must be depends on the objective of the study; that is, we may be only concerned with general trends and, therefore, some variables that are thought to have only a minor influence in the problem may be neglected for simplicity. For most engineering problems (including areas outside of fluid mechanics), pertinent variables can be classified into three general groups—geometry, material properties, and external effects. Geometry. The geometric characteristics can usually be described by a series of lengths and angles. In most problems the geometry of the system plays an important role, and a sufficient number of geometric variables must be included to describe the system. These variables can usually be readily identified. Material Properties. Since the response of a system to applied external effects such as forces, pressures, and changes in temperature is dependent on the nature of the materials involved in the system, the material properties that relate the external effects and the responses must be included as variables. For example, for Newtonian fluids the viscosity of the fluid is the property that relates the applied forces to the rates of deformation of the fluid. As the material behavior becomes more complex, such as would be true for non-Newtonian fluids, the determination of material properties becomes difficult, and this class of variables can be troublesome to identify. External Effects. This terminology is used to denote any variable that produces, or tends to produce, a change in the system. For example, in structural mechanics, forces (either concentrated or distributed) applied to a system tend to change its geometry, and such forces would need to be considered as pertinent variables. For fluid mechanics, variables in this class would be related to pressures, velocities, or gravity. The above general classes of variables are intended as broad categories that should be helpful in identifying variables. It is likely, however, that there will be important variables that do not fit easily into one of the above categories and each problem needs to be carefully analyzed. Since we wish to keep the number of variables to a minimum, it is important that all variables are independent. For example, if in a given problem we know that the moment of inertia of the area of a circular plate is an important variable, we could list either the moment of inertia or the plate diameter as the pertinent variable. However, it would be unnecessary to include both moment of inertia and diameter, assuming that the diameter enters the problem only through the moment of inertia. In more general terms, if we have a problem in which the variables are f(p, q, r, . . . , u, υ, w, . . .) = 0 (7.3) and it is known that there is an additional relationship among some of the variables, for example, q = f1 (u, υ, w, . . .) (7.4) then q is not required and can be omitted. Conversely, if it is known that the only way the variables u, υ, w, . . . enter the problem is through the relationship expressed by Eq. 7.4, then the variables u, υ, w, . . . can be replaced by the single variable q, therefore reducing the number of variables. In summary, the following points should be considered in the selection of variables: 1. Clearly define the problem. What is the main variable of interest (the dependent variable)? 2. Consider the basic laws that govern the phenomenon. Even a crude theory that describes the essential aspects of the system may be helpful. 7.4 Some Additional Comments about Dimensional Analysis 357 3. Start the variable selection process by grouping the variables into three broad classes: geometry, material properties, and external effects. 4. Consider other variables that may not fall into one of the above categories. For example, time will be an important variable if any of the variables are time dependent. 5. Be sure to include all quantities that enter the problem even though some of them may be held constant (e.g., the acceleration of gravity, g). For a dimensional analysis it is the dimensions of the quantities that are important—not specific values! 6. Make sure that all variables are independent. Look for relationships among subsets of the variables. 7.4.2 Determination of Reference Dimensions Typically, in fluid mechanics, the required number of reference dimensions is three, but in some problems only one or two are required. For any given problem it is obviously desirable to reduce the number of pi terms to a minimum; therefore, we wish to reduce the number of variables to a minimum—that is, we certainly do not want to include extraneous variables. It is also important to know how many reference dimensions are required to describe the variables. As we have seen in the preceding examples, F, L, and T appear to be a convenient set of basic dimensions for characterizing fluid-mechanical quantities. There is, however, really nothing “fundamental” about this set and, as previously noted, M, L, and T would also be suitable. Actually any set of measurable quantities could be used as basic dimensions provided that the selected combination can be used to describe all secondary quantities. However, the use of FLT or MLT as basic dimensions is the simplest, and these dimensions can be used to describe fluid-mechanical phenomena. Of course, in some problems only one or two of these are required. In addition, we occasionally find that the number of reference dimensions needed to describe all variables is smaller than the number of basic dimensions. This point is illustrated in Example 7.2. Interesting discussions, both practical and philosophical, relative to the concept of basic dimensions can be found in the books by Huntley (Ref. 4) and by Isaacson and Isaacson (Ref. 12). E X AMPLE 7.2 Determination of Pi Terms GIVEN An open, cylindrical paint can having a diameter D is filled to a depth h with paint having a specific weight γ. The vertical deflection, δ, of the center of the bottom is a function of D, h, d, γ, and E, where d is the thickness of the bottom and E is the modulus of elasticity of the bottom material. FIND Determine the functional relationship between the vertical deflection, δ, and the independent variables using dimensional analysis. γ SOLUTION From the statement of the problem h δ = f(D, h, d, γ, E) and the dimensions of the variables are δ≐L D≐L h≐L d≐L γ ≐ FL−3 ≐ ML−2T−2 E ≐ FL−2 ≐ ML−1T −2 where the dimensions have been expressed in terms of both the FLT and MLT systems. E, d δ D ■ Figure E7.2 We now apply the pi theorem to determine the required number of pi terms. First, let us use F, L, and T as our system of basic dimensions. There are six variables and two reference dimensions 358 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling Π 2 = h Da γb the MLT system of basic dimensions? The answer is that we cannot, and a closer look at the dimensions of the variables listed above reveals that actually only two reference dimensions, MT −2 and L, are required. This is an example of the situation in which the number of reference dimensions differs from the number of basic dimensions. It does not happen very often and can be detected by looking at the dimensions of the variables (regardless of the systems used) and making sure how many reference dimensions are actually required to describe the variables. Once the number of reference dimensions has been determined, we can proceed as before. Since the number of repeating variables must equal the number of reference dimensions, it follows that two reference dimensions are still required and we could again use D and γ as repeating variables. The pi terms would be determined in the same manner. For example, the pi term containing E would be developed as and following the same procedure as above, a = −1, b = 0 so that Π 4 = EDaγ b (ML−1T −2 ) (L) a (ML−2T −2 ) b ≐ (MT −2 ) 0L0 (F and L) required so that four pi terms are needed. For repeating variables, we can select D and γ so that = δ D a γb (L) (L) (FL ) ≐ F0L0 a Π1 −3 b and 1 + a − 3b = 0  (for L) b = 0  (for F) Therefore, a = −1, b = 0, and Π1 = δ D Similarly, 1 + b = 0  (for MT −2 ) h Π2 = D −1 + a − 2b = 0  (for L) The remaining two pi terms can be found using the same procedure, with the result Π4 = d E Π 3 =   Π 4 = D Dγ Thus, this problem can be studied by using the relationship h d E δ = ϕ( , , ) D D D Dγ and, therefore, a = −1, b = −1 so that (Ans) COMMENTS Let us now solve the same problem using the MLT system. Although the number of variables is obviously the same, it would seem that there are three reference dimensions required, rather than two. If this were indeed true, it would certainly be fortuitous, since we would reduce the number of required pi terms from four to three. Does this seem right? How can we reduce the number of required pi terms by simply using E Dγ which is the same as Π 4 obtained using the FLT system. The other pi terms would be the same, and the final result is the same; that is, δ h d E = ϕ( , , ) (Ans) D D D Dγ This will always be true—you cannot affect the required number of pi terms by using M, L, and T instead of F, L, and T, or vice versa. A handy “rule of thumb” is to check the number of reference dimensions in both the FLT system and the MLT system. The smaller number of reference dimensions is subtracted from the number of variables to determine the number of pi terms. 7.4.3 Uniqueness of Pi Terms A little reflection on the process used to determine pi terms by the method of repeating variables reveals that the specific pi terms obtained depend on the somewhat arbitrary selection of repeating variables. For example, in the problem of studying the pressure drop in a pipe, we selected D, V, and ρ as repeating variables. This led to the formulation of the problem in terms of pi terms as ΔpℓD ρVD = ϕ( 2 μ ) ρV (7.5) What if we had selected D, V, and μ as repeating variables? A quick check will reveal that the pi term involving Δpℓ becomes ΔpℓD2 Vμ 7.5 D Δpℓ Determination of Pi Terms by Inspection 359 and the second pi term remains the same. Thus, we can express the final result as ρV 2 ΔpℓD2 ρVD = ϕ1 ( μ ) Vμ ρ VD ____ μ D2 Δpℓ Vμ ρ____ VD Both results are correct, and both would lead to the same final relationship for Δpℓ . Note, however, that the functions ϕ and ϕ1 in Eqs. 7.5 and 7.6 will be different because the dependent pi terms are different for the two relationships. As shown by the figure in the margin, the resulting graph of dimensionless data will be different for the two formulations. However, when extracting the physical variable, Δpℓ , from the two results, the values will be the same. We can conclude from this illustration that there is not a unique set of pi terms which arises from a dimensional analysis. However, the required number of pi terms is fixed, and once a correct set is determined, all other possible sets can be developed from this set by combinations of products of powers of the original set. Thus, if we have a problem involving, say, three pi terms, Π 1 = ϕ(Π 2, Π 3 ) μ Once a correct set of pi terms is obtained, any other set can be obtained by manipulation of the original set. (7.6) we could always form a new set from this one by combining the pi terms. For example, we could form a new pi term, Π 2′, by letting Π 2′ = Π a2 Π b3 where a and b are arbitrary exponents. Then the relationship could be expressed as Π 1 = ϕ1 (Π 2′, Π 3 ) or Π 1 = ϕ2 (Π 2, Π 2′ ) All of these would be correct. It should be emphasized, however, that the required number of pi terms cannot be reduced by this manipulation; only the form of the pi terms is altered. By using this technique, we see that the pi terms in Eq. 7.6 could be obtained from those in Eq. 7.5; that is, we multiply Π 1 in Eq. 7.5 by Π 2 so that ΔpℓD2 ΔpℓD ρVD = ( ρV2 )( μ ) Vμ which is the Π 1 of Eq. 7.6. 7.5 Determination of Pi Terms by Inspection The method of repeating variables for forming pi terms has been presented in Section 7.3. This method provides a step-by-step procedure that if executed properly will provide a correct and complete set of pi terms. Although this method is simple and straightforward, it is rather tedious, particularly for problems in which large numbers of variables are involved. Since the only restrictions placed on the pi terms are that they be (1) correct in number, (2) dimensionless, and (3) independent, it is possible to simply form the pi terms by inspection, without resorting to the more formal procedure. To illustrate this approach, we again consider the pressure drop per unit length along a smooth pipe. Regardless of the technique to be used, the starting point remains the same—determine the variables, which in this case are Δpℓ = f (D, ρ, μ, V) Next, the dimensions of the variables are listed: Δpℓ ≐ FL−3 D≐L ρ ≐ FL−4T 2 μ ≐ FL−2T V ≐ LT −1 360 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling Pi terms can be formed by inspection by simply making use of the fact that each pi term must be dimensionless. and subsequently the number of reference dimensions determined. The application of the pi theorem then tells us how many pi terms are required. In this problem, since there are five variables and three reference dimensions, two pi terms are needed. Thus, the required number of pi terms can be easily obtained. The determination of this number should always be done at the beginning of the analysis. Once the number of pi terms is known, we can form each pi term by inspection, simply making use of the fact that each pi term must be dimensionless. We will always let Π 1 contain the dependent variable, which in this example is Δpℓ. Since this variable has the dimensions FL−3, we need to combine it with other variables so that a nondimensional product will result. One possibility is to first divide Δpℓ by ρ so that Δpℓ (FL−3 ) L ≐ ≐ 2   (cancels F) ρ (FL−4T 2 ) T The dependence on F has been eliminated, but Δpℓ ρ is obviously not dimensionless. To eliminate the dependence on T, we can divide by V 2 so that Δpℓ 1 L 1 1 ( ρ ) V 2 ≐ ( T 2 ) (LT−1 ) 2 ≐ L   (cancels T) Finally, to make the combination dimensionless we multiply by D so that Δpℓ 1 0 ( ρV 2 ) D ≐ ( L ) (L) ≐ L   (cancels L) Thus, Π1 = ΔpℓD ρV 2 Next, we will form the second pi term by selecting the variable that was not used in Π 1, which in this case is μ. We simply combine μ with the other variables to make the combination dimensionless (but do not use Δpℓ in Π 2, since we want the dependent variable to appear only in Π 1 ). For example, divide μ by ρ (to eliminate F), then by V (to eliminate T), and finally by D (to eliminate L). Thus, Π2 = μ (FL−2T) ≐ ≐ F 0 L0 T 0 ρVD (FL−4T 2 )(LT −1 )(L) and, therefore, ΔpℓD μ = ϕ( 2 ρVD ) ρV which is, of course, the same result we obtained by using the method of repeating variables. Sometimes some of the pi terms are obvious, such as when two variables have the same dimensions. For flow in a circular pipe, both the pipe length L and the pipe diameter D appear and their ratio, L /D, is an obvious pi term. An additional concern, when one is forming pi terms by inspection, is to make certain that they are all independent. In the pipe flow example, Π 2 contains μ, which does not appear in Π 1 , and therefore these two pi terms are obviously independent. In a more general case a pi term would not be independent of the others in a given problem if it can be formed by some combination of the others. For example, if Π 2 can be formed by a combination of say Π 3, Π 4, and Π 5 such as Π2 = Π 23 Π 4 Π5 then Π 2 is not an independent pi term. We can ensure that each pi term is independent of those preceding it by incorporating a new variable in each pi term. Although forming pi terms by inspection is essentially equivalent to the repeating variable method, it is less structured. With a little practice the pi terms can be readily formed by inspection, and this method offers an alternative to more formal procedures. 7.6 7.6 A useful physical interpretation can often be given to dimensionless groups. Common Dimensionless Groups in Fluid Mechanics 361 Common Dimensionless Groups in Fluid Mechanics Above, in Section 7.4, we learned two important facts: (1) The number of dimensional variables necessary to describe any problem in fluid mechanics is limited, and (2) the number of different dimensionless groups (pi terms) that could be formed from the dimensional variables is unlimited. If we consider only the field of fluid mechanics, including both compressible and incompressible flows, a rather complete list of geometric and dimensional variables emerges: lengths (ℓ1, ℓ2, ℓ3 . . .), angles (α1, α2 , α3 . . .), surface roughness (ε), density (ρ), viscosity (μ), surface tension (σ), modulus of elasticity (Ev), speed of sound (c), gravity (g), fluid velocity (V), boundary velocity (U), frequency of oscillation or boundary angular velocity (ω), pressure (p or Δp), wall shear stress (τw), head loss (hL), lift (ℒ ), and drag (𝒟). Of course not all of these variables appear in every problem. Rather than perform a dimensional analysis for every new problem, we could do the analysis once and for all using this list of variables to create a set of dimensionless groups that could be used for (almost) any fluid mechanics problem. If we were to do so, we would end up with a complete set of pi terms but, for the most part, we would have different pi terms from those that are commonly used in fluid mechanics. Table 7.1 shows common dimensionless groups that are used in fluid mechanics. With a few exceptions, each dimensionless group is either a “number” or a “coefficient.” The numbers are named in honor of people from the history of fluid mechanics (see Section 1.10) and usually represent independent variables such as velocity or viscosity. The “coefficients” generally represent dependent variables such as pressure or drag. It is usually possible to provide a physical interpretation to the dimensionless groups, which can be helpful in assessing their influence in a particular application. For example, the Froude number is an index of the ratio of the mass times acceleration of a fluid particle (sometimes called the “inertia force”) to the force of gravity (weight). This can be demonstrated by considering a fluid particle moving along a streamline (Fig. 7.3). The magnitude of the component of inertia force FI along the streamline can be expressed as FI = asm, where as is the magnitude of the acceleration along the streamline for a particle having a mass m. From our study of particle motion along a curved path (see Section 3.1) we know that as = dVs dVs = Vs dt ds where s is measured along the streamline. If we write the velocity, Vs, and length, s, in dimensionless form, that is, V *s =   Vs s   s* = V ℓ where V and ℓ represent some characteristic velocity and length, respectively, then as = V 2 dV *s V* ℓ s ds*     and FI = dV*s V2 V*s m ℓ ds* Vs e lin m S a tre mg ■ Figure 7.3 The force of gravity acting on a fluid particle moving along a streamline. 362 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling ■ Table 7.1 Some Common Variables and Dimensionless Groups in Fluid Mechanics Variables: Acceleration of gravity, g; Bulk modulus, Eυ; Characteristic length, ℓ; Density, ρ; Frequency of oscillating flow or rotation of boundary, ω; Pressure, p (or Δp); Speed of sound, c; Surface tension, σ; Velocity, V; Viscosity, μ Boundary velocity, U; Surface roughness, ε; Wall shear stress, τw; head loss, hL; Drag, 𝒟; Lift, ℒ Dimensionless Group Interpretation (Index of Force Ratio Indicated) Types of Applications Reynolds number, Re inertia force viscous force Generally of importance in all types of fluid dynamics problems Froude number, Fr inertia force gravitational force Flow with a free surface Euler number, Eu pressure force inertia force Problems in which pressure, or pressure differences, are of interest ρV2 Eυ Cauchy number,a Ca inertia force compressibility force Flows in which the compressibility of the fluid is important V c Mach number,a Ma inertia force compressibility force Flows in which the compressibility of the fluid is important ωℓ V Strouhal number, St inertia (local) force inertia (convective) force Unsteady flow with a characteristic frequency of oscillation or a rotating boundary ρV2ℓ σ Weber number, We inertia force surface tension force Problems in which surface tension is important V U Velocity Ratio, Φ fluid velocity boundary velocity Flows with a moving boundary ε ℓ Relative Roughness surface roughness characteristic length Turbulent flow over a rough surface ℓ1 ℓ Length Ratio length reference length Flow with two (or more) relevant dimensions α Angle of Attack Angle between approach Flow over immersed bodies velocity and reference axis Loss Coefficient, K energy loss kinetic energy Internal flow, especially through valves and fittings Pressure Coefficient, Cp pressure dynamic pressure Euler number with ½ in honor of Bernoulli Friction Coefficient, Cf shear force inertia force Laminar or turbulent internal and external flow with viscous stress 𝒟 12ρV 2ℓ2 Drag Coefficient, CD drag force inertia force Flow over immersed bodies ℒ 12ρV2ℓ2 Lift Coefficient, CL lift force inertia force Flow over airfoils, wings, and turbomachine blades ρVℓ μ V Special names along with physical interpretations are given to the most common dimensionless groups. Name √gℓ p ρV2 hL V22g Δp 12 ρV τw 2 12 ρV 2 a The Cauchy number and the Mach number are related, and either can be used as an index of the relative effects of inertia and compressibility. See accompanying discussion. 7.6 Common Dimensionless Groups in Fluid Mechanics 363 The magnitude of the weight of the particle, FG, is FG = gm, so the ratio of the inertia to the gravitational force is FI V 2 dV*s = V* FG gℓ s ds* Thus, the force ratio FIFG is proportional to V2 gℓ, and the square root of this ratio, V √gℓ, is called the Froude number. We see that a physical interpretation of the Froude number is that it is a measure of, or an index of, the relative importance of inertial forces acting on fluid particles to the weight of the particle. Note that the Froude number is not really equal to this force ratio, but is simply some type of average measure of the relative influence of these two forces. In a problem in which gravity (or weight) is not important, the Froude number would not appear as an important pi term. A similar interpretation in terms of indices of force ratios can be given to the other dimensionless groups, as indicated in Table 7.1, and a further discussion of the basis for this type of interpretation is given in the last section in this chapter. Some additional details about these important dimensionless groups are given below, and the types of application or problem in which they arise are briefly noted in the last column of Table 7.1. VIDEO V7.3 Reynolds number Reynolds Number. The Reynolds number is undoubtedly the most famous dimensionless parameter in fluid mechanics. It is named in honor of Osborne Reynolds (1842–1912), a British engineer who first demonstrated that this combination of variables could be used as a criterion to distinguish between laminar and turbulent flow. In most fluid flow problems there will be a characteristic length, ℓ, and a velocity, V, as well as the fluid properties density, ρ, and viscosity, μ, which are relevant variables in the problem. Thus, with these variables the Reynolds number Re = No separation Re ≈ 0.2 Laminar boundary layer, wide turbulent wake Re ≈ 20,000 arises naturally from the dimensional analysis. The Reynolds number is a measure of the ratio of the inertia force on an element of fluid to the viscous force on an element. When these two types of forces are important in a given problem, the Reynolds number will play an important role. However, if the Reynolds number is very small (Re ⪡ 1), this is an indication that the viscous forces are dominant in the problem, and it may be possible to neglect the inertial effects; that is, the density of the fluid will not be an important variable. Flows at very small Reynolds numbers are commonly referred to as “creeping flows” as discussed in Section 6.10. For large Reynolds number flows, viscous effects are small relative to inertial effects and for these cases one might expect that it may be possible to neglect the effect of viscosity and consider the problem as one involving a “nonviscous” fluid, as considered in detail in Sections 6.4 through 6.7. This is not always the case, however. At high Reynolds numbers, viscous effects are confined to a thin layer near a bounding surface. For accelerating flow, the boundary layer does have a small effect on the flow, and most of the flow can be assumed inviscid; but for decelerating flows the boundary layer can separate from the surface with profound effects, as shown by the figure in the margin. A high Reynolds number also warns us to expect turbulence, which is dominated by the inertial effects of the disorderly motion. Froude Number. The Froude number Fr = VIDEO V7.4 Froude number ρVℓ μ V √gℓ is distinguished from the other dimensionless groups in Table 7.1 in that it contains the acceleration of gravity, g. The acceleration of gravity becomes an important variable in a fluid dynamics problem in which the fluid weight is an important force. As discussed, the Froude number is a measure of the ratio of the inertia force on an element of fluid to the weight of the element. It will generally be important in problems involving flows with free surfaces since gravity principally affects this type of flow. Typical problems would include the study of the flow of water around ships (with the resulting wave action) or flow through rivers or open conduits. The Froude number is named in honor of William Froude (1810–1879), a British civil engineer, mathematician, and naval architect who pioneered the use of towing tanks for the study of ship design. It is to be noted that the Froude number is sometimes defined as the square of the Froude number listed in Table 7.1. 364 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling Euler Number. The Euler number Eu = p ρV2 can be interpreted as a measure of the ratio of pressure forces to inertial forces, where p is some characteristic pressure in the flow field. Very often the Euler number is written in terms of a pres1 sure difference, Δp, so that Eu = ΔpρV 2. Also, this combination expressed as Δp2 ρV 2 is called the pressure coefficient. Some form of the Euler number would normally be used in problems in which pressure or the pressure difference between two points is an important variable. The Euler number is named in honor of Leonhard Euler (1707–1783), a famous Swiss mathematician who pioneered work on many phenomena in science, engineering, and mathematics, including the relationship between pressure and flow. For problems in which cavitation is of concern, the dimension1 less group (pr − pυ )2 ρV 2 is commonly used, where pυ is the vapor pressure and pr is some reference pressure. Although this dimensionless group has the same form as the pressure coefficient, it is generally referred to as the cavitation number.     Cauchy Number and Mach Number. The Cauchy number Ca = The Mach number is a commonly used dimensionless parameter in compressible flow problems. ρV 2 Eυ and the Mach number Ma = V c are important dimensionless groups in problems in which fluid compressibility is a significant factor. Since the speed of sound, c, in a fluid is equal to c = √Eυ ρ (see Section 1.7.3), it follows that Ma = V ρ A Eυ and the square of the Mach number Ma2 = ρV 2 = Ca Eυ is equal to the Cauchy number. Thus, either number (but not both) may be used in problems in which fluid compressibility is important. Both numbers can be interpreted as representing an index of the ratio of inertial forces to forces resisting compression. When the Mach number is relatively small (say, less than 0.3), the inertial forces induced by the fluid motion are not sufficiently large to cause a significant change in the fluid density, and in this case the compressibility of the fluid can be neglected. The Mach number is the more commonly used parameter in compressible flow problems, particularly in the fields of gas dynamics and aerodynamics. The Cauchy number is named in honor of Augustin Louis de Cauchy (1789–1857), a French engineer, mathematician, and hydrodynamicist. The Mach number is named in honor of Ernst Mach (1838–1916), an Austrian physicist and philosopher. Strouhal Number. The Strouhal number St = VIDEO V7.5 Strouhal number ωℓ V is a dimensionless parameter that is likely to be important in unsteady, oscillating flow problems in which the frequency of the oscillation is ω. It represents a measure of the ratio of inertial forces due to the unsteadiness of the flow (local acceleration) to the inertial forces due to changes in velocity from point to point in the flow field (convective acceleration). This type of unsteady flow may develop when a fluid flows past a solid body (such as a wire or cable) placed in the moving stream. For example, in a certain Reynolds number range, a periodic flow will develop downstream from a 7.6 Common Dimensionless Groups in Fluid Mechanics 365 cylinder placed in a moving fluid due to a regular pattern of vortices that are shed from the body. (See Fig. 9.23.) This system of vortices, called a Kármán vortex street [named after Theodor von Kármán (1881–1963), a famous Hungarian-American fluid mechanician], creates an oscillating flow at a discrete frequency, ω, such that the Strouhal number can be closely correlated with the Reynolds number. When the frequency is in the audible range, a sound can be heard and the bodies appear to “sing.” In fact, the Strouhal number is named in honor of Vincenz Strouhal (1850–1922), who used this parameter in his study of “singing wires.” The most dramatic evidence of this phenomenon occurred in 1940 with the collapse of the Tacoma Narrows Bridge. The shedding frequency of the vortices coincided with the natural frequency of the bridge, thereby setting up a resonant condition that eventually led to the collapse of the bridge. There are, of course, other types of oscillating flows. For example, blood flow in arteries is periodic and can be analyzed by breaking up the periodic motion into a series of harmonic components (Fourier series analysis), with each component having a frequency that is a multiple of the fundamental frequency, ω (the pulse rate). Rather than use the Strouhal number in this type of problem, a dimensionless group formed by the product of St and Re is used; that is St × Re = ρωℓ2 μ The square root of this dimensionless group is often referred to as the frequency parameter. Something like the Strouhal number (actually, its reciprocal) appears in the field of fluid machinery, where ω represents the rotational speed of a rotor. Weber Number. The Weber number We = VIDEO V7.6 Weber number ρV 2ℓ σ may be important in problems in which there is an interface between two fluids. In this situation the surface tension may play an important role in the phenomenon of interest. The Weber number can be thought of as an index of the inertial force to the surface tension force acting on a fluid element. Common examples of problems in which this parameter may be important include the flow of thin films of liquid, or the formation of droplets or bubbles. Clearly, not all problems involving flows with an interface will require the inclusion of surface tension. The flow of water in a river is not affected significantly by surface tension, since inertial and gravitational effects are dominant (We ⪡ 1). However, as discussed in a later section, for river models (which may have small depths) caution is required so that surface tension does not become important in the model, whereas it is not important in the actual river. The Weber number is named after Moritz Weber (1871–1951), a German professor of naval mechanics who was instrumental in formalizing the general use of common dimensionless groups as a basis for similitude studies. The Coefficients. The remainder of the dimensionless groups in Table 7.1, most of which are “coefficients” rather than “numbers,” generally represent dependent variables. Some additional information about these groups will be discussed in Section 7.9 and their significance will be apparent in Chapter 8 and Chapter 9. Several of the groups in Table 7.1 will reappear (often in disguise) in Chapter 12 on turbomachinery. We close this discussion of common dimensionless groups with two interesting observations. First, it is usually possible to interpret the groups as ratios of energies as well as ratios of forces. The Reynolds number, for example, can be shown to represent the ratio of kinetic energy to the dissipation of energy associated with viscosity. Second, the list of dimensionless groups in Table 7.1 represents only a small fraction of the common dimensionless groups (if one is willing to stretch the meaning of “common” just a bit). The Land Chart of Dimensionless Numbers (Ref. 16) lists 154 dimensionless groups, most of which are named for a prominent engineer or scientist. 366 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling THE WIDE WORLD OF FLUIDS Slip at the micro scale A goal in chemical and biological analy- characterize this situation, a dimensionless number termed the ses is to miniaturize the experiment, which has many advantages Knudsen number, Kn = λ/ℓ , is commonly employed. Here λ is the including reduction in sample size. In recent years, there has been mean free path and ℓ is the characteristic length of the system. significant work on integrating these tests on a single microchip to If Kn is smaller than 0.01, then the flow can be described by form the “lab-on-a-chip” system. These devices are on the millime- the Navier–Stokes equations with no-slip at the walls. For ter scale with complex passages for fluid flow on the micron scale 0.01 < Kn < 0.3, the same equations can be used, but there can (or smaller). While there are advantages to miniaturization, care be slip between the fluid and the wall so the boundary conditions must be taken in moving to smaller and smaller flow regimes, need to be adjusted. For Kn > 10, the continuum assumption as you will eventually bump into the continuum assumption. To breaks down and the Navier–Stokes equations are no longer valid. 7.7 Correlation of Experimental Data One of the most important uses of dimensional analysis is as an aid in the efficient handling, interpretation, and correlation of experimental data. Since the field of fluid mechanics relies heavily on empirical data, it is not surprising that dimensional analysis is such an important tool in this field. As noted previously, a dimensional analysis cannot provide a complete answer to any given problem, since the analysis only provides the dimensionless groups describing the phenomenon, and not the specific relationship among the groups. To determine this relationship, suitable experimental data must be obtained. The degree of difficulty involved in this process depends on the number of pi terms, and the nature of the experiments (How hard is it to obtain the measurements?). The simplest problems are obviously those involving the fewest pi terms, and the following sections indicate how the complexity of the analysis increases with the increasing number of pi terms. 7.7.1 Problems with One Pi Term Application of the pi theorem indicates that if the number of variables minus the number of reference dimensions is equal to unity, then only one pi term is required to describe the phenomenon. The functional relationship that must exist for one pi term is Π1 = C If only one pi term is involved in a problem, it must be equal to a constant. where C is a constant. This is one situation in which a dimensional analysis reveals the specific form of the relationship and, as is illustrated by the following example, shows how the individual variables are related. The value of the constant, however, must still be determined by experiment, analysis, or computation. EXAMPLE 7.3 Flow with Only One Pi Term GIVEN As shown in Fig. E7.3, assume that the drag, 𝒟, acting 𝔇 = f(D,V, μ) on a spherical particle that falls very slowly through a viscous fluid, is a function of the particle diameter, D, the particle velocity, V, and the fluid viscosity, μ. FIND Determine, with the aid of dimensional analysis, how VIDEO D the drag depends on the particle velocity. SOLUTION From the information given, it follows that 𝒟 = f(D, V, μ) μ V ■ Figure E7.3 V7.7 Stokes flow 7.7 and the dimensions of the variables are 𝒟≐F D≐L V ≐ LT−1 μ ≐ FL−2T We see that there are four variables and three reference dimensions (F, L, and T) required to describe the variables. Thus, according to the pi theorem, one pi term is required. This pi term can be easily formed by inspection and can be expressed as Π1 = 𝒟 μVD Because there is only one pi term, it follows that 𝒟 =C μVD where C is a constant. Thus, 𝒟 = CμVD Thus, for a given particle and fluid, the drag varies directly with the velocity so that 𝒟∝V (Ans) Correlation of Experimental Data be emphasized that once the value of C is determined it is not necessary to run similar tests by using different spherical particles and fluids; that is, C is a universal constant as long as the drag is a function only of particle diameter, velocity, and fluid viscosity. An approximate solution to this problem can also be obtained theoretically, from which it is found that C = 3π so that 𝒟 = 3πμVD This equation is commonly called Stokes law and is used in the study of the settling of particles. Our experiments would reveal that this result is only valid for small Reynolds numbers (ρVD/μ ⪡ 1). This follows, since in the original list of variables, we have neglected inertial effects (fluid density is not included as a variable). The inclusion of an additional variable would lead to another pi term so that there would be two pi terms rather than one. Consider a free-body diagram of a sphere in Stokes flow; there would be a buoyant force in the same direction as the drag in Fig. E7.3, as well as a weight force in the opposite direction. As shown above, the drag force is proportional to the product of the diameter and fall velocity, 𝒟 ∝ VD. The weight and buoyant force are proportional to the diameter cubed, W and FB ∝ D3 . Given equilibrium conditions, the force balance can be written as 𝒟 = W − FB COMMENTS Actually, the dimensional analysis reveals that the drag not only varies directly with the velocity, but it also varies directly with the particle diameter and the fluid viscosity. We could not, however, predict the value of the drag, since the constant, C, is unknown. An experiment would have to be performed in which the drag and the corresponding velocity are measured for a given particle and fluid. Although in principle we would only have to run a single test, we would certainly want to repeat it several times to obtain a reliable value for C. It should 367 Based on the scaling laws for these terms, it follows that VD ∝ D3. Hence, the fall velocity will be proportional to the square of the diameter, V ∝ D2 . Therefore, for two spheres, one having twice the diameter of the other, and falling through the same fluid, the sphere with the larger diameter will fall four times faster (see Video V7.7). This problem demonstrates that dimensional analysis is helpful in theoretical studies as well as experimental studies. 7.7.2 Problems with Two or More Pi Terms If a given phenomenon can be described with two pi terms such that For problems involving only two pi terms, results of an experiment can be conveniently presented in a simple graph. Π 1 = ϕ(Π 2 ) the functional relationship among the variables can then be determined by varying Π 2 and measuring the corresponding values of Π 1. For this case the results can be conveniently presented in graphical form by plotting Π 1 versus Π 2 as is illustrated in Fig. 7.4. It should be emphasized that the curve shown in Fig. 7.4 would be a “universal” one for the particular phenomenon studied. This means that if the variables and the resulting dimensional analysis are correct, then there is only a single relationship between Π 1 and Π 2, as illustrated in Fig. 7.4. However, since this is an empirical Π1 k–r=2 Valid range ■ Figure 7.4 The graphical presentation of data for Π2 problems involving two pi terms, with an illustration of the potential danger of extrapolation of data. 368 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling relationship, we can only say that it is valid over the range of Π 2 covered by the experiments. It would be unwise to extrapolate beyond this range, since as illustrated with the dashed lines in the figure, the nature of the phenomenon could dramatically change as the range of Π 2 is extended. In addition to presenting the data graphically, it may be possible (and desirable) to obtain an empirical equation relating Π 1 and Π 2 by using a standard curve-fitting technique. EXAMPLE 7.4 Dimensionless Correlation of Experimental Data GIVEN The relationship between the pressure drop per unit length along a smooth-walled, horizontal pipe and the variables that affect the pressure drop is to be determined experimentally. In the laboratory the pressure drop was measured over a 5-ft length of smooth-walled pipe having an inside diameter of 0.496 in. The fluid used was water at 60 °F (μ = 2.34 × 10 −5 lb ∙ sft2, ρ = 1.94 slugsft3 ). Tests were run in which the velocity was varied and the corresponding pressure drop measured. The results of these tests are shown below: SOLUTION The first step is to perform a dimensional analysis during the planning stage before the experiments are actually run. As was discussed in Section 7.3, we will assume that the pressure drop per unit length, Δpℓ, is a function of the pipe diameter, D, fluid density, ρ, fluid viscosity, μ, and the velocity, V. Thus, Δpℓ = f (D, ρ, μ, V) and application of the pi theorem yields two pi terms Π1 = D Δpℓ ρV2  and Π 2 = ρVD μ 1.17 1.95 2.91 5.84 11.13 16.92 23.34 28.73 6.26 15.6 30.9 106 329 681 1200 1730 between the pressure drop per unit length and the other variables. should be in feet, not inches or meters. Note that since the Reynolds numbers are all greater than 2100, the flow in the pipe is turbulent (see Section 8.1.1). A plot of these two pi terms can now be made with the results shown in Fig. E7.4a. The correlation appears to be quite good, 0.022 ρVD = ϕ( μ ) DΔpℓρV2 ρVDμ 0.0195 0.0175 0.0155 0.0132 0.0113 0.0101 0.00939 0.00893 4.01 × 103 6.68 × 103 9.97 × 103 2.00 × 104 3.81 × 104 5.80 × 104 8.00 × 104 9.85 × 104 These are dimensionless groups so that their values are independent of the system of units used so long as a consistent system is used. For example, if the velocity is in fts, then the diameter 0.018 D Δpℓ ______ ρV2 To determine the form of the relationship, we need to vary the Reynolds number, Re = ρVDμ, and to measure the corresponding values of D ΔpℓρV2. The Reynolds number could be varied by changing any one of the variables, ρ, V, D, or μ, or any combination of them. However, the simplest way to do this is to vary the velocity, since this will allow us to use the same fluid and pipe. Based on the data given, values for the two pi terms can be computed, with the result: 0.020 0.016 0.014 0.012 0.010 0.008 0 20,000 40,000 60,000 80,000 100,000 ρ VD Re = _____ μ (a) 4 2 D Δpℓ ______ ρV2 ρV2 Pressure drop for 5-ft length (lb/ft2) FIND Make use of these data to obtain a general relationship Hence, D Δpℓ Velocity (ft/s) 10–2 8 6 4 × 10–3 103 2 4 6 8 104 ρ VD Re = _____ μ (b) ■ Figure E7.4 2 4 6 8 105 7.8 and if it was not, this would suggest that either we had large experimental measurement errors or that we had perhaps omitted an important variable. The curve shown in Fig. E7.4a represents the general relationship between the pressure drop and the other factors in the range of Reynolds numbers between 4.01 × 103 and 9.85 × 104. Thus, for this range of Reynolds numbers it is not necessary to repeat the tests for other pipe sizes or other fluids provided the assumed independent variables (D, ρ, μ, V) are the only important ones. Since the relationship between Π 1 and Π 2 is nonlinear, it is not immediately obvious what form of empirical equation might be used to describe the relationship. If, however, the same data are plotted on a logarithmic scale, as is shown in Fig. E7.4b, the data form a straight line, suggesting that a suitable equation is of the form Π 1 = AΠ n2 where A and n are empirical constants to be determined from the data by using a suitable curve-fitting technique, such as a nonlinear regression program. For the data given in this example, a good fit of the data is obtained with the equation Π 1 = 0.150 Π 2−0.25 (Ans) Modeling and Similitude 369 COMMENTS In 1911, H. Blasius (1883 – 1970), a German fluid mechanician, established a similar empirical equation that is used widely for predicting the pressure drop in smooth pipes in the range 4 × 103 < Re < 105 (Ref. 16). This equation can be expressed in the form D Δpℓ ρV 2 = 0.1582 ( ρVD −14 μ ) The so-called Blasius formula is based on numerous experimental results of the type used in this example. Flow in pipes is discussed in more detail in the next chapter, where it is shown how pipe roughness (which introduces another variable) may affect the results given in this example (which is for smooth-walled pipes). It is not really necessary to do a full dimensional analysis for this problem; we can use common dimensionless groups from Table 7.1. We would expect a Reynolds number and an Euler number to describe the problem. Since we are dealing with pressure drop per unit length, the pipe length ℓ is implicit in Δpℓ . Then Π 1 = (D/ℓ) × Eu. As the number of required pi terms increases, it becomes more difficult to display the results in a convenient graphical form and to determine a specific empirical equation that describes the phenomenon. For problems involving three pi terms Π 1 = ϕ(Π 2, Π 3 ) For problems in volving more than two or three pi terms, it is often necessary to use a model to predict specific characteristics. it is still possible to show data correlations on simple graphs by plotting families of curves as illustrated in Fig. 7.5. This is an informative and useful way of representing the data in a general way. It may also be possible to determine a suitable empirical equation relating the three pi terms. However, as the number of pi terms continues to increase, corresponding to an increase in the general complexity of the problem of interest, both the graphical presentation and the determination of a suitable empirical equation become intractable. For these more complicated problems, it is often more feasible to use models to predict specific characteristics of the system rather than to try to develop general correlations. Π1 k–r=3 Π3 = C1 (constant) Π 3 = C2 Π3 = C3 Π3 = C4 ■ Figure 7.5 The graphical presentation Π2 of data for problems involving three pi terms. 370 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling 7.8 Modeling and Similitude V ℓ Prototype Vm ℓm Model VIDEO V7.8 Model airplane Physical models are widely used in fluid mechanics. Major engineering projects involving structures, aircraft, ships, rivers, harbors, dams, air and water pollution, and so on, frequently involve the use of models. Although the term model is used in many different contexts, the “engineering model” generally conforms to the following definition. A model is a representation of a physical system that may be used to predict the behavior of the system in some desired respect. The physical system for which the predictions are to be made is called the prototype. Although mathematical or computer models may also conform to this definition, our interest will be in physical models, that is, models that resemble the prototype but are generally of a different size, may involve different fluids, and often operate under different conditions (pressures, velocities, etc.). As shown by the figure in the margin, usually a model is smaller than the prototype. Therefore, it is more easily handled in the laboratory and less expensive to construct and operate than a large prototype (it should be noted that variables or pi terms without a subscript will refer to the prototype, whereas the subscript m will be used to designate the model variables or pi terms). Occasionally, if the prototype is very small, it may be advantageous to have a model that is larger than the prototype so that it can be more easily studied. For example, large models have been used to study the motion of red blood cells, which are approximately 8 μm in diameter. With the successful development of a valid model, it is possible to predict the behavior of the prototype under a certain set of conditions. We may also wish to examine a priori the effect of possible design changes that are proposed for a hydraulic structure or fluid flow system. There is, of course, an inherent danger in the use of models in that predictions can be made that are in error and the error not detected until the prototype is found not to perform as predicted. It is, therefore, imperative that the model be properly designed and tested and that the results be interpreted correctly. In the following sections we will develop the procedures for designing models so that the model and prototype will behave in a similar fashion. 7.8.1 Theory of Models The theory of models can be readily developed by using the principles of dimensional analysis. It has been shown that any given problem can be described in terms of a set of pi terms as Π 1 = ϕ(Π 2, Π 3, . . . , Π n ) (7.7) In formulating this relationship, only a knowledge of the general nature of the physical phenomenon, and the variables involved, is required. Specific values for variables (size of components, fluid properties, and so on) are not needed to perform the dimensional analysis. Thus, Eq. 7.7 applies to any system that is governed by the same variables. If Eq. 7.7 describes the behavior of a particular prototype, a similar relationship can be written for a model of this prototype; that is, Π 1m = ϕ(Π 2m, Π 3m, . . . , Π nm ) (7.8) where the form of the function will be the same as long as the same phenomenon is involved in both the prototype and the model. Variables, or pi terms, without a subscript will refer to the prototype, whereas the subscript m will be used to designate the model variables or pi terms. The pi terms can be developed so that Π 1 contains the variable that is to be predicted from observations made on the model. Therefore, if the model is designed and operated under the following conditions, Π 2m = Π 2 Π 3m = Π 3 (7.9) ⋮ Π nm = Π n The similarity requirements for a model can be readily obtained with the aid of dimensional analysis. then with the presumption that the form of ϕ is the same for model and prototype, it follows that Π 1 = Π 1m (7.10) Equation 7.10 is the desired prediction equation and indicates that the measured value of Π 1m obtained with the model will be equal to the corresponding Π 1 for the prototype as long as the other pi terms are equal. The conditions specified by Eqs. 7.9 provide the model design conditions, also called similarity requirements or modeling laws. 7.8 Modeling and Similitude 371 As an example of the procedure, consider the problem of determining the drag, 𝒟, on a thin rectangular plate (w × h in size) placed normal to a fluid with velocity, V, as shown by the figure in the margin. The dimensional analysis of this problem was performed in Example 7.1, where it was assumed that 𝒟 = f (w, h, μ, ρ, V) Application of the pi theorem yielded 𝒟 w ρVw = ϕ( , 2 h μ ) w ρV 2 w V ρ, μ h (7.11) (Do you recognize a drag coefficient, a length ratio, and a Reynolds number?) We are now concerned with designing a model that could be used to predict the drag on a certain prototype (which presumably has a different size than the model). Since the relationship expressed by Eq. 7.11 applies to both prototype and model, Eq. 7.11 is assumed to govern the prototype, with a similar relationship 𝒟m 2 wm ρmV 2m = ϕ( wm ρmVmwm , μm ) hm (7.12) for the model. The model design conditions, or similarity requirements, are therefore ρmVmwm ρVw wm w =   and   = μm μ hm h The size of the model is obtained from the first requirement, which indicates that hm wm = w (7.13) h We are free to establish the height ratio between the model and prototype, hmh, but then the model plate width, wm, is fixed in accordance with Eq. 7.13. The second similarity requirement indicates that the model and prototype must be operated at the same Reynolds number. Thus, the required velocity for the model is obtained from the relationship Vm = μm ρ w V μ ρm wm (7.14) Note that this model design requires not only geometric scaling, as specified by Eq. 7.13, but also the correct scaling of the velocity in accordance with Eq. 7.14. This result is typical of most model designs—there is more to the design than simply scaling the geometry! With the foregoing similarity requirements satisfied, the prediction equation for the drag is 𝒟m 𝒟 = 2 w2ρV 2 wm ρmV 2m or w 2 ρ V 2 𝒟 = ( ) ( )( ) 𝒟m wm ρm Vm Similarity between a model and a prototype is achieved by equating pi terms. Thus, a measured drag on the model, 𝒟m, must be multiplied by the ratio of the square of the plate widths, the ratio of the fluid densities, and the ratio of the square of the velocities to obtain the predicted value of the prototype drag, 𝒟. Generally, as is illustrated in this example, to achieve similarity between model and prototype behavior, all the corresponding pi terms must be equated between model and prototype. Usually, one or more of these pi terms will involve ratios of important lengths (such as wh in the foregoing example); that is, they are purely geometrical. Thus, when we equate the pi terms involving length ratios, we are requiring that there be complete geometric similarity between the model and prototype. This means that the model must be a scaled version of the prototype. Geometric scaling should extend to the finest features of the system, such as surface roughness, or small protuberances on a structure, since these kinds of geometric features may significantly influence the flow. Any deviation from complete geometric similarity for a model must be carefully considered. Sometimes 372 VIDEO Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling V7.9 Environmental models © Stanford University (with permission) complete geometric scaling may be difficult to achieve, particularly when dealing with surface roughness, since roughness is difficult to characterize and control. Another group of typical pi terms (such as the Reynolds number in the foregoing example) involves force ratios as noted in Table 7.1. The equality of these pi terms requires the ratio of like forces in model and prototype to be the same. Thus, for flows in which the Reynolds numbers are equal, the ratio of viscous forces in model and prototype is equal to the ratio of inertia forces. This is further illustrated by the photographs in the margin. The top photograph is of flow past a cylinder with diameter, D, and velocity, Vo, while the velocity is doubled and the diameter halved in the bottom photograph (image shown in 2x enlargement). Since the Reynolds numbers match, the flows look and behave the same. If other pi terms are involved, such as the Froude number or Weber number, a similar conclusion can be drawn; that is, the equality of these pi terms requires the ratio of like forces in model and prototype to be the same. Thus, when these types of pi terms are equal in model and prototype, we have dynamic similarity between model and prototype. It follows that with both geometric and dynamic similarity the streamline patterns will be the same and corresponding velocity ratios (VmV) and acceleration ratios (ama) are constant throughout the flow field. Thus, kinematic similarity exists between model and prototype. To have complete similarity between model and prototype, we must maintain geometric, kinematic, and dynamic similarity between the two systems. This will automatically follow if all the important variables are included in the dimensional analysis, and if all the similarity requirements based on the resulting pi terms are satisfied. THE WIDE WORLD OF FLUIDS Modeling parachutes in a water tunnel The first use of a tested in a water tunnel. The model parachutes are reduced in parachute with a free-fall jump from an aircraft occurred in 1914, size by a factor of 30 to 60 times. Various types of tests can be although parachute jumps from hot-air balloons had occurred performed, ranging from the study of the velocity fields in the since the late 1700s. In more modern times parachutes are com- wake of the canopy with a steady free-stream velocity to the monly used by the military and for safety and sport. It is not study of conditions during rapid deployment of the canopy. surprising that there remains interest in the design and charac- According to the researchers, the advantage of using water as teristics of parachutes, and researchers at the Worcester the working fluid, rather than air, is that the velocities and Polytechnic Institute have been studying various aspects of deployment dynamics are slower than in the atmosphere, thus the aerodynamics associated with parachutes. An unusual part providing more time to collect detailed experimental data. (See of their study is that they are using small-scale parachutes Problem 7.53.) EXAMPLE 7.5 Prediction of Prototype Performance from Model Data GIVEN A long structural component of a bridge has an ellip- tical cross section shown in Fig. E7.5. It is known that when a steady wind blows past this type of bluff body, vortices may develop on the downwind side that are shed in a regular fashion at some definite frequency. Since these vortices can create harmful periodic forces acting on the structure, it is important to determine the shedding frequency. For the specific structure of interest, D = 0.1 m, H = 0.3 m, and a representative wind velocity is 50 kmhr. Standard air can be assumed. The shedding frequency is to be determined through the use of a small-scale model that is to be tested in a water tunnel. For the model Dm = 20 mm and the water temperature is 20 °C. FIND Determine the model dimension, Hm, and the velocity at which the test should be performed. If the shedding frequency for the model is found to be 49.9 Hz, what is the corresponding frequency for the prototype? D V H ■ Figure E7.5 VIDEO V7.10 Flow past an ellipse 7.8 Modeling and Similitude 373 SOLUTION We expect the shedding frequency, ω, to depend on the lengths D and H, the approach velocity, V, and the fluid density, ρ, and viscosity, μ. Thus, 10 −3 kgm ˙ s, ρ = 998 kgm3. The fluid velocity for the prototype is V= ω = f(D, H, V, ρ, μ) where (50 × 103 mhr) = 13.9 ms (3600 shr) The required velocity can now be calculated from Eq. 1 as −1 ω≐T D≐L H≐L V ≐ LT −1 ρ ≐ ML−3 μ ≐ ML−1T −1 Vm = [1.79 × 10−5 kg(m ∙ s)] (998 kgm3 ) (0.1 m) × (13.9 ms) (20 × 10−3 m) Vm = 4.79 ms Since there are six variables and three reference dimensions (MLT), three pi terms are required. Application of the pi theorem yields ωD D ρVD = ϕ( , V H μ ) We recognize the pi term on the left as the Strouhal number, and the dimensional analysis indicates that the Strouhal number is a function of the length ratio, DH, and the Reynolds number. Thus, to maintain similarity between model and prototype Dm D = Hm H Hm = (Ans) The second similarity requirement indicates that the Reynolds number must be the same for model and prototype so that the model velocity must satisfy the condition μm ρ D V μ ρm D m and the predicted prototype vortex shedding frequency is ω= V Dm ω Vm D m (13.9 ms) (20 × 10−3 m) (49.9 Hz) (4.79 ms) (0.1 m) (Ans) COMMENT This same model could also be used to predict From the first similarity requirement Vm = ωm Dm ωD = V Vm ω = 29.0 Hz ρmVmDm ρVD = μm μ (Ans) This is a reasonable velocity that could be readily achieved in a water tunnel. With the two similarity requirements satisfied, it follows that the Strouhal numbers for prototype and model will be the same so that = and Dm H D (20 × 10−3 m) = (0.3 m) (0.1 m) Hm = 60 × 10−3 m = 60 mm [1.00 × 10−3 kg(m ∙ s)] (1.23 kgm3 ) (1) the drag per unit length, 𝒟ℓ (lbft or Nm), on the prototype, since the drag would depend on the same variables as those used for the frequency. Thus, the similarity requirements would be the same and with these requirements satisfied it follows that the drag coefficient would be equal in model and prototype. The measured drag per unit length on the model could then be related to the corresponding drag per unit length on the prototype through the relationship 𝒟ℓ = ρ D V 2 ( Dm )( ρm )( Vm ) 𝒟ℓm Of course it is not necessary to do the formal dimensional analysis; one could simply choose Strouhal number, Reynolds number, and length ratio from Table 7.1. For air at standard conditions, μ = 1.79 × 10 −5 kgm ˙ s, ρ = 1.23 kgm3, and for water at 20 °C, μ = 1.00 × 7.8.2 Model Scales It is clear from the preceding section that the ratio of like quantities for the model and prototype naturally arises from the similarity requirements. For example, if in a given problem there are two length variables ℓ1 and ℓ2, the resulting similarity requirement based on a length ratio is ℓ1m ℓ1 = ℓ2 ℓ2m 374 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling so that ℓ1m ℓ2m = ℓ1 ℓ2 The ratio of a model variable to the corresponding prototype variable is called the scale for that variable. We define the ratio ℓ1mℓ1 or ℓ2mℓ2 as the length scale. For true models there will be only one length scale, and all lengths are fixed in accordance with this scale. There are, however, other scales such as the velocity scale, VmV, density scale, ρmρ, viscosity scale, μmμ, and so on. In fact, we can define a scale for each of the variables in the problem. Thus, it is actually meaningless to talk about a “scale” of a model without specifying which scale. We will designate the length scale as λℓ, and other scales as λV, λρ, λμ, and so on, where the subscript indicates the particular scale. Also, we will take the ratio of the model value to the prototype value as the scale (rather than the inverse). Length scales are often specified, for example, as 1 1 : 10 or as a 10 scale model. The meaning of this specification is that the model is one-tenth the size of the prototype, and the tacit assumption is that all relevant lengths are scaled accordingly so the model is geometrically similar to the prototype. THE WIDE WORLD OF FLUIDS “Galloping Gertie” One of the most dramatic bridge collapses interference effects due to two bridges in close proximity, wind occurred in 1940 when the Tacoma Narrows Bridge, located near tunnel tests were run in a 9 m × 9 m wind tunnel operated by Tacoma, Washington, failed due to aerodynamic instability. The the National Research Council of Canada. Models of the two bridge had been nicknamed “Galloping Gertie” due to its ten- side-by-side bridges, each having a length scale of 1:211, were dency to sway and move in high winds. On the fateful day of the tested under various wind conditions. Since the failure of the collapse the wind speed was 65 km/hr. This particular combina- original Tacoma Narrows Bridge, it is now common practice to tion of a high wind and the aeroelastic properties of the bridge use wind tunnel model studies during the design process to created large oscillations leading to its failure. The bridge was evaluate any bridge that is to be subjected to wind-induced replaced in 1950, and a second bridge parallel to the existing vibrations. (See Problem 7.74.) structure was opened in 2007. To determine possible wind 7.8.3 Practical Aspects of Using Models VIDEO V7.11 Model of fish hatchery pond Validation of Model Design. Most model studies involve simplifying assumptions with regard to the variables to be considered. Although the assumptions are frequently less stringent than required for mathematical models, they nevertheless introduce some uncertainty in the model design. It is, therefore, desirable to check the design experimentally whenever possible. In some situations the purpose of the model is to predict the effects of certain proposed changes in a given prototype, and in this instance some actual prototype data may be available. The model can be designed, constructed, and tested, and the model prediction can be compared with these data. If the agreement is satisfactory, then the model can be changed in the desired manner, and the corresponding effect on the prototype can be predicted with increased confidence. Another useful and informative procedure is to run tests with a series of models of different sizes, where one of the models can be thought of as the prototype and the others as “models” of this prototype. With the models designed and operated on the basis of the proposed design, a necessary condition for the validity of the model design is that an accurate prediction be made between any pair of models, since one can always be considered as a model of the other. Although suitable agreement in validation tests of this type does not unequivocally indicate a correct model design (e.g., the length scales between laboratory models may be significantly different than required for actual prototype prediction), it is certainly true that if agreement between models cannot be achieved in these tests, there is no reason to expect that the same model design can be used to predict prototype behavior correctly. Distorted Models. Although the general idea behind establishing similarity requirements for models is straightforward (we simply equate pi terms), it is not always possible to satisfy all the known requirements. If one or more of the similarity requirements are not met, for example, if Π 2m ≠ Π 2, then it follows that the prediction equation Π 1 = Π 1m is not true; that is, Π 1 ≠ Π 1m. Models for which one or more of the similarity requirements are not satisfied are called distorted models. 7.8 Models for which one or more similarity requirements are not satisfied are called distorted models. Modeling and Similitude 375 Distorted models are rather commonplace, and they are used for a variety of reasons. For example, perhaps a suitable fluid cannot be found for the model. The classic example of a distorted model occurs in the study of open channel or free-surface flows. Typically, in these problems both the Reynolds number, ρVℓμ, and the Froude number, V √gℓ, are involved. Froude number similarity requires Vm √gmℓm = V √gℓ If the model and prototype are operated in the same gravitational field, then the required velocity scale is Vm ℓm = = √λℓ V Bℓ Reynolds number similarity requires ρmVmℓm ρVℓ = μm μ and the velocity scale is μm ρ ℓ Vm = μ ρm ℓm V VIDEO V7.12 Distorted river model Since the velocity scale must be equal to the square root of the length scale, it follows that μmρm νm = (λℓ ) 32 = ν μρ (7.15) where the ratio μρ is the kinematic viscosity, ν. Although in principle it may be possible to satisfy this design condition, it may be quite difficult, if not impossible, to find a suitable model fluid, particularly for small length scales. For problems involving rivers, spillways, and harbors, for which the prototype fluid is water, the models are also relatively large so that the only practical model fluid is water. However, in this case (with the kinematic viscosity scale equal to unity) Eq. 7.15 will not be satisfied, and a distorted model will result. Generally, hydraulic models of this type are distorted and are designed on the basis of the Froude number, with the Reynolds number different between model and prototype. Distorted models can be successfully used, but the interpretation of results obtained with this type of model is obviously more difficult than the interpretation of results obtained with true models for which all similarity requirements are met. There are no general rules for handling distorted models, and essentially each problem must be considered on its own merits. The success of using distorted models depends to a large extent on the skill and experience of the investigator responsible for the design of the model and in the interpretation of experimental data obtained from the model. Distorted models are widely used, and additional information can be found in the references at the end of the chapter. References 14 and 15 contain detailed discussions of several practical examples of distorted fluid flow and hydraulic models. THE WIDE WORLD OF FLUIDS Old Man River in (large) miniature One of the world’s largest embankments were sculpted by hand to match the actual con- scale models, a Mississippi River model, resides near Jackson, tours. The main purpose of the model was to predict floods. Mississippi. It is a detailed, complex model that covers many This was done by supplying specific amounts of water at pre- acres and replicates the 1,250,000-acre Mississippi River basin. scribed locations along the model and then measuring the Built by the Army Corps of Engineers and used from 1943 to water depths up and down the model river. Because of the 1973, today it has mostly gone to ruin. As with many hydraulic length scale, there is a difference in the time taken by the cor- models, this is a distorted model, with a horizontal scale of responding model and prototype events. Although it takes 1:2000 and a vertical scale of 1:100. One step along the model days for the actual floodwaters to travel from Sioux City, Iowa, river corresponds to one mile along the river. All essential river to Omaha, Nebraska, it would take only minutes for the simu- basin elements such as geological features, levees, and railroad lated flow in the model. 376 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling 7.9 Some Typical Model Studies Models are used to investigate many different types of fluid mechanics problems, and it is difficult to characterize in a general way all necessary similarity requirements, since each problem is unique. We can, however, broadly classify many of the problems on the basis of the general nature of the flow and subsequently develop some general characteristics of model designs in each of these classifications. In the following sections we will consider models for the study of (1) flow through closed conduits, (2) flow around immersed bodies, and (3) flow with a free surface. Turbomachine models are considered in Chapter 12. 7.9.1 Flow through Closed Conduits Geometric and Reynolds number similarity is usually required for models involving flow through closed conduits. Common examples of this type of flow include pipe flow and flow through valves, fittings, and metering devices. Although the conduit cross sections are often circular, they could have other shapes as well and may contain expansions or contractions. Since there are no fluid interfaces or free surfaces, the dominant forces are inertial and viscous so that the Reynolds number is an important similarity parameter. For low Mach numbers (Ma < 0.3), compressibility effects are usually negligible for both the flow of liquids and gases. For this class of problems, geometric similarity between model and prototype must be maintained. Generally the geometric characteristics can be described by a series of length terms, ℓ1, ℓ2, ℓ3, . . . , ℓi, and ℓ, where ℓ is some particular length dimension for the system. Such a series of length terms leads to a set of length ratios ℓi ℓ where i = 1, 2, . . . , and so on. In addition to the basic geometry of the system, the roughness of the internal surface in contact with the fluid may be important. If the average height of surface roughness elements is defined as ε, then the pi term representing roughness will be the relative roughness, εℓ. This parameter indicates that for complete geometric similarity, surface roughness would also have to be scaled. Note that this implies that for length scales less than 1, the model surfaces should be smoother than those in the prototype since εm = λℓε. To further complicate matters, the pattern of roughness elements in model and prototype would have to be similar. These are conditions that are virtually impossible to satisfy exactly. Fortunately, in some problems the surface roughness plays a minor role and can be neglected. However, in other problems (such as turbulent flow through pipes) roughness can be very important. It follows from this discussion that for flow in closed conduits at low Mach numbers, any dependent pi term (the one that contains the particular variable of interest, such as pressure drop) can be expressed as ℓi ε ρVℓ Dependent pi term = ϕ ( , , ℓ ℓ μ ) (7.16) This is a general formulation for this type of problem. The first two pi terms of the right side of Eq. 7.16 lead to the requirement of geometric similarity so that εm ℓim ℓi ε =    = ℓm ℓ ℓm ℓ or ℓm ℓim εm = = = λℓ ε ℓi ℓ This result indicates that the investigator is free to choose a length scale, λℓ, but once this scale is selected, all other pertinent lengths must be scaled in the same ratio. The additional similarity requirement arises from the equality of Reynolds numbers ρmVmℓm ρVℓ = μm μ 7.9 Accurate predictions of flow behavior require the correct scaling of velocities. Some Typical Model Studies 377 From this condition the velocity scale is established so that μm ρ ℓ Vm = μ ρm ℓm V (7.17) and the actual value of the velocity scale depends on the viscosity and density scales, as well as the length scale. Different fluids can be used in model and prototype. However, if the same fluid is used (with μm = μ and ρm = ρ), then Vm ℓ = V ℓm Thus, Vm = Vλℓ, which indicates that the fluid velocity in the model will be larger than that in the prototype for any length scale less than 1. Since length scales are typically much less than unity, Reynolds number similarity may be difficult to achieve because of the large model velocities required. With these similarity requirements satisfied, it follows that the dependent pi term will be equal in model and prototype. For example, if the dependent variable of interest is the pressure differential,3 Δp, between two points along a closed conduit, then the dependent dimensionless group could be the Euler number Eu = Δp ρV2 The prototype pressure drop would then be obtained from the relationship Δp = ρ V 2 Δpm ρm ( Vm ) so that from a measured pressure differential in the model, Δpm, the corresponding pressure differential for the prototype could be predicted. Note that in general Δp ≠ Δpm. E X AMPLE 7.6 Reynolds Number Similarity GIVEN Model tests are to be performed to study the flow through a large check valve having a 2-ft- diameter inlet and carrying water at a flowrate of 30 cfs as shown in Fig. E7.6a. The working fluid in the model is water at the same temperature as that in the prototype. Complete geometric similarity exists between model and prototype, and the model inlet diameter is 3 in. FIND Q = 30 cfs Determine the required flowrate in the model. (Qm = ?) SOLUTION To ensure dynamic similarity, the model tests should be run so that Rem = Re or VmDm VD = νm ν where V and D correspond to the inlet velocity and diameter, respectively. Since the same fluid is to be used in model and D = 2 ft (Dm = 3 in.) ■ Figure E7.6a prototype, ν = νm, and therefore Vm D = V Dm 3 In some previous examples the pressure differential per unit length, Δpℓ, was used. This is appropriate for flow in long pipes or conduits in which the pressure would vary linearly with distance. However, in the more general situation the pressure may not vary linearly with position so that it is necessary to consider the pressure differential, Δp, as the dependent variable. In this case the distance between pressure taps is an additional variable (as well as the distance of one of the taps measured from some reference point within the flow system). 378 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling The discharge, Q, is equal to VA, where A is the inlet area, so [(π4)D2m] 2 ) VmAm Qm D = =( Q VA Dm [(π4)D ] Dm = D say one with D = 1 in., the required model velocity is 229 fts, a very high velocity that would be difficult to achieve. These results are indicative of one of the difficulties encountered in maintaining Reynolds number similarity—the required model velocities may be impractical to obtain. and for the data given 25 (312 ft) (30 ft3s) (2 ft) Qm = 3.75 cfs Qm = 20 (Ans) tain Reynolds number similarity using the same fluid in model and prototype, the required velocity scale is inversely proportional to the length scale, that is, VmV = (DmD) −1 . This strong influence of the length scale on the velocity scale is shown in Fig. E7.6b. For this particular example, DmD = 0.125, and the corresponding velocity scale is 8 (see Fig. E7.6b). Thus, with the prototype velocity equal to V = (30 ft3/s)(π/4) (2 ft) 2 = 9.50 ft/s, the required model velocity is Vm = 76.4 ft/s. Although this is a relatively large velocity, it could be attained in a laboratory facility. It is to be noted that if we tried to use a smaller model, In some problems Reynolds number similarity may be relaxed. 15 Vm /V COMMENT As indicated by the foregoing analysis, to main- 10 (0.125, 8) 5 0 0 0.2 0.4 0.6 0.8 1 Dm /D ■ Figure E7.6b Two additional points should be made with regard to modeling flows in closed conduits. First, for large Reynolds numbers, inertial forces are much larger than viscous forces, and in this case it may be possible to neglect viscous effects (this would imply turbulent flow, not inviscid flow). The important practical consequence of this is that it would not be necessary to maintain Reynolds number similarity between model and prototype. However, both model and prototype would have to operate at large Reynolds numbers. Since we do not know, a priori, what a “large Reynolds number” is, the effect of Reynolds numbers would have to be determined from the model. This could be accomplished by varying the model Reynolds number to determine the range (if any) over which the dependent pi term ceases to be affected by changes in Reynolds number. The second point relates to the possibility of cavitation in flow through closed conduits. For example, flow through the complex passages that may exist in valves may lead to local regions of high velocity (and thus low pressure), which can cause the fluid to cavitate. If the model is to be used to study cavitation phenomena, then the vapor pressure, pυ, becomes an important variable and 1 an additional similarity requirement such as equality of the cavitation number (pr − pυ )2ρV2 is required, where pr is some reference pressure. The use of models to study cavitation is complicated, since it is not fully understood how vapor bubbles form and grow. The initiation of bubbles seems to be influenced by the microscopic particles that exist in most liquids, and how this aspect of the problem influences model studies is not clear. Additional details can be found in Ref. 18. 7.9.2 Flow around Immersed Bodies Geometric and Reynolds number similarity is usually required for models involving flow around bodies. Models have been widely used to study the flow characteristics associated with bodies that are completely immersed in a moving fluid. Examples include flow around aircraft, submarines, automobiles, golf balls, and buildings. (These types of models are usually tested in wind tunnels as is illustrated in Fig. 7.6.) Modeling laws for these problems are similar to those described in the preceding section; that is, geometric and Reynolds number similarity is required. Since there are no fluid interfaces, surface tension (and therefore the Weber number) is not important. Also, gravity will not affect the flow patterns, so the Froude number need not be considered. The Mach number will be important for high-speed flows in which compressibility becomes an important factor, but for incompressible flows (such as those in liquids or gases at relatively low speeds) the Mach 7.9 Some Typical Model Studies 379 ■ Figure 7.6 Model of the National Bank of Commerce, San Antonio, Texas, for measurement of peak, rms, and mean pressure distributions. The model is located in a long-test-section, meteorological wind tunnel. (Photograph courtesy of CPP Wind Engineering; www.cppwind.com.) number can be omitted as a similarity requirement. In this case, a general formulation for these problems is ℓi ε ρVℓ Dependent pi term = ϕ ( , , ℓ ℓ μ ) VIDEO V7.13 Wind engineering models (7.18) where ℓ is some characteristic length of the system and ℓi represents other pertinent lengths, εℓ is the relative roughness of the surface (or surfaces), and ρVℓμ is the Reynolds number. Frequently, the dependent variable of interest for this type of problem is the drag, 𝒟, developed on the body, and in this situation the dependent dimensionless group would usually be expressed in the form of a drag coefficient, CD = 𝒟 1 2 2 2 ρV ℓ ℓ2 is usually taken as some representative area of the object. Thus, drag studies can be undertaken with the formulation ℓi ε ρVℓ = CD = ϕ ( , , ℓ ℓ μ ) 𝒟 1 2 2 2 ρV ℓ (7.19) It is clear from Eq. 7.19 that geometric similarity εm ℓim ℓi ε =    = ℓm ℓ ℓm ℓ as well as Reynolds number similarity ρmVmℓm ρVℓ = μm μ For flow around bodies, drag is often the dependent variable of interest. must be maintained. If these conditions are met, then 𝒟 or 1 2 2 2 ρV ℓ 𝒟= = 𝒟m 1 2 2 2 ρmV mℓm ρ V 2 ℓ 2 𝒟 ρm ( Vm ) ( ℓm ) m Measurements of model drag, 𝒟m, can then be used to predict the corresponding drag, 𝒟, on the prototype from this relationship. As was discussed in the previous section, one of the common difficulties with models is related to the Reynolds number similarity requirement, which establishes the model velocity as Vm = μm ρ ℓ V μ ρm ℓm (7.20) νm ℓ V ν ℓm (7.21) or Vm = 380 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling where νm ν is the ratio of kinematic viscosities. If the same fluid is used for model and prototype so that νm = ν, then Vm = VIDEO V7.14 Model airplane test in water VIDEO V7.15 Large scale wind tunnel ℓ V ℓm and, therefore, the required model velocity will be higher than the prototype velocity for ℓℓm greater than 1. Since this ratio is often relatively large, the required value of Vm may be large. For 1 example, for a 10 length scale, and a prototype velocity of 50 mph, the required model velocity is 500 mph. This is a value that is unreasonably high to achieve with liquids, and for gas flows this would be in the range where compressibility would be important in the model (but not in the prototype). As an alternative, we see from Eq. 7.21 that Vm could be reduced by using a different fluid in the model such that νm ν < 1. For example, the ratio of the kinematic viscosity of water to that 1 of air is approximately 10, so that if the prototype fluid were air, tests might be run on the model using water. This would reduce the required model velocity, but it still may be difficult to achieve the necessary velocity in a suitable test facility, such as a water tunnel. Another possibility for wind tunnel tests would be to increase the air pressure in the tunnel so that ρm > ρ, thus reducing the required model velocity as specified by Eq. 7.20. Fluid viscosity is not strongly influenced by pressure. Although pressurized tunnels have been built, they are obviously complicated and expensive. The required model velocity can also be reduced if the length scale is modest; that is, the model is relatively large. For wind tunnel testing, this requires a large test section, which greatly increases the cost of the facility. However, large wind tunnels suitable for testing very large models (or prototypes) are in use. One such tunnel, located at the NASA Ames Research Center, Moffett Field, California, has a test section that is 40 ft by 80 ft and can accommodate test speeds to 345 mph. Such a large and expensive test facility is obviously not feasible for university or industrial laboratories, so most model testing has to be accomplished with relatively small models. EXAM PLE 7.7 Model Design Conditions and Predicted Prototype Performance GIVEN The drag on the airplane shown in Fig. E7.7 cruising at 240 mph in standard air is to be determined from tests on a 1:10 scale model placed in a pressurized wind tunnel. To minimize compressibility effects, the airspeed in the wind tunnel is also to be 240 mph. V = 240 mph FIND Determine (a) the required air pressure in the tunnel (assuming the same air temperature for model and prototype) and (b) the drag on the prototype corresponding to a measured force of 1 lb on the model. ■ Figure E7.7 (Photograph courtesy of NASA) SOLUTION (a) From Eq. 7.19 it follows that drag can be predicted from a geometrically similar model if the Reynolds numbers in model and prototype are the same. Thus, ρVℓ ρm Vmℓm = μm μ 1 For this example, Vm = V and ℓmℓ = 10 so that ρm μm V ℓ = μ Vm ℓm ρ μm = (1) (10) μ and therefore ρm μm = 10 μ ρ This result shows that the same fluid with ρm = ρ and μm = μ cannot be used if Reynolds number similarity is to be maintained. One possibility is to pressurize the wind tunnel to increase the density of the air. We assume that an increase in pressure does not significantly change the viscosity so that the required increase in density is given by the relationship ρm = 10 ρ 7.9 For an ideal gas, p = ρRT so that 381 (b) The drag could be obtained from Eq. 7.19 so that 𝒟 pm ρm = p ρ 1 2 for constant temperature (T = Tm ). Therefore, the wind tunnel would need to be pressurized so that ρV ℓ 2 2 = 𝒟m 1 2 ρmV 2mℓ2m or ρ V 2 ℓ 2 𝒟 ρm ( Vm ) ( ℓm ) m 1 = ( ) (1) 2 (10) 2𝒟m 10 𝒟= pm = 10 p Since the prototype operates at standard atmospheric pressure, the required pressure in the wind tunnel is 10 atmospheres or pm = 10 (14.7 psia) = 147 psia Some Typical Model Studies (Ans) = 10𝒟m Thus, for a drag of 1 lb on the model the corresponding drag on the prototype is 𝒟 = 10 lb COMMENT Thus, we see that a high pressure would be (Ans) required and this could not be achieved easily or inexpensively. However, under these conditions, Reynolds similarity would be attained. CD = 𝔇⁄ ½ AρV 2 VIDEO V7.16 Full Scale Race Car Model 100 10 1 0.1 10 –1 101 103 ρVd Re = ____ μ 105 At high Reynolds numbers the drag is often essentially independent of the Reynolds number. 𝔇 V 107 Fortunately, in many situations the flow characteristics are not strongly influenced by the Reynolds number over the operating range of interest. In these cases we can avoid the rather stringent similarity requirement of matching Reynolds numbers. To illustrate this point, consider the variation in the drag coefficient with the Reynolds number for a smooth sphere of diameter d placed in a uniform stream with approach velocity, V. Some typical data are shown in the figure in the margin (a larger version is shown in Fig. 9.21c). For Reynolds numbers between approximately 103 and 2 × 105 the drag coefficient is relatively constant and does not strongly depend on the specific value of the Reynolds number. Thus, exact Reynolds number similarity is not required in this range. For other geometric shapes we would typically find that for high Reynolds numbers, inertial forces are dominant (rather than viscous forces), and the drag is essentially independent of the Reynolds number. Finally, consider the interpretation of trends in the dimensional data when plotting pi terms. For example, if ρ, μ, and d remain constant, then an increase in Re comes from an increase in V. Intuitively, it would seem that if V increases, the drag would increase. However, as shown in the figure, the drag coefficient generally decreases with increasing Re. When interpreting data, one needs to be aware if the variables are nondimensional. In this case, the physical drag force is proportional to the drag coefficient times the velocity squared. Thus, as shown by the figure in the margin, the drag force does, as expected, increase with increasing velocity. The exception occurs in the Reynolds number range 2 × 105 < Re < 4 × 105 where the drag coefficient decreases dramatically with increasing Reynolds number. This phenomenon is discussed in Section 9.3. For problems involving high velocities in which the Mach number is greater than about 0.3, the influence of compressibility, and therefore the Mach number (or Cauchy number), becomes significant. In this case complete similarity requires not only geometric and Reynolds number similarity but also Mach number similarity so that Vm V = cm c (7.22) This similarity requirement, when combined with that for Reynolds number similarity (Eq.7.21), yields c ν ℓm = cm νm ℓ (7.23) Clearly the same fluid with c = cm and ν = νm cannot be used in model and prototype unless the length scale is unity (which means that we are running tests on the prototype). In high-speed 382 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling aerodynamics the prototype fluid is usually air, and it is difficult to satisfy Eq. 7.23 for reasonable length scales. Thus, models involving high-speed flows are often distorted with respect to Reynolds number similarity, but Mach number similarity is maintained. 7.9.3 Flow with a Free Surface Froude number similarity is usually required for models involving freesurface flows. Flows in canals, rivers, spillways, and stilling basins, as well as flow around ships, are all examples of flow phenomena involving a free surface. For this class of problems, both gravitational and inertial forces are important and, therefore, the Froude number becomes an important similarity parameter. Also, since there is a free surface with a liquid–air interface, forces due to surface tension may be significant, and the Weber number becomes another similarity parameter that needs to be considered along with the Reynolds number. Geometric variables will obviously still be important. Thus, a general formulation for problems involving flow with a free surface can be expressed as ℓi ε ρVℓ V ρV2ℓ , Dependent pi term = ϕ ( , , , ℓ ℓ μ √gℓ σ ) (7.24) As discussed previously, ℓ is some characteristic length of the system, ℓi represents other pertinent lengths, and εℓ is the relative roughness of the various surfaces. Since gravity is the driving force in these problems, Froude number similarity is definitely required so that Vm √gmℓm VIDEO V7.17 River flow model = V √gℓ The model and prototype are expected to operate in the same gravitational field (gm = g), and therefore it follows that Vm ℓm = = √λℓ V Bℓ (7.25) Thus, when models are designed on the basis of Froude number similarity, the velocity scale is determined by the square root of the length scale. As is discussed in Section 7.8.3, to simultaneously have Reynolds and Froude number similarity it is necessary that the kinematic viscosity scale be related to the length scale as νm = (λℓ ) 32 ν (7.26) The working fluid for the prototype is normally either freshwater or seawater and the length scale is small. Under these circumstances it is virtually impossible to satisfy Eq. 7.26, so models involving free-surface flows are usually distorted. The problem is further complicated if an attempt is made to model surface tension effects, since this requires the equality of Weber numbers, which leads to the condition σmρm = (λℓ ) 2 σρ VIDEO V7.18 Boat model (7.27) for the kinematic surface tension (σρ). It is again evident that the same fluid cannot be used in model and prototype if we are to have similitude with respect to surface tension effects for λℓ ≠ 1. Fortunately, in many problems involving free-surface flows, both surface tension and viscous effects are small and consequently strict adherence to Weber and Reynolds number similarity is not required. Certainly, surface tension is not important in large hydraulic structures and rivers. Our only concern would be if in a model the depths were reduced to the point where surface tension becomes an important factor, whereas it is not in the prototype. This is of particular importance in the design of river models, since the length scales are typically small (so that the width of the model is reasonable), but with a small length scale the required model depth may be very small. To overcome this problem, different horizontal and vertical length scales are often used for river models. Although this approach eliminates surface tension effects in the model, it introduces geometric 7.9 Some Typical Model Studies 383 ■ Figure 7.7 A scale hydraulic model (1:197) of the Guri Dam in Venezuela which is used to simulate the characteristics of the flow over and below the spillway and the erosion below the spillway. (Photograph courtesy of St. Anthony Falls Hydraulic Laboratory.) VIDEO V7.19 Dam model distortion that must be accounted for empirically, usually by increasing the model surface roughness. It is important in these circumstances that verification tests with the model be performed (if possible) in which model data are compared with available prototype river flow data. Model roughness can be adjusted to give satisfactory agreement between model and prototype, and then the model subsequently used to predict the effect of proposed changes on river characteristics (such as velocity patterns or surface elevations). For large hydraulic structures, such as dam spillways, the Reynolds numbers are large so that viscous forces are small in comparison to the forces due to gravity and inertia. In this case, Reynolds number similarity is not maintained and models are designed on the basis of Froude number similarity. Care must be taken to ensure that the model Reynolds numbers are also large, but they are not required to be equal to those of the prototype. This type of hydraulic model is usually made as large as possible so that the Reynolds number will be large. A spillway model is shown in Fig. 7.7 Also, for relatively large models the geometric features of the prototype can be accurately scaled, as well as surface roughness. Note that εm = λℓε, which indicates that the model surfaces must be smoother than the corresponding prototype surfaces for λℓ < 1. THE WIDE WORLD OF FLUIDS Ice engineering Various types of models have been studied piers, ships, and offshore platforms can be performed. Ambient in wind tunnels, water tunnels, and towing tanks for many temperatures can be controlled as low as −20 °F, and at this years. But another type of facility is needed to study ice and temperature a 2-mm-per-hour ice growth rate can be achieved. ice-related problems. The U.S. Army Cold Regions Research It is also possible to control the mechanical properties of the and Engineering Laboratory has developed a unique complex ice to properly match the physical scale of the model. Tests that houses research facilities for studies related to the run in the recirculating flume can simulate river processes dur- mechanical behavior of ice and ice–structure interactions. The ing ice formation. And in the large research area, scale models laboratory contains three separate cold-rooms—a test basin, a of lakes and rivers can be built and operated to model ice flume, and a general research area. In the test basin, large- interactions with various types of engineering projects. (See scale model studies of ice forces on structures such as dams, Problem 7.75.) E X AMPLE 7.8 Froude Number Similarity GIVEN The spillway for the dam shown in Fig. E7.8a is 20 m wide and is designed to carry 125 m3/s at flood stage. A 1:15 model is constructed to study the flow characteristics through the spillway. The effects of surface tension and viscosity are to be neglected. FIND (a) Determine the required model width and flowrate. (b) What operating time for the model corresponds to a 24-hr period in the prototype? ■ Figure E7.8a (Photo by Marty Melchior.) 384 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling SOLUTION The width, wm, of the model spillway is obtained from the length scale, λℓ, so that wm 1 = λℓ = w 15 20 m = 1.33 m 15 (Ans) Of course, all other geometric features (including surface roughness) of the spillway must be scaled in accordance with the same length scale. With the neglect of surface tension and viscosity, Eq. 7.24 indicates that dynamic similarity will be achieved if the Froude numbers are equal between model and prototype. Thus, Vm √gmℓm = ℓm tm V ℓm = = = √λℓ t Vm ℓ Bℓ This result indicates that time intervals in the model will be smaller than the corresponding intervals in the prototype if 1 λℓ < 1. For λℓ = 15 and a prototype time interval of 24 hr Therefore, wm = or V √gℓ and for gm = g ℓm Vm = V Bℓ tm = √151 (24 hr) = 6.20 hr COMMENT As indicated by the foregoing analysis, the time scale varies directly as the square root of the length scale. Thus, as shown in Fig. E7.8b, the model time interval, tm , corresponding to a 24-hr prototype time interval can be varied by changing the length scale, λℓ . The ability to scale times may be very useful, since it is possible to “speed up” events in the model that may occur over a relatively long time in the prototype. There is, of course, a practical limit to how small the length scale (and the corresponding time scale) can become. For example, if the length scale is too small then surface tension effects may become important in the model whereas they are not in the prototype. In such a case the present model design, based simply on Froude number similarity, would not be adequate. Since the flowrate is given by Q = VA, where A is an appropriate cross-sectional area, it follows that Qm Vm A m ℓm ℓm 2 = = Q VA B ℓ(ℓ) = (λℓ ) 52 20 16   tm , hr 12 where we have made use of the relationship AmA = (ℓmℓ) 2. 1 For λℓ = 15 and Q = 125 m3s 8 (1/15, 6.20 hr) (Ans) 4 The time scale can be obtained from the velocity scale, since the velocity is distance divided by time (V = ℓt), and therefore 0 Qm = 1 ( 15 ) 52 (125 m s) = 0.143 m s 3 V ℓ tm = Vm t ℓm VIDEO V7.20 Testing of large yacht model (Ans) 3 0 0.1 0.2 0.3 0.4 0.5 ℓm ___ ℓ ■ Figure E7.8b There are, unfortunately, problems involving flow with a free surface in which viscous, inertial, and gravitational forces are all important. The drag on a ship as it moves through water is due to the viscous shearing stresses that develop along its hull, as well as a pressure-induced component of drag caused by both the shape of the hull and wave action. The shear drag is a function of the Reynolds number, whereas the pressure drag is a function of the Froude number. Since both Reynolds number and Froude number similarity cannot be simultaneously achieved by using water as the model fluid (which is the only practical fluid for ship models), some technique other than a straightforward model test must be employed. One common approach is to measure the total drag on a small, geometrically similar model as it is towed through a model basin at Froude numbers matching those of the prototype. The shear drag on the model is calculated using analytical techniques of the type described in Chapter 9. This calculated value is then subtracted from the total drag to obtain pressure drag, and using Froude number scaling the pressure drag on the prototype can then be predicted. The experimentally determined value can then be combined with a calculated 7.10 Similitude Based on Governing Differential Equations 385 ■ Figure 7.8 Instrumented, smallwaterplane-area, twin hull (SWATH) model suspended from a towing carriage. (Photograph courtesy of the U.S. Navy’s David W. Taylor Research Center.) value of the shear drag (again using analytical techniques) to provide the desired total drag on the ship. Ship models are widely used to study new designs, but the tests require extensive facilities (see Fig. 7.8). It is clear from this brief discussion of various types of models involving free-surface flows that the design and use of such models requires considerable ingenuity, as well as a good understanding of the physical phenomena involved. This is generally true for most model studies. Modeling is both an art and a science. Motion picture producers make extensive use of model ships, fires, explosions, and the like. It is interesting to attempt to observe the flow differences between these distorted model flows and the real thing. 7.10 Similarity laws can be directly developed from the equations governing the phenomenon of interest. Similitude Based on Governing Differential Equations In the preceding sections of this chapter, dimensional analysis has been used to obtain similarity laws. This is a simple, straightforward approach to modeling, which is widely used. The use of dimensional analysis requires only a knowledge of the variables that influence the phenomenon of interest. Although the simplicity of this approach is attractive, it must be recognized that omission of one or more important variables may lead to serious errors in the model design. An alternative approach is available if the equations (usually differential equations) governing the phenomenon are known. In this situation similarity laws can be developed from the governing equations, even though it may not be possible to obtain analytic solutions to the equations. To illustrate the procedure, consider the flow of an incompressible Newtonian fluid. For simplicity we will restrict our attention to two-dimensional flow, although the results are applicable to the general three-dimensional case. From Chapter 6 we know that the governing equations are the continuity equation ∂u ∂υ + =0 ∂x ∂y (7.28) and the Navier–Stokes equations ρ( ∂p ∂u ∂u ∂u ∂ 2u ∂ 2u +u + υ ) = − + μ( 2 + 2 ) ∂t ∂x ∂y ∂x ∂x ∂y (7.29) ρ( ∂p ∂υ ∂ 2υ ∂υ ∂υ ∂ 2υ +u + υ ) = − − ρg + μ ( 2 + 2 ) ∂t ∂x ∂y ∂y ∂x ∂y (7.30) where the y axis is vertical, so that the gravitational body force, ρg, only appears in the “y equation.” To continue the mathematical description of the problem, boundary conditions are required. For example, velocities on all boundaries may be specified; that is, u = uB and υ = υB at all boundary points x = xB and y = yB. In some types of problems it may be necessary to specify the pressure over some part of the boundary. For time-dependent problems, initial conditions would also have to be provided, which means that the values of all dependent variables would be given at some time (usually taken at t = 0). 386 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling Once the governing equations, including boundary and initial conditions, are known, we are ready to proceed to develop similarity requirements. The next step is to define a new set of variables that are dimensionless. To do this we select a reference quantity for each type of variable. In this problem the variables are u, υ, p, x, y, and t so we will need a reference velocity, V, a reference pressure, p0, a reference length, ℓ, and a reference time, τ. These reference quantities should be parameters that appear in the problem. For example, ℓ may be a characteristic length of a body immersed in a fluid or the width of a channel through which a fluid is flowing. The velocity, V, may be the freestream velocity or the inlet velocity. The new dimensionless (starred) variables can be expressed as Each variable is made dimensionless by dividing by an appropriate reference quantity. as shown in the figure in the margin. The governing equations can now be rewritten in terms of these new variables. For example, y u=V ρ, μ υ=0 p = p0 p u υ   υ* =   p* = p0 V V y x t x* =   y* =   t* = τ ℓ ℓ u* = ∂u ∂Vu* ∂x* V ∂u* = = ∂x ∂x* ∂x ℓ ∂x* x=ℓ x Actual and y* u* = 1 υ* = 0 p* = 1 ∂ 2u V ∂ ∂u* ∂x* V ∂ 2u* = = ℓ ∂x* ( ∂x* ) ∂x ∂x2 ℓ2 ∂x*2 Re x* = 1 x* Dimensionless The other terms that appear in the equations can be expressed in a similar fashion. Thus, in terms of the new variables the governing equations become ∂u* ∂υ* + =0 ∂x* ∂y* (7.31) and p0 ∂p* ρV ∂u* ρV2 μV ∂ 2u* ∂u* ∂u* ∂ 2u* + u* + υ* = − + + 2 2 [ τ ] ∂t* [ ℓ ]( ∂x* [ ℓ ] ∂x* [ ℓ ]( ∂x* ∂y* ) ∂y*2 ) (7.32) ρV ∂υ* ρV2 ∂υ* ∂υ* + [ τ ] ∂t* [ ℓ ](u* ∂x* + υ* ∂y* ) FIℓ FIc p0 ∂p* μV ∂ 2υ* ∂ 2υ* = −[ ] − [ρg] + [ 2 ]( + ℓ ∂y* ℓ ∂x*2 ∂y*2 ) FP FG (7.33) FV The terms appearing in brackets contain the reference quantities and can be interpreted as indices of the various forces (per unit volume) that are involved. Thus, as is indicated in Eq. 7.33, FIℓ = inertia (local) force, FIc = inertia (convective) force, Fp = pressure force, FG = gravitational force, and FV = viscous force. As the final step in the nondimensionalization process, we will divide each term in Eqs. 7.32 and 7.33 by one of the bracketed quantities. Although any one of these quantities could be used, it is conventional to divide by the bracketed quantity ρV2 ℓ, which is the index of the convective inertia force. The final nondimensional form then becomes p0 ∂p* μ ℓ ∂u* ∂u* ∂u* ∂ 2u* ∂ 2u* + u* + υ* = − + + 2 2 [ τV ] ∂t* [ ρV ] ∂x* [ ρVℓ ]( ∂x* ∂x* ∂y* ∂y*2 ) (7.34) p0 ∂p* μ gℓ ℓ ∂υ* ∂υ* ∂υ* ∂ 2υ* ∂ 2υ* + u* + υ* = − − + + 2 2 2 [ τV ] ∂t* [ ρV ] ∂y* [ V ] [ ρVℓ ]( ∂x* ∂x* ∂y* ∂y*2 ) (7.35) We see that bracketed terms are standard dimensionless groups (or their reciprocals) which were developed from dimensional analysis; that is, ℓτV is a form of the Strouhal number, p0 ρV2 7.10 Governing equations expressed in terms of dimensionless variables lead to the appropriate dimensionless groups. Similitude Based on Governing Differential Equations 387 the Euler number, gℓV2 the reciprocal of the square of the Froude number, and μρVℓ the reciprocal of the Reynolds number. From this analysis it is now clear how the dimensionless groups can be interpreted as the ratio of two forces, and how these groups arise naturally from the governing equations. Although we really have not helped ourselves with regard to obtaining an analytical solution to these equations (they are still complicated and not amenable to an analytical solution), the dimensionless forms of the equations, Eqs. 7.31, 7.34, and 7.35, can be used to establish similarity requirements. From these equations it follows that if two systems are governed by these equations, then the solutions (in terms of u*, υ*, p*, x*, y*, and t*) will be the same if the four parameters ℓτV, p0 ρV2, V2 gℓ, and ρVℓμ are equal for the two systems. The two systems will be dynamically similar. Of course, boundary and initial conditions expressed in dimensionless form must also be equal for the two systems, and this will require complete geometric similarity. These are the same similarity requirements that would be determined by a dimensional analysis if the same variables were considered. However, the advantage of working with the governing equations is that the variables appear naturally in the equations, and we do not have to worry about omitting an important one, provided the governing equations are correctly specified. We can thus use this method to deduce the conditions under which two solutions will be similar even though one of the solutions will most likely be obtained experimentally. In the foregoing analysis we have considered a general case in which the flow may be unsteady, and both the actual pressure level, p0, and the effect of gravity are important. A reduction in the number of similarity requirements can be achieved if one or more of these conditions is removed. For example, if the flow is steady, the dimensionless group, ℓτV, can be eliminated. The actual pressure level will only be of importance if we are concerned with cavitation. If not, the flow patterns and the pressure differences will not depend on the pressure level. In this case, 1 p0 can be taken as ρV2 (or 2ρV2 ), and the Euler number can be eliminated as a similarity requirement. However, if we are concerned about cavitation (which will occur in the flow field if the pressure at certain points reaches the vapor pressure, pυ), then the actual pressure level is important. Usually, in this case, the characteristic pressure, p0, is defined relative to the vapor pressure such that p0 = pr − pυ where pr is some reference pressure within the flow field. With p0 defined in this manner, the similarity parameter p0 ρV2 becomes (p r − pυ )ρV2. This parameter is frequently 1 written as ( pr − pυ )2ρV2, and in this form, as was noted previously in Section 7.6, is called the cavitation number. Thus we can conclude that if cavitation is not of concern we do not need a similarity parameter involving p0 , but if cavitation is to be modeled, then the cavitation number becomes an important similarity parameter. The Froude number, which arises because of the inclusion of gravity, is important for problems in which there is a free surface. Examples of these types of problems include the study of rivers, flow through hydraulic structures such as spillways, and the drag on ships. In these situations the shape of the free surface is influenced by gravity, and therefore the Froude number becomes an important similarity parameter. However, if there are no free surfaces, the only effect of gravity is to superimpose a hydrostatic pressure distribution on the pressure distribution created by the fluid motion. The hydrostatic distribution can be eliminated from the governing equation (Eq. 7.30) by defining a new pressure, p′ = p − ρgy, and with this change the Froude number does not appear in the nondimensional governing equations. We conclude from this discussion that for the steady flow of an incompressible fluid without free surfaces, dynamic and kinematic similarity will be achieved if (for geometrically similar systems) Reynolds number similarity exists. If free surfaces are involved, Froude number similarity must also be maintained. For free-surface flows we have tacitly assumed that surface tension is not important. We would find, however, that if surface tension is included, its effect would appear in the free-surface boundary condition, and the Weber number, ρV 2ℓσ, would become an additional similarity parameter. In addition, if the governing equations for compressible fluids are considered, the Mach number, Vc, would appear as an additional similarity parameter. It is clear that all the common dimensionless groups that we previously developed by using dimensional analysis appear in the governing equations that describe fluid motion when these equations are expressed in terms of dimensionless variables. Thus, use of the governing equations to obtain similarity laws provides an alternative to dimensional analysis. This approach has the       388 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling advantage that the variables are known and the assumptions involved are clearly identified. In addition, a physical interpretation of the various dimensionless groups can often be obtained. 7.11 similitude dimensionless product basic dimensions pi term Buckingham pi theorem method of repeating variables common dimensionless group model modeling laws prototype prediction equation model design conditions similarity requirements modeling laws length scale distorted model true model Chapter Summary and Study Guide Many practical engineering problems involving fluid mechanics require experimental data for their solution. Thus, laboratory studies and experimentation play a significant role in this field. It is important to develop good procedures for the design of experiments so they can be efficiently completed with as broad applicability as possible. To achieve this end the concept of similitude is often used in which measurements made in the laboratory can be utilized for predicting the behavior of other similar systems. In this chapter, dimensional analysis is used for designing such experiments, as an aid for correlating experimental data, and as the basis for the design of physical models. As the name implies, dimensional analysis is based on a consideration of the dimensions required to describe the variables in a given problem. A discussion of the use of dimensions and the concept of dimensional homogeneity (which forms the basis for dimensional analysis) was included in Chapter 1. Essentially, dimensional analysis simplifies a given problem described by a certain set of variables by reducing the number of variables that need to be considered. In addition to being fewer in number, the new variables are dimensionless products of the original variables. Typically these new dimensionless variables are much simpler to work with in performing the desired experiments. The Buckingham pi theorem, which forms the theoretical basis for dimensional analysis, is introduced. This theorem establishes the framework for reducing a given problem described in terms of a set of variables to a new set of fewer dimensionless variables. A simple method, called the repeating variable method, is described for actually forming the dimensionless variables (often called pi terms). Forming dimensionless variables by inspection is also considered. It is shown how the use of dimensionless variables can be of assistance in planning experiments and as an aid in correlating experimental data. The repeating variable method (or any other method for forming dimensionless groups) can be applied to a generic fluid mechanics problem in which nearly all of the pertinent dimensional variables appear. The result is a set of common dimensionless groups (called “Numbers” and “Coefficients”) that can be used to describe almost any problem. These common dimensionless groups represent the ratios of relevant forces or energies in a flow. For problems in which there are a large number of variables, the use of physical models is described. Models are used to make specific predictions from laboratory tests rather than formulating a general relationship for the phenomenon of interest. The correct design of a model is obviously imperative for the accurate predictions of other similar, but usually larger, systems. It is shown how dimensional analysis can be used to establish a valid model design. An alternative approach for establishing similarity requirements using governing equations (usually differential equations) is presented. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ ■ ■ ■ ■ ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. use the Buckingham pi theorem to determine the number of independent dimensionless variables needed for a given flow problem. form a set of dimensionless variables using the method of repeating variables. form a set of dimensionless variables by inspection. recognize common dimensionless groups used in fluid mechanics and list the physical phenomena they represent use dimensionless variables as an aid in interpreting and correlating experimental data. use dimensional analysis to establish a set of similarity requirements (and prediction equation) for a model to be used to predict the behavior of another similar system (the prototype). rewrite a given governing equation in a suitable nondimensional form and deduce similarity requirements from the nondimensional form of the equation. References 389 Some of the important equations in this chapter are: ρVℓ μ Reynolds number Re = Froude number Fr = Euler number Eu = p ρV2 Cauchy number Ca = ρV 2 Ev Mach number Ma = V c St = ωℓ V Strouhal number Weber number Relative roughness We = V √gℓ ρV 2ℓ σ ε ℓ K= hL V 2g Pressure coefficient Cp = Δp 12ρV 2 Friction coefficient Cf = τw 12ρV 2 Loss coefficient 2 Drag coefficient CD = 𝒟 12ρV 2ℓ2 Lift coefficient CL = ℒ 12ρV 2ℓ2 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Bridgman, P. W., Dimensional Analysis, Yale University Press, New Haven, Conn., 1922. Murphy, G., Similitude in Engineering, Ronald Press, New York, 1950. Langhaar, H. L., Dimensional Analysis and Theory of Models, Wiley, New York, 1951. Huntley, H. E., Dimensional Analysis, Macdonald, London, 1952. Duncan, W. J., Physical Similarity and Dimensional Analysis: An Elementary Treatise, Edward Arnold, London, 1953. Sedov, K. I., Similarity and Dimensional Methods in Mechanics, Academic Press, New York, 1959. Ipsen, D. C., Units, Dimensions, and Dimensionless Numbers, McGraw-Hill, New York, 1960. Kline, S. J., Similitude and Approximation Theory, McGraw-Hill, New York, 1965. Skoglund, V. J., Similitude—Theory and Applications, International Textbook, Scranton, Pa., 1967. Baker, W. E., Westline, P. S., and Dodge, F. T., Similarity Methods in Engineering Dynamics—Theory and Practice of Scale Modeling, Hayden (Spartan Books), Rochelle Park, N. J., 1973. Taylor, E. S., Dimensional Analysis for Engineers, Clarendon Press, Oxford, UK, 1974. Isaacson, E. de St. Q., and Isaacson, M. de St. Q., Dimensional Methods in Engineering and Physics, Wiley, New York, 1975. 390 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling 13. 14. 15. 16. 17. 18. Schuring, D. J., Scale Models in Engineering, Pergamon Press, New York, 1977. Yalin, M. S., Theory of Hydraulic Models, Macmillan, London, 1971. Sharp, J. J., Hydraulic Modeling, Butterworth, London, 1981. Omega Engineering Inc., The Land Chart of Dimensionless Numbers, Model DLN-1, 1991. Schlichting, H., Boundary-Layer Theory, 7th Ed., McGraw-Hill, New York, 1979. Knapp, R. T., Daily, J. W., and Hammitt, F. G., Cavitation, McGraw-Hill, New York, 1970. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are not unique answers to these problems. Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. Section 7.1 Dimensional Analysis 7.1 A mixing basin in a sewage filtration plant is stirred by ˙ (ft ∙ lb/s). mechanical agitation (paddlewheel) with a power input W The degree of mixing of fluid particles is measured by a “velocity gradient” G given by G= ˙ W B μV where μ is the fluid viscosity in lb ∙ s/ft2 and V is the basin volume in ft3. Find the units of the velocity gradient. 7.2 An equation used to evaluate vacuum filtration is Q= ΔpA2 , α(VRw + ARf ) Where Q ≐ L3/T is the filtrate volume flow rate, Δp ≐ F/L2 the vacuum pressure differential, A ≐ L2 the filter area, α the filtrate “viscosity,” V ≐ L3 the filtrate volume, R ≐ L/F the sludge specific resistance, w ≐ F/L3 the weight of dry sludge per unit volume of filtrate, and Rf the specific resistance of the filter medium. What are the dimensions of Rf and and α? 7.3 Verify the left-hand side of Eq. 7.2 is dimensionless using the MLT system. 7.4 The Reynolds number, ρVD/μ, is a very important parameter in fluid mechanics. Verify that the Reynolds number is dimensionless, using both the FLT system and the MLT system for basic dimensions, and determine its value for ethyl alcohol flowing at a velocity of 3 m/s through a 2-in.-diameter pipe. 7.5 For the flow of a thin film of a liquid with a depth h and a free surface, two important dimensionless parameters are the Froude number, V/ √gh, and the Weber number, ρV2h/σ. Determine the value of these two parameters for glycerin (at 20 °C) flowing with a velocity of 0.7 m/s at a depth of 3 mm. 7.6 The Mach number for a body moving through a fluid with velocity V is defined as V/c, where c is the speed of sound in the fluid. This dimensionless parameter is usually considered to be important in fluid dynamics problems when its value exceeds 0.3. What would be the velocity of a body at a Mach number of 0.3 if the fluid is (a) air at standard atmospheric pressure and 20 °C, and (b) water at the same temperature and pressure? Section 7.3 Determination of Pi Terms 7.7 A mixing basin in a sewage filtration plant is stirred by a ˙ ≐ F ∙ L/T. Other parammechanical agitator with a power input W eters describing the performance of the mixing process are the fluid absolute viscosity μ ≐ F ∙ T/L2 , the basin volume V ≐ L3 , and the velocity gradient G ≐ 1/T. Determine the form of the dimensionless relationship. 7.8 The excess pressure inside a bubble (discussed in Chapter 1) is known to be dependent on bubble radius and surface tension. After finding the pi terms, determine the variation in excess pressure if we (a) double the radius and (b) double the surface tension. 7.9 At a sudden contraction in a pipe the diameter changes from D1 to D2. The pressure drop, Δp, which develops across the contraction is a function of D1 and D2, as well as the velocity, V, in the larger pipe, and the fluid density, ρ, and viscosity, μ. Use D1, V, and μ as repeating variables to determine a suitable set of dimensionless parameters. Why would it be incorrect to include the velocity in the smaller pipe as an additional variable? Problems 7.10 Water sloshes back and forth in a tank as shown in Fig. P7.10. The frequency of sloshing, ω, is assumed to be a function of the acceleration of gravity, g, the average depth of the water, h, and the length of the tank, ℓ. Develop a suitable set of dimensionless parameters for this problem using g and ℓ as repeating variables. 391 SPEED LIMIT 40 ω ω h ℓ ■ Figure P7.10 7.11 Assume that the flowrate, Q, of a gas from a smokestack is a function of the density of the ambient air, ρa, the density of the gas, ρg, within the stack, the acceleration of gravity, g, and the height and diameter of the stack, h and d, respectively. Use ρa, d, and g as repeating variables to develop a set of pi terms that could be used to describe this problem. 7.12 The pressure rise, Δp, across a pump can be expressed as Δp = f(D, ρ, ω, Q) ■ Figure P7.15 7.16 A cone and plate viscometer consists of a cone with a very small angle α that rotates above a flat surface as shown in Fig. P7.16. The torque, 𝒯, required to rotate the cone at an angular velocity ω is a function of the radius, R, the cone angle, α, and the fluid viscosity, μ, in addition to ω. With the aid of dimensional analysis, determine how the torque will change if both the viscosity and angular velocity are doubled. where D is the impeller diameter, ρ the fluid density, ω the rotational speed, and Q the flowrate. Determine a suitable set of dimensionless parameters. R ω 𝒯 7.13 A thin elastic wire is placed between rigid supports. A fluid flows past the wire, and it is desired to study the static deflection, δ, at the center of the wire due to the fluid drag. Assume that δ = f(ℓ, d, ρ, μ, V, E) where ℓ is the wire length, d the wire diameter, ρ the fluid density, μ the fluid viscosity, V the fluid velocity, and E the modulus of elasticity of the wire material. Develop a suitable set of pi terms for this problem. 7.14 Because of surface tension, it is possible, with care, to support an object heavier than water on the water surface as shown in Fig. P7.14. (See Video V1.9.) The maximum thickness, h, of a square of material that can be supported is assumed to be a function of the length of the side of the square, ℓ, the density of the material, ρ, the acceleration of gravity, g, and the surface tension of the liquid, σ. Develop a suitable set of dimensionless parameters for this problem. ℓ h Fluid α ■ Figure P7.16 7.17 The pressure drop, Δp, along a straight pipe of diameter D has been experimentally studied, and it is observed that for laminar flow of a given fluid and pipe, the pressure drop varies directly with the distance, ℓ, between pressure taps. Assume that Δp is a function of D and ℓ, the velocity, V, and the fluid viscosity, μ. Use dimensional analysis to deduce how the pressure drop varies with pipe diameter. 7.18 A cylinder with a diameter D floats upright in a liquid as shown in Fig. P7.18. When the cylinder is displaced slightly along its vertical axis it will oscillate about its equilibrium position with a frequency, ω. Assume that this frequency is a function of the diameter, D, the mass of the cylinder, m, and the specific weight, γ, Cylinder diameter = D ■ Figure P7.14 7.15 Under certain conditions, wind blowing past a rectangular speed limit sign can cause the sign to oscillate with a frequency ω. (See Fig. P7.15 and Video V9.9.) Assume that ω is a function of the sign width, b, sign height, h, wind velocity, V, air density, ρ, and an elastic constant, k, for the supporting pole. The constant, k, has dimensions of FL. Develop a suitable set of pi terms for this problem. Liquid ■ Figure P7.18 392 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling of the liquid. Determine, with the aid of dimensional analysis, how the frequency is related to these variables. If the mass of the cylinder were increased, would the frequency increase or decrease? †7.19 Consider a typical situation involving the flow of a fluid that you encounter almost every day. List what you think are the important physical variables involved in this flow and determine an appropriate set of pi terms for this situation. Section 7.5 Determination of Pi Terms by Inspection 7.27 A viscous fluid is poured onto a horizontal plate as shown in Fig. P7.27. Assume that the time, t, required for the fluid to flow a certain distance, d, along the plate is a function of the volume of fluid poured, V , acceleration of gravity, g, fluid density, ρ, and fluid viscosity, μ. Determine an appropriate set of pi terms to describe this process. Form the pi terms by inspection. 7.20 The weir shown in Fig. P7.20 is used to measure the volume flowrate Q. The height H is a measure of this flowrate. The weir has length L (perpendicular to the paper). Select and include relevant fluid properties and find the appropriate dimensionless parameters. Volume V H d Q ■ Figure P7.27 g P Density = ρ Absolute viscosity = μ ■ Figure P7.20 7.21 Experiments are conducted on a washing machine agitator. The relevant dimensional parameters are the driving torque, 𝒯, the oscillation frequency, f, the angular velocity, ω, the number of paddles, N, the paddle height, H, and the paddle width, w. Specify the relevant fluid properties and find the appropriate dimensionless parameters. ˙ , to a large industrial fan depends on the 7.22 The input power, W fan impeller diameter D, fluid viscosity μ, fluid density ρ, volumetric flow Q, and blade rotational speed ω. What are the appropriate dimensionless parameters? 7.23 Develop the appropriate dimensionless parameters for the period τ of transverse vibration of a turbine rotor of mass m connected to a shaft of stiffness k ≐ F/L and length ℓ ≐ L. Other relevant dimensional parameters are the eccentricity ε ≐ L of the center of mass of the rotor, the speed of rotation N ≐ 1/T of the shaft, and the amplitude A ≐ L of the vibration. 7.24 A vapor bubble rises in a liquid. The relevant dimensional parameters are the liquid specific weight, γℓ , the vapor specific weight, γν , bubble velocity, V, bubble diameter, d, surface tension, σ, and liquid viscosity, μ. Find appropriate dimensionless parameters. 7.25 The speed of deep ocean waves depends on the wave length and gravitational acceleration. What are the appropriate dimensionless parameters? 7.26 A coach has been trying to evaluate the accuracy of a baseball pitcher. After two years of studying, he proposes a function that can be presented as the accuracy of any pitcher: Acc = f (V, a, m, ρ, p, z), where Acc is the dimensionless accuracy, V is the velocity of the ball, a is the age of the pitcher, m is the mass of the pitcher, ρ is the density of air where the game is played, p is the pressure where the game is played (which varies with elevation), and z is the elevation above sea level. Find the dimensionless groups for this function. 7.28 The velocity, c, at which pressure pulses travel through arteries (pulse-wave velocity) is a function of the artery diameter, D, and wall thickness, h, the density of blood, ρ, and the modulus of elasticity, E, of the arterial wall. Determine a set of nondimensional parameters that can be used to study experimentally the relationship between the pulse-wave velocity and the variables listed. Form the nondimensional parameters by inspection. 7.29 As shown in Fig. P7.29 and Video V5.6, a jet of liquid directed against a block can tip over the block. Assume that the velocity, V, needed to tip over the block is a function of the fluid density, ρ, the diameter of the jet, D, the weight of the block, 𝒲, the width of the block, b, and the distance, d, between the jet and the bottom of the block. (a) Determine a set of dimensionless parameters for this problem. Form the dimensionless parameters by inspection. (b) Use the momentum equation to determine an equation for V in terms of the other variables. (c) Compare the results of parts (a) and (b). b V ρ D 𝒲 d ■ Figure P7.29 7.30 By inspection, arrange the following dimensional parameters into dimensionless parameters: (a) kinematic viscosity, v, length, ℓ, and time, t; and (b) volume flow rate, Q, pump diameter, D, and pump impeller rotation speed, N. 7.31 A screw propeller has the following relevant dimensional parameters: axial thrust, F, propeller diameter, D, fluid kinematic viscosity, v, fluid density, ρ, gravitational acceleration, g, advance velocity, V, and rotational speed, N. Find appropriate dimensionless parameters to present the test data. Problems Section 7.7 Correlation of Experimental Data Section 7.6 Common Dimensionless Groups in Fluid Mechanics 7.32 Shown in the following table are several flow situations and the associated characteristic velocity, size, and fluid kinematic viscosity. Determine the Reynolds number for each of the flows and indicate for which ones the inertial effects are small relative to viscous effects. Flow Airplane Mosquito Syrup on pancake Blood in capillary 393 Velocity ft/s Size ft Kinematic viscosity ft2/s 400 0.02 60 0.015 4.3 × 10−4 1.6 × 10−4 0.03 0.12 2.5 × 10−3 0.001 5 × 10−5 1.2 × 10−5 *7.38 The pressure rise, Δp = p2 − p1, across the abrupt expansion of Fig. P7.38 through which a liquid is flowing can be expressed as Δp = f (A1, A2, ρ, V1 )     where A1 and A2 are the upstream and downstream cross-sectional areas, respectively, ρ is the fluid density, and V1 is the upstream velocity. Some experimental data obtained with A2 = 1.25 ft2, V1 = 5.00 ft/s, and using water with ρ = 1.94 slugs/ft3 are given in the following table: A1 (ft2) Δp (lb/ft ) 2 7.33 A large, hot plate hangs vertically in a room. Heat transfer from the plate to the air near it causes the air density to decrease. This lighter air rises upward with a velocity proportional to the group 0.10 0.25 3.25 7.85 0.37 10.3 0.52 11.6 0.61 12.3 Plot the results of these tests using suitable dimensionless parameters. With the aid of a standard curve fitting program determine a general equation for Δp and use this equation to predict Δp for water flowing through an abrupt expansion with an area ratio A1/A2 = 0.35 at a velocity V1 = 3.75 ft/s. p2 [αT gL(Tp − Tf )]1/2 , p1 −1 where αT is the coefficient of thermal expansion [units = (°R) or K−1], g is the acceleration of gravity, L is the plate length, and Tp and Tf are the temperatures of the plate and fluid, respectively, in °R or K. The resulting heat transfer process depends on the Grashof Number, defined by G= ρ2αTg(Tp − Tf )L3 μ2 , V1 Area = A1 Area = A2 ■ Figure P7.38 where ρ is density and μ is dynamic viscosity. Show that G is dimensionless. How is G related to the common dimensionless parameters of fluid mechanics? Is it equivalent to a combination, *7.39 The pressure drop, Δp, over a certain length of horizontal power, or product of one or more of the common dimensionless pipe is assumed to be a function of the velocity, V, of the fluid in parameters listed in Table 7.1? the pipe, the pipe diameter, D, and the fluid density and viscosity, 7.34 Develop the Weber number by starting with estimates for the ρ and μ. (a) Show that this flow can be described in2 dimensionless form as a “pressure coefficient,” Cp = Δp/(0.5 ρV ) that depends inertia and surface tension forces. on the Reynolds number, Re = ρVD/μ. (b) The following data were 7.35 Develop the Froude number by starting with estimates of the obtained in an experiment involving a fluid with ρ = 2 slugs/ft3, fluid kinetic energy and fluid potential energy. μ = 2 × 10−3 lb ∙ s/ft2, and D = 0.1 ft. Plot a dimensionless graph and use a power law equation to determine the functional relation7.36 The following dimensionless groups are often used to pre- ship between the pressure coefficient and the Reynolds number. Q sent data on centrifugal pumps: flow coefficient φ = , head (c) What are the limitations on the applicability of your equation ωD3 obtained in part (b)? ˙ gH W coefficient ψ = 2 2 , power coefficient ξ = 3 5 , efficiency η = ωD ωD V, ft/s Δp, lb/ft2 ω√Q ω√Q ρgQH , specific speed Ns = , specific diameter D = . s ˙ 3 192 (gh) 3/4 (gh) 1/4 W 11 704 Show that the last three groups can be formed from combinations 17 1088 of the first three groups. 20 1280 7.37 The dimensional parameters used to describe the operation of a ship or airplane propeller (sometimes called a screw propeller) are rotational speed, ω, diameter, D, fluid density, ρ, speed of the propeller relative to the fluid, V, and thrust devel- *7.40 The pressure drop across a short hollowed plug placed in a oped, T. The common dimensionless groups are called the thrust circular tube through which a liquid is flowing (see Fig. P7.40) coefficient and the advance ratio. Propose appropriate definitions can be expressed as for these groups. Δp = f (ρ, V, D, d)     394 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling where ρ is the fluid density, and V is the mean velocity in the tube. Some experimental data obtained with D = 0.2 ft, ρ = 2.0 slugs/ft3, and V = 2 ft/s are given in the following table: d (ft) 0.06 Δp (lbft ) 2 0.08 493.8 0.10 156.2 64.0 0.15 12.6 Plot the results of these tests, using suitable dimensionless parameters, on a log–log scale. Use a standard curve-fitting technique to determine a general equation for Δp. What are the limits of applicability of the equation? Δp V D d ■ Figure P7.40 *7.41 As shown in Fig. 2.26. Fig. P7.41, and Video V2.12, a rectangular barge floats in a stable configuration provided the distance between the center of gravity, CG, of the object (boat and load) and the center of buoyancy, C, is less than a certain amount, H. If this distance is greater than H, the boat will tip over. Assume H is a function of the boat’s width, b, length, ℓ, and draft, h. (a) Put this relationship into dimensionless form. (b) The results of a set of experiments with a model barge with a width of 1.0 m are shown in the table. Plot these data in dimensionless form and determine a power-law equation relating the dimensionless parameters. ℓ, m h, m H, m 2.0 4.0 2.0 4.0 2.0 4.0 0.10 0.10 0.20 0.20 0.35 0.35 0.833 0.833 0.417 0.417 0.238 0.238 CG h For these tests ℓ = 45 mm, D = 67 mm, and γ = 9.60 kN/m3. (a) Perform a dimensional analysis, and based on the data given, determine if variables used for this problem appear to be correct. Explain how you arrived at your answer. (b) If possible, determine an equation relating the pouring time and viscosity for the cylinder and liquids used in these tests. If it is not possible, indicate what additional information is needed. μ (N ∙ s/m2) 11 17 39 61 107 t(s) 15 23 53 83 145 7.43 In order to maintain uniform flight, smaller birds must beat their wings faster than larger birds. It is suggested that the relationship between the wingbeat frequency, ω, beats per second, and the bird’s wingspan, ℓ, is given by a power law relationship, ω ∼ ℓn. (a) Use dimensional analysis with the assumption that the wingbeat frequency is a function of the wingspan, the specific weight of the bird, γ, the acceleration of gravity, g, and the density of the air, ρa , to determine the value of the exponent n. (b) Some typical data for various birds are given in the following table. Do these data support your result obtained in part (a)? Provide appropriate analysis to show how you arrived at your conclusion. Bird Purple martin Robin Mourning dove Crow Canada goose Great blue heron Wingspan, m Wingbeat frequency, beats/s 0.28 0.36 0.46 1.00 1.50 1.80 5.3 4.3 3.2 2.2 2.6 2.0 1 7.44 A 250-m-long ship has a wetted area of 8000 m2. A 100 -scale model is tested in a towing tank with the prototype fluid, and the results are: Model velocity (m/s) 0.57 1.02 1.40 Model drag (N) 0.50 1.02 1.65. Calculate the prototype drag at 7.5 m/s and 12.0 m/s. 7.45 A student is interested in the aerodynamic drag on spheres. She conducts a series of wind tunnel tests on a 10-cm-diameter sphere. The air in the wind tunnel is at 50 °C and 101.3 kPa. She presents the results of her tests in the form of a correlation, 𝒟 = 0.0015V1.92 , H b C ■ Figure P7.41 7.42 The time, t, it takes to pour a certain volume of liquid from a cylindrical container depends on several factors, including the viscosity of the liquid. (See Video V1.3.) Assume that for very viscous liquids the time it takes to pour out two-thirds of the initial volume depends on the initial liquid depth, ℓ, the cylinder diameter, D, the liquid viscosity, μ, and the liquid specific weight, γ. The data shown in the following table were obtained in the laboratory. where 𝒟 is the drag force in Newtons and V is the velocity in m/s. Another student realizes that the results would have been more effectively presented in terms of dimensionless parameters, say, Π 1 = f ( Π 2 ) = C Π a2 , where Π 1 and Π 2 are dimensionless parameters and C and a are constants. What are the most appropriate dimensionless parameters (Π 1 and Π 2 ) and the corresponding values of C and a? What would be the drag on a 2-cm-diameter sphere placed in a 1-m/s, 20 °C water stream? Section 7.8 Modeling and Similitude 7.46 Air at 80 °F is to flow through a 2-ft pipe at an average velocity of 6 ft/s. What size pipe should be used to move water at Problems Particle 60 °F and average velocity of 3 ft/s if Reynolds number similarity is enforced? 7.47 You are to conduct wind tunnel testing of a new football design that has a smaller lace height than previous designs (see Videos V6.1 and V6.2). It is known that you will need to maintain Re and St similarity for the testing. Based on standard college quarterbacks, the prototype parameters are set at V = 40 mph and ω = 300 rpm, where V and ω are the speed and angular velocity of the football. The prototype football has a 7-in. diameter. Due to instrumentation required to measure pressure and shear stress on the surface of the football, the model will require a length scale of 2:1 (the model will be larger than the prototype). Determine the required model freestream velocity and model angular velocity. 7.48 A model of a submarine, 1:15 scale, is to be tested at 180 ft/s in a wind tunnel with standard sea-level air, while the prototype will be operated in seawater. Determine the speed of the prototype to ensure Reynolds number similarity. 7.49 The drag characteristics of a torpedo are to be studied in a water tunnel using a 1:5 scale model. The tunnel operates with fresh water at 20 °C, whereas the prototype torpedo is to be used in seawater at 15.6 °C. To correctly simulate the behavior of the prototype moving with a velocity of 30 m/s, what velocity is required in the water tunnel? 7.50 For a certain fluid flow problem it is known that both the Froude number and the Weber number are important dimensionless parameters. If the problem is to be studied by using a 1:15 scale model, determine the required surface tension scale if the density scale is equal to 1. The model and prototype operate in the same gravitational field. 7.51 The fluid dynamic characteristics of an airplane flying 240 mph at 10,000 ft are to be investigated with the aid of a 1:20 scale model. If the model tests are to be performed in a wind tunnel using standard air, what is the required air velocity in the wind tunnel? Is this a realistic velocity? 7.52 If an airplane travels at a speed of 1120 km/hr at an altitude of 15 km, what is the required speed at an altitude of 8 km to satisfy Mach number similarity? Assume the air properties correspond to those for the U.S. standard atmosphere. 7.53 (See The Wide World of Fluids article “Modeling Parachutes in a Water Tunnel,” Section 7.8.1.) Flow characteristics for a 30-ftdiameter prototype parachute are to be determined by tests of a 1-ft-diameter model parachute in a water tunnel. Some data collected with the model parachute indicate a drag of 17 lb when the water velocity is 4 ft/s. Use the model data to predict the drag on the prototype parachute falling through air at 10 ft/s. Assume the drag to be a function of the velocity, V, the fluid density, ρ, and the parachute diameter, D. 7.54 When small particles of diameter d are transported by a moving fluid having a velocity V, they settle to the ground at some distance ℓ after starting from a height h as shown in Fig. P7.54. The variation in ℓ with various factors is to be studied with a model 1 having a length scale of 10 . Assume that ℓ = f (h, d, V, γ, μ) where γ is the particle specific weight and μ is the fluid viscosity. The same fluid is to be used in both the model and the prototype, but γ (model) = 9 × γ (prototype). (a) If V = 50 mph, at what velocity should the model tests be run? (b) During a certain model test it was found that ℓ (model) = 0.8 ft. What would be the predicted ℓ for this test? 395 V h ℓ ■ Figure P7.54 7.55 A thin layer of an incompressible fluid flows steadily over a horizontal smooth plate as shown in Fig. P7.55. The fluid surface is open to the atmosphere, and an obstruction having a square cross section is placed on the plate as shown. A model 1 with a length scale of 4 and a fluid density scale of 1.0 is to be designed to predict the depth of fluid, y, along the plate. Assume that inertial, gravitational, surface tension, and viscous effects are all important. What are the required viscosity and surface tension scales? Free surface y V ■ Figure P7.55 7.56 During a storm, a snow drift is formed behind a snow fence as shown in Fig. P7.56. (See Video V9.9.) Assume that the height of the drift, h, is a function of the number of inches of snow deposited by the storm, d, height of the fence, H, width of slats in the fence, b, wind speed, V, acceleration of gravity, g, air density, ρ, and specific weight of snow, γs . (a) If this problem is to be studied with a model, determine the similarity requirements for the model and the relationship between the drift depth for model and prototype (prediction equation). (b) A storm with winds of 30 mph 1 deposits 16 in. of snow having a specific weight of 5.0 lb/ft3. A 2 -sized scale model is to be used to investigate the effectiveness of a proposed snow fence. If the air density is the same for the model and the storm, determine the required specific weight for the model snow and required wind speed for the model. b Fence b Drift V H h H ■ Figure P7.56 7.57 Air bubbles discharge from the end of a submerged tube as shown in Fig. P7.57. The bubble diameter, D, is assumed to be a function of the air flowrate, Q, the tube diameter, d, the acceleration of gravity, g, the density of the liquid, ρ, and the surface tension of the liquid, σ. (a) Determine a suitable set of dimensionless variables for this problem. (b) Model tests are to be run on the Earth for a prototype that is to be operated on a planet where the acceleration of gravity is 10 times greater than that on Earth. The model and prototype are to use the same fluid, 396 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling and the prototype tube diameter is 0.25 in. Determine the tube diameter for the model and the required model flowrate if the prototype flowrate is to be 0.001 ft3/s. the prototype blast with energy release E and a model blast with 1/1000th the energy release (Em = 0.001 E). At what distance from the model blast will the pressure rise be the same as that at a distance of 1 mile from the prototype blast? 7.62 An incompressible fluid oscillates harmonically (V = V0 sin ωt, where V is the velocity) with a frequency of 10 rad/s in a 1 4-in.-diameter pipe. A 4 scale model is to be used to determine the pressure difference per unit length, Δpℓ (at any instant) along the pipe. Assume that Q d Δpℓ = f (D, V0, ω, t, μ, ρ)     D where D is the pipe diameter, ω the frequency, t the time, μ the fluid viscosity, and ρ the fluid density. (a) Determine the similarity requirements for the model and the prediction equation for Δpℓ . (b) If the same fluid is used in the model and the prototype, at what frequency should the model operate? ■ Figure P7.57 7.58 For a certain model study involving a 1:5 scale model it is known that Froude number similarity must be maintained. The possibility of cavitation is also to be investigated, and it is assumed that the cavitation number must be the same for model and prototype. The prototype fluid is water at 30 °C, and the model fluid is water at 70 °C. If the prototype operates at an ambient pressure of 101 kPa (abs), what is the required ambient pressure for the model system? 7.59 A model hydrofoil is to be tested. Is it practical to satisfy both the Reynolds number and the Froude number for the hydrofoil when it is operating near the water surface? Support your decision. 7.60 A thin layer of particles rests on the bottom of a horizontal tube as shown in Fig. P7.60. When an incompressible fluid flows through the tube, it is observed that at some critical velocity the particles will rise and be transported along the tube. A model is to be used to determine this critical velocity. Assume the critical velocity, Vc, to be a function of the pipe diameter, D, particle diameter, d, the fluid density, ρ, and viscosity, μ, the density of the particles, ρp, and the acceleration of gravity, g. (a) Determine the similarity requirements for the model, and the relationship between the critical velocity for model and prototype (the prediction equa1 tion). (b) For a length scale of 2 and a fluid density scale of 1.0, what will be the critical velocity scale (assuming all similarity requirements are satisfied)? Vc ■ Figure P7.60 7.61 The pressure rise, Δp, across a blast wave, as shown in Fig. P7.61, and Video V11.8, is assumed to be a function of the amount of energy released in the explosion, E, the air density, ρ, the speed of sound, c, and the distance from the blast, d. (a) Put this relationship in dimensionless form. (b) Consider two blasts: Air (ρ , c) Δp = p2 – p1 (2) (1) d ■ Figure P7.61 7.63 As shown in Fig. P7.63, a “noisemaker” B is towed behind a minesweeper A to set off enemy acoustic mines such as at C. The drag force of the noisemaker is to be studied in a water tunnel at a 1 4 scale model (model ¼ the size of the prototype). The drag force is assumed to be a function of the speed of the ship, the density and viscosity of the fluid, and the diameter of the noisemaker. (a) If the prototype towing speed is 3 m/s, determine the water velocity in the tunnel for the model tests. (b) If the model tests of part (a) produced a model drag of 900 N, determine the drag expected on the prototype. A B C ■ Figure P7.63 7.64 The drag characteristics for a newly designed automobile having a maximum characteristic length of 20 ft are to be determined through a model study. The characteristics at both low speed (approximately 20 mph) and high speed (90 mph) are of interest. For a series of projected model tests, an unpressurized wind tunnel that will accommodate a model with a maximum characteristic length of 4 ft is to be used. Determine the range of air velocities that would be required for the wind tunnel if Reynolds number similarity is desired. Are the velocities suitable? Explain. 7.65 The drag characteristics of an airplane are to be determined by model tests in a wind tunnel operated at an absolute pressure of 1300 kPa. If the prototype is to cruise in standard air at 385 km/hr, and the corresponding speed of the model is not to differ by more than 20% from this (so that compressibility effects may be ignored), what range of length scales may be used if Reynolds number similarity is to be maintained? Assume the viscosity of air is unaffected by pressure, and the temperature of air in the tunnel is equal to the temperature of the air in which the airplane will fly. 7.66 The drag on a sphere moving in a fluid is known to be a function of the sphere diameter, the velocity, and the fluid viscosity and density. Laboratory tests on a 4-in.-diameter sphere were performed in a water tunnel and some model data are plotted in Fig. P7.66. Problems For these tests the viscosity of the water was 2.3 × 10−5 lb ∙ s/ft2 and the water density was 1.94 slugs/ft3 . Estimate the drag on an 8-ft-diameter balloon moving in air at a velocity of 3 ft/s. Assume the air to have a viscosity of 3.7 × 10−7 lb ∙ s/ft2 and a density of 2.38 × 10−3 slugs/ft3. 397 D = 0.05 m Infant formula ρ = 1036.6 kg/m3 μ = 1.7 × 10−3 N·s/m2 Q = 16.8 m3/hr at average formula temperature 1.4 m Model drag, lb 6 4 30% glycol solution ■ Figure P7.71 2 0 0 2 4 6 8 10 12 Model velocity, ft/s Section 7.9 Some Typical Model Studies 7.72 The pressure rise, Δp, across a centrifugal pump of a given shape can be expressed as ■ Figure P7.66 Δp = f (D, ω, ρ, Q)     7.68 A flagpole has a diameter of 1.0 ft and is 45 ft. long. The manufacturer wants to find the bending moment at the base of the pole for a wind speed of 60 mph. To do so, a scale model of the flagpole is built with a diameter of 1.0 in. and a length of 45 in. (a) What must be the air speed at which the model is tested? (b) If the wind velocity over the flagpole in part (a) is reduced by 50%, how much would the bending moment at the base be reduced? 7.69 A very small needle valve is used to control the flow of air 1 in a 8 -in. air line. The valve has a pressure drop of 4.0 psi at a flow rate of 0.005 ft3/s of 60 °F air. Tests are performed on a large, geometrically similar valve and the results are used to predict the performance of the smaller valve. How many times larger can the model valve be if 60 °F water is used in the test and the water flow rate is limited to 7.0 gal/min? 7.70 In low-speed external flow over a bluff object, vortices are shed from the object (as shown in Fig. P7.70). The frequency of vortex shedding, f ≐ 1/T, depends on ρ, μ, V, and d. Make a dimensional analysis of this phenomenon. Transform the dimensionless parameters that you developed into “standard” parameters. What are their names? Two identically shaped objects with a size ratio 2:1 are tested in air flow. What velocity ratio is necessary for dynamic similarity? What is the ratio of vortex-shedding frequencies? V d ■ Figure P7.70 7.71 In a nutritional products plant, a plate heat exchanger cools infant formula by approximately 25 °C from an inlet temperature of 68 °C as it flows through the plates shown in Fig. P7.71. The process engineer would like to be able to cool the infant formula by as much as 35 °C as it passes through the plates. The plate heat 1 exchanger is currently in use, so he would like to build a 2 -scale model for testing purposes. In the model, water will be used instead of infant formula. What flow rate is needed in the model? 8 Model data ωm = 40π rad/s Dm = 8 in. 6 Δ pm (psi) 7.67 A dam spillway is 40 ft long and has fluid velocity of 10 ft/s. Considering Weber number effects as minor, calculate the corresponding model fluid velocity for a model length of 5 ft. where D is the impeller diameter, ω the angular velocity of the impeller, ρ the fluid density, and Q the volume rate of flow through the pump. A model pump having a diameter of 8 in. is tested in the laboratory using water. When operated at an angular velocity of 40π rad/s the model pressure rise as a function of Q is shown in Fig. P8.72. Use this curve to predict the pressure rise across a geometrically similar pump (prototype) for a prototype flowrate of 6 ft3/s. The prototype has a diameter of 12 in. and operates at an angular velocity of 60π rad/s. The prototype fluid is also water. 4 2 0 0 0.5 1.0 1.5 2.0 Qm (ft3/s) ■ Figure P7.72 7.73 At a large fish hatchery the fish are reared in open, water-filled tanks. Each tank is approximately square in shape with curved corners, and the walls are smooth. To create motion in the tanks, water is supplied through a pipe at the edge of the tank. The water is drained from the tank through an opening at the center. (See Video V7.9.) A model with a length scale of 1:13 is to be used to determine the velocity, V, at various locations within the tank. Assume that V = f (ℓ, ℓi , ρ, μ, g, Q) where ℓ is some characteristic length such as the tank width, ℓi represents a series of other pertinent lengths, such as inlet pipe diameter, fluid depth, etc., ρ is the fluid density, μ is the fluid viscosity, g is the acceleration of gravity, and Q is the discharge through the tank. (a) Determine a suitable set of dimensionless parameters for this problem and the 398 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling prediction equation for the velocity. If water is to be used for the model, can all of the similarity requirements be satisfied? Explain and support your answer with the necessary calculations. (b) If the flowrate into the full-sized tank is 250 gpm, determine the required value for the model discharge assuming Froude number similarity. What model depth will correspond to a depth of 32 in. in the fullsized tank? 7.74 (See The Wide World of Fluids article titled “Galloping Gertie,” Section 7.8.2.) The Tacoma Narrows Bridge failure is a dramatic example of the possible serious effects of wind-induced vibrations. As a fluid flows around a body, vortices may be created that are shed periodically, creating an oscillating force on the body. If the frequency of the shedding vortices coincides with the natural frequency of the body, large displacements of the body can be induced as was the case with the Tacoma Narrows Bridge. To illustrate this type of phenomenon, consider fluid flow past a circular cylinder. (See Video V7.5.) Assume the frequency, n, of the shedding vortices behind the cylinder is a function of the cylinder diameter, D, the fluid velocity, V, and the fluid kinematic viscosity, v. (a) Determine a suitable set of dimensionless variables for this problem. One of the dimensionless variables should be the Strouhal number, nD/V. (b) Some results of experiments in which the shedding frequency of the vortices (in Hz) was measured, using a particular cylinder and Newtonian, incompressible fluid, are shown in Fig. P7.74. Is this a “universal curve” that can be used to predict the shedding frequency for any cylinder placed in any fluid? Explain. (c) A certain structural component in the form of a 1-in.-diameter, 12-ft-long rod acts as a cantilever beam with a natural frequency of 19 Hz. Based on the data in Fig. P7.72, estimate the wind speed that may cause the rod to oscillate at its natural frequency. Hint: Use a trial-and-error solution. 7.76 As illustrated in Video V7.9, models are commonly used to study the dispersion of a gaseous pollutant from an exhaust stack located near a building complex. Similarity requirements for the pollutant source involve the following independent variables: the stack gas speed, V, the wind speed, U, the density of the atmospheric air, ρ, the difference in densities between the air and the stack gas, ρ − ρs, the acceleration of gravity, g, the kinematic viscosity of the stack gas, vs, and the stack diameter, D. (a) Based on these variables, determine a suitable set of similarity requirements for modeling the pollutant source. (b) For this type of model a typical length scale might be 1:200. If the same fluids were used in model and prototype, would the similarity requirements be satisfied? Explain and support your answer with the necessary calculations. 7.77 As winds blow past buildings, complex flow patterns can develop due to various factors such as flow separation and interactions between adjacent buildings. (See Video V7.13.) Assume that the local gage pressure, p, at a particular location on a building is a function of the air density, ρ, the wind speed, V, some characteristic length, ℓ, and all other pertinent lengths, ℓi , needed to characterize the geometry of the building or building complex. (a) Determine a suitable set of dimensionless parameters that can be used to study the pressure distribution. (b) An eight-story building that is 100 ft tall is to be modeled in a wind tunnel. If a length scale of 1:300 is to be used, how tall should the model building be? (c) How will a measured pressure in the model be related to the corresponding prototype pressure? Assume the same air density in model and prototype. Based on the assumed variables, does the model wind speed have to be equal to the prototype wind speed? Explain. 7.78 Assume that the pump performance characteristics shown in Fig. P7.78 are for a pump running at 1200 rpm. Assuming that the pump head, hp, the pump impeller diameter, D, and the flowrate, Q, through the pump are the other relevant parameters, calculate the pump performance characteristics at speeds of 1800 rpm and 2400 rpm. Neglect viscous effects. 0.22 0.20 0.18 T = 55°F 200 0.16 150 Head (ft) St = nD/ V thickness and velocity would be required if the length scale is to be 110? (c) If the model and prototype ice have the same density, can the model ice have the same strength properties as that of the prototype ice? Explain. 0.14 hp 100 50 0.12 10 100 Re = VD/v 1,000 10,000 ■ Figure P7.74 0 0 25 50 75 100 125 150 Volume flowrate, Q (gal/min) ■ Figure P7.78 7.75 (See The Wide World of Fluids article titled “Ice Engineering,” Section 7.9.3.) A model study is to be developed to determine the force exerted on bridge piers due to floating chunks of ice in a river. The piers of interest have square cross sections. Assume that the force, R, is a function of the pier width, b, the thickness of the ice, d, the velocity of the ice, V, the acceleration of gravity, g, the density of the ice, ρi, and a measure of the strength of the ice, Ei , where Ei has the dimensions FL−2. (a) Based on these variables determine a suitable set of dimensionless variables for this problem. (b) The prototype conditions of interest include an ice thickness of 12 in. and an ice velocity of 6 ft/s. What model ice 7.79 A stream of atmospheric air is used to keep a ping-pong ball aloft by blowing the air upward over the ball. The ping-pong ball has a mass of 2.5 g and a diameter D1 = 3.8 cm, and the air stream has an upward velocity of V1 = 0.942 m/s. This system is to be modeled by pumping water upward with a velocity V2 over a solid ball of diameter D2 and density ρb2 = 2710 kg/m3 . In both cases, the net weight of the ball Wb is equal to the air drag, Wb = CDρAV2 , 2 Problems where CD = 0.60, ρ is the fluid density, A the ball’s projected area, and V the velocity of the fluid upstream from the ball. Determine all possible combinations of V2 and D2. [Hint: A force balance involving the drag on the ball, the buoyant force on the ball, and the weight of the ball is needed.] 399 x = 0 and dydx = 0 at x = 0. (a) Rewrite the equation and boundary conditions in dimensionless form using the beam length, ℓ, as the reference length. (b) Based on the results of part (a), what are the similarity requirements and the prediction equation for a model to predict deflections? 1 7.80 A 10 -scale model of an airplane is tested in a wind tunnel at 70°F and 14.40 psia. The model test results are: Velocity (mph) 0 50 100 150 200 Drag (lb) 0 5 21 46 85. Find the corresponding airplane velocities and drags if only fluid compressibility is important and the airplane is flying in the U.S. Standard Atmosphere at 30,000 ft. Assume that the air is an ideal gas. 7.81 A company manufactures geometrically similar airplane propellers up to 4.0 m in diameter. Wind tunnel tests are run on geometrically similar propellers up to 0.33 m in diameter. In the test of a 0.33-m model of a 4.0-m propeller, the air velocity was 50 m/s. The model rotational speed was 2000 rpm, the thrust 100 N, and the propeller input torque 20 N ∙ m. Calculate the corresponding prototype rotational speed, thrust, and input torque for an airplane speed of 100 m/s. The Reynolds numbers and Mach numbers are of minor importance. 7.82 A breakwater is a wall built around a harbor so the incoming waves dissipate their energy against it. The significant dimensionless parameters are the Froude number Fr and the Reynolds number Re. A particular breakwater measuring 450 m long and 20 m deep 1 is hit by waves 5 m high and velocities up to 30 m/s. In a 100 -scale model of the breakwater, the wave height and velocity can be controlled. Can complete similarity be obtained using water for the model test? y P x ℓ ■ Figure P7.84 7.85 A liquid is contained in a pipe that is closed at one end as shown in Fig. P7.85. Initially the liquid is at rest, but if the end is suddenly opened the liquid starts to move. Assume the pressure p1 remains constant. The differential equation that describes the resulting motion of the liquid is ρ ∂υz ∂t = ∂ 2υz l ∂υz p1 + μ( 2 + ℓ r ∂r ) ∂r where υz is the velocity at any radial location, r, and t is time. Rewrite this equation in dimensionless form using the liquid density, ρ, the viscosity, μ, and the pipe radius, R, as reference parameters. p1 r End initially closed vz z R ℓ Section 7.10 Similitude Based on Governing Differential Equations ■ Figure P7.85 7.83 A viscous fluid is contained between wide, parallel plates spaced a distance h apart as shown in Fig. P7.83. The upper plate is fixed, and the bottom plate oscillates harmonically with a velocity amplitude U and frequency ω. The differential equation for the velocity distribution between the plates is 7.86 A student drops two spherical balls of different diameters and different densities. She has a stroboscopic photograph showing the positions of each ball as a function of time. However, she wants to express the velocity of each as a function of time in dimensionless form. Develop the dimensionless group. The equation of motion for each ball is ρ ∂u ∂ 2u =μ 2 ∂t ∂y where u is the velocity, t is time, and ρ and μ are fluid density and viscosity, respectively. Rewrite this equation in a suitable nondimensional form using h, U, and ω as reference parameters. Fixed plate y h u mg − where m is ball mass, g is acceleration of gravity, CD is a dimensionless and constant drag coefficient, ρ is air mass density, A is ball cross-sectional area (= πD2/4) with D ball diameter, V is ball velocity, and t is time. 7.87 The basic equation that describes the motion of the fluid above a large oscillating flat plate is ∂ 2u ∂u = v 2, ∂t ∂y x u = Ucos ω t ■ Figure P7.83 7.84 The deflection of the cantilever beam of Fig. P7.84 is governed by the differential equation. EI d 2y 2 dx CD dV ρAV 2 = m , 2 dt = P(x − ℓ) where E is the modulus of elasticity and I is the moment of inertia of the beam cross section. The boundary conditions are y = 0 at where u is the fluid velocity component parallel to the plate, t is time, y is the spatial coordinate perpendicular to the plate, and v is the fluid kinematic viscosity. The plate oscillating velocity is given by U = U0 sin ωt. Find appropriate dimensionless parameters and the dimensionless differential equation. 7.88 The dimensionless parameters for a ball released and falling from rest in a fluid are CD,   ρ gt2 Vt ,   ,  and  , ρb D D 400 Chapter 7 ■ Dimensional Analysis, Similitude, and Modeling where CD is a drag coefficient (assumed to be constant at 0.4), g is the acceleration of gravity, D is the ball diameter, t is the time after it is released, ρ is the density of the fluid in which it is dropped, and ρb is the density of the ball. Ball 1, an aluminum ball (ρb1 = 2710 kg/m3 ) having a diameter D1 = 1.0 cm, is dropped in water (ρ = 1000 kg/m3). The ball velocity V1 is 0.733 m/s at t1 = 0.10 s. Find the corresponding velocity V2 and time t2 for ball 2 having D2 = 2.0 cm and ρb2 = ρb1 . Next, use the computed value of V2 and the equation of motion, CD dV ρAV2 = ρb V , ρbVg − 2 dt where V is the volume of the ball and A is its cross-sectional area, to verify the value of t2. Should the two values of t2 agree? ■ Lifelong Learning Problems 7.1 LL Microfluidics is the study of fluid flow in fabricated devices at the micro scale. Advances in microfluidics have enhanced the ability of scientists and engineers to perform laboratory experiments using miniaturized devices known as a “lab-ona-chip.” Obtain information about a lab-on-a-chip device that is available commercially and investigate its capabilities. Summarize your findings in a brief report. 7.2 LL For some types of aerodynamic wind tunnel testing, it is difficult to simultaneously match both the Reynolds number and Mach number between model and prototype. Engineers have developed several potential solutions to the problem including pressurized wind tunnels and lowering the temperature of the flow. Obtain information about cryogenic wind tunnels and explain the advantages and disadvantages. Summarize your findings in a brief report. 7.3 LL As mentioned in the text, there is a large number of common, named, dimensionless groups. Do some research and find at least 20 common dimensionless groups that have not been considered in this chapter. Report the group name, group definition, physical significance of the group, and whether the group is similar to another group or can be formed by combining other groups. 8 Viscous Flow in Pipes Learning Objectives VIDEO V8.1 Turbulent jet Pipe flow is very important in our daily operations. After completing this chapter, you should be able to: ■ identify and explain various characteristics of the flow in pipes. ■ discuss the main properties of laminar and turbulent pipe flow and appreciate their differences. ■ calculate losses in straight portions of pipes as well as those in various pipe system components. ■ apply appropriate equations and principles to analyze a variety of pipe flow situations. ■ predict the flowrate in a pipe by use of common flowmeters. In the previous chapters we have considered a variety of topics concerning the motion of fluids. The basic governing principles concerning mass, momentum, and energy were developed and applied, in conjunction with rather severe assumptions, to numerous flow situations. In this chapter we will apply the basic principles to a specific, important topic—the incompressible flow of viscous fluids in pipes and ducts. The transport of a fluid (liquid or gas) in a closed conduit (commonly called a pipe if it is of round cross section or a duct if it is not round) is extremely important in our daily operations. A brief consideration of the world around us will indicate that there is a wide variety of applications of pipe flow. Such applications range from the large, man-made Alaskan pipeline that carries crude oil almost 800 miles across Alaska, to the more complex (and certainly not less useful) natural systems of “pipes” that carry blood throughout our body and air into and out of our lungs. Other examples include the water pipes in our homes and the distribution system that delivers the water from the city well to the house. Numerous hoses and pipes carry hydraulic fluid or other fluids to various components of vehicles and machines. The air quality within our buildings is maintained at comfortable levels by the distribution of conditioned (heated, cooled, humidifieddehumidified) air through a maze of pipes and ducts. Although all of these systems are different, the fluid mechanics principles governing the fluid motions are common. The purpose of this chapter is to understand the basic processes involved in such flows. Some of the basic components of a typical pipe system are shown in Fig. 8.1. They include the pipes themselves (perhaps of more than one diameter), the various fittings used to connect the individual pipes to form the desired system, the flowrate control devices (valves), and the pumps or turbines that add energy to or remove energy from the fluid. Even the most simple pipe systems are actually quite complex when they are viewed in terms of rigorous analytical considerations. We will use an “exact” analysis of the simplest pipe flow topics (such as laminar flow in long, straight, constant diameter pipes) and dimensional analysis considerations combined with experimental 401 402 Chapter 8 ■ Viscous Flow in Pipes Outlet Pipe Pump Elbow Tee Valve Inlet ■ Figure 8.1 Typical pipe system components. results for the other pipe flow topics. Such an approach is not unusual in fluid mechanics investigations. When “real-world” effects are important (such as viscous effects in pipe flows), it is often difficult or “impossible” to use only theoretical methods to obtain the desired results. A judicious combination of experimental data with theoretical considerations and dimensional analysis often provides the desired results. The flow in pipes discussed in this chapter is an example of such an analysis. 8.1 The pipe is assumed to be completely full of the flowing fluid. General Characteristics of Pipe Flow Before we apply the various governing equations to pipe flow examples, we will discuss some of the basic concepts of pipe flow. With these ground rules established, we can then proceed to formulate and solve various important flow problems. Although not all conduits used to transport fluid from one location to another are round in cross section, most of the common ones are. These include typical water pipes, hydraulic hoses, and other conduits that are designed to withstand a considerable pressure difference across their walls without undue distortion of their shape. Typical conduits of noncircular cross section include heating and air conditioning ducts that are often of rectangular cross section. Normally the pressure difference between the inside and outside of these ducts is relatively small. Most of the basic principles involved are independent of the cross-sectional shape, although the details of the flow may be dependent on it. Unless otherwise specified, we will assume that the conduit is round, although we will show how to account for other shapes. For all flows involved in this chapter, we assume that the pipe is completely filled with the fluid being transported as is shown in Fig. 8.2a. Thus, we will not consider a concrete pipe through which rainwater flows without completely filling the pipe, as is shown in Fig.8.2b. Such flows, called open-channel flow, are treated in Chapter 10. The difference between open-channel flow and the pipe flow of this chapter is in the fundamental mechanism that drives the flow. For open-channel p2 ≠ p1 (1) (1) Q Q (2) p1 = p2 (2) (a) ■ Figure 8.2 (a) Pipe flow. (b) Open-channel flow. (b) 8.1 General Characteristics of Pipe Flow 403 Turbulen Turbulent Dye Pipe D Q = VA Dye streak Transitional Smooth, well-rounded entrance Laminar (b) (a) ■ Figure 8.3 (a) Experiment to illustrate type of flow. (b) Typical dye streaks. flow, gravity alone is the driving force—the water flows down a hill. For pipe flow, gravity may be important (the pipe need not be horizontal), but the main driving force is likely to be a pressure gradient along the pipe. If the pipe is not full, it is not possible to maintain this pressure difference, p1 − p2 . 8.1.1 Laminar or Turbulent Flow VIDEO V8.2 Laminar/ turbulent pipe flow A flow may be laminar, transitional, or turbulent. The flow of a fluid in a pipe may be laminar flow or it may be turbulent flow. Osborne Reynolds (1842–1912), a British scientist and mathematician, was the first to distinguish the difference between these two classifications of flow by using a simple apparatus as shown by the figure in the margin, which is a sketch of Reynolds’ dye experiment. Reynolds injected dye into a pipe in which water flowed due to gravity. The entrance region of the pipe is depicted in Fig. 8.3a. If water runs through a pipe of diameter D with an average velocity V, the following characteristics are observed by injecting neutrally buoyant dye as shown. For “small enough flowrates” the dye streak (a streakline) will remain as a well-defined line as it flows along, with only slight blurring due to molecular diffusion of the dye into the surrounding water. For a somewhat larger “intermediate flowrate” the dye streak fluctuates in time and space, and intermittent bursts of irregular behavior appear along the streak. On the other hand, for “large enough flowrates” the dye streak almost immediately becomes blurred and spreads across the entire pipe in a random fashion. These three characteristics, denoted as laminar, transitional, and turbulent flow, respectively, are illustrated in Fig. 8.3b. The curves shown in Fig. 8.4 represent the x component of the velocity as a function of time at a point A in the flow. The random fluctuations of the turbulent flow (with the associated particle mixing) are what disperse the dye throughout the pipe and cause the blurred appearance illustrated in Fig. 8.3b. For laminar flow in a pipe there is only one component of velocity, V = uiˆ. For turbulent flow the predominant component of velocity is also along the pipe, but it is unsteady uA Turbulent Q A x Transitional Laminar t ■ Figure 8.4 Time dependence of fluid velocity at a point. 404 Chapter 8 ■ Viscous Flow in Pipes Pipe flow characteristics are dependent on the value of the Reynolds number. VIDEO V8.3 Intermittent turbulent burst in pipe flow (random) and accompanied by random components normal to the pipe axis, V = uiˆ + υjˆ + wkˆ . Such motion in a typical flow occurs too fast for our eyes to follow. Slow-motion pictures of the flow can more clearly reveal the irregular, seemingly random, turbulent nature of the flow. As was discussed in Chapter 7, we should not label dimensional quantities as being “large” or “small,” such as “small enough flowrates” in the preceding paragraphs. Rather, the appropriate dimensionless quantity should be identified and the “small” or “large” character attached to it. A quantity is “large” or “small” only relative to a reference quantity. The ratio of those quantities results in a dimensionless quantity. For pipe flow the most important dimensionless parameter is the Reynolds number, Re—the ratio of the inertia to viscous effects in the flow. Hence, in the previous paragraph the term flowrate should be replaced by Reynolds number, Re = ρVDμ, where V is the average velocity in the pipe. That is, the flow in a pipe is laminar, transitional, or turbulent provided the Reynolds number is “small enough,” “intermediate,” or “large enough.” It is not only the fluid velocity that determines the character of the flow—its density, viscosity, and the pipe size are of equal importance. These parameters combine to produce the Reynolds number. The distinction between laminar and turbulent pipe flow and its dependence on an appropriate dimensionless quantity was first pointed out by Osborne Reynolds in 1883. The Reynolds number ranges for which laminar, transitional, or turbulent pipe flows are obtained cannot be precisely given. The actual transition from laminar to turbulent flow may take place at various Reynolds numbers, depending on how much the flow is disturbed by vibrations of the pipe, roughness of the entrance region, and the like. For general engineering purposes (i.e., without undue precautions to eliminate such disturbances), the following values are appropriate: The flow in a round pipe is laminar if the Reynolds number is less than approximately 2100. The flow in a round pipe is turbulent if the Reynolds number is greater than approximately 4000. For Reynolds numbers between these two limits, the flow may switch between laminar and turbulent conditions in an apparently random fashion (transitional flow). THE WIDE WORLD OF FLUIDS Nanoscale flows The term nanoscale generally refers to flows (for example, the fact that a fluid sticks to a solid bound- objects with characteristic lengths from atomic dimensions up ary) may not be valid for nanoscale flows. Also, ultratiny to a few hundred nanometers (nm). (Recall that 1 nm = 10−9 m.) mechanical pumps and valves are difficult to manufacture and Nanoscale fluid mechanics research has recently uncovered may become clogged by tiny particles such as biological mole- many surprising and useful phenomena. No doubt many more cules. As a possible solution to such problems, researchers remain to be discovered. For example, in the future researchers have investigated the possibility of using a system that does envision using nanoscale tubes to push tiny amounts of water- not rely on mechanical parts. It involves using light-sensitive soluble drugs to exactly where they are needed in the human molecules attached to the surface of the tubes. By shining light body. Because of the tiny diameters involved, the Reynolds onto the molecules, the light-responsive molecules attract numbers for such flows are extremely small and the flow is defi- water and cause motion of water through the tube. (See nitely laminar. In addition, some standard properties of everyday Problem 8.8.) EXAMPLE 8.1 Laminar or Turbulent Flow GIVEN Water at a temperature of 50 °F flows through a pipe of diameter D = 0.73 in. and into a glass as shown in Fig. E8.1a. V D μ, ρ Q FIND Determine (a) the minimum time taken to fill a 12-oz glass (volume = 0.0125 ft3) with water if the flow in the pipe is to be laminar. Repeat the calculations if the water temperature is 140 °F. (b) the maximum time taken to fill the glass if the flow is to be turbulent. Repeat the calculations if the water temperature is 140 °F. ■ Figure E8.1a 8.1 General Characteristics of Pipe Flow 405 SOLUTION (a) If the flow in the pipe is to remain laminar, the minimum time to fill the glass will occur if the Reynolds number is the maximum allowed for laminar flow, typically Re = ρVDμ = 2100. Thus, V = 2100 μρD, where from Table B.1, ρ = 1.94 slugs /ft3 and μ = 2.73 × 10−5 lb ∙ s /ft2 at 50 °F, while ρ = 1.91 slugs /ft3 and μ = 0.974 × 10−5 lb ∙ s /ft2 at 140 °F. Thus, the maximum average velocity for laminar flow in the pipe is         2100μ 2100(2.73 × 10−5 lb ∙ s /ft2 ) = ρD (1.94 slugs /ft3 ) (0.7312 ft) = 0.486 lb ∙ s /slug = 0.486 ft /s V=       Similarly, V = 0.176 ft /s at 140 °F. With V = volume of glass and V = Qt we obtain common fluids (water, gasoline, air). By repeating the calculations at various water temperatures, T (i.e., with different densities and viscosities), the results shown in Fig. E8.1b are obtained. As the water temperature increases, the kinematic viscosity, ν = μ/ρ, decreases and the corresponding times to fill the glass increase as indicated. (Temperature effects on the viscosity of gases are the opposite; increase in temperature causes an increase in viscosity.) If the flowing fluid had been honey with a kinematic viscosity (ν = μ/ρ) 3000 times greater than that of water, the velocities given earlier would be increased by a factor of 3000 and the times reduced by the same factor. As shown in the following sections, the pressure needed to force a very viscous fluid through a pipe at such a high velocity may be unreasonably large. 4(0.0125 ft3 ) V V = = 2 Q (π4)D V (π[0.7312]2ft2 ) (0.486 ft /s) = 8.85 s at T = 50 °F (Ans) t= Similarly, t = 24.4 s at 140 °F. To maintain laminar flow, the less viscous hot water requires a lower flowrate than the cold water. 40 (b) If the flow in the pipe is to be turbulent, the maximum time to fill the glass will occur if the Reynolds number is the minimum allowed for turbulent flow, Re = 4000. Thus, V = 4000μ ρD = 0.925 ft /s and 30 t = 4.65 s at 50 °F t, s 20 Note that because water is “not very viscous,” the velocity must be “fairly small” to maintain laminar flow. In general, turbulent flows are encountered more often than laminar flows because of the relatively small viscosity of most turbulent flow (50˚F, 8.85 s) (Ans) Similarly, V = 0.335 ft /s and t = 12.8 s at 140 °F. COMMENTS laminar flow 10 0 (50˚F, 4.65 s) 0 50 100 150 200 250 T,˚F ■ Figure E8.1b 8.1.2 Entrance Region and Fully Developed Flow Any fluid flowing in a pipe had to enter the pipe at some location. The region of flow near where the fluid enters the pipe is termed the entrance region and is illustrated in Fig. 8.5. It may be the first few feet of a pipe connected to a tank or the initial portion of a long run of a hot air duct coming from a furnace. As is shown in Fig. 8.5, the fluid typically enters the pipe with a nearly uniform velocity profile at section (1). As the fluid moves through the pipe, viscous effects cause it to stick to the pipe wall (the no-slip boundary condition). This is true whether the fluid is relatively inviscid air or a very viscous oil. Thus, a boundary layer in which viscous effects are important is produced along the pipe wall such that the initial velocity profile changes with distance along the pipe, x. (See also Fig. 6.14b.) The boundary layer grows in thickness as the fluid proceeds downstream. For fluid outside the boundary layer [within the inviscid core surrounding the centerline from (1) to 2)], viscous effects are negligible. Since the fluid velocity within the boundary layer is reduced, the inviscid core fluid must accelerate as it moves downstream to maintain constant mass flow at all cross sections. Once the fluid reaches the end of the entrance length, section (2), the velocity profile does not vary with x. The boundary layer has grown in thickness to completely fill the pipe. 406 Chapter 8 ■ Viscous Flow in Pipes Fully developed flow Entrance region flow D Boundary layer Inviscid core r (1) x (2) (3) ℓe (6) (5) (4) x6 – x5 Fully developed flow x5 – x4 Developing flow ■ Figure 8.5 Entrance region, developing flow, and fully developed flow in a pipe system. The shape of the velocity profile in the pipe depends on whether the flow is laminar or turbulent, as does the length of the entrance region, ℓe. As with many other properties of pipe flow, the dimensionless entrance length, ℓe D, correlates quite well with the Reynolds number. Analytic and experimental investigations have demonstrated that typical entrance lengths are given by The entrance length is a function of the Reynolds number. 100 ℓe D 10 1 0.1 100 102 104 Re 106 ℓe = 0.06 Re for laminar flow D (8.1) ℓe = 4.4 (Re) 16 for turbulent flow D (8.2) and For very low Reynolds number flows the entrance length can be quite short (ℓe = 0.6D if Re = 10), whereas for large Reynolds number flows it may take a length equal to many pipe diameters before the end of the entrance region is reached (ℓe = 120D for Re = 2000). For many practical engineering problems, 104 < Re < 105 so that as shown by the figure in the margin, 20D < ℓe < 30D. Calculation of the velocity profile and pressure distribution within the entrance region is quite complex. However, once the fluid reaches the end of the entrance region, section (2) of Fig. 8.5, the flow is simpler to describe because the velocity is a function of only the distance from the pipe centerline, r, and independent of x. This is true until the character of the pipe changes in some way, such as a change in diameter, or the fluid flows through a bend, valve, or some other component at section (3). The flow between (2) and (3) is termed fully developed flow. Beyond the interruption of the fully developed flow [at section (4)], the flow gradually begins its return to its fully developed character [section (5)] and continues with this profile until the next pipe system component is reached [section (6)]. In many cases the pipe is long enough so that there is a considerable length of fully developed flow compared with the developing flow length [(x3 − x2 ) ⪢ ℓe and (x6 − x5 ) ⪢ (x5 − x4 )]. In other cases the distance between one component (bend, tee, valve, etc.) of the pipe system and the next component is so short that fully developed flow is never achieved. 8.1.3 Pressure and Shear Stress Fully developed steady flow in a constant diameter pipe may be driven by gravity and /or pressure forces. For horizontal pipe flow, gravity has no effect except for a hydrostatic pressure variation across the pipe, γD, that is usually negligible. It is the pressure difference, Δp = p1 − p2, between one section of the horizontal pipe and another which forces the fluid through the pipe. Viscous effects provide the restraining force that exactly balances the pressure force, thereby allowing the fluid to flow through the pipe with no acceleration. If viscous effects were absent in such flows, the pressure would be constant throughout the pipe, except for the hydrostatic variation. 8.2 p Fully Developed Laminar Flow 407 Fully developed flow: ∂p/ ∂ x = constant Entrance flow Entrance pressure drop Δp x3 – x2 = ℓ x1 = 0 x2 = ℓe x3 x ■ Figure 8.6 Pressure distribution along a horizontal pipe. Laminar flow characteristics are different than those for turbulent flow. 8.2 In non-fully developed flow regions, such as the entrance region of a pipe, the fluid accelerates or decelerates as it flows (the velocity profile changes from a uniform profile at the entrance of the pipe to its fully developed profile at the end of the entrance region). Thus, in the entrance region there is a balance between pressure, viscous, and inertia (acceleration) forces. The result is a pressure distribution along the horizontal pipe as shown in Fig. 8.6. The magnitude of the pressure gradient, ∂p∂x, is larger in the entrance region than in the fully developed region, where it is a constant, ∂p∂x = −Δpℓ < 0. The fact that there is a nonzero pressure gradient along the horizontal pipe is a result of viscous effects. As is discussed in Chapter 3, if the viscosity were zero, the pressure would not vary with x. The need for the pressure drop can be viewed from two different standpoints. In terms of a force balance, the pressure force is needed to overcome the viscous forces generated. In terms of an energy balance, the work done by the pressure force is needed to overcome the viscous dissipation of energy throughout the fluid. If the pipe is not horizontal, the pressure gradient along it is due in part to the component of weight in that direction. As is discussed in Section 8.2.1, this contribution due to the weight either enhances or retards the flow, depending on whether the flow is downhill or uphill. The nature of the pipe flow is strongly dependent on whether the flow is laminar or turbulent. This is a direct consequence of the differences in the nature of the shear stress in laminar and turbulent flows. As is discussed in some detail in Section 8.3.3, the shear stress in laminar flow is a direct result of momentum transfer among the randomly moving molecules (a microscopic phenomenon). The shear stress in turbulent flow is largely a result of momentum transfer among the randomly moving, finite-sized fluid particles (a macroscopic phenomenon). The net result is that the physical properties of the shear stress are quite different for laminar flow than for turbulent flow. Fully Developed Laminar Flow As is indicated in the previous section, the flow in long, straight, constant diameter sections of a pipe becomes fully developed. That is, the velocity profile is the same at any cross section of the pipe. Although this is true whether the flow is laminar or turbulent, the details of the velocity profile (and other flow properties) are quite different for these two types of flow. As will be seen in the remainder of this chapter, knowledge of the velocity profile can lead directly to other useful information such as pressure drop, head loss, flowrate, and the like. Thus, we begin by developing the equation for the velocity profile in fully developed laminar flow. If the flow is not fully developed, a theoretical analysis becomes much more complex and is outside the scope of this text. If the flow is turbulent, a rigorous theoretical analysis is as yet not possible. Although most flows are turbulent rather than laminar, and many pipes are not long enough to allow the attainment of fully developed flow, a theoretical treatment and full understanding of 408 Chapter 8 ■ Viscous Flow in Pipes Element at time t + δ t Fluid element at time t Velocity profile r D ^ x ℓ V = u(r)i (1) ■ Figure 8.7 Motion of a cylindrical fluid element within a pipe. (2) fully developed laminar flow is of considerable importance. First, it represents one of the few theoretical viscous analyses that can be carried out “exactly” (within the framework of quite general assumptions) without using other ad hoc assumptions or approximations. An understanding of the method of analysis and the results obtained provides a foundation from which to carry out more complicated analyses. Second, there are many practical situations involving the use of fully developed laminar pipe flow. There are numerous ways to derive important results pertaining to fully developed laminar flow. Three alternatives include: (1) from F = ma applied directly to a fluid element, (2) from the Navier–Stokes equations of motion, and (3) from dimensional analysis methods. 8.2.1 From F = ma Applied Directly to a Fluid Element Steady, fully developed pipe flow experiences no acceleration. Velocity profiles Streamlines We consider the fluid element at time t as is shown in Fig. 8.7. It is a circular cylinder of fluid of length ℓ and radius r centered on the axis of a horizontal pipe of diameter D. Because the velocity is not uniform across the pipe, the initially flat ends of the cylinder of fluid at time t become distorted at time t + δt when the fluid element has moved to its new location along the pipe as shown in the figure. If the flow is fully developed and steady, the distortion on each end of the fluid element is the same, and no part of the fluid experiences any acceleration as it flows, as shown by the figure in the margin. The local acceleration is zero (∂V∂t = 0) because the flow is steady, and the convective acceleration is zero (V ∙ 𝛁 V = u ∂u∂x ˆi = 0) because the flow is fully developed. Thus, every part of the fluid merely flows along its streamline parallel to the pipe walls with constant velocity, although neighboring particles have slightly different velocities. The velocity varies from one pathline to the next. This velocity variation, combined with the fluid viscosity, produces the shear stress. If gravitational effects are neglected, the pressure is constant across any vertical cross section of the pipe, although it varies along the pipe from one section to the next. Thus, if the pressure is p = p1 at section (1), it is p2 = p1 − Δp at section (2) where Δp is the pressure drop between sections (1) and (2). We anticipate the fact that the pressure decreases in the direction of flow so that Δp > 0. A shear stress, τ, acts on the surface of the cylinder of fluid. This viscous stress is a function of the radius of the cylinder, τ = τ(r). As was done in fluid statics analysis (Chapter 2), we isolate the cylinder of fluid as is shown in Fig. 8.8 and apply Newton’s second law, Fx = max. In this case, even though the fluid is moving, it is not accelerating, so that ax = 0. Thus, fully developed horizontal pipe flow is merely a balance between pressure and viscous forces—the pressure difference acting on the end of the cylinder of area πr2, and the shear stress acting on the lateral surface of the cylinder of area 2πrℓ. This force balance can be written as (p1 )πr2 − (p1 − Δp)πr2 − (τ)2πrℓ = 0 τ 2 π rℓ r p1 π r2 ℓ (p1 – Δp) π r 2 x ■ Figure 8.8 Free-body diagram of a cylinder of fluid. 8.2 Fully Developed Laminar Flow 409 which can be simplified to give Δp 2τ = r ℓ (8.3) Equation 8.3 represents the basic balance in forces needed to drive each fluid particle along the pipe with constant velocity. Since neither Δp nor ℓ are functions of the radial coordinate, r, it follows that 2τr must also be independent of r. That is, τ = Cr, where C is a constant. At r = 0 (the centerline of the pipe) there is no shear stress (τ = 0). At r = D2 (the pipe wall) the shear stress is a maximum, denoted τw, the wall shear stress. Hence, C = 2τw D and the shear stress distribution throughout the pipe is a linear function of the radial coordinate τ= 2τwr D (8.4) as is indicated in Fig. 8.9. The linear dependence of τ on r is a result of the pressure force being proportional to r2 (the pressure acts on the end of the fluid cylinder; area = πr2 ) and the shear force being proportional to r (the shear stress acts on the lateral sides of the cylinder; area = 2πrℓ). If the viscosity were zero there would be no shear stress, and the pressure would be constant throughout the horizontal pipe ( Δp = 0). As is seen from Eqs. 8.3 and 8.4, the pressure drop and wall shear stress are related by Basic horizontal pipe flow is governed by a balance between viscous and pressure forces. Δp = 4ℓτw D (8.5) A small shear stress can produce a large pressure difference if the pipe is relatively long (ℓD ⪢ 1). Although we are discussing laminar flow, a closer consideration of the assumptions involved in the derivation of Eqs. 8.3, 8.4, and 8.5 reveals that these equations are valid for both laminar and turbulent flow. To carry the analysis further we must prescribe how the shear stress is related to the velocity. This is the critical step that separates the analysis of laminar from that of turbulent flow— from being able to solve for the laminar flow properties and not being able to solve for the turbulent flow properties without additional ad hoc assumptions. As is discussed in Section 8.3, the shear stress dependence for turbulent flow is very complex. However, for laminar flow of a Newtonian fluid, the shear stress is simply proportional to the velocity gradient, “τ = μ dudy” (see Section 1.6). In the notation associated with our pipe flow, this becomes τ = −μ du dr (8.6) The negative sign is included to give τ > 0 with dudr < 0 (the velocity decreases from the pipe centerline to the pipe wall). Equations 8.3 and 8.6 represent the two governing laws for fully developed laminar flow of a Newtonian fluid within a horizontal pipe. The one is Newton’s second law of motion, and the other is the definition of a Newtonian fluid. By combining these two equations, we obtain Δp du = −( r dr 2μℓ ) Ideal (inviscid) profile Laminar profile τ (D/2) = τ w τ (r) u(r) r x τ (0) = 0 τw Vc V = Vc /2 ■ Figure 8.9 Shear stress distribution within the fluid in a pipe (laminar or turbulent flow) and typical velocity profiles. 410 Chapter 8 ■ Viscous Flow in Pipes which can be integrated to give the velocity profile as follows: ∫ du = − Δp 2μℓ ∫ r dr or u = −( Δp 2 r + C1 4μℓ ) where C1 is a constant. Because the fluid is viscous it sticks to the pipe wall so that u = 0 at r = D2. Thus, C1 = (Δp16μℓ)D2. Hence, the velocity profile can be written as ΔpD2 2r 2 2r 2 u(r) = ( 1 − = V 1 − c[ (D) ] (D) ] 16μℓ )[ Under certain restrictions the velocity profile in a pipe is parabolic. R where Vc = ΔpD2 (16μℓ) is the centerline velocity. An alternative expression can be written by using the relationship between the wall shear stress and the pressure gradient (Eqs. 8.5 and 8.7) to give u(r) = τw D r 2 1 − (R) ] 4μ [   where R = D2 is the pipe radius. This velocity profile, plotted in Fig. 8.9, is parabolic in the radial coordinate, r, has a maximum velocity, Vc, at the pipe centerline, and a minimum velocity (zero) at the pipe wall. The volume flowrate through the pipe can be obtained by integrating the velocity profile across the pipe. Since the flow is axisymmetric about the centerline, the velocity is constant on small area elements consisting of rings of radius r and thickness dr, as shown in the figure in the margin. Thus, Q= r dr (8.7) ∫ u dA = ∫ r=R u(r)2πr dr = 2π Vc r=0 ∫ 0 R r 2 1 − [ ( R ) ]r dr or Q= dA = 2π r dr πR2Vc 2 By definition, the average velocity is the flowrate divided by the cross-sectional area, V = QA = QπR2, so that for this flow V= ΔpD2 πR2Vc Vc = = 2 32μℓ 2πR2 (8.8) πD4 Δp 128μℓ (8.9) and Q= 16Q0 Q ~ D4 Q Q0 D0 D 2D0 As is indicated in Eq. 8.8, the average velocity is one-half of the maximum velocity. In general, for velocity profiles of other shapes (such as for turbulent pipe flow), the average velocity is not merely the average of the maximum (Vc ) and minimum (0) velocities as it is for the laminar parabolic profile. The two velocity profiles indicated in Fig. 8.9 provide the same flowrate—one is the fictitious ideal (μ = 0) profile; the other is the actual laminar flow profile. The above results confirm the following properties of laminar pipe flow. For a horizontal pipe the flowrate is (a) directly proportional to the pressure drop, (b) inversely proportional to the viscosity, (c) inversely proportional to the pipe length, and (d) proportional to the pipe diameter to the fourth power. With all other parameters fixed, an increase in diameter by a factor of 2 will increase the flowrate by a factor of 24 = 16—the flowrate is very strongly dependent on pipe size. This dependence is shown by the figure in the margin. Likewise, a small error in pipe diameter can cause a relatively large error in flowrate. For example, a 2% error in diameter gives an 8% error in flowrate (Q ∼ D4 or δQ ∼ 4D3 δD, so that δQQ = 4 δDD). This flow, the properties of which were first 8.2 411 Fully Developed Laminar Flow Q pπ r 2 τ 2π r𝓵 Fluid cylinder x 𝓵 θ r 𝓵 (p + Δ p) π r 2 θ 𝒲 𝒲 sin θ = γ π r2𝓵 sin θ (a) (b) ■ Figure 8.10 Free-body diagram of a fluid cylinder for flow in a nonhorizontal pipe. established experimentally by two independent workers, G. Hagen (1797–1884) in 1839 and J. Poiseuille (1799–1869) in 1840, is termed Hagen–Poiseuille flow. Equation 8.9 is commonly referred to as Poiseuille’s law. Recall that all of these results are restricted to laminar flow (those with Reynolds numbers less than approximately 2100) in a horizontal pipe. The adjustment necessary to account for nonhorizontal pipes, as shown in Fig. 8.10, can be easily included by replacing the pressure drop, Δp, by the combined effect of pressure and gravity, Δp − γℓ sin θ, where θ is the angle between the pipe and the horizontal. (Note that θ > 0 if the flow is uphill, while θ < 0 if the flow is downhill.) This can be seen from the force balance in the x direction (along the pipe axis) on the cylinder of fluid shown in Fig. 8.10b. The method is exactly analogous to that used to obtain the Bernoulli equation (Eq. 3.6) when the streamline is not horizontal. The net force in the x direction is a combination of the pressure force in that direction, Δpπr 2, and the component of weight in that direction, −γπr2ℓ sin θ. The result is a slightly modified form of Eq. 8.3 given by Poiseuille’s law is valid for laminar flow only. Δp − γℓ sin θ 2τ = r ℓ (8.10) Thus, all of the results for the horizontal pipe are valid provided the pressure gradient is adjusted for the elevation term, that is, Δp is replaced by Δp − γℓ sin θ so that V= (Δp − γℓ sin θ)D2 32μℓ (8.11) Q= π(Δp − γℓ sin θ)D4 128μℓ (8.12) and It is seen that the driving force for pipe flow can be either a pressure drop in the flow direction, Δp, or the component of weight in the flow direction, −γℓ sin θ. If the flow is downhill, gravity helps the flow (a smaller pressure drop is required; sin θ < 0). If the flow is uphill, gravity works against the flow (a larger pressure drop is required; sin θ > 0). Note that γℓ sin θ = γΔz (where Δz is the change in elevation) is a hydrostatic type pressure term. If there is no flow, V = 0 and Δp = γℓ sin θ = γΔz, as expected for fluid statics. E X AMPLE 8.2 Laminar Pipe Flow GIVEN An oil with a viscosity of μ = 0.40 N ∙ s /m2 and density ρ = 900 kg /m3 flows in a pipe of diameter D = 0.020 m.   (a) What pressure drop, p1 − p2, is needed to produce a flowrate of Q = 2.0 × 10−5 m3/s if the pipe is horizontal with x1 = 0 and x2 = 10 m? FIND (b) How steep a hill, θ, must the pipe be on if the oil is to flow through the pipe at the same rate as in part (a), but with p1 = p2? (c) For the conditions of part (b), if p1 = 200 kPa, what is the pressure at section x3 = 5 m, where x is measured along the pipe? 412 Chapter 8 ■ Viscous Flow in Pipes SOLUTION (a) If the Reynolds number is less than 2100 the flow is laminar and the equations derived in this section are valid. Since the average velocity is V = QA = (2.0 × 10−5 m3/s) [π(0.020) 2m24] = 0.0637 m /s, the Reynolds number is Re = ρVDμ = 2.87 < 2100. Hence, the flow is laminar and from Eq. 8.9 with ℓ = x2 − x1 = 10 m, the pressure drop is   Δp = p1 − p2 = = 128μℓQ πD4 128(0.40 N ∙ s /m2 ) (10.0 m) (2.0 × 10−5 m3/s)   π(0.020 m) 4 or Δp = 20,400 N /m2 = 20.4 kPa   (Ans) (b) If the pipe is on a hill of angle θ such that Δp = p1 − p2 = 0, Eq. 8.12 gives sin θ = − 128μQ (1) πρgD4 (2.31 m) = 20,400 N /m2, which is equivalent to that needed for the horizontal pipe. For the horizontal pipe it is the work done by the pressure forces that overcomes the viscous dissipation. For the zero-pressure-drop pipe on the hill, it is the change in potential energy of the fluid “falling” down the hill that is converted to the energy lost by viscous dissipation. Note that if it is desired to increase the flowrate to Q = 1.0 × 10−4 m3/s with p1 = p2, the value of θ given by Eq. 1 is sin θ = −1.15. Since the sine of an angle cannot be greater than 1, this flow would not be possible. The weight of the fluid would not be large enough to offset the viscous force generated for the flowrate desired. A larger-diameter pipe would be needed.   (c) With p1 = p2 the length of the pipe, ℓ, does not appear in the flowrate equation (Eq. 1). This is a statement of the fact that for such cases the pressure is constant all along the pipe (provided the pipe lies on a hill of constant slope). This can be seen by substituting the values of Q and θ from case (b) into Eq. 8.12 and noting that Δp = 0 for any ℓ. For example, Δp = p1 − p3 = 0 if ℓ = x3 − x1 = 5 m. Thus, p1 = p2 = p3 so that p3 = 200 kPa or sin θ = −128(0.40 N ∙ s /m2 ) (2.0 × 10−5 m3/s)   3 π(900 kg /m ) (9.81 m /s2 ) (0.020 m) 4   (Ans)   Thus, θ = −13.34°. COMMENT This checks with the previous horizontal result as is seen from the fact that a change in elevation of Δz = ℓ sin θ = (10 m) sin(−13.34°) = −2.31 m is equivalent to a pressure change of Δp = ρg Δz = (900 kg /m3 ) (9.81 m /s2 )   (Ans) COMMENT Note that if the fluid were gasoline (μ = 3.1 × 10−4 N ∙ sm2 and ρ = 680 kgm3 ), the Reynolds number would be Re = 2790, the flow would probably not be laminar, and use of Eqs. 8.9 and 8.12 would give incorrect results. Also note from Eq. 1 that the kinematic viscosity, ν = μρ, is the important viscous parameter. This is a statement of the fact that with constant pressure along the pipe, it is the ratio of the viscous force (∼μ) to the weight force (∼γ = ρg) that determines the value of θ. 8.2.2 From the Navier–Stokes Equations Poiseuille’s law can be obtained from the Navier– Stokes equations. In the previous section we obtained results for fully developed laminar pipe flow by applying Newton’s second law and the assumption of a Newtonian fluid to a specific portion of the fluid—a cylinder of fluid centered on the axis of a long, round pipe. When this governing law and assumptions are applied to a general fluid flow (not restricted to pipe flow), the result is the Navier–Stokes equations as discussed in Chapter 6. In Section 6.9.3 these equations were solved for the specific geometry of fully developed laminar flow in a round pipe. The results are the same as those given in Eq. 8.7. We will not repeat the detailed steps used to obtain the laminar pipe flow from the Navier– Stokes equations (see Section 6.9.3) but will indicate how the various assumptions used and steps applied in the derivation correlate with the analysis used in the previous section. General motion of an incompressible Newtonian fluid is governed by the continuity equation (conservation of mass, Eq. 6.31) and the momentum equation (Eq. 6.127), which are rewritten here for convenience: 𝛁∙V=0 (8.13) 𝛁p ∂V + g + ν𝛁2 V + (V ∙ 𝛁)V = − ρ ∂t (8.14) For steady, fully developed flow in a pipe, the velocity contains only an axial component, which is a function of only the radial coordinate [V = u(r)iˆ]. For such conditions, the left side of the 8.2 Fully Developed Laminar Flow 413 Eq. 8.14 is zero. This is equivalent to saying that the fluid experiences no acceleration as it flows along. The same constraint was used in the previous section when considering F = ma for the fluid cylinder. Thus, with g = −gkˆ the Navier–Stokes equations become 𝛁∙V=0 𝛁p + ρgkˆ = μ𝛁 2 V (8.15) The flow is governed by a balance of pressure, weight, and viscous forces in the flow direction, similar to that shown in Fig. 8.10 and Eq. 8.10. If the flow were not fully developed (as in an entrance region, for example), it would not be possible to simplify the Navier–Stokes equations to that form given in Eq. 8.15 (the nonlinear term (V ∙ 𝛁)V would not be zero), and the solution would be very difficult to obtain. Because of the assumption that V = u(r)iˆ , the continuity equation, Eq. 8.13, is automatically satisfied. This conservation of mass condition was also automatically satisfied by the incompressible flow assumption in the derivation in the previous section. The fluid flows across one section of the pipe at the same rate that it flows across any other section. When it is written in terms of polar coordinates (as was done in Section 6.9.3), the component of Eq. 8.15 along the pipe becomes The governing differential equations can be simplified by appropriate assumptions. ∂p 1 ∂ ∂u + ρg sin θ = μ r r ∂r ( ∂r ) ∂x (8.16) Since the flow is fully developed, u = u(r) and the right side is a function of, at most, only r. The left side is a function of, at most, only x. It was shown that this leads to the condition that the pressure gradient in the x direction is a constant—∂p∂x = −Δpℓ. The same condition was used in the derivation of the previous section (Eq. 8.3). It is seen from Eq. 8.16 that the effect of a nonhorizontal pipe enters into the Navier–Stokes equations in the same manner as was discussed in the previous section. The pressure gradient in the flow direction is coupled with the effect of the weight in that direction to produce an effective pressure gradient of −Δpℓ + ρg sin θ. The velocity profile is obtained by integration of Eq. 8.16. Since it is a second-order equation, two boundary conditions are needed—(1) the fluid sticks to the pipe wall (as was also done in Eq. 8.7) and (2) either of the equivalent forms that the velocity remains finite throughout the flow (in particular u < ∞ at r = 0) or, because of symmetry, that dudr = 0 at r = 0. In the derivation of the previous section, only one boundary condition (the no-slip condition at the wall) was needed because the equation integrated was a first-order equation. The other condition (dudr = 0 at r = 0) was automatically built into the analysis because of the fact that τ = −μ dudr and τ = 2τw rD = 0 at r = 0. The results obtained by either applying F = ma to a fluid cylinder (Section 8.2.1) or solving the Navier–Stokes equations (Section 6.9.3) are exactly the same. Similarly, the basic assumptions regarding the flow structure are the same. This should not be surprising because the two methods are based on the same principle—Newton’s second law. One is restricted to fully developed laminar pipe flow from the beginning (the drawing of the free-body diagram), and the other starts with the general governing equations (the Navier–Stokes equations) with the appropriate restrictions concerning fully developed laminar flow applied as the solution process progresses. 8.2.3 From Dimensional Analysis (1) (2) D V μ 𝓵 Δp = p1 – p2 = F(V, 𝓵, D, μ ) Although fully developed laminar pipe flow is simple enough to allow the rather straightforward solutions discussed in the previous two sections, it is worthwhile to consider this flow from a dimensional analysis standpoint as will be necessary for turbulent flow in Sec. 8.4. Thus, we assume that the pressure drop in the horizontal pipe, Δp, is a function of the average velocity of the fluid in the pipe, V, the length of the pipe, ℓ, the pipe diameter, D, and the viscosity of the fluid, μ, as shown by the figure in the margin. We have not included the density or the specific weight of the 414 Chapter 8 ■ Viscous Flow in Pipes fluid as parameters because for such flows they are not important parameters. There is neither mass (density) times acceleration nor a component of weight (specific weight times volume) in the flow direction involved. (Note that density or specific weight should be included for non-horizontal pipe flow.) Thus, Δp = F(V, ℓ, D, μ) Five variables can be described in terms of three reference dimensions (M, L, T ). According to the results of dimensional analysis (Chapter 7), this flow can be described in terms of k − r = 5 − 3 = 2 dimensionless groups. One such representation is D Δp ℓ = ϕ( ) μV D (8.17) where ϕ(ℓD) is an unknown function of the length to diameter ratio of the pipe. Although this is as far as dimensional analysis can take us, it seems reasonable to impose a further assumption that the pressure drop is directly proportional to the pipe length. That is, it takes twice the pressure drop to force fluid through a pipe if its length is doubled. The only way that this can be true is if ϕ(ℓD) = CℓD, where C is a constant. Thus, Eq. 8.17 becomes D Δp Cℓ = μV D which can be rewritten as Δp Cμ V = ℓ D2 or Q = AV = Dimensional analysis can be used to put pipe flow para-meters into dimensionless form. (π4C) ΔpD4 μℓ (8.18) The basic functional dependence for laminar pipe flow given by Eq. 8.18 is the same as that obtained by the analysis of the two previous sections. The value of C must be determined by theory (as done in the previous two sections) or experiment. For a round pipe, C = 32. For ducts of other cross-sectional shapes, the value of C is different (see Section 8.4.3). It is usually advantageous to describe a process in terms of dimensionless quantities. To this end we rewrite the pressure drop equation for laminar horizontal pipe flow, Eq. 8.8, as Δp = 32μℓVD2 and divide both sides by the dynamic pressure, ρV 2 2, to obtain the dimensionless form as Δp 1 2 2 ρV = (32μℓVD2 ) 1 2 2 ρV = 64( μ ℓ 64 ℓ = ρVD )( D ) Re ( D ) This is often written as Δp = f ℓ ρV 2 D 2 where the dimensionless quantity f = Δp(Dℓ)(ρV2 2) 10 Laminar flow 1 f is termed the friction factor, or sometimes the Darcy friction factor [H. P. G. Darcy (1803–1858)]. (This parameter should not be confused with the less-used Fanning friction factor, which is defined to be f4. In this text we will use only the Darcy friction factor.) Thus, the friction factor for laminar fully developed round pipe flow is simply 0.1 0.01 10 f= 100 1000 Re as shown by the figure in the margin. 64 Re (8.19) 8.2 Fully Developed Laminar Flow 415 By substituting the pressure drop in terms of the wall shear stress (Eq. 8.5), we obtain an alternate expression for the friction factor as a dimensionless wall shear stress f= 8τw ρV2 (8.20) Knowledge of the friction factor will allow us to obtain a variety of information regarding pipe flow. For turbulent flow the dependence of the friction factor on the Reynolds number is much more complex than that given by Eq. 8.19 for laminar flow. This is discussed in detail in Section 8.4. 8.2.4 Energy Considerations In the previous three sections we derived the basic laminar flow results from application of F = ma or dimensional analysis considerations. It is equally important to understand the implications of energy considerations of such flows. To this end we consider the energy equation for incompressible, steady flow between two locations as is given in Eq. 5.89 and presented here without shaft work head p1 p2 V21 V22 + z1 = + z2 + hL + α1 + α2 γ γ 2g 2g Recall that the kinetic energy coefficients, α1 and α2, compensate for the fact that the velocity profile across the pipe is not uniform. For uniform velocity profiles, α = 1, whereas for any nonuniform profile, α > 1. The head loss term, hL, accounts for any energy loss associated with the flow. This loss is a direct consequence of the viscous dissipation that occurs throughout the fluid in the pipe. For the ideal (inviscid) cases discussed in previous chapters, α1 = α2 = 1, hL = 0, and the energy equation reduces to the familiar Bernoulli equation discussed in Chapter 3 (Eq. 3.7). Even though the velocity profile in viscous pipe flow is not uniform, for fully developed flow it does not change from section (1) to section (2) so that α1 = α2. Thus, the kinetic energy is the same at any section (α1 V 21 2 = α2 V 22 2) and the energy equation becomes hL p1 γ p2 γ (2) (1) z1 (8.21) z2 p1 p2 ( γ + z1) − ( γ + z2) = hL (8.22) The energy dissipated by the viscous forces within the fluid is supplied by the excess work done by the pressure and gravity forces as shown by the figure in the margin. A comparison of Eqs. 8.22 and 8.10 shows that the head loss is given by hL = 2τℓ γr (recall p1 = p2 + Δp and z2 − z1 = ℓ sin θ), which, by use of Eq. 8.4, can be rewritten in the form The head loss in a pipe is a result of the viscous shear stress on the wall. hL = 4ℓτw γD (8.23) It is the shear stress at the wall (which is directly related to the viscosity and the shear stress throughout the fluid) that is responsible for the head loss. A closer consideration of the assumptions involved in the derivation of Eq. 8.23 will show that it is valid for both laminar and turbulent flow. 416 Chapter 8 ■ Viscous Flow in Pipes EXAMPLE 8.3 Laminar Pipe Flow Properties GIVEN The flowrate, Q, of corn syrup through the horizontal FIND (a) Plot K(T ) versus T for 60 °F ≤ T ≤ 160 °F. (b) Determine the wall shear stress and the pressure drop, Δp = p1 − p2, for Q = 0.5 ft3/s and T = 100 °F. (c) For the conditions of part (b), determine the net pressure force, (πD24) Δp, and the net shear force, πDℓτw , on the fluid within the pipe between the sections (1) and (2). SOLUTION (a) 6 ft Q (2) (1) 3-in. diameter (a) 100 10–1 K, ft5/(lbts) pipe shown in Fig. E8.3a is to be monitored by measuring the pressure difference between sections (1) and (2). It is proposed that Q = K Δp, where the calibration constant, K, is a function of temperature, T, because of the variation of the syrup’s viscosity and density with temperature. These variations are given in Table E8.3. 10–2 10–3 If the flow is laminar, it follows from Eq. 8.9 that 3 Q= π( 12 ft) 4 Δp πD4 Δp = 128μℓ 128μ(6 ft) 10–4 60 100 or Q = K Δp = 140 T, °F 180 (b) 1.60 × 10−5 Δp μ (1) ■ Figure E8.3 where the units on Q, Δp, and μ are ft3/s, lb /ft2, and lb ∙ s /ft2, respectively. Thus 1.60 × 10−5 K= (Ans) μ ■ Table E8.3   where the units of K are ft5/lb ∙ s. By using values of the viscosity from Table E8.3, the calibration curve shown in Fig. E8.3b is obtained. This result is valid only if the flow is laminar. COMMENT As shown in Section 8.5, for turbulent flow the flowrate is not linearly related to the pressure drop so it would not be possible to have Q = K Δp. Note also that the value of K is independent of the syrup density (ρ was not used in the calculations) since laminar pipe flow is governed by pressure and viscous effects; inertia is not important. (b) For T = 100 °F, the viscosity is μ = 3.8 × 10−3 lb ∙ s /ft2 so that with a flowrate of Q = 0.5 ft3/s the pressure drop (according to Eq. 8.9) is Δp = = 128μℓQ πD4 128(3.8 × 10−3 lb ∙ s /ft2 ) (6 ft) (0.5 ft3/s) = 119 lb /ft ρ (slugsft 3 ) μ (lb ∙ sft2 ) 60 80 100 120 140 160 2.07 2.06 2.05 2.04 2.03 2.02 4.0 × 10−2 1.9 × 10−2 3.8 × 10−3 4.4 × 10−4 9.2 × 10−5 2.3 × 10−5 Hence, the flow is laminar. From Eq. 8.5 the wall shear stress is 3 τw = π( 12 ft) 4 (Ans)   provided the flow is laminar. For this case (119 lb /ft2 ) ( 12 ft) ΔpD = = 1.24 lb /ft2 4ℓ 4(6 ft)     (Ans) (c) For the conditions of part (b), the net pressure force, Fp, on the fluid within the pipe between sections (1) and (2) is 3 2 T (°F) Fp = 2 π 2 π 3 D Δp = ( ft) (119 lb /ft2 ) = 5.84 lb 4 4 12   (Ans) Similarly, the net viscous force, Fυ, on that portion of the fluid is Q 0.5 ft3/s = 10.2 ft /s V= = π 3 2 A ( 12 ft) 4 so that 3 (2.05 slugs /ft3 ) (10.2 ft /s) ( 12 ft) ρVD Re = = μ (3.8 × 10−3 lb ∙ s /ft2 ) = 1380 < 2100 D Fυ = 2π ( ) ℓτw 2 3 = 2π [ ft (6 ft) (1.24 lb /ft2 ) = 5.84 lb 2(12) ]   (Ans) COMMENT Note that the values of these two forces are the same. The net force is zero; there is no acceleration. 8.3 8.3 Fully Developed Turbulent Flow 417 Fully Developed Turbulent Flow In the previous section various properties of fully developed laminar pipe flow were discussed. Since turbulent pipe flow is actually more likely to occur than laminar flow in practical situations, it is necessary to obtain similar information for turbulent pipe flow. However, turbulent flow is a very complex process. Numerous persons have devoted considerable effort in attempting to understand the variety of baffling aspects of turbulence. Although a considerable amount of knowledge about the topic has been developed, the field of turbulent flow still remains the least understood area of fluid mechanics. In this book we can provide only some of the very basic ideas concerning turbulence. The interested reader should consult some of the many books available for further reading (Refs. 1–3). 8.3.1 Transition from Laminar to Turbulent Flow Random, turbulent fluctuations Turbulent 3 Turbulent bursts 2 4000 Transitional 2000 1 Laminar 0 t, sec ■ Figure 8.11 Transition from laminar to turbulent flow in a pipe. 0 Re = VD/v u, ft/s Turbulent flows involve randomly fluctuating parameters. When viscous effects are present in a flow, the flow may be laminar, turbulent, or in transition between the two. Whether a flow will be laminar or turbulent can be inferred from the value of the Reynolds number. The value that distinguishes between the two, called the critical value, depends on the specific flow situation involved. For example, flow in a pipe and flow along a flat plate (boundary layer flow, as is discussed in Section 9.2.4) can be laminar or turbulent, depending on the value of the Reynolds number. As a general rule for pipe flow, the value of the Reynolds number must be less than approximately 2100 for laminar flow and greater than approximately 4000 for turbulent flow. For flow along a flat plate the transition between laminar and turbulent flow occurs at a Reynolds number of approximately 500,000 (see Section 9.2.4). The value is much larger for the flat plate because the length term in the Reynolds number is the distance measured from the leading edge of the plate. Consider a long section of pipe that is initially filled with a fluid at rest. As the valve is opened to start the flow, the flow velocity and, hence, the Reynolds number increase from zero (no flow) to their maximum steady-state flow values, as is shown in Fig. 8.11. Assume this transient process is slow enough so that unsteady effects are negligible (quasi-steady flow). For an initial time period the Reynolds number is small enough for laminar flow to occur. At some time the Reynolds number reaches 2100, and the flow begins its transition to turbulent conditions. Intermittent spots or bursts of turbulence appear. As the Reynolds number is increased, the entire flow field becomes turbulent. The flow remains turbulent as long as the Reynolds number exceeds approximately 4000. A typical trace of the axial component of velocity measured at a given location in the flow, u = u(t), is shown in Fig. 8.12. Its irregular, seemingly random nature is the distinguishing feature of turbulent flow. (It is currently unknown whether turbulent flow properties are truly random. Since they appear random, we will use that descriptor in this text.) The character of many of the 418 Chapter 8 ■ Viscous Flow in Pipes u u' u(t) _ u = time-averaged (or mean) value T tO tO + T t ■ Figure 8.12 The time-averaged, u, and fluctuating, u′, description of a parameter for turbulent flow. VIDEO V8.4 Laminar and turbulent mixing important properties of the flow depends strongly on the existence and nature of the turbulent fluctuations or randomness indicated. Calculation of the heat transfer, pressure drop, and many other parameters would not be possible without inclusion of the seemingly small, but very important, effects associated with the irregular motion of the flow. Also shown in Fig. 8.12 is the time-averaged velocity, u. This is the velocity most commonly needed for pipe flows and many other engineering applications. Consider flow in a pan of water placed on a stove. With the stove turned off, the fluid is stationary. The initial sloshing has died out because of viscous dissipation within the water. With the stove turned on, a temperature gradient in the vertical direction, ∂T∂z, is produced. The water temperature is greatest near the pan bottom and decreases toward the top of the fluid layer. If the temperature difference is very small, the water will remain stationary, even though the water density is smallest near the bottom of the pan because of the decrease in density with an increase in temperature. A further increase in the temperature gradient will cause a buoyancy-driven instability that results in fluid motion—the light, warm water rises to the top, and the heavy, cold water sinks to the bottom. This slow, regular “turning over” increases the heat transfer from the pan to the water and promotes mixing within the pan. As the temperature gradient increases still further, the fluid motion becomes more vigorous and eventually turns into a chaotic, random, turbulent flow with considerable mixing, vaporization (boiling), and greatly increased heat transfer rate. The flow has progressed from a stationary fluid, to laminar flow, and finally to turbulent, multiphase (liquid and vapor) flow. Mixing processes and heat and mass transfer processes are considerably enhanced in turbulent flow compared to laminar flow. This is due to the macroscopic scale of the randomness in turbulent flow. We are all familiar with the “rolling,” vigorous eddy-type motion of the water in a pan being heated on the stove (even if it is not heated to boiling). Such finite-sized random mixing is very effective in transporting energy and mass throughout the flow field, thereby increasing the various rate processes involved. Laminar flow, on the other hand, can be thought of as very small but finite-sized fluid particles flowing smoothly in layers, one over another. The only randomness and mixing take place on the molecular scale and result in relatively small heat, mass, and momentum transfer rates. Without turbulence it would be virtually impossible to carry out life as we now know it. Mixing is one positive application of turbulence, as discussed above, but there are other situations where turbulent flow is desirable. To transfer the required heat between a solid and an adjacent fluid (such as in the cooling coils of an air conditioner or a boiler of a power plant) would require an enormously large heat exchanger if the flow were laminar. Similarly, the required mass transfer of a liquid state to a vapor state (such as is needed in the evaporated cooling system associated with sweating) would require very large surfaces if the fluid flowing past the surface were laminar rather than turbulent. As shown in Chapter 9, turbulence can also aid in delaying flow separation to reduce drag forces. 8.3 Fully Developed Turbulent Flow 419 THE WIDE WORLD OF FLUIDS Smaller heat exchangers Automobile radiators, air condition- termed “turbulators,” which provide additional turbulent mixing ers, and refrigerators contain heat exchangers that transfer motion than would normally occur. Such enhancement mecha- energy from (to) the hot (cold) fluid within the heat exchanger nisms include internal fins, spiral wire or ribbon inserts, and ribs tubes to (from) the colder (hotter) surrounding fluid. These units or grooves on the inner surface of the tube. While these mech- can be made smaller and more efficient by increasing the heat anisms can increase the heat transfer rate by 1.5 to 3 times transfer rate across the tubes’ surfaces. If the flow through over that for a bare tube at the same flowrate, they also the tubes is laminar, the heat transfer rate is relatively small. increase the pressure drop (and therefore the power) needed Significantly larger heat transfer rates are obtained if the flow to produce the flow within the tube. Thus, a compromise involv- within the tubes is turbulent. Even greater heat transfer rates ing increased heat transfer rate and increased power consump- can be obtained by the use of turbulence promoters, sometimes tion is often needed. VIDEO V8.5 Stirring cream into coffee VIDEO V8.6 Stirring color into paint Turbulence is also of importance in the mixing of fluids. Smoke from a stack would continue for miles as a ribbon of pollutant without rapid dispersion within the surrounding air if the flow were laminar rather than turbulent. Under certain atmospheric conditions this is observed to occur. Although there is mixing on a molecular scale (laminar flow), it is several orders of magnitude slower and less effective than the mixing on a macroscopic scale (turbulent flow). It is considerably easier to mix cream into a cup of coffee (turbulent flow) than to thoroughly mix two colors of a viscous paint (laminar flow). In other situations laminar (rather than turbulent) flow is desirable. The pressure drop in pipes (hence, the power requirements for pumping) can be considerably lower if the flow is laminar rather than turbulent. Fortunately, the blood flow through a person’s arteries is normally laminar, except in the largest arteries with high blood flowrates. The aerodynamic drag on an airplane wing can be considerably smaller with laminar flow past it than with turbulent flow. 8.3.2 Turbulent Shear Stress The fundamental difference between laminar and turbulent flow lies in the chaotic, random behavior of the various fluid parameters. Such variations occur in the three components of velocity, the pressure, the shear stress, the temperature, and any other variable that has a field description. Turbulent flow is characterized by random, three-dimensional vorticity (i.e., fluid particle rotation or spin; see Section 6.1.3). As is indicated in Fig. 8.12, such flows can be described in terms of their mean values (denoted with an overbar) on which are superimposed the fluctuations (denoted with a prime). Thus, if u = u(x, y, z, t) is the x component of instantaneous velocity, then its time mean (or time-average) value, u, is Turbulent flow parameters can be described in terms of mean and fluctuating portions. u= 1 T ∫ t0 +T (8.24) u(x, y, z, t) dt t0 where the time interval, T, is considerably longer than the period of the longest fluctuations, but considerably shorter than any unsteadiness of the average velocity. This is illustrated in Fig. 8.12. The fluctuating part of the velocity, u′, is that time-varying portion that differs from the average value u = u + u′ or u′ = u − u (8.25) The structure and characteristics of turbulence may vary from one flow situation to another. One turbulent flow may have very large velocity fluctuations and another may have quite small fluctuations. Unfortunately, simply calculating the time average of the fluctuations will not tell us the general magnitude of the fluctuations as it will always be zero, since u′ = = 1 T t0 +T ∫ (u − u ) dt = t0 1 (T u − T u) = 0 T     1 T( ∫ t0 t0 +T u dt − u ∫ t0 t0 +T dt) 420 Chapter 8 ■ Viscous Flow in Pipes u' (u')2 u' or (u')2 (u')2 > 0 0 t u' = 0 ■ Figure 8.13 Average of the fluctuations and average of the square of the fluctuations. The fluctuations are equally distributed on either side of the average. Since the square of a fluctuation quantity cannot be negative [(u′) 2 ≥ 0], as indicated in Fig. 8.13, its average value is positive. Thus, (u′ ) 2 = 1 T ∫ t0 +T (u′) 2 dt > 0 t0 This concept can be applied to the other fluctuating velocity components, v′ or w′, as well. On the other hand, it may be that the average of products of the fluctuations, such as u′υ′, are zero or nonzero (either positive or negative). The time average of square of velocity fluctuation is used to measure the turbulence intensity (or level of the turbulence), which may be larger in a very gusty wind than it is in a relatively steady (although turbulent) wind. The turbulence intensity, ℐ, is often defined as the square root of the mean square of the fluctuating velocity divided by the time-averaged velocity, or ℐ= V, ft/s 15 10 5 0 0 20 40 t, min The relationship between fluid motion and shear stress is very complex for turbulent flow. √ (u′) 2 = u 1 [T ∫ t0 t0 +T 12 (u′) 2 dt] u The larger the turbulence intensity, the larger the fluctuations of the velocity (and other flow parameters). Well-designed wind tunnels have typical values of ℐ ≈ 0.01, although with extreme care, values as low as ℐ = 0.0002 have been obtained. On the other hand, values of ℐ ≳ 0.1 are found for the flow in the atmosphere and rivers. A typical atmospheric wind-speed graph is shown in the figure in the margin. Another turbulence parameter that is different from one flow situation to another is the period of the fluctuations—the time scale of the fluctuations shown in Fig. 8.12. In many flows, such as the flow of water from a faucet, typical frequencies are on the order of 10, 100, or 1000 cycles per second (cps). For other flows, such as the Gulf Stream current in the Atlantic Ocean or flow of the atmosphere of Jupiter, characteristic random oscillations may have a period on the order of hours, days, or more. It is tempting to extend the concept of viscous shear stress for laminar flow (τ = μ dudy) to that of turbulent flow by replacing u, the instantaneous velocity, by u, the time-averaged velocity. However, numerous experimental and theoretical studies have shown that such an approach leads to completely incorrect results. That is, τ ≠ μ d udy. A physical explanation for this behavior can be found in the concept of what produces a shear stress. Laminar flow is modeled as fluid particles that flow smoothly along in layers, gliding past the slightly slower or faster ones on either side. As is discussed in Chapter 1, the fluid actually consists of numerous molecules darting about in an almost random fashion as is indicated in Fig. 8.14a. The motion is not entirely random—a slight bias in one direction produces the flowrate we 8.3 y Fully Developed Turbulent Flow 421 y Average velocity profile, u = u(y) Velocity profile, u = u(y) (2) A A A A u u u1 < u2 (1) Turbulent eddies (a) (b) ■ Figure 8.14 (a) Laminar flow shear stress caused by random motion of molecules. (b) Turbulent flow as a series of random, three-dimensional eddies. Turbulent flow shear stress is larger than laminar flow shear stress because of the irregular, random motion. (© P. Dimotakis. Dimotakis, P.E., R.C. Miake-Lye, and D.A. Papantoniou 1983 Structure and Dynamics of Round Turbulent Jets. Phys. Fluids 26(11):3185–3192.) associate with the motion of fluid particles, u. As the molecules dart across a given plane (plane A–A, for example), the ones moving upward have come from an area of smaller average x component of velocity than the ones moving downward, which have come from an area of larger velocity. The momentum flux in the x direction across plane A–A of Fig. 8.14a gives rise to a drag (to the left) of the lower fluid on the upper fluid and an equal but opposite effect of the upper fluid on the lower fluid. The sluggish molecules moving upward across plane A–A must be accelerated by the fluid above this plane. The rate of change of momentum in this process produces (on the macroscopic scale) a shear force. Similarly, the more energetic molecules moving down across plane A–A must be slowed down by the fluid below that plane. This shear force is present only if there is a gradient in u = u(y), otherwise the average x component of velocity (and momentum) of the upward and downward molecules is exactly the same. In addition, there are attractive forces between molecules. By combining these effects we obtain the well-known Newton viscosity law: τ = μ dudy, where on a molecular basis μ is related to the mass and speed (temperature) of the random motion of the molecules. Although the above random motion of the molecules is also present in turbulent flow, there is another factor that is generally more pronounced and therefore more important. A simplistic way of thinking about turbulent flow is to consider it as consisting of a series of random, three-dimensional eddy type motions as is depicted (in one dimension only) in Fig. 8.14b and apparent in the photograph in the margin. These eddies range in size from very small diameter (on the order of the size of a fluid particle) to fairly large diameter (on the order of the size of the object or flow geometry considered). They move about randomly, conveying mass with an average velocity u = u(y). This eddy structure greatly promotes mixing within the fluid. It also greatly increases the transport of x momentum across plane A–A. That is, finite “chunks” of fluid (not merely individual molecules as in laminar flow) are randomly transported across this plane, resulting in a relatively large (when compared with laminar flow) shear force. These “chunks” vary in size but are much larger than molecules. THE WIDE WORLD OF FLUIDS Listen to the flowrate Sonar systems are designed to listen to on the other hand, work for both laminar and turbulent flow. See transmitted and reflected sound waves in order to locate sub- Sec. 8.6.1.) The flowmeters contain a sonar-based array that lis- merged objects. They have been used successfully for many tens to and interprets pressure fields generated by the turbu- years to detect and track underwater objects such as subma- lent motion in pipes. By listening to the pressure fields associ- rines and aquatic animals. Recently, sonar techniques have ated with the movement of the turbulent eddies, the device can been refined so that they can be used to determine the flowrate determine the speed at which the eddies travel past an array of in pipes. These new flowmeters work for turbulent, not laminar, sensors. The flowrate is determined by using a calibration proce- pipe flows because their operation depends strictly on the exis- dure that links the speed of the turbulent structures to the volu- tence of turbulent eddies within the flow. (Ultrasonic flowmeters, metric flowrate. 422 Chapter 8 ■ Viscous Flow in Pipes The random velocity components that account for this momentum transfer (hence, the shear force) are u′ (for the x component of velocity) and υ′ (for the rate of mass transfer crossing the plane). A more detailed consideration of the processes involved will show that the apparent shear stress on plane A–A is given by the following (Ref. 2): τ=μ VIDEO V8.7 Turbulence in a bowl The shear stress is the sum of a laminar portion and a turbulent portion. du − ρu′υ′ = τlam + τturb dy (8.26) Note that if the flow is laminar, u′ = υ′ = 0, so that u′υ′ = 0 and Eq. 8.26 reduces to the customary random molecule-motion-induced laminar shear stress, τlam = μ d udy. For turbulent flow it is found that the turbulent shear stress, τturb = −ρu′υ′, is positive (a positive υ′ is usually correlated with a negative u′). Hence, the shear stress is greater in turbulent flow than in laminar flow. Note the units on τturb are (density)(velocity) 2 = (slugsft3 )(fts) 2 = (slugs ∙ fts2 )ft2 = lbft2, or Nm2, as expected. Terms of the form −ρu′υ′ (or −ρυ′w′, etc.) are called Reynolds stresses in honor of Osborne Reynolds who first discussed them in 1895. It is seen from Eq. 8.26 that the shear stress in turbulent flow is not merely proportional to the gradient of the time-averaged velocity, u(y). It also contains a contribution due to the random fluctuations of the x and y components of velocity. The density is involved because of the momentum transfer of the fluid within the random eddies. Although the relative magnitude of τlam compared to τturb is a complex function dependent on the specific flow involved, typical measurements indicate the structure shown in Fig. 8.15a. (Recall from Eq. 8.4 that the shear stress is proportional to the distance from the centerline of the pipe.) In a very narrow region near the wall (the viscous sublayer), the laminar shear stress is dominant. Away from the wall (in the outer layer) the turbulent portion of the shear stress is dominant. The transition between these two regions occurs in the overlap layer. The corresponding typical velocity profile is shown in Fig. 8.15b. The scale of the sketches shown in Fig. 8.15 is not necessarily correct. Typically the value of τturb is 100 to 1000 times greater than τlam in the outer region, while the converse is true in the viscous sublayer. A correct modeling of turbulent flow is strongly dependent on an accurate knowledge of τturb. This, in turn, requires an accurate knowledge of the fluctuations u′ and υ′, or ρu′υ′. As yet it is not possible to solve the governing equations (the Navier–Stokes equations) for these details of the flow, although numerical techniques (see Appendix A) using the largest and fastest computers available have produced important information about some of the characteristics of turbulence. Considerable effort has gone into the study of turbulence. Much remains to be learned. Perhaps studies in the new areas of chaos and fractal geometry will provide the tools for a better understanding of turbulence (see Section 8.3.5). The vertical scale of Fig. 8.15 is also distorted. The viscous sublayer is usually a very thin layer adjacent to the wall. For example, for water flow in a 3-in.-diameter pipe with an average velocity of 10 fts, the viscous sublayer is approximately 0.002 in. thick. Since the fluid motion within this thin layer is critical in terms of the overall flow (the no-slip condition and the wall shear stress occur in this layer), it is not surprising to find that turbulent pipe flow properties can be quite Viscous sublayer Pipe wall R R τ lam Overlap layer τ turb τ r r Outer layer Pipe centerline 0 0 τw 0 0 Vc τ (r) u(r) (a) (b) ■ Figure 8.15 Structure of turbulent flow in a pipe. (a) Shear stress. (b) Average velocity. 8.3 Fully Developed Turbulent Flow 423 dependent on the roughness of the pipe wall, unlike laminar pipe flow, which is independent of roughness. Small roughness elements (scratches, rust, sand or dirt particles, etc.) can easily disturb this viscous sublayer (see Section 8.4), thereby affecting the entire flow. A model for the shear stress for turbulent flow is given in terms of the eddy viscosity, η, where τturb = η Various ad hoc assumptions have been used to approximate turbulent shear stresses. du dy (8.27) This extension of laminar flow terminology was introduced by J. Boussinesq, a French scientist, in 1877. Although the concept of an eddy viscosity is intriguing, in practice it is not an easy parameter to use. Unlike the absolute viscosity, μ, which is a known value for a given fluid, the eddy viscosity is a function of both the fluid and the flow conditions. That is, the eddy viscosity of water cannot be looked up in handbooks—its value changes from one turbulent flow condition to another and from one point in a turbulent flow to another. The inability to accurately determine the Reynolds stress, −ρu′υ′, is equivalent to not knowing the eddy viscosity. Several semiempirical theories have been proposed (Ref. 3) to determine approximate values of η. L. Prandtl (1875–1953), a German physicist and aerodynamicist, proposed that the turbulent process could be viewed as the random transport of bundles of fluid particles over a certain distance, ℓm, the mixing length, from a region of one velocity to another region of a different velocity. By the use of some ad hoc assumptions and physical reasoning, it was concluded that the eddy viscosity was given by η = ρℓ2m | | du dy Thus, the turbulent shear stress is τturb = ρℓ2m( du 2 dy ) (8.28) The problem is thus shifted to that of determining the mixing length, ℓm. Further considerations indicate that ℓm is not a constant throughout the flow field. Near a solid surface the turbulence is dependent on the distance from the surface. Thus, additional assumptions are made regarding how the mixing length varies throughout the flow. The net result is that as yet there is no general, all-encompassing, useful model that can accurately predict the shear stress throughout a general incompressible, viscous turbulent flow. Without such information it is impossible to integrate the force balance equation to obtain the turbulent velocity profile and other useful information, as was done for laminar flow to obtain Eq. 8.7. Further information on models of turbulence is presented in Appendix A. 8.3.3 Turbulent Velocity Profile VIDEO V8.8 Turbulent channel flow Considerable information concerning turbulent velocity profiles has been obtained through the use of dimensional analysis, experimentation, numerical simulations, and semiempirical theoretical efforts. As is indicated in Fig. 8.15, fully developed turbulent flow in a pipe can be broken into three regions which are characterized by their distances from the wall: the viscous sublayer very near the pipe wall, the overlap region, and the outer turbulent layer throughout the center portion of the flow. Within the viscous sublayer the viscous shear stress is dominant compared with the turbulent (or Reynolds) stress, and the random, eddying nature of the flow is essentially absent. In the outer turbulent layer the Reynolds stress is dominant, and there is considerable mixing and randomness to the flow. The character of the flow within these two regions is entirely different. For example, within the viscous sublayer the fluid viscosity is an important parameter; the density is unimportant. In the outer layer the opposite is true. By a careful use of dimensional analysis arguments for the flow in each layer and by a matching of the results in the common overlap layer, it has been possible to obtain the following conclusions about the turbulent velocity profile in a smooth pipe (Ref. 5). In the viscous sublayer the velocity profile can be written in dimensionless form as yu* u = ν u* (8.29) 424 Chapter 8 ■ Viscous Flow in Pipes 25 Experimental data 20 Eq. 8.29 Pipe centerline Eq. 8.30 15 u ___ u* Turbulent effects 10 Viscous and turbulent effects 5 Viscous sublayer 0 1 10 102 103 104 yu* ____ v ■ Figure 8.16 Typical structure of the turbulent velocity profile in a pipe. where y = R − r is the distance measured from the wall, u is the time-averaged x component of velocity, and u* = (τw ρ) 12 is termed the friction velocity. Note that u* is not an actual velocity of the fluid—it is merely a quantity that has dimensions of velocity. As is indicated in Fig. 8.16, Eq. 8.29 (commonly called the law of the wall) is valid very near the smooth wall, for 0 ≤ yu*ν ≲ 5. Dimensional analysis arguments indicate that in the overlap region the velocity should vary as the logarithm of y. Thus, the following expression has been proposed: A turbulent flow velocity profile can be divided into various regions. VIDEO V8.9 Laminar to turbulent flow from a pipe yu* u + 5.0 = 2.44 ln( ν ) u* (8.30) where the constants 2.44 and 5.0 have been determined experimentally. As is indicated in Fig. 8.16, for regions not too close to the smooth wall, but not all the way out to the pipe center, Eq. 8.30 gives a reasonable correlation with the experimental data. Note that the horizontal scale is a logarithmic scale. This tends to exaggerate the size of the viscous sublayer relative to the remainder of the flow. As is shown in Example 8.4, the viscous sublayer is usually quite thin. Similar results can be obtained for turbulent flow past rough walls (Ref. 17). A number of other correlations exist for the velocity profile in turbulent pipe flow. In the central region (the outer turbulent layer) the expression (Vc − u)u* = 2.5 ln(Ry), where Vc is the centerline velocity, is often suggested as a good correlation with experimental data. Another often-used (and relatively easy to use) correlation is the empirical power-law velocity profile u r 1n = (1 − ) Vc R (8.31) In this representation, the value of n is a function of the Reynolds number, as is indicated in Fig. 8.17. The one-seventh power-law velocity profile (n = 7) is often used as a reasonable approximation for many practical flows. Typical turbulent velocity profiles based on this power-law representation are shown in Fig. 8.18. A closer examination of Eq. 8.31 shows that the power-law profile cannot be valid near the wall, since according to this equation the velocity gradient is infinite there. In addition, Eq. 8.31 cannot be precisely valid near the centerline because it does not give d udr = 0 at r = 0. However, it does provide a reasonable approximation to the measured velocity profiles across most of the pipe. Note from Fig. 8.18 that the turbulent profiles are much “flatter” than the laminar profile and that this flatness increases with Reynolds number (i.e., with n). The reason for this flatness can be inferred from the discussion concerning Fig. 8.14 in Sec. 8.3.2. Increased momentum transport due to 8.3 Fully Developed Turbulent Flow 425 11 10 9 n 8 7 6 5 104 105 106 ρVD Re = ____ μ ■ Figure 8.17 Exponent, n, for power-law velocity profiles. (Adapted from Ref. 1.) 1.0 n = 10 n=6 Laminar VIDEO V8.10 Laminar/ turbulent velocity profiles r __ R n=8 0.5 Turbulent 0 0 0.5 _ u __ 1.0 Vc ■ Figure 8.18 Typical laminar flow and turbulent flow velocity profiles. turbulence, in essence, homogenizes the velocity profile; high velocities interchange places with low velocities resulting in a flattened time average profile. Recall from Chapter 3 that reasonable approximate results are often obtained by using the inviscid Bernoulli equation and by assuming a fictitious uniform velocity profile. Since most flows are turbulent and turbulent flows tend to have nearly uniform velocity profiles, the usefulness of the Bernoulli equation and the uniform profile assumption is not unexpected. Of course, many properties of the flow cannot be accounted for without including viscous effects. E X AMPLE 8.4 Turbulent Pipe Flow Properties Water at 20 °C (ρ = 998 kgm3 and ν = 1.004 × 10 m s) flows through a horizontal pipe of 0.1-m diameter with a flowrate of Q = 4 × 10−2 m3s and a pressure gradient of 2.59 kPa/m. GIVEN −6 FIND 2 (a) Determine the approximate thickness of the viscous sublayer. (b) Determine the approximate centerline velocity, Vc. (c) Determine the ratio of the turbulent to laminar shear stress, τturb τlam, at a point midway between the centerline and the pipe wall (i.e., at r = 0.025 m). 426 Chapter 8 ■ Viscous Flow in Pipes SOLUTION (a) According to Fig. 8.16, the thickness of the viscous sublayer, δs, is approximately obtained by integration of the power-law velocity profile as follows. Since the flow is axisymmetric, δsu* =5 ν Q = AV = ∫ u dA = V ∫ r=R c r=0 r 1n 1 − (2πr) dr ( R) which can be integrated to give Therefore, δs = 5 ν u* Q = 2πR2Vc where n2 (n + 1) (2n + 1) Thus, since Q = πR2V, we obtain 12 τw ρ) (1) 2n2 V = Vc (n + 1) (2n + 1) The wall shear stress can be obtained from the pressure drop data and Eq. 8.5, which is valid for either laminar or turbulent flow. Thus, With n = 8.4 in the present case, this gives u* = ( D Δp (0.1 m) (2.59 × 103 N /m2 ) τw = = = 64.8 N/m2 4ℓ 4(1 m) Hence, from Eq. 1 we obtain Vc = (n + 1) (2n + 1) 2n2 = 6.04 m /s Recall that Vc = 2V for laminar pipe flow.   τ= 2τwr 2(64.8 N /m2 ) (0.025 m) = D (0.1 m)   or δs = τ = τlam + τturb = 32.4 N/m2   (Ans) COMMENT As stated previously, the viscous sublayer is very thin. Minute imperfections on the pipe wall will protrude into this sublayer and affect some of the characteristics of the flow (i.e., wall shear stress and pressure drop). where τlam = −μ d udr. From the power-law velocity profile (Eq. 8.31) we obtain the gradient of the average velocity as Vc du r (1−n)n = − (1 − ) dr nR R which gives (b) The centerline velocity can be obtained from the average velocity and the assumption of a power-law velocity profile as follows. For this flow with Q 0.04 m3/s = = 5.09 m /s A π(0.1 m) 24 (6.04 m /s) 0.025 m (1−8.4)8.4 du =− 1 − dr 8.4(0.05 m) ( 0.05 m ) = −26.5/s Thus, du du = −(νρ) dr dr = −(1.004 × 10−6 m2/s) (998 kg /m3 ) (–26.5/s) = 0.0266 N/m2 τlam = −μ the Reynolds number is Re = (Ans)   so that V=   (c) From Eq. 8.4, which is valid for laminar or turbulent flow, the shear stress at r = 0.025 m is 64.8 N/m2 12 = 0.255 m /s u* = ( 998 kg /m3 ) 5(1.004 × 10−6 m2/s) 0.255 m /s = 1.97 × 10−5 m ≃ 0.02 mm V = 1.186V = 1.186 (5.09 m /s) (5.09 m /s) (0.1 m) VD = 5.07 × 105 = ν (1.004 × 10−6 m2/s) Thus, from Fig. 8.17, n = 8.4 so that Thus, the ratio of turbulent to laminar shear stress is given by 18.4 r u ≈ 1− ) Vc ( R To determine the centerline velocity, Vc, we must know the relationship between V (the average velocity) and Vc. This can be τ − τlam τturb 32.4 − 0.0266 = 1220 = = τlam τlam 0.0266 (Ans) COMMENT As expected, most of the shear stress at this loca- tion in the turbulent flow is due to the turbulent shear stress. The turbulent flow characteristics discussed in this section are not unique to turbulent flow in round pipes. Many of the characteristics introduced (i.e., the Reynolds stress, the viscous sublayer, the overlap layer, the outer layer, the general characteristics of the velocity profile, etc.) are found in other turbulent flows. In particular, turbulent pipe flow and turbulent flow past a solid wall (boundary layer flow) share many of these common traits. Such ideas are discussed more fully in Chapter 9. 8.4 Dimensional Analysis of Pipe Flow 427 8.3.4 Turbulence Modeling Although it is not yet possible to theoretically predict the random, irregular details of turbulent flows, it would be useful to be able to predict the time-averaged flow fields (pressure, velocity, etc.) directly from the basic governing equations. To this end one can time average the governing Navier–Stokes equations (Eqs. 6.31 and 6.127) to obtain equations for the average velocity and pressure. However, because the Navier–Stokes equations are nonlinear, the resulting time-averaged differential equations contain not only the desired average pressure and velocity as variables, but also averages of products of the fluctuations—terms of the type that one tried to eliminate by averaging the equations! For example, the Reynolds stress −ρu′υ′ (see Eq. 8.26) occurs in the time-averaged momentum equation. Thus, it is not possible to merely average the basic differential equations and obtain governing equations involving only the desired averaged quantities. This is the reason for the variety of ad hoc assumptions that have been proposed to provide “closure” to the equations governing the average flow. That is, the set of governing equations must be a complete or closed set of equations—the same number of equations as unknowns. Various attempts have been made to solve this closure problem (Refs. 1, 32). Such schemes involving the introduction of an eddy viscosity or the mixing length (as introduced in Section 8.3.2) are termed algebraic or zero-equation models. Other methods include the one-equation and twoequation models, where zero-, one- and two- are the number of differential equations that describe the model. Because of the complexity of these turbulence models,as well as the Navier-Stokes equations themselves, the only viable path to solving the equations is to apply numerical methods on a computer. This field, called Computational Fluid Dynamics, is discussed briefly in Appendix A. Turbulence modeling is an important topic. Although considerable progress has been made, much remains to be done in this area. 8.3.5 Chaos and Turbulence Chaos theory may eventually provide a deeper understanding of turbulence. VIDEO 8.4 V8.11 2D grid turbulence in soap film. Chaos theory is a relatively new branch of mathematical physics that may provide insight into the complex nature of turbulence. This method combines mathematics and numerical (computer) techniques to provide a new way to analyze certain problems. Chaos theory, which is quite complex and is currently under development, involves the behavior of nonlinear dynamical systems and their response to initial and boundary conditions. The flow of a viscous fluid, which is governed by the nonlinear Navier–Stokes equations (Eq. 6.127), may be such a system. To solve the Navier–Stokes equations for the velocity and pressure fields in a viscous flow, one must specify the particular flow geometry being considered (the boundary conditions) and the condition of the flow at some particular time (the initial conditions). If, as some researchers predict, the Navier–Stokes equations allow chaotic behavior, then the state of the flow at times after the initial time may be very sensitive to the initial conditions. A slight variation to the initial flow conditions may cause the flow at later times to be quite different than it would have been with the original, only slightly different initial conditions. When carried to the extreme, the flow may be “chaotic,” “random,” or perhaps (in current terminology), “turbulent.” The occurrence of such behavior would depend on the value of the Reynolds number. For example, it may be found that for sufficiently small Reynolds numbers the flow is not chaotic (i.e., it is laminar), while for large Reynolds numbers it is chaotic with turbulent characteristics. Thus, with the advancement of chaos theory it may be found that the numerous ad hoc turbulence ideas mentioned in previous sections (i.e., eddy viscosity, mixing length, law of the wall, etc.) may not be needed. It may be that chaos theory can provide the turbulence properties and structure directly from the governing equations. As of now we must wait until this exciting topic is developed further. The interested reader is encouraged to consult Ref. 4 for a general introduction to chaos or Ref. 33 for additional material. Dimensional Analysis of Pipe Flow As noted previously, turbulent flow can be a very complex, difficult topic—one that as yet has defied a rigorous theoretical treatment. Thus, most turbulent pipe flow analyses are based on experimental data and semi-empirical formulas. These data are expressed conveniently in dimensionless form. 428 Chapter 8 ■ Viscous Flow in Pipes It is often necessary to determine the head loss, hL, that occurs in a pipe flow so that the mechanical energy equation, Eq. 5.84 or 5.89, can be used in the analysis of pipe flow problems. As shown in Fig. 8.1, a typical pipe system usually consists of various lengths of straight pipe interspersed with various types of components (valves, elbows, etc.). The overall head loss for the pipe system consists of the head loss due to viscous effects in the straight pipes, termed the major loss and denoted hL major, and the head loss in the various pipe components, termed the minor loss and denoted hL minor. That is, hL = hL major + hL minor The head loss designations of “major” and “minor” do not necessarily reflect the relative importance of each type of loss. For a pipe system that contains many components and a relatively short length of pipe, the minor loss may actually be larger than the major loss. 8.4.1 Major Losses Δp = p1 – p2 (1) (2) ρ,μ D V ε 𝓵 Unlike laminar pipe flow characteristics, turbulent pipe flow characteristics depend on the fluid density and the pipe roughness. A dimensional analysis treatment of pipe flow provides the most convenient base from which to consider turbulent, fully developed pipe flow. An introduction to this topic was given in Section 8.3. As is discussed in Sections 8.2.1 and 8.2.4, the pressure drop and head loss in a pipe are dependent on the wall shear stress, τw, between the fluid and pipe surface. A fundamental difference between laminar and turbulent flow is that the shear stress for turbulent flow is a function of the density of the fluid, ρ. For laminar flow, the shear stress is independent of the density, leaving the viscosity, μ, as the only important fluid property. Thus, as indicated by the figure in the margin, the pressure drop, Δp, for steady, incompressible turbulent flow in a horizontal round pipe of diameter D can be written in functional form as Δp = F(V, D, ℓ, ε, μ, ρ) (8.32) where V is the average velocity, ℓ is the pipe length, and ε is a measure of the roughness of the pipe wall. It is clear that Δp should be a function of V, D, and ℓ. The dependence of Δp on the fluid properties μ and ρ is expected because of the dependence of τ on these parameters. Although the pressure drop for laminar pipe flow is found to be independent of the roughness of the pipe, it is necessary to include this parameter when considering turbulent flow. As is discussed in Section 8.3.3 and illustrated in Fig. 8.19, for turbulent flow there is a relatively thin viscous sublayer formed in the fluid near the pipe wall. In many instances this layer is very thin; δsD ⪡ 1, where δs is the sublayer thickness. If a typical wall roughness element protrudes sufficiently far into (or even through) this layer, the structure and properties of the viscous sublayer (along with Δp and τw) will be different than if the wall were smooth. Thus, for turbulent flow the pressure drop is expected to be a function of the wall roughness. For laminar flow there is no thin viscous layer— viscous effects are important across the entire pipe. Thus, relatively small roughness elements have completely negligible effects on laminar pipe flow. Of course, for pipes with very large wall “roughness” (εD ≳ 0.1), such as that in corrugated pipes, the flowrate may be a function of the “roughness.” We will consider only typical constant diameter pipes with relative roughnesses in the range 0 ≤ εD ≲ 0.05. Analysis of flow in corrugated pipes does not fit into the standard constant diameter pipe category, although experimental results for such pipes are available (Ref. 30). The list of parameters given in Eq. 8.32 is apparently a complete one. That is, experiments have shown that other parameters (such as surface tension, vapor pressure, etc.) do not affect the pressure drop for the conditions stated (steady, incompressible flow; round, horizontal pipe). Since there are seven variables (k = 7) that can be written in terms of the three reference dimensions MLT (r = 3), Eq. 8.32 can be written in dimensionless form in terms of k − r = 4 dimensionless groups. As was discussed in Section 7.9.1, one such representation is Δp 1 2 2 ρV ~ ρVD ℓ ε , , = ϕ( μ D D) This result differs from that used for laminar flow (see Eq. 8.17) in two ways. First, we have chosen to make the pressure dimensionless by dividing by the dynamic pressure, ρV2 2, rather than a characteristic viscous shear stress, μVD. This convention was chosen in recognition of the fact that the shear stress for turbulent flow is normally dominated by τturb, which is a stronger function of the 8.4 429 Velocity profile, u = u(y) R = D/2 y Dimensional Analysis of Pipe Flow δs x Viscous sublayer or ε δs ε ■ Figure 8.19 Flow in the viscous sublayer Rough wall near rough and smooth walls. Smooth wall density than it is of viscosity. Second, we have introduced two additional dimensionless parameters, the Reynolds number, Re = ρVDμ, and the relative roughness, εD, which are not present in the laminar formulation because the two parameters ρ and ε are not important in fully developed laminar pipe flow. As was done for laminar flow, the functional representation can be simplified by imposing the reasonable assumption that the pressure drop should be proportional to the pipe length. (Such a step is not within the realm of dimensional analysis. It is suggested by theory and experiments.) The only way that this can be true is if the ℓD dependence is factored out as Δp 1 2 2 ρV = ℓ ε ϕ(Re, ) D D As was discussed in Section 8.2.3, the quantity ΔpD(ℓρV2 2) is termed the friction factor, f. Thus, for a horizontal pipe Δp = f ℓ ρV2 D 2 (8.33) where f = ϕ(Re, ε D) For laminar fully developed flow, the value of f is simply f = 64Re, independent of εD. For turbulent flow, the functional dependence of the friction factor on the Reynolds number and the relative roughness, f = ϕ(Re, εD), is a rather complex one that cannot, as yet, be obtained from a theoretical analysis. The results are obtained from an exhaustive set of experiments and usually presented in terms of a curve-fitting formula or the equivalent graphical form. From Eq. 5.89 the head form of the mechanical energy equation for steady incompressible flow with no shaft work head is p1 p2 V21 V22 + z1 = + z2 + hL + α1 + α2 γ γ 2g 2g The major head loss in pipe flow is given in terms of the friction factor. where hL is the head loss between sections (1) and (2). With the assumption of a constant diameter (D1 = D2 so that V1 = V2 ), horizontal (z1 = z2 ) pipe with fully developed flow (α1 = α2 ), this becomes Δp = p1 − p2 = γhL, which can be combined with Eq. 8.33 to give hL major = f ℓ V2 D 2g (8.34) 430 Chapter 8 ■ Viscous Flow in Pipes Equation 8.34, called the Darcy–Weisbach equation, is valid for any fully developed, steady, incompressible pipe flow—whether the pipe is horizontal or on a hill. On the other hand, Eq. 8.33 is valid only for horizontal pipes. In general, with V1 = V2 the energy equation gives p1 − p2 = γ(z2 − z1 ) + γhL = γ(z2 − z1 ) + f (Photograph courtesy of CorrView) 0.08 Completely turbulent flow 0.06 0.04 ε D 0.1 0.01 0.001 0 0.0001 0.02 0.00001 f ℓ ρV2 D 2 Part of the pressure change is due to the elevation change and part is due to the head loss associated with frictional effects, which are given in terms of the friction factor, f. It is not easy to determine the functional dependence of the friction factor on the Reynolds number and relative roughness. Much of this information is a result of experiments conducted by J. Nikuradse in 1933 (Ref. 6) and amplified by many others since then. One difficulty lies in the determination of the roughness of the pipe. Nikuradse used artificially roughened pipes produced by gluing sand grains of known size onto pipe walls to produce pipes with sandpaper-type surfaces. The pressure drop needed to produce a desired flowrate was measured, and the data were converted into the friction factor for the corresponding Reynolds number and relative roughness. The tests were repeated numerous times for a wide range of Re and εD to determine the f = ϕ(Re, εD) dependence. In commercially available pipes the roughness is not as uniform and well defined as in the artificially roughened pipes used by Nikuradse. However, it is possible to obtain a measure of the effective relative roughness of typical pipes and thus to obtain the friction factor. Typical roughness values for various pipe surfaces are given in Table 8.1. Figure 8.20 shows the functional dependence of f on Re and εD and is called the Moody chart in honor of L. F. Moody, who, along with C. F. Colebrook, correlated the original data of Nikuradse in terms of the relative roughness of commercially available pipe materials. It should be noted that the values of εD do not necessarily correspond to the actual values obtained by a microscopic determination of the average height of the roughness of the surface. They do, however, provide the correct correlation for f = ϕ(Re, εD). It is important to observe that the values of relative roughness given pertain to new, clean pipes. After considerable use, most pipes (because of a buildup of corrosion or scale) may have a relative roughness that is considerably larger (perhaps by an order of magnitude) than that given. As shown by the upper figure in the margin, very old pipes may have enough scale buildup to not only alter the value of ε but also to change their effective diameter by a considerable amount. The following characteristics are observed from the data of Fig. 8.20. For laminar flow, f = 64Re, which is independent of relative roughness. For turbulent flows with very large Reynolds numbers, f = ϕ(εD), which, as shown by the figure in the margin, is independent of the Reynolds number. For such flows, commonly termed completely turbulent flow (or wholly turbulent flow), the laminar sublayer is so thin (its thickness decreases with increasing Re) that the surface roughness completely dominates the character of the flow near the wall. Hence, the pressure drop ■ Table 8.1 Equivalent Roughness for New Pipes [Adapted from Moody (Ref. 7) and Colebrook (Ref. 8)] Equivalent Roughness, 𝜺 Pipe Feet Millimeters Riveted steel Concrete Wood stave Cast iron Galvanized iron Commercial steel or wrought iron Drawn tubing Plastic, glass 0.003–0.03 0.001–0.01 0.0006–0.003 0.00085 0.0005 0.9–9.0 0.3–3.0 0.18–0.9 0.26 0.15 0.00015 0.000005 0.0 (smooth) 0.045 0.0015 0.0 (smooth) 8.4 431 Dimensional Analysis of Pipe Flow 0.1 0.09 Wholly turbulent flow 0.08 Laminar flow 0.05 0.07 Transition range 0.04 0.06 0.03 0.05 0.02 0.015 0.04 0.01 0.008 0.006 0.03 f 0.004 ε __ D 0.025 0.002 0.02 0.001 0.0008 0.0006 0.0004 0.015 0.0002 Smooth 0.0001 0.00005 0.01 0.009 0.008 2(103) 103 4 6 2(104) 8 104 4 6 2(105) 8 105 4 6 2(106) 8 106 4 6 2(107) 8 4 6 8 0.00001 107 ρ VD Re = _____ μ ■ Figure 8.20 Friction factor as a function of Reynolds number and relative roughness for round pipes—the Moody chart. (Data from Ref. 7 with permission.) For any pipe, even smooth ones, the head loss is not zero. required is a result of an inertia-dominated turbulent shear stress rather than the viscosity-dominated laminar shear stress normally found in the viscous sublayer. For flows with moderate values of Re, the friction factor is indeed dependent on both the Reynolds number and relative roughness— f = ϕ(Re, εD). The gap in the figure for which no values of f are given (the 2100 < Re < 4000 range) is a result of the fact that the flow in this transition range may be laminar or turbulent (or an unsteady mix of both) depending on the specific circumstances involved. Note that even for smooth pipes (ε = 0) the friction factor is not zero. That is, there is a head loss in any pipe, no matter how smooth the surface is made. This is a result of the no-slip boundary condition that requires any fluid to stick to any solid surface it flows over. There is always some microscopic surface roughness that produces the no-slip behavior (and thus f ≠ 0) on the molecular level, even when the roughness is considerably less than the viscous sublayer thickness. Such pipes are called hydraulically smooth. Various investigators have attempted to obtain an analytical expression for f = ϕ(Re, εD). Note that the Moody chart covers an extremely wide range in flow parameters. The nonlaminar region covers more than four orders of magnitude in Reynolds number—from Re = 4 × 103 to Re = 108. Obviously, for a given pipe and fluid, typical values of the average velocity do not cover this entire range. However, because of the large variety in pipes (D), fluids (ρ and μ), and velocities (V), such a wide range in Re is needed to accommodate nearly all applications of pipe flow. In many cases the particular pipe flow of interest is confined to a relatively small region of the Moody chart, and simple semiempirical expressions can be developed for those conditions. For example, a company that manufactures cast iron water pipes with diameters between 2 and 12 in. may use a simple equation valid for their conditions only. The Moody chart, on the other hand, is universally valid for all steady, fully developed, incompressible pipe flows. 432 Chapter 8 ■ Viscous Flow in Pipes The following equation from Colebrook is valid for the entire nonlaminar range of the Moody chart The turbulent portion of the Moody chart is represented by the Colebrook formula. 1 = −2.0 log (   √f εD 2.51 + 3.7 Re√f ) (8.35a) In fact, the Moody chart is a graphical representation of this equation, which is an empirical fit of the pipe flow pressure drop data. Equation 8.35 is called the Colebrook formula. A difficulty with its use is that it is implicit in the dependence of f. That is, for given conditions (Re and εD), it is not possible to solve for f without some sort of iterative scheme. With the use of modern computers and calculators, such calculations are not difficult. A word of caution is in order concerning the use of the Moody chart or the equivalent Colebrook formula. Because of various inherent inaccuracies involved (uncertainty in the relative roughness, uncertainty in the experimental data used to produce the Moody chart, etc.), the use of several place accuracy (i.e., more than about two significant digits) in pipe flow problems is usually not justified. As a rule of thumb, a 10% accuracy is the best expected. It is possible to obtain an equation that adequately approximates the ColebrookMoody chart relationship but does not require an iterative scheme. For example, the Haaland equation (Ref. 34), which is easier to use, is given by 1 √f = −1.8 log [( ε/D 1.11 6.9 + 3.7 ) Re ] (8.35b) where one can solve for f explicitly. EXAMPLE 8.5 Comparison of Laminar or Turbulent Pressure Drop GIVEN Air under standard conditions flows through a 4.0-mm-diameter drawn tubing with an average velocity of V = 50 ms. For such conditions the flow would normally be turbulent. However, if precautions are taken to eliminate disturbances to the flow (the entrance to the tube is very smooth, the air is dust free, the tube does not vibrate, etc.), it may be possible to maintain laminar flow. FIND (a) Determine the pressure drop in a 0.1-m section of the tube if the flow is laminar. (b) Repeat the calculations if the flow is turbulent. SOLUTION Under standard temperature and pressure conditions the density and viscosity are ρ = 1.23 kg /m3 and μ = 1.79 × 10−5 N ∙ s /m2. Thus, the Reynolds number is   COMMENT   Δp = (1.23 kg /m3 ) (50 m /s) (0.004 m) ρVD Re = = 13,700 = μ 1.79 × 10−5 N ∙ s /m2 = which would normally indicate turbulent flow. (a) If the flow were laminar, then f = 64Re = 6413,700 = 0.00467, and the pressure drop in a 0.1-m-long horizontal section of the pipe would be Δp = f   ℓ1 2 ρV D2 = (0.00467) (0.1 m) 1 (1.23 kg /m3 ) (50 m /s) 2 (0.004 m) 2 or Note that the same result is obtained from Eq. 8.8: 32μℓ V D2 32(1.79 × 10−5 N ∙ s /m2 ) (0.1 m) (50 m /s)     (0.004 m) 2 = 179 N/m2 (b) If the flow were turbulent, then f = ϕ(Re, ε/D), where from Table 8.1, ε = 0.0015 mm so that εD = 0.0015 mm 4.0 mm = 0.000375. From the Moody chart with Re = 1.37 × 104 and εD = 0.000375 we obtain f = 0.028. Thus, the pressure drop in this case would be approximately Δp = f   (0.1 m) 1 ℓ 1 2 ρV = (0.028) (1.23 kgm3 ) (50 ms) 2 D2 (0.004 m) 2 or Δp = 0.179 kPa (Ans) Δp = 1.076 kPa (Ans) 8.4 COMMENT A considerable savings in effort to force the fluid through the pipe could be realized (0.179 kPa rather than 1.076 kPa) if the flow could be maintained as laminar flow at this Reynolds number. In general this is very difficult to do, although laminar flow in pipes has been maintained up to Re ≈ 100,000 in rare instances. An alternate method to determine the friction factor for the turbulent flow would be to use the Colebrook formula, Eq. 8.35a. Thus, Dimensional Analysis of Pipe Flow 433 often-used equation, commonly referred to as the Blasius formula, for turbulent flow in smooth pipes (εD = 0) with Re < 105 is f= 0.316 Re14 For our case this gives f = 0.316(13,700) −0.25 = 0.0292 which is in agreement with the previous results. Note that the value of f is relatively insensitive to εD for this particular situa√f tion. Whether the tube was smooth glass (εD = 0) or the drawn or tubing (εD = 0.000375) would not make much difference in the pressure drop. For this flow, an increase in relative roughness −4 1.83 × 10 1 = −2.0 log(1.01 × 10−4 + (1) by a factor of 30 to εD = 0.0113 (equivalent to a commercial ) √f √f steel surface; see Table 8.1) would give f = 0.043. This would By using a root-finding technique on a computer or calculator, represent an increase in pressure drop and head loss by a factor the solution to Eq. 1 is determined to be f = 0.0291, in agree- of 0.0430.0291 = 1.48 compared with that for the original ment (within the accuracy of reading the graph) with the Moody drawn tubing. chart method of f = 0.028. The pressure drop of 1.076 kPa in a length of 0.1 m of pipe Equation 8.35b provides an alternate form to the Colebrook corresponds to a change in absolute pressure [assuming p = formula that can be used to solve for the friction factor directly. 101 kPa (abs) at x = 0] of approximately 1.076101 = 0.0107, or about 1%. Thus, the incompressible flow assumption on which 1 ε/D 1.11 6.9 0.000375 1.11 6.9 = −1.8 log [( + = −1.8 log + the above calculations (and all of the formulas in this chapter) are 4 [( 3.7 ) 3.7 ) Re ] 1.37 × 10 ] √f based is reasonable. However, if the pipe were 2 m long the pres= 0.0289 sure drop would be 21.5 kPa, approximately 20% of the original This agrees with the Colebrook formula and Moody chart values pressure. In this case the density would not be approximately constant along the pipe, and a compressible flow analysis would obtained above. Numerous other empirical formulas can be found in the litera- be needed. Such considerations are discussed in Chapter 11. ture (Ref. 5) for portions of the Moody chart. For example, an 1 = −2.0 log ( εD 2.51 0.000375 2.51 + = −2.0 log ( + 3.7 3.7 Re√f ) 1.37 × 104 √f ) 8.4.2 Minor Losses Losses due to pipe system components are given in terms of loss coefficients. As discussed in the previous section, the head loss in long, straight sections of pipe, the major losses, can be calculated by use of the friction factor obtained from either the Moody chart or the Colebrook equation. Most pipe systems, however, consist of considerably more than straight pipes. These additional components (valves, bends, tees, and the like) add to the overall head loss of the system. Such losses are generally termed minor losses, with the corresponding head loss denoted hL minor. In this section we indicate how to determine the various minor losses that commonly occur in pipe systems. The head loss associated with flow through a valve is a common minor loss. The purpose of a valve is to provide a means to regulate the flowrate. This is accomplished by changing the geometry of the system (i.e., closing or opening the valve alters the flow pattern through the valve), which in turn alters the losses associated with the flow through the valve. The flow resistance or head loss through the valve may be a significant portion of the resistance in the system. In fact, with the valve closed, the resistance to the flow is infinite—the fluid cannot flow. Such minor losses may be very important indeed. With the valve wide open the extra resistance due to the presence of the valve may or may not be negligible. The flow pattern through a typical component such as a valve is shown in Fig. 8.21. It is not difficult to realize that a theoretical analysis to predict the details of such flows to obtain the head loss for these components is not, as yet, possible. Thus, the head loss information for essentially all components is given in dimensionless form and based on experimental data. The most common method used to determine these head losses or pressure drops is to specify the loss coefficient, KL, which is defined as KL = Δp hL minor =1 2 2 (V 2g) 2 ρV 434 Chapter 8 ■ Viscous Flow in Pipes Q Q (b) (a) ■ Figure 8.21 Flow through a valve. so that 1 Δp = KL 2 ρV2 or hL minor = KL hL, minor hL, minor ~ V 2 V For most flows the loss coefficient is independent of the Reynolds number. V2 2g (8.36) The pressure drop across a component that has a loss coefficient of KL = 1 is equal to the dynamic pressure, ρV2 2. As shown by Eq. 8.36 and the figure in the margin, for a given value of KL the head loss is proportional to the square of the velocity. The actual value of KL is strongly dependent on the geometry of the component considered. It may also be dependent on the fluid properties. That is, KL = ϕ(geometry, Re) where Re = ρVDμ is the pipe Reynolds number. For many practical applications the Reynolds number is large enough so that the flow through the component is dominated by inertia effects, with viscous effects being of secondary importance. This is true because of the relatively large accelerations and decelerations experienced by the fluid as it flows along a rather curved, variable area (perhaps even torturous) path through the component (see Fig. 8.21). In a flow that is dominated by inertia effects rather than viscous effects, it is usually found that pressure drops and head losses correlate directly with the dynamic pressure. This is why the friction factor for very large Reynolds number, fully developed pipe flow is independent of the Reynolds number. The same condition is found to be true for flow through pipe components. Thus, in most cases of practical interest the loss coefficients for components are a function of geometry only, KL = ϕ(geometry). Minor losses are sometimes given in terms of an equivalent length, ℓeq. In this terminology, the head loss through a component is given in terms of the equivalent length of pipe that would produce the same head loss as the component. That is, hL minor = KL ℓeq V2 V2 =f 2g D 2g or ℓeq = KLD f 8.4 Dimensional Analysis of Pipe Flow (a) (b) (c) (d) 435 ■ Figure 8.22 Entrance flow conditions and loss coefficient (Data from Refs. 28, 29). (a) Reentrant, KL = 0.8, (b) sharp-edged, KL = 0.5, (c) slightly rounded, KL = 0.2 (see Fig. 8.24), (d) well-rounded, KL = 0.04 (see Fig. 8.24). Minor head losses are often a result of the dissipation of kinetic energy. VIDEO V8.12 Entrance/exit flows where D and f are based on the pipe containing the component. The head loss of the pipe system is the same as that produced in a straight pipe whose length is equal to the pipes of the original system plus the sum of the additional equivalent lengths of all of the components of the system. Most pipe flow analyses, including those in this book, use the loss coefficient method rather than the equivalent length method to determine the minor losses. Many pipe systems contain various transition sections in which the pipe diameter changes from one size to another. Such changes may occur abruptly or rather smoothly through some type of area change section. Any change in flow area contributes losses that are not accounted for by friction alone. By continuity, the area change causes a velocity change in incompressible flow. Thus the V2/2g terms do not cancel from the energy equation given in Eq. 5.84. The extreme cases involve flow into a pipe from a reservoir (an entrance) or out of a pipe into a reservoir (an exit). A fluid may flow from a reservoir into a pipe through any number of differently shaped entrance regions as are sketched in Fig. 8.22. Each geometry has an associated loss coefficient. A typical flow pattern for flow entering a pipe through a square-edged entrance is sketched in Fig. 8.23a. As was discussed in Chapter 3, a vena contracta region may result because the fluid cannot turn a sharp right-angle corner. The flow is said to separate from the sharp corner. The maximum velocity at section (2) is greater than that in the pipe at section (3), and the pressure there is lower. If this high-speed fluid could slow down efficiently, the kinetic energy could be converted into pressure (the Bernoulli effect), and the ideal pressure distribution indicated in Fig. 8.23b would result. The head loss for the entrance would be essentially zero. Such is not the case. Although a fluid may be accelerated very efficiently, it is very difficult to slow down (decelerate) a fluid efficiently. Thus, the extra kinetic energy of the fluid at section (2) is partially lost because of viscous dissipation, so that the pressure does not return to the ideal value. An entrance head loss (pressure drop) is produced as is indicated in Fig. 8.23b. The majority of this loss is due to inertia effects that are eventually dissipated by the shear stresses within the fluid. Only a small portion of the loss is due to the wall shear stress within the entrance region. The net effect is that the loss coefficient for a square-edged entrance is approximately KL = 0.50. One-half of a velocity head is lost as the fluid enters the pipe. If the pipe protrudes into the tank (a reentrant entrance) as is shown in Fig. 8.22a, the losses are even greater. An obvious way to reduce the entrance loss is to round the entrance region as is shown in Fig. 8.22c, thereby reducing or eliminating the vena contracta effect. Typical values for the loss coefficient 436 Chapter 8 ■ Viscous Flow in Pipes Separated flow Vena contracta (1) V1 = 0 (2) V > V 2 3 (3) x Flow separation at corner (a) p1 ρV 22 ____ 2 ρV 23 KL ____ p Actual x1 ρV 23 ____ 2 Ideal full recovery of kinetic energy x2 p3 x3 2 x (b) ■ Figure 8.23 Flow pattern and pressure distribution for a sharp-edged entrance. 0.6 r Sharp-edged entrance 0.5 0.4 D KL 0.3 0.2 0.1 0 0 0.05 0.1 D for entrances with various amounts of rounding of the lip are shown in Fig. 8.24. A significant reduction in KL can be obtained with only slight rounding. A head loss (the exit loss) is also produced when a fluid flows from a pipe into a tank as is shown in the figure in the margin. In these cases the entire kinetic energy of the exiting fluid (velocity V1) is dissipated through viscous effects as the stream of fluid mixes with the fluid in the tank and KL = 1.0 eventually comes to rest (V2 = 0). The exit loss from points (1) and (2) is therefore equivalent to one velocity head, or KL = 1. No matter what its shape, the loss coefficient for an exit is 1.0. A word of caution: Exit loss should be considered only when the downstream point in the analysis lies a sufficient distance beyond the exit to ensure that the exit kinetic energy has been dissipated. Consider calculation of flow through a fire hose and nozzle as shown in the margin. In the energy equation, the exit kinetic energy will be included as V222, so it must not be included in the losses. Losses also occur because of a change in pipe diameter as is shown in Figs. 8.25 and 8.26. The sharp-edged entrance and exit flows discussed in the previous paragraphs are limiting cases of this type of flow with either A1 A2 = ∞, or A1 A2 = 0, respectively. The loss coefficient for a sudden contraction, KL = hL(V22 2g), is a function of the area ratio, A2 A1, as is shown in (2) (1) The exit loss coefficient is always 1.0 regardless of shape. Nozzle Hose 2 1 ■ Figure 8.24 Entrance loss coefficient as a function of rounding of the inlet edge (Data from Ref. 9). r __ 8.4 Dimensional Analysis of Pipe Flow 437 0.6 A1 A2 V22 hL = KL ___ 2g 0.4 KL 0.2 0 0 0.2 0.4 0.6 0.8 1.0 A2/A1 ■ Figure 8.25 Loss coefficient for a sudden contraction (Ref. 10). 1.0 A1 0.8 A2 V12 hL = KL ___ 2g 0.6 KL 0.4 0.2 0 0 0.2 0.4 0.6 0.8 1.0 A1/A2 ■ Figure 8.26 Loss coefficient for a sudden expansion (Ref. 10). Fig. 8.25. The value of KL changes gradually from one extreme of a sharp-edged entrance (A2 A1 = 0 with KL = 0.50) to the other extreme of no area change (A2 A1 = 1 with KL = 0). In many ways, the flow in a sudden expansion is similar to exit flow. As is indicated in Fig. 8.27, the fluid leaves the smaller pipe and initially forms a jet-type structure as it enters the larger pipe. Within a few diameters downstream of the expansion, the jet becomes dispersed across the pipe, and fully developed flow becomes established again. In this process [between sections (2) and (3)] a portion of the kinetic energy of the fluid is dissipated as a result of viscous effects. A squareedged exit is the limiting case with A1 A2 = 0. A sudden expansion is one of the few components (perhaps the only one) for which the loss coefficient can be obtained by means of a simple analysis. To do this we consider the continuity and momentum equations for the control volume shown in Fig. 8.27 and the energy equation applied between (2) and (3). We assume that the flow is uniform at sections (1), (2), and (3) and the pressure is constant across the left side of the control volume (pa = pb = pc = p1 ). The resulting three governing equations (mass, momentum, and energy) are A1V1 = A3V3 p1A3 − p3A3 = ρA3V3 (V3 − V1 ) The loss coefficient for a sudden expansion can be theoretically calculated. and p1 p3 V 23 V 21 = + hL + + γ γ 2g 2g These can be combined to give the loss coefficient, KL = hL(V21 2g), as KL = (1 − A1 2 A2 ) where we have used the fact that A2 = A3. This result, plotted in Fig. 8.26, is in good agreement with experimental data. As with so many minor loss situations, it is not the viscous effects directly 438 Chapter 8 ■ Viscous Flow in Pipes Control volume a b V1 (1) V1 V3 c (2) (3) ■ Figure 8.27 Control volume used to calculate the loss coefficient for a sudden expansion. 1.4 1.2 1.0 KL ____________ V8.13 Separated flow in a diffuser (1 – A1/A2)2 VIDEO (i.e., the wall shear stress) that cause the loss. Rather, it is the dissipation of kinetic energy (another type of viscous effect) as the fluid decelerates inefficiently. The losses may be quite different if the contraction or expansion is gradual. Typical results for a conical diffuser with a given area ratio, A2 A1, are shown in Fig. 8.28. (A diffuser is a device shaped to decelerate a fluid.) Clearly the included angle of the diffuser, θ, is a very important parameter. For very small angles, the diffuser is excessively long and most of the head loss is due to the wall shear stress as in fully developed flow. For moderate or large angles, the flow separates from the walls and the losses are due mainly to a dissipation of the kinetic energy of the jet leaving the smaller diameter pipe. In fact, for moderate or large values of θ (i.e., θ > 35° for the case shown in Fig. 8.28), the conical diffuser is, perhaps unexpectedly, less efficient than a sharp-edged expansion which has KL = (1 − A1 A2 ) 2. There is an optimum angle (θ ≈ 8° for the case illustrated) for which the loss coefficient is a minimum. The relatively small value of θ for the minimum KL results in a long diffuser and is an indication of the fact that it is difficult to efficiently decelerate a fluid. It must be noted that the conditions indicated in Fig. 8.28 represent typical results only. Flow through a diffuser is very complicated and may be strongly dependent on the area ratio A2 A1, specific details of the geometry, and the Reynolds number. The data are often presented in terms of a pressure recovery coefficient, Cp = (p2 − p1 )(ρV21 2), which is the ratio of the static pressure rise across the diffuser to the inlet dynamic pressure. Considerable effort has gone into understanding this important topic (Refs. 11, 12). Flow in a conical contraction (a nozzle; reverse the flow direction shown in Fig. 8.28) is less complex than that in a conical expansion. Typical loss coefficients based on the downstream 0.8 V21 hL = KL ___ 2g 0.6 V1 0.4 A2 ___ fixed A1 V2 θ 0.2 0 0 30 60 90 θ , degrees 120 150 ■ Figure 8.28 Loss coefficient for a typical conical diffuser (Ref. 5). 180 8.4 Dimensional Analysis of Pipe Flow 439 1.0 b b Separated flow 0.8 a D a Secondary flow ℛ 0.6 90° KL Primary flow 0.4 ε __ D = 0.01 0.002 0.2 0.001 0 0 0 2 4 6 8 10 12 ℛ/D ■ Figure 8.29 Character of the flow in a 90° bend and the associated loss coefficient (Ref. 5). VIDEO V8.14 Car exhaust system (high-speed) velocity can be quite small, ranging from KL = 0.02 for θ = 30°, to KL = 0.07 for θ = 60°, for example. It is relatively easy to accelerate a fluid efficiently. Bends in pipes produce a greater head loss than if the pipe were straight. The losses are due to the separated region of flow near the inside of the bend (especially if the bend is sharp) and the swirling secondary flow that occurs because of the imbalance of centripetal forces as a result of the curvature of the pipe centerline. These effects and the associated values of KL for large Reynolds number flows through a 90° bend are shown in Fig. 8.29. The friction loss due to the axial length of the pipe bend must be calculated and added to that given by the loss coefficient of Fig. 8.29. For situations in which space is limited, a flow direction change is often accomplished by use of miter bends, as is shown in Fig. 8.30, rather than smooth bends. The considerable losses in such bends can be reduced by the use of carefully designed guide vanes that help direct the flow with less unwanted swirl and disturbances. Another important category of pipe system components is that of commercially available pipe fittings such as elbows, tees, reducers, valves, and filters. The values of KL for such components depend strongly on the shape of the component and only very weakly on the Reynolds number for typical large Re flows. Thus, the loss coefficient for a 90° elbow depends on whether the pipe joints Guide vanes Q Q Separated flow KL ≈ 1.1 KL ≈ 0.2 (a) ■ Figure 8.30 Character of the flow in a 90° (b) mitered bend and the associated loss coefficient: (a) without guide vanes, (b) with guide vanes. 440 Chapter 8 ■ Viscous Flow in Pipes ■ Table 8.2 Loss Coefficients for Pipe Components (hL = KL Component KL a. Elbows Regular 90°, flanged Regular 90°, threaded Long radius 90°, flanged Long radius 90°, threaded Long radius 45°, flanged Regular 45°, threaded 0.3 1.5 0.2 0.7 0.2 0.4 V2 (Data from Refs. 5, 10, 27) 2g ) b. 180° return bends 180° return bend, flanged 180° return bend, threaded 0.2 1.5 c. Tees Line flow, flanged Line flow, threaded Branch flow, flanged Branch flow, threaded 0.2 0.9 1.0 2.0 d. Union, threaded 0.08 V 90° elbow V 45° elbow V 180° return bend V Tee V Tee * e. Valves Globe, fully open Angle, fully open Gate, fully open 1 Gate, 4 closed 10 2 0.15 0.26 1 2.1 3 4 17 2 ∞ 0.05 5.5 Gate, 2 closed Gate, closed Swing check, forward flow Swing check, backward flow Ball valve, fully open 1 Ball valve, 3 closed 2 Ball valve, 3 closed V Union 210 * See Fig. 8.31 for typical valve geometry. Extensive tables are available for loss coefficients of standard pipe components. are threaded or flanged but is, within the accuracy of the data, fairly independent of the pipe diameter, flowrate, or fluid properties (the Reynolds number effect). Typical values of KL for such components are given in Table 8.2. These typical components are designed more for ease of manufacturing and costs than for reduction of the head losses that they produce. The flowrate from a faucet in a typical house is sufficient whether the value of KL for an elbow is the typical KL = 1.5, or it is reduced to KL = 0.2 by use of a more expensive long-radius, gradual bend (Fig. 8.29). Valves control the flowrate by providing a means to adjust the overall system loss coefficient to the desired value. When the valve is closed, the value of KL is infinite and no fluid flows. Opening of the valve reduces KL, producing the desired flowrate. Typical cross sections of various types of valves are shown in Fig. 8.31. Some valves (such as the conventional globe valve) are designed for general use, providing convenient control between the extremes of fully closed and 8.4 Dimensional Analysis of Pipe Flow 441 ■ Figure 8.31 Internal structure of various valves: (a) globe valve, (b) gate valve, (c) swing check valve, (d) stop check valve. (Courtesy of Crane Co., Fluid Handling Division.) V1 V3 >> V1 (1) (2) V2 = V1 ■ Figure 8.32 Head loss in a valve is due to dissipation of the kinetic energy of the large-velocity fluid near the valve seat. fully open. Others (such as a needle valve) are designed to provide very fine control of the flowrate. The check valve provides a diode type operation that allows fluid to flow in one direction only. Loss coefficients for typical valves are given in Table 8.2. As with many system components, the head loss in valves is mainly a result of the dissipation of kinetic energy of a high-speed portion of the flow. This high speed, V3, is illustrated in Fig. 8.32. With the exception of sudden expansions and exits, KL values throughout this section are representative of typical values based on experimental measurements. The uncertainty can range from about 10 percent to more than 50 percent. In addition, they apply to components in isolation with either uniform or fully developed flow upstream. If two components are close together, the resulting losses usually are greater than the sum of the losses of the components in isolation. 442 Chapter 8 ■ Viscous Flow in Pipes EXAMPLE 8.6 Minor Losses GIVEN The closed-circuit wind tunnel shown in Fig. E8.6a is a smaller version of that depicted in Fig. E8.6b in which air at standard conditions is to flow through the test section [between sections (5) and (6)] with a velocity of 200 ft/s. The flow is driven by a fan that essentially increases the static pressure by the amount p1 − p9 that is needed to overcome the head losses experienced by the fluid as it flows around the circuit. Estimate the value of p1 − p9 and the horsepower supplied to the fluid by the fan. FIND V5 = 200 ft/s (6) (5) (4) (7) Test section Flow-straightening screens (3) (8) ■ Figure E8.6a (© SMART BLADE GmbH.) Q (1) (9) (2) SOLUTION Fan The maximum velocity within the wind tunnel occurs in the test section (smallest area; see Table E8.6). Thus, the maximum Mach number of the flow is Ma5 = V5c5, where V5 = 200 ft /s and from Eq. 1.20 the speed of sound is c5 = (kRT5 ) 12 = {1.4 (1716 ft ∙ lb /slug ∙ °R) [(460 + 59)°R]}12 = 1117 ft /s. Thus, Ma5 = 2001117 = 0.179. As was indicated in Chapter 3 and discussed fully in Chapter 11, most flows can be considered as incompressible if the Mach number is less than about 0.3. Hence, we can use the incompressible formulas for this problem. The purpose of the fan in the wind tunnel is to provide the necessary energy to overcome the net head loss experienced by the air as it flows around the circuit. This can be found from the energy equation between points (1) and (9) as p9 V 29 p1 V 21 + z1 = + z9 + hL1 − 9 + + γ γ 2g 2g where hL1 − 9 is the total head loss from (1) to (9). With z1 = z9 and V1 = V9 this gives p9 p1 − = hL1 − 9 γ γ (1) Similarly, by writing the energy equation (Eq. 5.84) across the fan, from (9) to (1), we obtain p9 V 29 p1 V 21 + z9 + hp = + z1 + + γ γ 2g 2g where hp is the actual head rise supplied by the pump (fan) to the air. Again since z9 = z1 and V9 = V1 this, when combined with Eq. 1, becomes hp = (p1 − p9 ) = hL1 − 9 γ The actual power supplied to the air (horsepower, 𝒫a ) is obtained from the fan head by 𝒫a = γQhp = γA5V5hp = γA5V5hL1 − 9 (2) ■ Figure E8.6b Thus, the power that the fan must supply to the air depends on the head loss associated with the flow through the wind tunnel. To obtain a reasonable, approximate answer we make the following assumptions. We treat each of the four turning corners as a mitered bend with guide vanes so that from Fig. 8.31 KLcorner = 0.2. Thus, for each corner hLcorner = KL V2 V2 = 0.2 2g 2g where, because the flow is assumed incompressible, V = V5 A5A. The values of A and the corresponding velocities throughout the tunnel are given in Table E8.6. We also treat the enlarging sections from the end of the test section (6) to the beginning of the nozzle (4) as a conical diffuser with a loss coefficient of KLdif = 0.6. This value is larger than that of a well-designed diffuser (see Fig. 8.28, for example). Since the wind tunnel diffuser is interrupted by the four turning corners ■ Table E8.6 Location Area (ft2 ) Velocity (ft/s) 1 2 3 4 5 6 7 8 9 22.0 28.0 35.0 35.0 4.0 4.0 10.0 18.0 22.0 36.4 28.6 22.9 22.9 200.0 200.0 80.0 44.4 36.4 8.4 and the fan, it may not be possible to obtain a smaller value of KLdif for this situation. Thus, 200 V 26 V 26 = 0.6 2g 2g The loss coefficients for the conical nozzle between sections (4) and (5) and the flow-straightening screens are assumed to be KLnoz = 0.2 and KLscr = 4.0 (Ref. 13), respectively. We neglect the head loss in the test section, although this may not be practical with a test model present. Finally, we neglect major losses (friction) in the relatively short straight tunnel sections. Thus, the total head loss is hL1 − 9 = hLcorner7 + hLcorner8 + hLcorner2 + hLcorner3 + hLdif + hLnoz + hLscr hL1 − 9 = [0.2(V27 + V28 + V22 + V23 ) + 0.6 V 26 + 0.2V 25 + 4.0V 24]2g = [0.2(80.02 + 44.42 + 28.62 + 22.92 ) + 0.6(200) 2 + 0.2(200) 2 + 4.0(22.9) 2] ft2s2[2(32.2 fts2 )] or hL1 − 9 = 560 ft Hence, from Eq. 1 we obtain the pressure rise across the fan as     (Ans) From Eq. 2 we obtain the power added to the fluid as 𝒫a = (0.0765 lb /ft3 ) (4.0 ft2 ) (200 ft /s) (560 ft) = 34,300 ft ∙ lb /s     or 𝒫a = 34,300 ft ∙ lb /s = 62.3 hp 550 (ft ∙ lb /s)hp   150 100 (200 ft/s, 62.3 hp) 50 0 0 50 100 150 200 250 300 V5, ft/s ■ Figure E8.6c or p1 − p9 = γhL1 − 9 = (0.0765 lb /ft3 ) (560 ft) = 42.8 lb /ft2 = 0.298 psi 443 250 𝒫a, hp hLdif = KLdif Dimensional Analysis of Pipe Flow (Ans)   COMMENTS By repeating the calculations with various test section velocities, V5 , the results shown in Fig. E8.6c are obtained. Since the head loss varies as V 25 and the power varies as head loss times V5 , it follows that the power varies as the cube of the velocity. Thus, doubling the wind tunnel speed requires an eightfold increase in power. With a closed-return wind tunnel of this type, all of the power required to maintain the flow is dissipated through viscous effects, with the energy remaining within the closed tunnel. If heat transfer across the tunnel walls is negligible, the air temperature within the tunnel will increase in time. For steady-state operations of such tunnels, it is often necessary to provide some means of cooling to maintain the temperature at acceptable levels. It should be noted that the actual size of the motor that powers the fan must be greater than the calculated 62.3 hp because the fan is not 100% efficient. The power calculated above is that needed by the fluid to overcome losses in the tunnel, excluding those in the fan. If the fan were 60% efficient, it would require a shaft power of 𝒫 = 62.3 hp(0.60) = 104 hp to run the fan. Determination of fan (or pump) efficiencies can be a complex problem that depends on the specific geometry of the fan. Introductory material about fan performance is presented in Chapter 12; additional material can be found in various references (Refs. 14, 15, 16, for example). It should also be noted that the above results are only approximate. Clever, careful design of the various components (corners, diffuser, etc.) may lead to improved (i.e., lower) values of the various loss coefficients, and hence lower power requirements. Since hL is proportional to V 2, the components with the larger V tend to have the larger head loss. Thus, even though KL = 0.2 for each of the four corners, the head loss for corner (7) is (V7V3 ) 2 = (8022.9) 2 = 12.2 times greater than it is for corner (3). 8.4.3 Noncircular Conduits Many of the conduits that are used for conveying fluids are not circular in cross section. Although the details of the flows in such conduits depend on the exact cross-sectional shape, many round pipe results can be carried over, with slight modification, to flow in conduits of other shapes. Theoretical results can be obtained for fully developed laminar flow in noncircular ducts, although the detailed mathematics often becomes rather cumbersome. For an arbitrary cross section, as is shown in Fig. 8.33, the velocity profile is a function of both y and z [V = u(y, z)iˆ]. This means that the governing equation from which the velocity profile is obtained (either the Navier– Stokes equations of motion or a force balance equation similar to that used for circular pipes, Eq. 8.6) is a partial differential equation rather than an ordinary differential equation. Although the equation is linear (for fully developed flow the convective acceleration is zero), its solution is not 444 Chapter 8 ■ Viscous Flow in Pipes z z A = cross-sectional V = u(y,z) area x y P = perimeter of pipe Dh = 4A/P = hydraulic diameter (a) (b) ■ Figure 8.33 Noncircular duct. The hydraulic diameter is used for noncircular duct calculations. as straightforward as for round pipes. Typically the velocity profile is given in terms of an infinite series representation (Ref. 17). Practical, easy-to-use results can be obtained as follows. Regardless of the cross-sectional shape, there are no inertia effects in fully developed laminar pipe flow. Thus, the friction factor can be written as f = CReh, where the constant C depends on the particular shape of the duct, and Reh is the Reynolds number, Reh = ρVDh μ, based on the hydraulic diameter. The hydraulic diameter defined as Dh = 4AP is four times the ratio of the cross-sectional flow area divided by the wetted perimeter, P, of the pipe as is illustrated in Fig. 8.33. It represents a characteristic length that defines the size of a cross section of a specified shape. The factor of 4 is included in the definition of Dh so that for round pipes the diameter and hydraulic diameter are equal [Dh = 4AP = 4(πD2 4)(πD) = D]. The hydraulic diameter is also used in the definition of the friction factor, hL = f(ℓDh )V2 2g, and the relative roughness, εDh. The values of C = f Reh for laminar flow have been obtained from theory and/or experiment for various shapes. Typical values are given in Table 8.3 along with the hydraulic diameter. Note that the value of C is relatively insensitive to the shape of the conduit. Unless the cross section is very “thin” in some sense, the value of C is not too different from its circular pipe value, C = 64. Once the friction factor is obtained, the calculations for noncircular conduits are identical to those for round pipes. ■ Table 8.3 Friction Factors for Laminar Flow in Noncircular Ducts (Data from Ref. 18) Shape I. Concentric Annulus Dh = D2 – D1 Parameter C = f Reh D1D2 0.0001 0.01 0.1 0.6 1.00 71.8 80.1 89.4 95.6 96.0 ab 0 0.05 0.10 0.25 0.50 0.75 1.00 96.0 89.9 84.7 72.9 62.2 57.9 56.9 D1 D2 II. Rectangle 2ab Dh = _____ a+b a b 8.4 The Moody chart, developed for round pipes, can also be used for noncircular ducts. Dimensional Analysis of Pipe Flow 445 Calculations for fully developed turbulent flow in ducts of noncircular cross section are usually carried out by using the Moody chart data (or the Colebrook formula) for round pipes with the diameter replaced by the hydraulic diameter and the Reynolds number based on the hydraulic diameter. Such calculations are usually accurate to within about 15%. If greater accuracy is needed, a more detailed analysis based on the specific geometry of interest is needed. It is important to note that actual cross-sectional area is used and not (π/4)(Dh2) when relating average velocity and flowrate (i.e., Q = VA). E X AMPLE 8.7 Noncircular Conduit GIVEN Air at a temperature of 120 °F and standard pressure flows from a furnace through an 8-in.-diameter pipe with an average velocity of 10 ft/s. It then passes through a transition section similar to the one shown in Fig. E8.7 and into a square duct whose side is of length a. The pipe and duct surfaces are smooth (ε = 0). The head loss per foot is to be the same for the pipe and the duct. FIND Determine the duct size, a. SOLUTION V We first determine the head loss per foot for the pipe, hL ℓ = (fD) V22g, and then size the square duct to give the same value. For the given pressure and temperature we obtain (from Table B.3) ν = 1.89 × 10−4 ft2s so that 8 Re = (10 ft /s) ( 12 ft) VD = 35,300 = ν 1.89 × 10−4 ft2/s ■ With this Reynolds number and with εD = 0 we obtain the friction factor from Fig. 8.20 as f = 0.022 so that hL 0.022 (10 ft /s) 2 = 8 = 0.0512 2 ℓ ( 12 ft) 2(32.2 ft /s ) Thus, for the square duct we must have f V2s hL = = 0.0512 ℓ Dh 2g (1) where Dh = 4AP = 4a24a = a and 2 π 8 ft (10 ft /s) Q 4 ( 12 ) 3.49 Vs = = = 2 A a2 a (2) is the velocity in the duct. By combining Eqs. 1 and 2 we obtain We have three unknowns (a, f, and Reh) and three equations— Eqs. 3, 4, and either in graphical form the Moody chart (Fig. 8.20) or the Colebrook equation (Eq. 8.35a). If we use the Moody chart, we can use a trial-and-error solution as follows. As an initial attempt, assume the friction factor for the duct is the same as for the pipe. That is, assume f = 0.022. From Eq. 3 we obtain a = 0.606 ft, while from Eq. 4 we have Reh = 3.05 × 104. From Fig. 8.20, with this Reynolds number and the given smooth duct we obtain f = 0.023, which does not quite agree with the assumed value of f. Hence, we do not have the solution. We try again, using the latest calculated value of f = 0.023 as our guess. The calculations are repeated until the guessed value of f agrees with the value obtained from Fig. 8.20. The final result (after only two iterations) is f = 0.023, Reh = 3.03 × 104, and a = 0.611 ft = 7.34 in. (Ans) f (3.49a2 ) 2 0.0512 = a 2(32.2) COMMENTS Alternatively, we can use the Colebrook equa- a = 1.30 f 15 tion or the Haaland equation (rather than the Moody chart) to obtain the solution as follows. For a smooth pipe (ε/Dh = 0) the Colebrook equation, Eq. 8.35a, becomes or (3) where a is in feet. Similarly, the Reynolds number based on the hydraulic diameter is Reh = Figure E8.7 Vs D h (3.49a2 )a 1.85 × 104 = = −4 ν a 1.89 × 10 (4) εDh 2.51 + 3.7 √f Reh √f ) 2.51 = −2.0 log( Re √f ) 1 = −2.0 log( h (5) 446 Chapter 8 ■ Viscous Flow in Pipes where from Eq. 3, f = 0.269 a 5 (6) If we combine Eqs. 4, 5, and 6 and simplify, Eq. 7 is obtained for a. 1.928 a−5/2 = −2 log (2.62 × 10−4 a−3/2 ) (7) By using a root-finding technique on a computer or calculator, the solution to Eq. 7 is determined to be a = 0.614 ft, in agreement (given the accuracy of reading the Moody chart) with that obtained by the trial-and-error method given earlier. 8.5 Note that the length of the side of the equivalent square duct is aD = 7.348 = 0.918, or approximately 92% of the diameter of the equivalent duct. It can be shown that this value, 92%, is a very good approximation for any pipe flow—laminar or turbulent. The cross-sectional area of the duct (A = a2 = 53.9 in.2 ) is greater than that of the round pipe (A = πD24 = 50.3 in.2 ). Also, it takes less material to form the round pipe (perimeter = πD = 25.1 in.) than the square duct (perimeter = 4a = 29.4 in.). Circles are very efficient shapes. Pipe Flow Examples Pipe systems may contain a single pipe with components or multiple interconnected pipes. In the previous sections of this chapter, we discussed concepts concerning flow in pipes and ducts. The purpose of this section is to apply these ideas to the solutions of various practical problems. The application of the pertinent equations is straightforward, with rather simple calculations that give answers to problems of engineering importance. The main idea involved is to apply the energy equation between appropriate locations within the flow system, with the head loss written in terms of the friction factor and the minor loss coefficients. We will consider two classes of pipe systems: those containing a single pipe (whose length may be interrupted by various components), and those containing multiple pipes in parallel, series, or network configurations. THE WIDE WORLD OF FLUIDS New hi-tech fountains Ancient Egyptians used fountains in to produce coherent, laminar streams of water that look like glass their palaces for decorative and cooling purposes. Current use of rods flying through the air. By using fast-acting control valves in fountains continues but with a hi-tech flair. Although the basic a synchronized manner it is possible to produce mesmerizing fountain still consists of a typical pipe system (i.e., pump, pipe, three-dimensional patterns of water droplets. The possibilities regulating valve, nozzle, filter, and basin), recent use of computer- are nearly limitless. With the initial artistic design of the fountain controlled devices has led to the design of innovative foun- established, the initial engineering design (i.e., the capacity and tains with special effects. For example, by using several rows of pressure requirements of the nozzles and the size of the pipes multiple nozzles, it is possible to program and activate control and pumps) can be carried out. It is often necessary to modify the valves to produce water jets that resemble symbols, letters, or artistic and/or engineering aspects of the design in order to the time of day. Other fountains use specially designed nozzles obtain a functional, pleasing fountain. (See Problem 8.55.) 8.5.1 Single Pipes II: Q = ? III: D = ? I: Δ p = ? The nature of the solution process for pipe flow problems can depend strongly on which of the various parameters are independent parameters (the “given”) and which is the dependent parameter (the “determine”). The three most common types of problems are shown in Table 8.4 in terms of the parameters involved. We assume the pipe system is defined in terms of the length of pipe sections used and the number of elbows, bends, and valves needed to convey the fluid between the desired locations. In all instances we assume the fluid properties are given. In a Type I problem we specify the desired flowrate or average velocity and determine the necessary pressure difference or head loss. For example, if a flowrate of 2.0 galmin is required for a dishwasher that is connected to the water heater by a given pipe system as shown by the figure in the margin, what pressure is needed in the water heater? In a Type II problem we specify the applied driving pressure (or, alternatively, the head loss) and determine the flowrate. For example, how many galmin of hot water are supplied to the dishwasher if the pressure within the water heater is 60 psi and the pipe system details (length, diameter, roughness of the pipe; number of elbows; etc.) are specified? 8.5 447 Pipe Flow Examples ■ Table 8.4 Pipe Flow Types Pipe flow problems can be categorized by what parameters are given and what is to be calculated. Variable Type I a. Fluid Density Viscosity b. Pipe Diameter Length Roughness c. Flow Flowrate or Average Velocity d. Pressure Pressure Drop or Head Loss Type II Type III Given Given Given Given Given Given Given Given Given Given Given Given Determine Given Given Given Determine Given Determine Given Given In a Type III problem we specify the pressure drop and the flowrate and determine the diameter of the pipe needed. For example, what diameter of pipe is needed between the water heater and dishwasher if the pressure in the water heater is 60 psi (determined by the city water system) and the flowrate is to be not less than 2.0 galmin (determined by the manufacturer)? Several examples of these types of problems follow. E X AMPLE 8.8 Type I, Determine Pressure Drop KL = 2 based on GIVEN Water at 60 °F flows from the basement to the second floor through the 0.75-in. (0.0625-ft)-diameter copper pipe (a drawn tubing) at a rate of Q = 12.0 gal/min = 0.0267 ft3/s and exits through a faucet of diameter 0.50 in. as shown in Fig. E8.8a. FIND (a) 10 ft 0.75-in.-diameter copper pipe Determine the pressure at point (1) if (6) (4) (b) the only losses included are major losses, or (c) Q= 12.0 gal/min all losses are included. SOLUTION (5) 10 ft 0.50-in. diameter g (3) Threaded 90° elbows (1) Since the fluid velocity in the pipe is given by V1 = Q/A1 = Q(πD24) = (0.0267 ft3/s)[π(0.0625 ft) 24] = 8.70 ft /s, and the fluid properties are ρ = 1.94 slugs /ft3 and μ = 2.34 × 10−5 lb ∙ sft2 (see Table B.1), it follows that Re = ρVDμ = (1.94 slugs /ft3 ) (8.70 ft /s) (0.0625 ft)(2.34 × 10−5 lb ∙ s /ft2 ) = 45,000. Thus, the flow is turbulent. The governing equation for either case (a), (b), or (c) is the energy equation given by Eq. 8.21, (2) Wide open globe valve 10 ft 5 ft all losses are neglected, pipe velocity 10 ft (7) (8) 15 ft ■ Figure E8.8a   1   p1 = γz2 + 2ρ(V22 − V21 ) + γhL (1) where the head loss is different for each of the three cases. (a) p1 p2 V 21 V 22 + z1 = + z2 + hL + α1 + α2 γ γ 2g 2g If all losses are neglected (hL = 0), Eq. 1 gives p1 = (62.4 lb /ft3 ) (20 ft) 1.94 slugs /ft3 ft 2 ft 2 19.6 − 8.70 + [( ( s) s) ] 2 = (1248 + 299) lb /ft2 = 1547 lb /ft2     where z1 = 0, z2 = 20 ft, p2 = 0 (free jet), γ = ρg = 62.4 lb/ft3, and the outlet velocity is V2 = QA2 = (0.0267 ft3/s)[π(0.50 12) 2ft24] = 19.6 ft /s. We assume that the kinetic energy coefficients α1 and α2 are unity. This is reasonable because turbulent velocity profiles are nearly uniform across the pipe. Thus,     or p1 = 10.7 psi (Ans) 448 Chapter 8 ■ Viscous Flow in Pipes COMMENT Note that for this pressure drop, the amount due 30.5 psi to elevation change (the hydrostatic effect) is γ(z2 − z1 ) = 8.67 psi and the amount due to the increase in kinetic energy is ρ(V22 − V21 )2 = 2.07 psi. 30 27.1 (b) If the only losses included are the major losses, the head loss is Pressure loss 21.0 20.2 19.3 18.5 20 ℓ V21 D 2g p, psi hL = f (a) No losses (c) Including all losses 27.8 From Table 8.1 the roughness for a 0.75-in.-diameter copper pipe (drawn tubing) is ε = 0.000005 ft so that εD = 8 × 10−5. With this εD and the calculated Reynolds number (Re = 45,000), the value of f is obtained from the Moody chart as f = 0.0215. Note that the Colebrook equation (Eq. 8.35a) would give the same value of f. Hence, with the total length of the pipe as ℓ = (15 + 5 + 10 + 10 + 20) ft = 60 ft and the elevation and kinetic energy portions the same as for part (a), Eq. 1 gives 1 ℓ V 21 ρ(V 22 − V 21 ) + ρf 2 D 2 = (1248 + 299) lb /ft2 (8.70 ft /s) 2 60 ft + (1.94 slugs /ft3 ) (0.0215) ( ) 0.0625 ft 2 = (1248 + 299 + 1515) lb /ft2 = 3062 lb /ft2 10.7 9.93 10 Elevation and kinetic energy 0 0 6.37 3.09 6.37 4.84 2.07 2.07 10 20 30 p2 = 0 40 50 60 Distance along pipe from point (1), ft Location: (1) p1 = γz2 + 11.7 12.4 10.7 (3) (4) (5) (6) (7) (8) (2) ■ Figure E8.8b   an attaching reservoir where the kinetic energy is zero. Thus, by combining Eqs. 2 and 3 we obtain the entire pressure drop as     p1 = (21.3 + 9.17) psi = 30.5 psi   or p1 = 21.3 psi (Ans) If major and minor losses are included, Eq. 1 becomes 1 ℓ V 21 V2 ρ(V 22 − V 21 ) + fγ + ∑ ρKL 2 D 2g 2 or p1 = 21.3 psi + ∑ ρKL V2 2 (2) where the 21.3 psi contribution is due to elevation change, kinetic energy change, and major losses [part (b)], and the last term represents the sum of all of the minor losses. The loss coefficients of the components (KL = 1.5 for each elbow and KL = 10 for the wide-open globe valve) are given in Table 8.2 (except for the loss coefficient of the faucet, which is given in Fig. E8.8a as KL = 2). Thus, (8.70 ft) 2 V2 ∑ ρKL = (1.94 slugs /ft3 ) [10 + 4(1.5) + 2] 2 2 = 1321 lb /ft2     or 2 V ∑ ρKL = 9.17 psi 2 (3) Note that we did not include an entrance or exit loss because points (1) and (2) are located within the fluid streams, not within pressure distribution along the pipe is as illustrated in Fig. E8.8b for cases (a) and (c)—neglecting all losses or including all losses. Note that not all of the pressure drop, p1 − p2, is a “pressure loss.” The pressure change due to the elevation and velocity changes is completely reversible. The portion due to the major and minor losses is irreversible. This flow can be illustrated in terms of the energy line and hydraulic grade line concepts introduced in Section 3.7. As is shown in Fig. E8.8c, for case (a) there are no losses and the energy line (EL) is horizontal, one velocity head (V22g) above the hydraulic grade line (HGL), which is one pressure head (γz) above the pipe itself. For cases (b) or (c) the energy line is not horizontal. Each bit of friction in the pipe or loss in a component reduces the available energy, thereby lowering the energy line. 80 H, elevation to energy line, ft friction is approximately (21.3 − 10.7) psi = 10.6 psi. p1 = γz2 + This pressure drop calculated by including all losses should be the most realistic answer of the three cases considered. COMMENTS More detailed calculations will show that the COMMENT Of this pressure drop, the amount due to pipe (c) (Ans) Slope due to pipe friction Sharp drop due to component loss 60 Energy line including all losses, case (c) 40 20 0 Energy line with no losses, case (a) 0 10 20 30 40 50 Distance along pipe from point (1), ft ■ Figure E8.8c 60 8.5 Thus, for case (a) the total head remains constant throughout the flow with a value of H7 =     For case (c) the energy line starts at H1 = =   3 (62.4 lb /ft ) +   (8.70 ft /s) = p7 V 27 + z7 + γ 2g (9.93 × 144) lb /ft2   (62.4 lb /ft3 ) = 44.1 ft   + (8.70 ft /s) 2 2(32.2 ft /s2 ) + 20 ft The head loss per foot of pipe is the same all along the pipe. That is, V 21 p1 + z1 + γ 2g (30.5 × 144)lb /ft2 449 The elevation of the energy line can be calculated at any point along the pipe. For example, at point (7), 50 ft from point (1), p1 V 21 (1547 lb /ft2 ) (8.70 ft /s) 2 + z1 = + +0 + 3 γ 2g (62.4 lb /ft ) 2(32.2 ft /s2 ) = 26.0 ft. p2 p3 V33 V22 = + z2 = + z3 = … + + γ γ 2g 2g H= Pipe Flow Examples hL 0.0215(8.70 ft /s) 2 V2 =f = = 0.404 ft /ft ℓ 2gD 2(32.2 ft /s2 ) (0.0625 ft) 2 2(32.2 ft /s2 ) + 0 = 71.6 ft and falls to a final value of p2 V 22 (19.6 ft /s) 2 + z2 = 0 + + 20 ft + γ 2g 2(32.2 ft /s2 ) = 26.0 ft H2 = Thus, the energy line is a set of straight-line segments of the same slope separated by steps whose height equals the head loss of the minor component at that location. As is seen from Fig. E8.8c, the globe valve produces the largest of all the minor losses. Although the governing pipe flow equations are quite simple, they can provide very reasonable results for a variety of applications, as is shown in the next example. EX AMPLE 8.9 GIVEN Type I, Determine Head Loss A utility worker is fixing power lines in a manhole space at a busy intersection as shown in Fig. E8.9a. A major portion of car exhaust is carbon dioxide (CO2), which has a density of 3.55 × 10−3 slugs/ft3. Since the density of air at standard conditions is 2.38 × 10−3 slugs/ft3, the carbon dioxide will tend to settle into the bottom of the manhole space and displace the oxygen, posing a very relevant danger of suffocation to utility workers. To avoid suffocation, a fan is used to supply fresh air from the surface into the manhole space as shown in Fig. E8.9b. The air is routed through an 8-in.-diameter plastic hose that is 30 ft in length. FIND Determine the horsepower needed by the fan to over- come the head loss with a required flowrate of 600 cfm, when (a) neglecting minor losses and (b) including minor losses for one sharp-edged entrance, one exit, and one 90° miter bend (no guide vanes). (1) Fan V (2) ■ Figure E8.9a (© TIM MCCAIG/iStockphoto.) ■ Figure E8.9b 450 Chapter 8 ■ Viscous Flow in Pipes SOLUTION The purpose of the fan is to deliver air at a certain flowrate, which also requires the fan to overcome the net head loss within the pipe. This can be found from the energy equation between points (1) and (2) as p1 p2 V 21 V 22 + z1 + hp = + z2 + hL + + γ γ 2g 2g where hL is the total head loss from (1) to (2) and hp is the head rise supplied by the fan to the air. Points (1) and (2) are assumed to be sufficiently away from the entrance and exit of the plumbing so that p1 = p2 = V1 = V2 = 0. Also, since the fluid is air, the change in elevation is negligible. Thus, the energy equation reduces to hp = hL. (a) Neglecting minor losses it follows that hP = f ℓ V2 D 2g (1) where V is the velocity within the hose. Therefore, we must use the Moody chart to find the friction factor, f. Since Q 600 ft3/min = = 1719 ft/min = 28.6 ft /s π A 2 (8/12 ft) 4 it follows that (2.38 × 10 −3 slugs /ft3 ) (28.6 ft /s) ( 12 ft) ρVD = μ 3.74 × 10 −7 lb ∙ s/ft2 5 = 1.21 × 10 8 From Table 8.1 the equivalent roughness, ε, for plastic is 0.0. Therefore, ε/D = 0. Using this information with the Moody chart, we find the friction factor, f, to be 0.017. We now have enough information to solve Eq. 1. (28.6 ft /s)2 30 ft hP = (0.017) [ (8 /12) ft ][ 2(32.2 ft /s2) ] = 9.72 ft   The actual power the fan must supply to the air (horsepower, 𝒫a) can be calculated by 𝒫a = γQhP (2) 1 min (9.72 ft) 60 s ) ft ∙ lb s 1 hp ft ∙ lb = 1.35 × 10 −2 hp 𝒫a = 7.44 ( s 550 ft ∙ lb/s ) (28.6 ft /s) 2 V2 = (0.5 + 1.1 + 1.0) 2g 2(32.2 ft /s2 ) = 33.0 ft hL minor = ∑ KL hL minor The power required to account for both the major and the minor losses is 𝒫a = 1.35 × 10 −2 hp + (7.65 × 10 −2 lb /ft3 ) (600 ft3/min) 1 hp 1 min (33 ft) ( ×( 60 s ) 550 ft ∙ lb /s ) = 5.94 × 10 −2 hp (Ans) the horsepower required is more than four times what is needed for a straight pipe. It is easy to see that in this situation neglecting minor losses would provide an underestimation of the power required to pump 600 cfm of air. Obviously, “minor” losses are not always minor! The table that follows contains different power requirements for a few different pipe situations using the preceding flow properties. Re =   (b) For the additional minor losses, we need to include the entrance, exit, and bend. The loss coefficient for an entrance with sharp edges in 0.5, the loss coefficient for a 90° bend with no guide vanes is 1.1, and the loss coefficient for the exit is 1.0. Therefore, the total minor loss is given as COMMENTS Note that with the inclusion of minor losses, V= 𝒫a = (7.65 × 10−2 lb /ft3 ) (600 ft3/min) ( we need to be cautious in making this assumption with a relatively short overall pipe length. Pipe Setup Straight pipe Entrance and exit Entrance, exit, 1 miter bend Entrance, exit, 2 miter bends Entrance, exit, 4 miter bends hL minor (ft) hL major (ft) hP (ft) 9.72 𝒫a (hp) 0 9.72 0.014 19.1 9.72 28.8 0.040 33.0 9.72 42.7 0.059 47.0 9.72 56.7 0.079 74.9 9.72 84.7 0.118 = 7.44 (Ans) COMMENTS Typically, flexible hosing is used to route air into confined spaces. Minor losses were neglected in part (a), but Some pipe flow problems require a trial-and-error solution technique. Caution also needs to be taken when placing the fan outside the confined space so that clean air is supplied and hazardous gases emitted from the manhole are not circulated back into the confined space. There is also the question of whether to have air forced into the confined space or have the contaminated air exhausted from the space and naturally draw in fresh air. Pipe flow problems in which it is desired to determine the flowrate for a given set of conditions (Type II problems) often require trial-and-error (i.e., iterative) or numerical root-finding techniques. This is because it is necessary to know the value of the friction factor to carry out the calculations, but the friction factor is a function of the unknown velocity (flowrate) in terms of the Reynolds number. The solution procedure is indicated in Example 8.10. 8.5 E X AMPLE 8.10 Pipe Flow Examples 451 Type II, Determine Flowrate GIVEN The fan shown in Fig. E8.10a is to provide airflow 8-in.-diameter galvanized iron duct Fan through the spray booth and fume hood so that workers are protected from harmful vapors and aerosols while mixing chemicals within the hood. For proper operation of the hood, the flowrate is to be between 6 ft3/s and 12 ft3/s. With the initial setup the flowrate is 9 ft3/s, the loss coefficient for the system is 5, and the duct is short enough so that major losses are negligible. It is proposed that when the factory is remodeled the 8-in.-diameter galvanized iron duct will be 100 ft long and the total loss coefficient will be 10. Q (2) FIND Determine if the flowrate will be within the required 6 ft3/s to 12 ft3/s after the proposed remodeling. Assume that the head added to the air by the fan remains constant. (1) ■ Figure E8.10a SOLUTION We can determine the head that the fan adds to the air by considering the initial situation (i.e., before remodeling). To do this we write the energy equation between section (1) in the room and section (2) at the exit of the duct as shown in Fig. E8.10a. p1 p2 V 21 V 22 + z1 + hp = + z2 + hL + + γ γ 2g 2g hp = 2g + hL 2(32.2 ft/s2 ) + 51.6 ft = 61.9 ft The energy equation now must be solved with the new configuration after remodeling. Using the same assumptions as before gives the same reduced energy equation as shown in Eq. 2. With the increase in duct length to 100 ft the duct is no longer short enough to neglect major losses. Thus, hp = V 22 ℓ V2 V2 +f + ∑ KL 2g D 2g 2g = 1+f ℓ + ∑ KL D 2(32.2 ft/s2 ) (61.9 ft) = R 1+f( 100 ft + 10 8/12 ft ) 3990 B 11 + 150 f (3) The value of f is dependent on Re, which is dependent on V, an unknown. (2.38 × 10 −3 slugs/ft3 ) (V) ( 12 ft) ρVD = μ 3.74 × 10 −7 lb ∙ s/ft2 8 (2)   (25.8 ft/s) 2 R Re = The diameter of the duct is given as 8 in., so the velocity at the exit can be calculated from the flowrate, where V = Q/A = (9 ft3/s)[π(8 /12) 24] = 25.8 ft /s. For the original configuration  the duct is short enough to neglect major losses and only minor losses contribute to the total head loss. This head loss can  be found from hL, minor = Σ KLV 2(2g) = 5(25.8 ft/s) 2 [2(32.2 ft/s2 )] = 51.6 ft. With this information the simplified energy equation, Eq. 2, can now be solved to find the head added to the air by the fan. hp = 2ghp V= (1) Since we are dealing with air, we can assume any change in elevation is negligible. We can also assume the pressure inside the room and at the exit of the duct is equal to atmospheric pressure and the air in the room has zero velocity. Therefore, Eq. 1 reduces to V 22 where V2 = V and ∑ KL = 10. We can now rearrange and solve for the velocity in ft/s. or Re = 4240 V (4) where again V is in feet per second. Also, since ε/D = (0.0005 ft)(8/12 ft) = 0.00075 (see Table 8.1 for the value of ε), we know which particular curve of the Moody chart is pertinent to this flow. Thus, we have three relationships (Eqs. 3, 4, and ε/D = 0.00075 curve of Fig. 8.20) from which we can solve for the three unknowns, f, Re, and V. This is done easily by an iterative scheme as follows. It is usually simplest to assume a value of f, calculate V from Eq. 3, calculate Re from Eq. 4, and look up the new value of f in the Moody chart for this value of Re. If the assumed f and the new f do not agree, the assumed answer is not correct—we do not have the solution to the three equations. Although values of f, V, or Re could be assumed as starting values, it is usually simplest to assume a value of f because the correct value often lies on the relatively flat portion of the Moody chart for which f is quite insensitive to Re. 452 Chapter 8 ■ Viscous Flow in Pipes Thus, we assume f = 0.019, approximately the large Re limit for the given relative roughness. From Eq. 3 we obtain 8.5 3990 = 17.0 ft/s B 11 + 150(0.019) 8 and from Eq. 4 Re = 4240(17.0) = 72,100 KL = 5 7 KL = 8 6 Minimum flowrate 5.5 needed 5 Proposed remodeled system 0 25 50 75 𝓵, ft KL = 10 100 125 π 8 ft = 5.83 ft3/s (Ans) 4 )( 12 ) proposed remodeling will not provide enough airflow to protect workers from inhalation hazards while mixing chemicals within the hood. By repeating the calculations for different duct lengths and different total minor loss coefficients, the results shown in Fig. E8.10b are obtained, which give flowrate as a function of duct length for various values of the minor loss coefficient. It will be necessary to redesign the remodeled system (e.g., larger fan, ■ Figure E8.10b shorter ducting, larger-diameter duct) so that the flowrate will be within the acceptable range. In many companies, teams of occupational safety and health experts and engineers work together during the design phase of remodeling (or when a new operation is being planned) to consider and prevent potential negative impacts on workers’ safety and health in an effort called “Prevention through Design.” They also may be required to ensure that exhaust from such a system exits the building away from human activity and into an area where it will not be drawn back inside the facility. Type II, Determine Flowrate GIVEN The turbine shown in Fig. E8.11 extracts 50 hp from z1 = 90 ft (1) the water flowing through it. The 1-ft-diameter, 300-ft-long pipe is assumed to have a friction factor of 0.02. Minor losses are negligible. 300-ft-long, 1-ft-diameter pipe FIND Determine the flowrate through the pipe and turbine. Turbine f = 0.02 Free jet z2 = 0 (2) ■ Figure E8.11 SOLUTION The energy equation (Eq. 8.21) can be applied between the surface of the lake [point (1)] and the outlet of the pipe as p2 p1 V 21 V 22 + z1 = + z2 + hL + hT + + γ γ 2g 2g (300 ft) ℓ V2 V2 = 0.02 hL = f = 0.0932 V 2 ft D 2g (1 ft) 2(32.2 ft /s2 ) where V is in fts. Also, the turbine head is hT = (1) where hT is the turbine head, p1 = V1 = p2 = z2 = 0, z1 = 90 ft, and V2 = V, the fluid velocity in the pipe. The head loss is given by   150 2 COMMENT It is seen that operation of the system after the EXAMPLE 8.11 7.5 6.5 With this Re and ε/D, Fig. 8.20 gives f = 0.022, which is not equal to the assumed solution of f = 0.019 (although it is close!). We try again, this time with the newly obtained value of f = 0.022, which gives V = 16.7 ft/s and Re = 70,800. With these values, Fig. 8.20 gives f = 0.022, which agrees with the assumed value. Thus, the solution is V = 16.7 ft/s, or Q = VA = (16.7 ft/s) ( Original system 9 Q, ft3/s V= 9.5 = 𝒫a 𝒫a = γQ γ(π4)D2V (50 hp)[(550 ft ∙ lb /s)hp]   3 2 (62.4 lb /ft )[(π4) (1 ft) V]   = 561 ft V Thus, Eq. 1 can be written as 90 = 561 V2 + 0.0932V 2 + 2(32.2) V 8.5 or 0.109V − 90V + 561 = 0 3 (2) where V is in ft /s. The velocity of the water in the pipe is found as the solution of Eq. 2. Surprisingly, there are two real, positive roots: V = 6.58 ft /s or V = 24.9 ft /s. The third root is negative (V = −31.4 ft /s) and has no physical meaning for this flow. Thus, the two acceptable flowrates are Q= π 2 π D V = (1 ft) 2 (6.58 ft /s) = 5.17 ft3/s 4 4 (Ans) or Q= π (1 ft) 2 (24.9 ft /s) = 19.6 ft3/s 4 (Ans) COMMENTS Either of these two flowrates gives the same power, 𝒫a = γQhT. The reason for two possible solutions can be seen from the following. With the low flowrate (Q = 5.17 ft3/s), 453 Pipe Flow Examples we obtain the head loss and turbine head as hL = 4.04 ft and hT = 85.3 ft. Because of the relatively low velocity there is a relatively small head loss and, therefore, a large head available for the turbine. With the large flowrate (Q = 19.6 ft3/s), we find hL = 57.8 ft and hT = 22.5 ft. The high-speed flow in the pipe produces a relatively large loss due to friction, leaving a relatively small head for the turbine. However, in either case the product of the turbine head times the flowrate is the same. That is, the power extracted (𝒫a = γQhT) is identical for each case. Although either flowrate will allow the extraction of 50 hp from the water, the details of the design of the turbine itself will depend strongly on which flowrate is to be used. Such information can be found in Chapter 12 and various references about turbomachines (Refs. 14, 19, 20). If the friction factor were not given, the solution to the problem would be lengthier. A trial-and-error solution similar to that in Example 8.10 would be required, along with the solution of a cubic equation. In pipe flow problems for which the diameter is the unknown (Type III), an iterative or numerical root-finding technique is required. This is, again, because the friction factor is a function of the diameter—through both the Reynolds number and the relative roughness. Thus, neither Re = ρVDμ = 4ρQπμD nor εD are known unless D is known. Commercial pipe and tubing are only available in specific sizes. Pipe diameter calculation seldom corresponds exactly to a standard size. Specifying a standard pipe size is usually more economical than ordering a custom fabrication just to match your carefully computed diameter. Specifying the standard pipe size that has the next largest diameter is customary. This practice of rounding upward implies that there is little need to be extremely accurate in pipe sizing calculations. Be aware that the loss and/or flowrate does not equal the originally specified values when the larger pipe is used, so a recalculation of loss and flowrate may be necessary. Examples 8.12 and 8.13 illustrate Type III problems. EX AMPLE 8.12 Type III without Minor Losses, Determine Diameter GIVEN Air at standard temperature and pressure flows through a horizontal, galvanized iron pipe (ε = 0.0005 ft) at a rate of 2.0 ft3/s. The pressure drop is to be no more than 0.50 psi per 100 ft of pipe. FIND Determine the minimum pipe diameter. SOLUTION We assume the flow to be incompressible with ρ = 0.00238 slugs /ft3 and μ = 3.74 × 10−7 lb ∙ s /ft2. Note that if the pipe were too long, the pressure drop from one end to the other, p1 − p2, would not be small relative to the pressure at the beginning, and compressible flow considerations would be required. For example, a pipe length of 200 ft gives (p1 − p2 )p1 = [(0.50 psi)(100 ft)](200 ft)14.7 psia = 0.068 = 6.8%, which is probably small enough to justify the incompressible assumption. With z1 = z2 and V1 = V2 the energy equation (Eq. 8.21) becomes p1 = p 2 + f ℓ ρV2 D 2 (1) where V = QA = 4Q(πD2 ) = 4(2.0 ft3/s)πD2, or V= 2.55 D2 where D is in feet. Thus, with p1 − p2 = (0.5 lb /in.2 ) (144 in.2/ft2 ) and ℓ = 100 ft, Eq. 1 becomes   p1 − p2 = (0.5) (144) lb /ft2 (100 ft) 1 2.55 ft 2 =f (0.00238 slugs /ft3) ( 2 ) D 2 D s   or D = 0.404 f 15 (2) 454 Chapter 8 ■ Viscous Flow in Pipes where D is in feet. Also Re = ρVDμ = (0.00238 slugs /ft3 ) [(2.55D2 ) ft /s]D(3.74 × 10 −7 lb ∙ s /ft2 ), or 0.25   0.20 (2 ft3/s, 0.196 ft) (3) D, ft 1.62 × 104 Re = D and 0.15 0.10 ε 0.0005 = D D (4) Thus, we have four equations [Eqs. 2, 3, 4, and either the Moody chart, the Colebrook equation (8.35a) or the Haaland equation (8.35b)] and four unknowns (f, D, εD, and Re) from which the solution can be obtained by trial-and-error methods. If we use the Moody chart, it is probably easiest to assume a value of f, use Eqs. 2, 3, and 4 to calculate D, Re, and εD, and then compare the assumed f with that from the Moody chart. If they do not agree, try again. Thus, we assume f = 0.02, a typical value, and obtain D = 0.404(0.02) 15 = 0.185 ft, which gives εD = 0.00050.185 = 0.0027 and Re = 1.62 × 1040.185 = 8.76 × 104. From the Moody chart we obtain f = 0.027 for these values of εD and Re. Since this is not the same as our assumed value of f, we try again. With f = 0.027, we obtain D = 0.196 ft, εD = 0.0026, and Re = 8.27 × 104, which in turn give f = 0.027, in agreement with the assumed value. Thus, the diameter of the pipe should be D = 0.196 ft (Ans) COMMENT If we use the Colebrook equation (Eq. 8.35a) with εD = 0.00050.404 f 15 = 0.00124f 15 and Re = 1.62 × 1040.404 f 15 = 4.01 × 104f 15, we obtain 1 √f = −2.0 log( 0.05 0 0 0.5 1 2 1.5 2.5 3 Q, ft3/s ■ Figure E8.12 or 1 √f = −2.0 log( 3.35 × 10−4 f 15 + 6.26 × 10−5 f 310 ) By using a root-finding technique on a computer or calculator, the solution to this equation is determined to be f = 0.027, and hence D = 0.196 ft, in agreement with the Moody chart method. By repeating the calculations for various values of the flowrate, Q, the results shown in Fig. E8.12 are obtained. Although an increase in flowrate requires a larger diameter pipe (for the given pressure drop), the increase in diameter is minimal. For example, if the flowrate is doubled from 1 ft3/s to 2 ft3/s, the diameter increases from 0.151 ft to 0.196 ft. Note that these exact diameters are probably not available, so size upward to the closest diameter. εD 2.51 + 3.7 Re√f ) In the previous example we only had to consider major losses. In some instances the inclusion of major and minor losses can cause a slightly lengthier solution procedure, even though the governing equations are essentially the same. This is illustrated in Example 8.13. EXAMPLE 8.13 Type III with Minor Losses, Determine Diameter GIVEN Water at 60 °F (ν = 1.21 × 10−5 ft2/s, see Table 1.5) is to flow from reservoir A to reservoir B through a pipe of length 1700 ft and roughness 0.0005 ft at a rate of Q = 26 ft3/s as shown in Fig. E8.13a. The system contains a sharp-edged entrance and four flanged 45° elbows. Elevation z1 = 44 ft (1) Total length = 1700 ft A Elevation z2 = 0 (2) FIND Determine the pipe diameter needed. D B ■ Figure E8.13a 8.5 455 Pipe Flow Examples SOLUTION The energy equation (Eq. 8.21) can be applied between two points on the surfaces of the reservoirs (p1 = p2 = V1 = V2 = z2 = 0) as follows: p2 p1 V 21 V 22 + z1 = + z2 + hL + + γ γ 2g 2g COMMENTS Alternatively, we can use the Colebrook equa- V ℓ f + ∑ KL ) 2g ( D 2   (1) where V = QA = 4QπD2 = 4(26 ft3/s)πD2, or V= 33.1 D2 (2) is the velocity within the pipe. (Note that the units on V and D are ft /s and ft, respectively.) The loss coefficients are obtained from Table 8.2 and Figs. 8.22 and 8.24 as KLent = 0.5, KLelbow = 0.2, and KLexit = 1. Thus, Eq. 1 can be written as 44 ft = V2 1700 f + [4(0.2) + 0.5 + 1]} 2(32.2 ft /s2 ) { D or, when combined with Eq. 2 to eliminate V, f = 0.00152 D5 − 0.00135 D (3) To determine D we must know f, which is a function of Re and εD, where Re = [(33.1)D2]D VD 2.74 × 106 = = ν D 1.21 × 10−5 (4) and ε 0.0005 = D D (5) where D is in feet. Again, we have four equations (Eqs. 3, 4, 5, and the Moody chart or the Colebrook equation) for the four unknowns D, f, Re, and εD. Consider the solution by using the Moody chart. Although it is often easiest to assume a value of f and make calculations to determine if the assumed value is the correct one, with the inclusion of minor losses this may not be the simplest method. For example, if we assume f = 0.02 and calculate D from Eq. 3, we would have to solve a fifth-order equation. With only major losses (see Example 8.12), the term proportional to D in Eq. 3 is absent, and it is easy to solve for D if f is given. With both major and minor losses included, this solution for D (given f ) would require more complication. Thus, for this type of problem it is perhaps easier to assume a value of D, calculate the corresponding f from Eq. 3, and with the values of Re and εD determined from Eqs. 4 and 5, look up the value of f in the Moody chart (or the Colebrook equation). tion rather than the Moody chart to solve for D. This is easily done by using the Colebrook equation (Eq. 8.35a) with f as a function of D obtained from Eq. 3 and Re and εD as functions of D from Eqs. 4 and 5. The resulting single equation for D can be solved by using a root-finding technique on a computer or calculator to obtain D = 1 .63 ft. This agrees with the solution obtained using the Moody chart. By repeating the calculations for various pipe lengths, ℓ, the results shown in Fig. E8.13b are obtained. As the pipe length increases it is necessary, because of the increased friction, to increase the pipe diameter to maintain the same flowrate. It is interesting to attempt to solve this example if all losses are neglected so that Eq. 1 becomes z1 = 0. Clearly from Fig. E8.13a, z1 = 44 ft. Obviously something is wrong. A fluid cannot flow from one elevation, beginning with zero pressure and velocity, and end up at a lower elevation with zero pressure and velocity unless energy is removed (i.e., a head loss or a turbine) somewhere between the two locations. If the pipe is short (negligible friction) and the minor losses are negligible, there is still the kinetic energy of the fluid as it leaves the pipe and enters the reservoir. After the fluid meanders around in the reservoir for some time, this kinetic energy is lost and the fluid is stationary. No matter how small the viscosity is, the exit loss cannot be neglected. The same result can be seen if the energy equation is written from the free surface of the upstream tank to the exit plane of the pipe, at which point the kinetic energy is still available to the fluid. In either case the energy equation becomes z1 = V22g in agreement with the inviscid results of Chapter 3 (the Bernoulli equation). 1.8 1.6 (1700 ft, 1.63 ft) 1.4 1.2 D, ft or z1 = The solution is obtained when the two values of f are in agreement. A few rounds of calculation will reveal that the solution is given by D ≈ 1.63 ft (Ans) 1.0 0.8 0.6 0.4 0.2 0.0 0 500 ■ Figure E8.13b 1000 𝓵, ft 1500 2000 456 Chapter 8 ■ Viscous Flow in Pipes 8.5.2 Multiple Pipe Systems Trachea Lung Bronchiole In many pipe systems there is more than one pipe involved. The complex system of tubes in our lungs (beginning as shown by the figure in the margin, with the relatively large-diameter trachea and ending in tens of thousands of minute bronchioles after numerous branchings) and the maze of pipes in a city’s water distribution system are typical of such systems. The governing mechanisms for the flow in multiple pipe systems are the same as for the single pipe systems discussed in this chapter. However, because of the numerous unknowns involved, additional complexities may arise in solving for the flow in multiple pipe systems. Some of these complexities are discussed in this section. THE WIDE WORLD OF FLUIDS Deepwater pipeline Pipelines used to transport oil and gas fixed platforms and shore. The steel pipe used is 28 in. in diam- are commonplace. But south of New Orleans, in deep waters of eter with a wall thickness of 1 18 in. The thick-walled pipe is the Gulf of Mexico, a not-so-common multiple pipe system was needed to withstand the large external pressure which is about recently installed. The new so-called Mardi Gras system of 3250 psi at a depth of 7300 ft. The pipe was installed in 240-ft pipes is situated in water depths of 4300 to 7300 ft. It trans- sections from a vessel the size of a large football stadium. The ports oil and gas from five deepwater fields with the interesting deepwater pipeline system has a total length of more than names of Holstein, Mad Dog, Thunder Horse, Atlantis, and Na 450 miles and the capability of transporting more than 1 million Kika. The deepwater pipelines connect with lines at intermedi- barrels of oil per day and 1.5 billion cu ft of gas per day. (See ate water depths to transport the oil and gas to shallow-water Problem 8.120.) The simplest multiple pipe systems can be classified into series or parallel flows, as are shown in Fig. 8.34. The nomenclature is similar to that used in electrical circuits. Indeed, an analogy between fluid and electrical circuits is often made as follows. In a simple electrical circuit, there is a balance between the voltage (e), current (i), and resistance (R) as given by Ohm’s law: e = iR. In a fluid circuit there is a balance between the pressure drop (Δp), the flowrate or velocity (Q or V), and the flow resistance as given in terms of the friction factor and minor loss coefficients ( f and KL ). ~ ~ For a simple flow [Δp = f (ℓD)(ρV2 2)], it follows that Δp = Q2R, where R, a measure of the resistance to the flow, is proportional to f. The main differences between the solution methods used to solve electrical circuit problems and those for fluid circuit problems lie in the fact that Ohm’s law is a linear equation (doubling the voltage doubles the current), while the fluid equations are generally nonlinear (doubling the pressure drop does not double the flowrate unless the flow is laminar). Thus, although some of the standard electrical engineering methods can be carried over to help solve fluid mechanics problems, others cannot. One of the simplest multiple pipe systems is that containing pipes in series, as is shown in Fig. 8.34a. Every fluid particle that passes through the system passes through each of the pipes. Q A V1 (1) V2 D1 V3 D2 (2) (3) D3 Q B (a) A D1 B V1 Q1 (1) V2 D2 Q2 (2) V3 D3 (3) Q3 ■ Figure 8.34 (a) Series and (b) parallel pipe (b) systems. 8.5 Pipe Flow Examples 457 Thus, the flowrate (but not the velocity) is the same in each pipe, and the head loss from point A to point B is the sum of the head losses in each of the pipes. The governing equations can be written as follows: Q1 = Q2 = Q3 and hLA–B = hL1 + hL2 + hL3 Series and parallel pipe systems are often encountered. (used in conjunction with the energy equation) where the subscripts refer to each of the pipes. In general, the friction factors will be different for each pipe because the Reynolds numbers (Rei = ρViDi μ) and the relative roughnesses (εi Di ) will be different. Thus L V2 V2 hL = ∑ hL,major + ∑ hL,minor = ∑ f ( )( ) + ∑ KL( ) D 2g 2g     or in terms of flowrate hL = ( ∑ f L KL Q2 + ∑ A2 D A2 ) 2g If the flowrate is given, it is a straightforward calculation to determine the head loss or pressure drop (Type I problem). If the pressure drop is given and the flowrate is to be calculated (Type II problem), an iteration scheme is needed. In this situation none of the friction factors, fi, are known, so the calculations may involve more trial-and-error attempts than for corresponding single pipe systems. The same is true for problems in which the pipe diameter (or diameters) is to be determined (Type III problems). Another common multiple pipe system contains pipes in parallel, as is shown in Fig. 8.34b. In this system a fluid particle traveling from A to B may take any of the paths available, with the total flowrate equal to the sum of the flowrates in each pipe. However, by writing the energy equation between points A and B it is found that the head loss experienced by any fluid particle traveling between these locations is the same, independent of the path taken. Thus, the governing equations for parallel pipes are Q = Q1 + Q2 + Q3 and hL1 = hL2 = hL3 Again, the method of solution of these equations depends on what information is given and what is to be calculated. Another type of multiple pipe system called a loop is shown in Fig. 8.35. In this case the flowrate through pipe (1) equals the sum of the flowrates through pipes (2) and (3), or Q1 = Q2 + Q3. As can be seen by writing the energy equation between the surfaces of each reservoir, the head loss for pipe (2) must equal that for pipe (3), even though the pipe sizes and flowrates may be different for each. That is, pA pB V2A V2B + zA = + zB + hL1 + hL2 + + γ γ 2g 2g A B Node, N D1 Q1 V2 D2 Q2 (2) V1 V3 (1) (3) D3 ■ Figure 8.35 Multiple-pipe loop system. Q3 458 Chapter 8 ■ Viscous Flow in Pipes A B D1, 𝓵1 (1) D2, 𝓵2 (2) C D3, 𝓵3 (3) ■ Figure 8.36 A threereservoir system. for a fluid particle traveling through pipes (1) and (2), while pA pB V 2A V 2B + zA = + zB + hL1 + hL3 + + γ γ 2g 2g For some pipe systems, the direction of flow is not known a priori. for fluid that travels through pipes (1) and (3). These can be combined to give hL2 = hL3. This is a statement of the fact that fluid particles that travel through pipe (2) and particles that travel through pipe (3) all originate from common conditions at the junction (or node, N ) of the pipes and all end up at the same final conditions. The flow in a relatively simple looking multiple pipe system may be more complex than it appears initially. The branching system termed the three-reservoir problem shown in Fig. 8.36 is such a system. Three reservoirs at known elevations are connected together with three pipes of known properties (length, diameter, and roughness). The problem is to determine the flowrates into or out of the reservoirs. If valve (1) were closed, the fluid would flow from reservoir B to C, and the flowrate could be easily calculated. Similar calculations could be carried out if valves (2) or (3) were closed with the others open. With all valves open, however, it is not necessarily obvious which direction the fluid flows. For the conditions indicated in Fig. 8.36, it is clear that fluid flows from reservoir A because the other two reservoir levels are lower. Whether the fluid flows into or out of reservoir B depends on the elevation of reservoirs B and C and the properties (length, diameter, roughness) of the three pipes. In general, the flow direction is not obvious, and the solution process must include the determination of this direction. This is illustrated in Example 8.14. EXAMPLE 8.14 Three-Reservoir, Multiple Pipe System GIVEN Three reservoirs are connected by three pipes as are shown in Fig. E8.14. For simplicity we assume that the diameter of each pipe is 1 ft, the friction factor for each is 0.02, and because of the large length-to-diameter ratio, minor losses are negligible. A Elevation = 100 ft D1 = 1 ft (1) 𝓵 = 1000 ft 1 FIND Determine the flowrate into or out of each reservoir. SOLUTION It is not obvious which direction the fluid flows in pipe (2). However, we assume that it flows out of reservoir B, write the governing equations for this case, and check our assumption. The continuity equation requires that Q1 + Q2 = Q3, which, since the diameters are the same for each pipe, becomes simply V1 + V 2 = V 3 V 2C B Elevation = 20 ft (2) D3 = 1 ft 𝓵3 = 400 ft (3) C Elevation = 0 ft ■ Figure E8.14 (1) The energy equation for the fluid that flows from A to C in pipes (1) and (3) can be written as V 2A D2 = 1 ft 𝓵2 = 500 ft ℓ1 V 21 ℓ3 V 23 pA pC + zA = + zC + f1 + f3 + + γ γ 2g 2g D1 2g D3 2g By using the fact that pA = pC = VA = VC = zC = 0, this becomes zA = f1 ℓ3 V 23 ℓ1 V 21 + f3 D1 2g D3 2g 8.5 For the given conditions of this problem we obtain 100 ft = (2) where V1 and V3 are in ft /s. Similarly the energy equation for fluid flowing from B and C is pB pC V 2C ℓ3 V 23 ℓ2 V 22 V 2B + zB = + zC + f2 + f3 + + γ γ 2g 2g D2 2g D3 2g + 258 = (V2 + V3 ) 2 + 0.5V 22 = (V2 + √160 + 1.25V 22 ) 2 + 0.5V 22 or 2V2√160 + 1.25V 22 = 98 − 2.75V 22 (3) Equations 1, 2, and 3 (in terms of the three unknowns V1, V2, and V3) are the governing equations for this flow, provided the fluid flows from reservoir B. It turns out, however, that there is no solution for these equations with positive, real values of the velocities. Although these equations do not appear to be complicated, there is no simple way to solve them directly. Thus, a trial-and-error solution is suggested. This can be accomplished as follows. Assume a value of V1 > 0, calculate V3 from Eq. 2, and then V2 from Eq. 3. It is found that the resulting V1, V2, V3 trio does not satisfy Eq. 1 for any value of V1 assumed. There is no solution to Eqs. 1, 2, and 3 with real, positive values of V1, V2, and V3. Thus, our original assumption of flow out of reservoir B must be incorrect. To obtain the solution, assume the fluid flows into reservoirs B and C and out of A. For this case the continuity equation becomes or (4) Application of the energy equation between points A and B and A and C gives zA = zB + f1 ℓ1 V 21 ℓ2 V 22 + f2 D1 2g D2 2g and ℓ3 V 23 ℓ1 V 21 + f3 zA = zC + f1 D1 2g D3 2g Pipe network problems can be solved using node and loop concepts. (7) which, upon squaring both sides, can be written as V 42 − 460 V22 + 3748 = 0 By using the quadratic formula, we can solve for V22 to obtain either V 22 = 452 or V 22 = 8.30. Thus, either V2 = 21.3 ft /s or V2 = 2.88 ft /s. The value V2 = 21.3 ft /s is not a root of the original equations. It is an extra root introduced by squaring Eq. 7, which with V2 = 21.3 becomes “1140 = −1140.” Thus, V2 = 2.88 ft /s and from Eq. 5, V1 = 15.9 ft /s. The corresponding flowrates are π 2 π D1V1 = (1 ft) 2 (15.9 ft /s) 4 4 = 12.5 ft3/s from A (Ans) π 2 π D2V2 = (1 ft) 2 (2.88 ft /s) 4 4 3 = 2.26 ft /s into B (Ans) Q3 = Q1 − Q2 = (12.5 − 2.26) ft3/s = 10.2 ft3/s into C (Ans) Q1 = A1V1 = Q2 = A2V2 = Q1 = Q2 + Q3 V1 = V2 + V3 (6) V3 = √160 + 1.25V 22 For the given conditions this can be written as 64.4 = 322 = V 21 + 0.4 V 23 Thus, Eq. 5 can be written as ℓ3 V 23 ℓ2 V 22 + f3 D2 2g D3 2g 0.4V 23 (5) Equations 4, 5, and 6 can be solved as follows. By subtracting Eq. 5 from 6 we obtain or 0.5V 22 258 = V 21 + 0.5 V 22 and or zB = f2 459 which, with the given data, become 0.02 1 [(1000 ft)V 21 + (400 ft)V 23] 2(32.2 fts2 ) (1 ft) 322 = V 21 + 0.4V 23 Pipe Flow Examples and Note the slight differences in the governing equations depending on the direction of the flow in pipe (2)—compare Eqs. 1, 2, and 3 with Eqs. 4, 5, and 6. COMMENT If the friction factors were not given, a trial-and- error procedure similar to that needed for Type II problems (see Section 8.5.1) would be required. The ultimate in multiple pipe systems is a network of pipes such as that shown in Fig. 8.37. Networks like these often occur in city water distribution systems and other systems that may have multiple “inlets” and “outlets.” The direction of flow in the various pipes is by no means obvious— in fact, it may vary in time, depending on how the system is used from time to time. The solution for pipe network problems is often carried out by use of node and loop equations similar in many ways to that done in electrical circuits. For example, the continuity equation requires that for each node (the junction of two or more pipes) the net flowrate is zero. What flows into a node must flow out at the same rate. In addition, the net pressure difference completely around a loop 460 Chapter 8 ■ Viscous Flow in Pipes ■ Figure 8.37 A general pipe network. (starting at one location in a pipe and returning to that location) must be zero. By combining these ideas with the usual head loss and pipe flow equations, the flow throughout the entire network can be obtained. Of course, trial-and-error solutions are usually required because the direction of flow and the friction factors may not be known. The addition of minor losses such as valves adds further complexity. Hardy Cross developed a clever calculation technique to perform the iterations (Ref. 35). Such a solution procedure using matrix techniques is ideally suited for computer use (Refs. 21, 22). 8.6 Pipe Flowrate Measurement It is often necessary to determine experimentally the flowrate in a pipe. In Chapter 3 we introduced various types of flow-measuring devices (Venturi meter, nozzle meter, orifice meter, etc.) and discussed their operation under the assumption that viscous effects were not important. In this section we will indicate how to account for the ever-present viscous effects in these flowmeters. We will also indicate other types of commonly used flowmeters. 8.6.1 Pipe Flowrate Meters Orifice, nozzle, and Venturi meters involve the concept “high velocity gives low pressure.” Three of the most common devices used to measure the instantaneous flowrate in pipes are the orifice meter, the nozzle meter, and the Venturi meter. As was discussed in Section 3.6.3, each of these meters operates on the principle that a decrease in flow area in a pipe causes an increase in velocity that is accompanied by a decrease in pressure. Correlation of the pressure difference with the velocity provides a means of measuring the flowrate. In the absence of viscous effects and under the assumption of a horizontal pipe, application of the Bernoulli equation (Eq. 3.7) between points (1) and (2) shown in Fig. 8.38 gave Qideal = A2V2 = A2 2(p1 − p2 ) B ρ(1 − β4 ) (8.37) where β = D2 D1. Based on the results of the previous sections of this chapter, we anticipate that there is a head loss between (1) and (2) so that the governing equations become Q = A1V1 = A2V2 and p1 p2 V21 V22 = + hL + + γ γ 2g 2g The ideal situation has hL = 0 and results in Eq. 8.37. The difficulty in including the head loss is that there is no accurate expression for it. The net result is that empirical coefficients are used in Q D1 V1 V2 (2) (1) D2 ■ Figure 8.38 Typical pipe flowmeter geometry. 8.6 461 Pipe Flowrate Measurement Pressure taps Q D1 = D (1) d A0 A2 A1 (2) D2 ■ Figure 8.39 Typical orifice meter construction. An orifice discharge coefficient is used to account for nonideal effects. the flowrate equations to account for the complex real-world effects brought on by the nonzero viscosity. The coefficients are discussed in this section. A typical orifice meter is constructed by inserting between two flanges of a pipe a flat plate with a hole, as shown in Fig. 8.39. The pressure at point (2) within the vena contracta is less than that at point (1). Nonideal effects occur for two reasons. First, the vena contracta area, A2, is less than the area of the hole, Ao, by an unknown amount. Thus, A2 = Cc Ao, where Cc is the contraction coefficient (Cc < 1). Second, the swirling flow and turbulent motion near the orifice plate introduce a head loss that cannot be calculated theoretically. Thus, an orifice discharge coefficient, Co, is used to take these effects into account. That is, Q = CoQideal = Co Ao Q ~ p1 − p2 Q p1 − p2 2(p1 − p2 ) B ρ(1 − β4 ) (8.38) where Ao = πd2 4 is the area of the hole in the orifice plate. The value of Co is a function of β = dD and the Reynolds number, Re = ρVDμ, where V = QA1. Typical values of Co are given in Fig. 8.40. As shown by Eq. 8.38 and the figure in the margin, for a given value of Co, the flowrate is proportional to the square root of the pressure difference. Note that the value of Co depends on the specific construction of the orifice meter (i.e., the placement of the pressure taps, whether the orifice plate edge is square or beveled, etc.). Very precise conditions governing the construction of standard orifice meters have been established to provide the greatest accuracy possible (Refs. 23, 24). Another type of pipe flowmeter that is based on the same principles used in the orifice meter is the nozzle meter, three variations of which are shown in Fig. 8.41. This device uses a contoured nozzle (typically placed between flanges of pipe sections) rather than a simple (and less expensive) plate with a hole as in an orifice meter. The resulting flow pattern for the nozzle meter is closer to ideal than the orifice meter flow. There is only a slight vena contracta and the secondary flow 0.66 D D 0.64 D __ 2 V d Co 0.62 d = 0.7 β = __ D 0.6 0.5 0.4 0.2 0.60 0.58 4 10 105 106 Re = ρVD/μ 107 108 ■ Figure 8.40 Orifice meter discharge coefficient (Data from Ref. 24). 462 Chapter 8 ■ Viscous Flow in Pipes d D Pressure taps (a) (b) (c) ■ Figure 8.41 Typical nozzle meter construction. The nozzle meter has less energy dissipated than the orifice meter. separation is less severe, but there still are viscous effects. These are accounted for by use of the nozzle discharge coefficient, Cn, where Q = CnQideal = CnAn 2(p1 − p2 ) B ρ(1 − β4 ) (8.39) with An = πd2 4. As with the orifice meter, the value of Cn is a function of the diameter ratio, β = dD, and the Reynolds number, Re = ρVDμ. Typical values obtained from experiments are shown in Fig. 8.42. Again, precise values of Cn depend on the specific details of the nozzle design. Accepted standards have been adopted (Ref. 24). Note that Cn > Co; the nozzle meter has less energy dissipated than the orifice meter. The most precise and most expensive of the three obstruction-type flowmeters is the Venturi meter shown in Fig. 8.43 [G. B. Venturi (1746–1822)]. Although the operating principle for this device is the same as for the orifice or nozzle meters, the geometry of the Venturi meter is designed to reduce head losses to a minimum. This is accomplished by providing a relatively streamlined contraction (which eliminates separation ahead of the throat) and a very gradual expansion downstream of the throat (which eliminates separation in this decelerating portion of the device). Most of the head loss that occurs in a well-designed Venturi meter is due to friction losses along the walls rather than losses associated with separated flows and the inefficient mixing motion that accompanies such flow. 1.00 0.2 0.4 0.6 0.98 d = 0.8 β = __ D Cn 0.96 D 0.94 4 10 105 V 106 d 107 Re = ρVD/μ D d Q ■ Figure 8.43 Typical Venturi meter construction. 108 ■ Figure 8.42 Nozzle meter discharge coefficient (Data from Ref. 24). 8.6 Pipe Flowrate Measurement 463 Thus, the flowrate through a Venturi meter is given by Q = CυQideal = CAT The Venturi discharge coefficient is a function of the specific geometry of the meter. 2(p1 − p2 ) B ρ(1 − β4 ) (8.40) where AT = πd 2 4 is the throat area. The throat-to-pipe diameter ratio (β = dD), the Reynolds number, and the shape of the converging and diverging sections of the meter are among the parameters that affect the value of Cυ, the Venturi discharge coefficient. These values can typically be calculated from relations available for the particular type of Venturi meter used. For example, the standard American Society of Mechanical Engineers Venturi discharge coefficients can be determined with Cv = 1.0054 − 6.88 √Red   103 < Red < 5 × 105 Cv = 1.0054 − 4864.9   5 × 105 < Red < 8 × 105 Red Cv = 1.0054 − 0.185 361,239 4/5 1 −   Red > 8 × 105 [ Red ] Re1/5 d where Red = Vdd/v, d is the throat diameter, and Vd is the throat velocity (Ref. 36). Except for the lower range of Reynolds numbers, Cv is usually between 0.97 and 0.99. Again, the precise values of Cn, Co, and Cυ depend on the specific geometry of the devices used. Considerable information concerning the design, use, and installation of standard flowmeters can be found in various books (Refs. 23, 24, 25, 26, 31). E X AMPLE 8.15 Nozzle Flowmeter GIVEN Ethyl alcohol flows through a pipe of diameter D = 60 mm in a refinery. The pressure drop across the nozzle meter used to measure the flowrate is to be Δp = 4.0 kPa when the flowrate is Q = 0.003 m3s. FIND Determine the diameter, d, of the nozzle. SOLUTION From Table 1.6 the properties of ethyl alcohol are ρ = 789 kg /m3 and μ = 1.19 × 10−3 N ∙ s /m2. Thus,     Re = = ρVD 4ρQ = μ πDμ 4(789 kg /m3 ) (0.003 m3/s) π(0.06 m) (1.19 × 10−3 N ∙ s /m2 ) d = (1.20 × 10−3 ) 12 = 0.0346 m Hence, with an initial guess of d = 0.0346 m or β = dD = 0.03460.06 = 0.577, we obtain from Fig. 8.42 (using Re = 42,200) a value of Cn = 0.972. Clearly this does not agree with our initial assumption of Cn = 1.0. Thus, we do not have the solution to Eq. 1 and Fig. 8.42. Next we assume β = 0.577 and Cn = 0.972 and solve for d from Eq. 1 to obtain = 42,200 From Eq. 8.39 the flowrate through the nozzle is Q = 0.003 m3/s = Cn 2(4 × 103 N/m2 ) π 2 d 4 B 789 kg /m3 (1 − β4 ) or 1.20 × 10−3 = Cnd2 √1 − β4 d=( (1) where d is in meters. Note that β = dD = d0.06. Equation 1 and Fig. 8.42 represent two equations for the two unknowns d and Cn that must be solved by trial and error. As a first approximation we assume that the flow is ideal, or Cn = 1.0, so that Eq. 1 becomes d = (1.20 × 10−3 √1 − β4)12 In addition, for many cases 1 − β4 ≈ 1, so that an approximate value of d can be obtained from Eq. 2 as (2) 12 1.20 × 10−3 √1 − 0.5774 ) 0.972 or d = 0.0341 m. With the new value of β = 0.03410.060 = 0.568 and Re = 42,200, we obtain (from Fig. 8.42) Cn ≈ 0.972 in agreement with the assumed value. Thus, d = 34.1 mm (Ans) COMMENTS If numerous cases are to be investigated, it may be much easier to replace the discharge coefficient data of Chapter 8 ■ Viscous Flow in Pipes 60 d=D 50 D d 40 Mercury Carbon tet Water 20 Alcohol 30 Gasoline Fig. 8.42 by the equivalent equation, Cn = ϕ(β, Re), and use a computer to iterate for the answer. Such equations are available in the literature (Ref. 24). This would be similar to using the Colebrook equation rather than the Moody chart for pipe friction problems. By repeating the calculations, the nozzle diameters, d, needed for the same flowrate and pressure drop but with different fluids are shown in Fig. E8.15. The diameter is a function of the fluid viscosity because the nozzle coefficient, Cn, is a function of the Reynolds number (see Fig. 8.42). In addition, the diameter is a function of the density because of this Reynolds number effect and, perhaps more importantly, because the density is involved directly in the flowrate equation, Eq. 8.39. These factors all combine to produce the results shown in the figure. d, mm 464 10 0 ■ Figure E8.15 There are many types of flowmeters. Numerous other devices are used to measure the flowrate in pipes. Many of these devices use principles other than the high-speed/low-pressure concept of the orifice, nozzle, and Venturi meters. A quite common, accurate, and relatively inexpensive flowmeter is the rotameter, or variable area meter as is shown in Fig. 8.44. In this device a float is contained within a tapered, transparent metering tube that is attached vertically to the pipeline. As fluid flows through the meter (entering at the bottom), the float will rise within the tapered tube and reach an equilibrium height that is a function of the flowrate. This height corresponds to an equilibrium condition for which the net force on the float (buoyancy, float weight, fluid drag) is zero. A calibration scale in the tube provides the relationship between the float position and the flowrate. Another useful pipe flowrate meter is a turbine meter as is shown in Fig. 8.45. A small, freely rotating propeller or turbine within the turbine meter rotates with an angular velocity that is Q VIDEO V8.15 Rotameter Float at large end of tube indicates maximum flowrate Position of edge of float against scale gives flowrate reading Tapered metering tube Metering float is freely suspended in process fluid Float at narrow end of tube indicates minimum flowrate Q ■ Figure 8.44 Rotameter-type flowmeter. (Courtesy of ABB.) 8.6 Pipe Flowrate Measurement 465 Magnetic sensor O T U Flow out IN Flow in FLOW Turbine ■ Figure 8.45 Turbine-type flowmeter. (Courtesy of FTI Flow Technology, Inc.) a function of (nearly proportional to) the average fluid velocity in the pipe. This angular velocity is picked up magnetically and calibrated to provide a very accurate measure of the flowrate through the meter. Non-intrusive flowmeters have become very common. These flowmeters operate outside of the pipe, therefore they induce no head loss and are relatively simple to install. One typical example is the ultrasonic flowmeter (available with either transit time or Doppler methods). They use sound wave reflection to measure the flow moving past them. One of the more popular flowmeters, which is not completely non-intrusive but has negligible head loss, is the magnetic flowmeter. These meters apply a magnetic field across a pipe section; when the fluid (an electrical conductor) passes through the magnetic field, a voltage is induced that is proportional to the flowrate. 8.6.2 Volume Flowmeters Volume flowmeters measure volume rather than volume flowrate. VIDEO V8.16 Water meter In many instances it is necessary to know the amount (volume or mass) of fluid that has passed through a pipe during a given time period, rather than the instantaneous flowrate. For example, we are interested in how many gallons of gasoline are pumped into the tank in our car rather than the rate at which it flows into the tank. Numerous quantity-measuring devices provide such information. The nutating disk meter shown in Fig. 8.46 is widely used to measure the net amount of water used in domestic and commercial water systems as well as the amount of gasoline delivered to your gas tank. This meter contains only one essential moving part and is relatively inexpensive and accurate. Its operating principle is very simple, but it may be difficult to understand its operation without actually inspecting the device firsthand. The device consists of a metering chamber with spherical sides and conical top and bottom. A disk passes through a central sphere and divides the chamber Pin Metering Chamber Flow In Flow Out Casing Diaphram Sphere ■ Figure 8.46 Nutating disk flowmeter. (Courtesy of Badger Meter, Inc. © 2011.) 466 Chapter 8 ■ Viscous Flow in Pipes The nutating disk meter has only one moving part; the bellows meter has a complex set of moving parts. into two portions. The disk is constrained to be at an angle not normal to the axis of symmetry of the chamber. A radial plate (diaphragm) divides the chamber so that the entering fluid causes the disk to wobble (nutate), with fluid flowing alternately above or below the disk. The fluid exits the chamber after the disk has completed one wobble, which corresponds to a specific volume of fluid passing through the chamber. During each wobble of the disk, the pin attached to the tip of the center sphere, normal to the disk, completes one circle. The volume of fluid that has passed through the meter can be obtained by counting the number of revolutions completed. Another quantity-measuring device that is used for gas flow measurements is the bellows meter. It contains a set of bellows that alternately fill and empty as a result of the pressure of the gas and the motion of a set of inlet and outlet valves somewhat like the chambers and valves of a heart. For each cycle a known volume of gas passes through the meter. The common household natural gas meter (see photograph in the margin) is of this type. The nutating disk meter (water meter) is an example of extreme simplicity—one cleverly designed moving part. The bellows meter (gas meter), on the other hand, is relatively complex—it contains many moving, interconnected parts. This difference is dictated by the application involved. One measures a common, safe-to-handle, relatively high-pressure liquid, whereas the other measures a relatively dangerous, low-pressure gas. Each device does its intended job very well. There are numerous devices used to measure fluid flow, only a few of which have been discussed here. The reader is encouraged to review trustworthy Internet resources and the literature to gain familiarity with other useful, clever devices (Refs. 25, 26). 8.7 Chapter Summary and Study Guide laminar flow transitional flow turbulent flow entrance length fully developed flow wall shear stress Poiseuille’s law friction factor turbulent shear stress major loss minor loss relative roughness Moody chart Colebrook formula loss coefficient hydraulic diameter multiple pipe systems orifice meter nozzle meter Venturi meter This chapter discussed the flow of a viscous fluid in a pipe. General characteristics of laminar, turbulent, fully developed, and entrance flows are considered. Poiseuille’s equation is obtained to describe the relationship among the various parameters for fully developed laminar flow. Various characteristics of turbulent pipe flow are introduced and contrasted to laminar flow. It is shown that the head loss for laminar or turbulent pipe flow can be written in terms of the friction factor (for major losses) and the loss coefficients (for minor losses). In general, the friction factor is obtained from the Moody chart or the Colebrook formula and is a function of the Reynolds number and the relative roughness. The minor loss coefficients are a function of the flow geometry for each system component. Analysis of noncircular conduits is carried out by use of the hydraulic diameter concept. Various examples involving flow in single pipe systems and flow in multiple pipe systems are presented. The inclusion of viscous effects and losses in the analysis of orifice, nozzle, and Venturi flowmeters is discussed. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed you should be able to ■ ■ ■ ■ ■ ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. determine which of the following types of flow will occur: entrance flow or fully developed flow; laminar flow or turbulent flow. use the Poiseuille equation in appropriate situations and understand its limitations. explain the main properties of turbulent pipe flow and how they are different from or similar to laminar pipe flow. use the Moody chart and the Colebrook equation to determine major losses in pipe systems. use minor loss coefficients to determine minor losses in pipe systems. determine the head loss in noncircular conduits. incorporate major and minor losses into the energy equation to solve a variety of pipe flow problems, including Type I problems (determine the pressure drop or head loss), Type II problems (determine the flow rate), and Type III problems (determine the pipe diameter). References ■ ■ 467 solve problems involving multiple pipe systems. determine the flowrate through orifice, nozzle, and Venturi flowmeters as a function of the pressure drop across the meter. Some of the important equations in this chapter are: Entrance length Pressure drop for fully developed laminar pipe flow Velocity profile for fully developed laminar pipe flow Volume flowrate for fully developed laminar horizontal pipe flow Friction factor for fully developed laminar pipe flow Energy equation for incompressible, steady flow Pressure drop for a horizontal pipe Head loss due to major losses Colebrook formula Haaland formula Head loss due to minor losses Volume flowrate for orifice, nozzle, or Venturi meter ℓe = 0.06 Re for laminar flow D (8.1) ℓe = 4.4 (Re) 16 for turbulent flow D (8.2) 4ℓτw D ΔpD2 2r 2 2r 2 u(r) = ( 1 − = V 1 − c[ (D) ] (D) ] 16μℓ )[ Δp = Q= πD4 Δp 128μℓ 64 Re p1 p2 V21 V22 + z1 = + z2 + hL + α1 + α2 γ γ 2g 2g hL major = f 1 √f 1 √f     (8.7) (8.9) f= Δp = f (8.5) (8.19) (8.21) ℓ ρV 2 D 2 (8.33) ℓ V2 D 2g (8.34) = −2.0 log ( εD 2.51 + 3.7 Re√f ) = −1.8 log [( hL minor = KL εD 1.11 6.9 + 3.7 ) Re ] V2 2g Q = Ci Ai 2(p1 − p2) B ρ(1 − β4) (8.35a) (8.35b) (8.36) (8.38, 8.39, 8.40) References 1. Bernard, P. S., and Wallace J. M., Turbulent Flow—Analysis, Measurement, and Prediction, Wiley, New York, 2002. 2. Panton, R. L., Incompressible Flow, 3rd Ed., Wiley, New York, 2005. 3. Schlichting, H., Boundary Layer Theory, 8th Ed., McGraw-Hill, New York, 2000. 4. Gleick, J., Chaos: Making a New Science, Viking Penguin, New York, 1987. 5. White, F. M., Fluid Mechanics, 6th Ed., McGraw-Hill, New York, 2008. 6. Nikuradse, J., “Stomungsgesetz in Rauhen Rohren,” VDI-Forschungsch, No. 361, 1933; or see NACA Tech Memo 1922. 7. Moody, L. F., “Friction Factors for Pipe Flow,” Transactions of the ASME, Vol. 66, 1944. 8. Colebrook, C. F., “Turbulent Flow in Pipes with Particular Reference to the Transition Between the Smooth and Rough Pipe Laws,” Journal of the Institute of Civil Engineers London, Vol. 11, 1939. 9. ASHRAE Handbook of Fundamentals, ASHRAE, Atlanta, 1981. 468 Chapter 8 ■ Viscous Flow in Pipes 10. Streeter, V. L., ed., Handbook of Fluid Dynamics, McGraw-Hill, New York, 1961. 11. Sovran, G., and Klomp, E. D., “Experimentally Determined Optimum Geometries for Rectilinear Diffusers with Rectangular, Conical, or Annular Cross Sections,” in Fluid Mechanics of Internal Flow, Sovran, G., ed., Elsevier, Amsterdam, 1967. 12. Runstadler, P. W., “Diffuser Data Book,” Technical Note 186, Creare, Inc., Hanover, N.H., 1975. 13. Laws, E. M., and Livesey, J. L., “Flow Through Screens,” Annual Review of Fluid Mechanics, Vol. 10, Annual Reviews, Inc., Palo Alto, Calif., 1978. 14. Balje, O. E., Turbomachines: A Guide to Design, Selection and Theory, Wiley, New York, 1981. 15. Wallis, R. A., Axial Flow Fans and Ducts, Wiley, New York, 1983. 16. Karassick, I. J. et al., Pump Handbook, 2nd Ed., McGraw-Hill, New York, 1985. 17. White, F. M., Viscous Fluid Flow, 3rd Ed., McGraw-Hill, New York, 2006. 18. Olson, R. M., Essentials of Engineering Fluid Mechanics, 4th Ed., Harper & Row, New York, 1980. 19. Dixon, S. L., Fluid Mechanics of Turbomachinery, 3rd Ed., Pergamon, Oxford, UK, 1978. 20. Finnemore, E. J., and Franzini, J. R., Fluid Mechanics, 10th Ed., McGraw-Hill, New York, 2002. 21. Streeter, V. L., and Wylie, E. B., Fluid Mechanics, 8th Ed., McGraw-Hill, New York, 1985. 22. Jeppson, R. W., Analysis of Flow in Pipe Networks, Ann Arbor Science Publishers, Ann Arbor, Mich., 1976. 23. Bean, H. S., ed., Fluid Meters: Their Theory and Application, 6th Ed., American Society of Mechanical Engineers, New York, 1971. 24. “Measurement of Fluid Flow by Means of Orifice Plates, Nozzles, and Venturi Tubes Inserted in Circular Cross Section Conduits Running Full,” Int. Organ. Stand. Rep. DIS-5167, Geneva, 1976. 25. Goldstein, R. J., ed., Flow Mechanics Measurements, 2nd Ed., Taylor and Francis, Philadelphia, 1996. 26. Benedict, R. P., Measurement of Temperature, Pressure, and Flow, 2nd Ed., Wiley, New York, 1977. 27. Hydraulic Institute, Engineering Data Book, 1st Ed., Cleveland Hydraulic Institute, 1979. 28. Harris, C. W., University of Washington Engineering Experimental Station Bulletin, 48, 1928. 29. Hamilton, J. B., University of Washington Engineering Experimental Station Bulletin, 51, 1929. 30. Miller, D. S., Internal Flow Systems, 2nd Ed., BHRA, Cranfield, UK, 1990. 31. Spitzer, D. W., ed., Flow Measurement: Practical Guides for Measurement and Control, Instrument Society of America, Research Triangle Park, N.C., 1991. 32. Wilcox, D. C., Turbulence Modeling for CFD, DCW Industries, Inc., La Canada, California, 1994. 33. Mullin, T., ed., The Nature of Chaos, Oxford University Press, Oxford, UK, 1993. 34. Haaland, S. E., “Simple and Explicit Formulas for the Friction-Factor in Turbulent Pipe Flow,” Transactions of the ASME, Journal of Fluids Engineering, Vol. 105, 1983. 35. Cross, H., “Analysis of Flow in Networks of Conduits or Conductors,” University of Illinois Bulletin No. 286, November 1936. 36. Performance Test Code 19.5—Flow Measurement, American Society of Mechanical Engineering, New York, 2013. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are not unique answers to these problems. Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. 469 Problems Section 8.1 General Characteristics of Pipe Flow †8.1 Under normal circumstances is the airflow though your trachea (your windpipe) laminar or turbulent? List all assumptions and show all calculations. Rainwater runoff from a parking lot flows through a 8.2 3-ft-diameter pipe, completely filling it. Whether flow in a pipe is laminar or turbulent depends on the value of the Reynolds number. (See Video V8.2.) Would you expect the flow to be laminar or turbulent? Support your answer with appropriate calculations. 8.3 Blue and yellow streams of paint at 60 °F (each with a density of 1.6 slugs/ft3 and a viscosity 1000 times greater than water) enter a pipe with an average velocity of 4 ft/s as shown in Fig. P8.3. Would you expect the paint to exit the pipe as green paint or separate streams of blue and yellow paint? Explain. Repeat the problem if the paint were “thinned” so that it is only 10 times more viscous than water. Assume the density remains the same. Yellow Green? 2 in. Splitter Blue 25 ft ■ Figure P8.3 8.4 Air at 200 °F flows at standard atmospheric pressure in a pipe at a rate of 0.08 lb/s. Determine the minimum diameter allowed if the flow is to be laminar. 8.5 To cool a given room it is necessary to supply 4 ft3/s of air through an 8-in.-diameter pipe. Approximately how long is the entrance length in this pipe? *8.6 The flow of water in a 3-mm-diameter pipe is to remain laminar. Plot a graph of the maximum flowrate allowed as a function of temperature for 0 < T < 100 °C. 8.7 The pressure distribution measured along a straight, horizontal portion of a 50-mm-diameter pipe attached to a tank is shown in the table below. Approximately how long is the entrance length? In the fully developed portion of the flow, what is the value of the wall shear stress? x (m) (±0.01 m) p (mm H2O) (±5 mm) 0 (tank exit) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 (pipe exit) 520 427 351 288 236 188 145 109 73 36 0 the average velocity is 10 diameters per second. (b) Repeat the calculations if the tube is a nanoscale tube with a diameter of D = 100 nm. Section 8.2 Fully Developed Laminar Flow 8.9 For fully developed laminar pipe flow in a circular pipe, the velocity profile is given by u(r) = 2 (1 − r2/R2) in m/s, where R is the inner radius of the pipe. Assuming that the pipe diameter is 4 cm, find the maximum and average velocities in the pipe as well as the volume flow rate. 8.10 A viscous fluid flows in a 0.10-m-diameter pipe such that its velocity measured 0.012 m away from the pipe wall is 0.8 m/s. If the flow is laminar, determine the centerline velocity and the flowrate. 8.11 The wall shear stress in a fully developed flow portion of a 12-in.-diameter pipe carrying water is 1.85 lb/ft2. Determine the pressure gradient, ∂p/∂x, where x is in the flow direction, if the pipe is (a) horizontal, (b) vertical with flow up, or (c) vertical with flow down. 8.12 The pressure drop needed to force water through a horizontal 1-in.-diameter pipe is 0.60 psi for every 12-ft length of pipe. Determine the shear stress on the pipe wall. Determine the shear stress at distances 0.3 and 0.5 in. away from the pipe wall. 8.13 Repeat Problem 8.12 if the pipe is on a 20° hill. Is the flow up or down the hill? Explain. 8.14 Water flows in a constant-diameter pipe with the following conditions measured: At section (a) pa = 32.4 psi and za = 56.8 ft; at section (b) pb = 29.7 psi and zb = 68.2 ft. Is the flow from (a) to (b) or from (b) to (a)? Explain. 8.15 For laminar flow in a round pipe of diameter D, at what distance from the centerline is the actual velocity equal to the average velocity? 8.16 Glycerin at 20 °C flows upward in a vertical 75-mm-diameter pipe with a centerline velocity of 1.0 m/s. Determine the head loss and pressure drop in a 10-m length of the pipe. 8.17 Water at 60 °F flows at a rate of 4.0 gal /min through a 6-in. I.D. plastic pipe. The pipe is 500 ft long and rises a vertical height of 40 ft over the 500 ft. Find the pressure drop. 8.18 At time t = 0 the level of water in tank A shown in Fig. P8.18 is 2 ft above that in tank B. Plot the elevation of the water in tank A as a function of time until the free surfaces in both tanks are at the same elevation. Assume quasisteady conditions— that is, the steady pipe flow equations are assumed valid at any time, even though the flowrate does change (slowly) in time. Neglect minor losses. Note: Verify and use the fact that the flow is laminar. 3 ft 2 ft at t = 0 3 ft 25 ft B 8.8 (See The Wide World of Fluids article titled “Nanoscale Flows,” Section 8.1.1.) (a) Water flows in a tube that has a diameter of D = 0.1 m. Determine the Reynolds number if A 0.1-in. diameter, galvanized iron ■ Figure P8.18 470 Chapter 8 ■ Viscous Flow in Pipes 8.19 A fluid flows through a horizontal 0.1-in.-diameter pipe. When the Reynolds number is 1500, the head loss over a 20-ft length of the pipe is 6.4 ft. Determine the fluid velocity. absolute viscosity is 6.6 × 10−71b ∙ sec/ft2. For the same pressure drop, what must the new pipe size be to carry the oil at 10.7 gal/min? 8.20 Asphalt at 120 °F, considered to be a Newtonian fluid with a viscosity 80,000 times that of water and a specific gravity of 1.09, flows through a pipe of diameter 2.0 in. If the pressure gradient is 1.6 psi/ft determine the flowrate assuming the pipe is (a) horizontal; (b) vertical with flow up. 8.24 A 3-in. schedule 40 commercial steel pipe (with an actual inside diameter of 3.068 in.) carries 210 °F SAE 40 crankcase oil at the rate of 6.0 gal/min. The oil specific gravity is 0.89, and the absolute viscosity is 6.6 × 10−7 1b ∙ sec/ft2. Calculate the pipe size required to carry the same flow rate at approximately one-half the pressure drop of the 3-in. pipe. Both pipes are horizontal. 8.21 Oil of SG = 0.87 and a kinematic viscosity ν = 2.2 × 10−4 m2/s flows through the vertical pipe shown in Fig. P8.21 at a rate of 4 × 10−4 m3/s. Determine the manometer reading, h. 8.25 Water at 20 °C flows down a vertical pipe with no pressure drop. Find the range of pipe diameters D (if any) for which the flow is definitely laminar. 8.26 A person is donating blood. The pint bag in which the blood is collected is initially flat and is at atmospheric pressure. Neglect the initial mass of air in the 1/8-in. I.D., 4 ft-long plastic tube carrying blood to the bag. The average blood pressure in the vein is 40 mm Hg above atmospheric pressure. Estimate the time required for the person to donate one pint of blood. Assume that blood has a specific gravity of 1.06 and a viscosity of 1.0 × 10−4 1b ∙ sec/ft2. The needle’s I.D. is 1/16 inch and the needle length is 2.0 in. The bag is 1.0 ft below the needle inlet and the vein’s I.D. is 1/ 8 in. Optional: Donate a pint of blood and check your answer. SG = 0.87 20 mm 4m Section 8.3 Fully Developed Turbulent Flow h SG = 1.3 Q ■ Figure P8.21 8.22 A liquid with SG = 0.96, μ = 9.2 × 10−4 N ∙ s/m2, and vapor pressure pν = 1.2 × 104 N/m2(abs) is drawn into the syringe as is indicated in Fig P8.22. What is the maximum flowrate if cavitation is not to occur in the syringe? 8.27 For oil (SG = 0.86, μ = 0.025 Ns/m2) flow of 0.2 m3/s through a round pipe with diameter of 500 mm, determine the Reynolds number. Is the flow laminar or turbulent? 8.28 As shown in Video V8.10 and Fig. P8.28, the velocity profile for laminar flow in a pipe is quite different from that for turbulent flow. With laminar flow the velocity profile is parabolic; with turbulent flow at Re = 10,000 the velocity profile can be approximated by the power-law profile shown in the figure. (a) For laminar flow, determine at what radial location you would place a Pitot tube if it is to measure the average velocity in the pipe. (b) Repeat part (a) for turbulent flow with Re = 10,000. Turbulent with Re = 10,000 r 1/5 u = 1 – __ __ 1.0 Vc [ R ] r __ R 10-mm-diameter Laminar with Re < 2100 u = 1 – __ r 2 __ (R) Vc 0.5 u r R 0.12 m 0.25-mm-diameter 0.10-m-long needle Vc 0 0.5 u __ 1.0 Vc patm = 101 kPa (abs) ■ Figure P8.28 ■ Figure P8.22 8.23 A 3-in. schedule 40 commercial steel pipe (with an actual inside diameter of 3.068 in.) carries 210 °F SAE 40 crankcase oil at the rate of 6.0 gal/min. The oil specific gravity is 0.89, and the 8.29 Water at 10 °C flows through a smooth 60-mm-diameter pipe with an average velocity of 8 m/s. Would a layer of rust of height 0.005 mm on the pipe wall protrude through the viscous sublayer? Justify your answer with appropriate calculations. Problems 8.30 When soup is stirred in a bowl, there is considerable turbulence in the resulting motion (see Video V8.7). From a very simplistic standpoint, this turbulence consists of numerous intertwined swirls, each involving a characteristic diameter and velocity. As time goes by, the smaller swirls (the fine scale structure) die out relatively quickly, leaving the large swirls that continue for quite some time. Explain why this is to be expected. 8.31 Water at 60 °F flows through a 6-in.-diameter pipe with an average velocity of 15 ft/s. Approximately what is the height of the largest roughness element allowed if this pipe is to be classified as smooth? Section 8.4.1 Major Losses 8.32 Water is pumped between two tanks as shown in Fig. P8.32. The energy line is as indicated. Is the fluid being pumped from A to B or B to A? Explain. Which pipe has the larger diameter: A to the pump or B to the pump? Explain. Energy line A P B ■ Figure P8.32 8.33 A person with no experience in fluid mechanics wants to estimate the friction factor for 1-in.-diameter galvanized iron pipe at a Reynolds number of 8,000. The person stumbles across the simple equation of f = 64/Re and uses this to calculate the friction factor. Explain the problem with this approach and estimate the error. 8.34 During a heavy rainstorm, water from a parking lot completely fills an 18-in.-diameter, smooth, concrete storm sewer. If the flowrate is 10 ft3/s, determine the pressure drop in a 100-ft horizontal section of the pipe. Repeat the problem if there is a 2-ft change in elevation of the pipe per 100 ft of its length. 471 8.38 Water flows through a horizontal 60-mm-diameter galvanized iron pipe at a rate of 0.02 m3/s. If the pressure drop is 135 kPa per 10 m of pipe, do you think this pipe is (a) a new pipe, (b) an old pipe with a somewhat increased roughness due to aging, or (c) a very old pipe that is partially clogged by deposits? Justify your answer. 8.39 Water flows at a rate of 10 gallons per minute in a new horizontal 0.75-in.-diameter galvanized iron pipe. Determine the pressure gradient, Δp/ℓ, along the pipe. 8.40 Carbon dioxide at a temperature of 0 °C and a pressure of 600 kPa (abs) flows through a horizontal 40-mm-diameter pipe with an average velocity of 2 m/s. Determine the friction factor if the pressure drop is 235 N/m2 per 10-m length of pipe. 8.41 Blood (assume μ = 4.5 × 10−5 lb ∙ s/ft2, SG = 1.0) flows through an artery in the neck of a giraffe from its heart to its head at a rate of 2.5 × 10−4 ft3/s. Assume the length is 10 ft and the diameter is 0.20 in. If the pressure at the beginning of the artery (outlet of the heart) is equivalent to 0.70 ft Hg, determine the pressure at the end of the artery when the head is (a) 8 ft above the heart, or (b) 6 ft below the heart. Assume steady flow. How much of this pressure difference is due to elevation effects, and how much is due to frictional effects? 8.42 A 40-m-long, 12-mm-diameter pipe with a friction factor of 0.020 is used to siphon 30 °C water from a tank as shown in Fig. P8.42. Determine the maximum value of h allowed if there is to be no cavitation within the hose. Neglect minor losses. 10 m 7m 3m 30 m 8.35 Water flows through a horizontal plastic pipe with a diameter of 0.2 m at a velocity of 10 cm/s. Determine the pressure drop per meter of pipe and the power lost to the friction per meter of pipe. h 8.36 Water flows downward through a vertical 10-mm-diameter galvanized iron pipe with an average velocity of 5.0 m/s and exits as a free jet. There is a small hole in the pipe 4 m above the outlet. Will water leak out of the pipe through this hole, or will air enter into the pipe through the hole? Repeat the problem if the average velocity is 0.5 m/s. 8.37 Air at standard conditions flows through an 8-in.-diameter, 14.6-ft-long, straight duct with the velocity versus head loss data indicated in the following table. Determine the average friction factor over this range of data. V (ft/min) 3950 3730 3610 3430 3280 3000 2700 h (in. water) 0.35 0.32 0.30 0.27 0.24 0.20 0.16 ■ Figure P8.42 8.43 Gasoline flows in a smooth pipe of 40-mm diameter at a rate of 0.001 m3/s. If it were possible to prevent turbulence from occurring, what would be the ratio of the head loss for the actual turbulent flow compared to that if it were laminar flow? 8.44 A 3-ft-diameter duct is used to carry ventilating air into a vehicular tunnel at a rate of 9000 ft3/min. Tests show that the pressure drop is 1.5 in. of water per 1500 ft of duct. What is the value of the friction factor for this duct and the approximate size of the equivalent roughness of the surface of the duct? 8.45 H. Blasius correlated data on turbulent friction factor in smooth pipes. His equation fsmooth ≈ 0.3164 Re−1/4 is reasonably accurate for Reynolds numbers between 4000 and 105. Use this 472 Chapter 8 ■ Viscous Flow in Pipes information for the following scenario. Water at 20 °C is to flow through a 3-cm I.D. plastic pipe at the rate of 0.001 m3/s. Find the incline angle of the pipe needed to make the static pressure constant along the pipe. 8.46 Von Karman suggested that the wholly turbulent friction factor be expressed by the equation f= 1 __ in. 2 Flow reducer washer (1) Q 50 holes of diameter 0.05 in. 1 , 4[0.57 − log(εD)]2 where ε is the absolute roughness of the pipe. Compare the values predicted by this equation and those indicated on the Moody chart. ■ Figure P8.52 8.47 The Swamee and Jain formula for the friction factor is 8.53 Water flows at a rate of 0.040 m3/s in a 0.12-m-diameter pipe that contains a sudden contraction to a 0.06-m-diameter pipe. Determine the pressure drop across the contraction section. How much of this pressure difference is due to losses and how much is due to kinetic energy changes? f= 0.25 . [log(ε3.7D + 5.74Re0.9 )]2 Compare this equation for ε/D = 0.00001, 0.0001, 0.001, and 0.01 and Reynolds numbers of 104, 105, 106, and 107 with the Moody chart and decide whether it is an acceptable replacement for the Colebrook formula. 8.48 The Haaland formula for the friction factor is f= 0.3086 {log[6.9Re + (ε3.7D) 1.11]}2 8.54 Water flows from the container shown in Fig. P8.54. Determine the loss coefficient needed in the valve if the water is to “bubble up” 3 in. above the outlet pipe. The entrance is slightly rounded. Vent . Compare this equation for f for εD = 0.00001, 0.0001, 0.001, and 0.01 and Reynolds numbers of 104, 105, 106, and 107 with the Moody chart and decide whether it is an acceptable replacement for the Colebrook formula. 27 in. 8.49 The Churchill formula for the friction factor is f = 8[( 1 -in.-diameter galvanized iron pipe 2 with threaded fittings 1/12 1 8 12 + , 1.5 ] ) Re (A + B) where 18 in. 0.9 7 A = {−2.457 ln [( ) Re 37,530 16 B=( . Re ) 3 in. 16 ε + ; 3.7D ]} Compare this equation for f for both the laminar and turbulent regimes for εD = 0.00001, 0.0001, 0.001, and 0.01 and Reynolds numbers of 10, 102, 103, 104, 105, 106, and 107 with the Moody chart and decide whether it is an acceptable replacement for the Colebrook formula. 2 in. 32 in. ■ Figure P8.54 8.55 (See The Wide World of Fluids article titled “New Hi-tech Fountains,” Section 8.5.) The fountain shown in Fig. P8.55 is designed to provide a stream of water that rises h = 10 ft to h = 20 ft above the nozzle exit in a periodic fashion. To do this the water from the pool enters a pump, passes through a pressure regulator Section 8.4.2 Minor Losses 8.50 Air at standard temperature and pressure flows through a 1-in.-diameter galvanized iron pipe with an average velocity of 8 ft/s. What length of pipe produces a head loss equivalent to (a) a flanged 90° elbow, (b) a wide-open angle valve, or (c) a sharp-edged entrance? h 8.51 Given 90° threaded elbows used in conjunction with copper pipe (drawn tubing) of 0.75-in. diameter, convert the loss for a single elbow to equivalent length of copper pipe for wholly turbulent flow. 8.52 To conserve water and energy, a “flow reducer” is installed in the shower head as shown in Fig. P8.52. If the pressure at point (1) remains constant and all losses except for that in the flow reducer are neglected, determine the value of the loss coefficient (based on the velocity in the pipe) of the flow reducer if its presence is to reduce the flowrate by a factor of 2. Neglect gravity. 4 ft Pump Flow control valve Pressure regulator ■ Figure P8.55 Problems that maintains a constant pressure ahead of the flow control valve. The valve is electronically adjusted to provide the desired water height. With h = 10 ft the loss coefficient for the valve is KL = 50. Determine the valve loss coefficient needed for h = 20 ft. All losses except for the flow control valve are negligible. The area of the pipe is 5 times the area of the exit nozzle. 8.56 Water flows through the screen in the pipe shown in Fig. P8.56 as indicated. Determine the loss coefficient for the screen. V = 20 ft/s the threaded elbows and fittings of the system. Do you agree or disagree with your boss? Support your answer with appropriate calculations. 6-in. length 6-in. length 90° threaded elbows 0.60-in. dia. Reducer 1-in. length Water Screen 473 4-in. length Q = 0.020 cfs Tee 6 in. SG = 3.2 Closed ball valve ■ Figure P8.56 ■ Figure P8.59 8.57 Air flows though the mitered bend shown in Fig. P8.57 at a rate of 5.0 cfs. To help straighten the flow after the bend, a set of 0.25-in.-diameter drinking straws is placed in the pipe as shown. Estimate the extra pressure drop between points (1) and (2) caused by these straws. (1) Tightly packed 0.25-in.-diameter, 12-in.-long straws (2) Section 8.4.3 Noncircular Conduits 8.60 Given two rectangular ducts with equal cross-sectional area, but different aspect ratios (width/height) of 2 and 4, which will have the greater frictional losses? Explain your answer. 8.61 A viscous oil with a specific gravity SG = 0.85 and a viscosity of 0.10 Pa∙s flows from tank A to tank B through the six rectangular slots indicated in Fig. P8.61. If the total flowrate is 30 mm3/s and minor losses are negligible, determine the pressure in tank A. 12 in. A ■ Figure P8.57 8.58 As shown in Fig. P8.58, water flows from one tank to another through a short pipe whose length is n times the pipe diameter. Head losses occur in the pipe and at the entrance and exit. (See Video V8.10.) Determine the maximum value of n if the major loss is to be no more than 10% of the minor loss and the friction factor is 0.02. D 𝓵 = nD Same elevation a a B Each slot 3 mm × 1 mm 0.6 m Section a−a ■ Figure P8.61 8.62 Air at standard temperature and pressure flows at a rate of 7.0 cfs through a horizontal, galvanized iron duct that has a rectangular cross-sectional shape of 12 in. by 6 in. Estimate the pressure drop per 200 ft of duct. 8.63 Water at 20 °C flows through a concentric annulus of inner diameter D1 = 2.0 cm and outer diameter D2 = 4.0 cm. The surface roughness is 0.002 cm. Calculate the frictional pressure loss per unit length for a flow rate of 0.01 m3/s. 8.64 Air at standard conditions flows through a horizontal 1 ft by 1.5 ft rectangular wooden duct at a rate of 5000 ft3/min. Determine the head loss, pressure drop, and power supplied by the fan to overcome the flow resistance in 500 ft of the duct. ■ Figure P8.58 Section 8.5.1 Single Pipes—Determine Pressure Drop 8.59 Water flows steadily through the 0.75-in.-diameter galvanized iron pipe system shown in Video V8.16 and Fig. P8.59 at a rate of 0.020 cfs. Your boss suggests that friction losses in the straight pipe sections are negligible compared to losses in Assume a car’s exhaust system can be approximated as 8.65 14 ft of 0.125-ft-diameter cast-iron pipe with the equivalent of six 90° flanged elbows and a muffler. (See Video V8.14.) The muffler acts as a resistor with a loss coefficient of KL = 8.5. Determine the 474 Chapter 8 ■ Viscous Flow in Pipes pressure at the beginning of the exhaust system if the flowrate is 0.10 cfs, the temperature is 250 °F, and the exhaust has the same properties as air. (1) 8.66 The pressure at section (2) shown in Fig. P8.66 is not to fall below 60 psi when the flowrate from the tank varies from 0 to 1.0 cfs and the branch line is shut off. Determine the minimum height, h, of the water tank under the assumption that (a) minor losses are negligible, (b) minor losses are not negligible. 3 in. 4 in. ■ Figure P8.71 10 ft h All pipe is 6-in.-diameter plastic (ε /D = 0), flanged fittings 6 ft Branch line 600 ft with 15 90° elbows (2) Main line 8.72 Water at 10 °C is pumped from a lake as shown in Fig. P8.72. If the flowrate is 0.011 m3/s, what is the maximum length inlet pipe, ℓ, that can be used without cavitation occurring? Length 𝓵 Elevation 650 m D = 0.07 m ε = 0.08 mm 900 ft Elevation 653 m Q= 0.011 m3/s ■ Figure P8.66 ■ Figure P8.72 8.67 Repeat Problem 8.66 with the assumption that the branch line is open so that half of the flow from the tank goes into the branch, and half continues in the main line. 1 2 -in.-diameter hose shown in Fig. P8.68 can withstand 8.68 The a maximum pressure of 200 psi without rupturing. Determine the maximum length, ℓ, allowed if the friction factor is 0.022 and the flowrate is 0.010 cfs. Neglect minor losses. Q= Nozzle tip diameter = 0.30 in. 0.010 cfs D = 0.50 in. 10 ft Water 𝓵 L 8.73 At a ski resort, water at 40 °F is pumped through a 3-in.diameter, 2000-ft-long steel pipe from a pond at an elevation of 4286 ft to a snow-making machine at an elevation of 4623 ft at a rate of 0.26 ft3/s. If it is necessary to maintain a pressure of 180 psi at the snow-making machine, determine the horsepower added to the water by the pump. Neglect minor losses. 8.74 Crude oil having a specific gravity of 0.80 and a viscosity of 6.0 × 10−5 ft2/sec flows through a pumping station at a rate of 10,000 barrels/hr. The oil then flows through 120,000 ft of 24-in. ID, horizontal, commercial steel pipe, enters another pumping station at 20 psig, and leaves at a discharge pressure equal to the discharge pressure of the previous pumping station. Calculate the discharge pressures and the power supplied to the crude oil at the second pumping station. See Fig. P8.74. 3 ft L = 120,000 ft Pump 4 ■ Figure P8.68 1 8.69 The hose shown in Fig. P8.69 will collapse if the pressure within it is lower than 10 psi below atmospheric pressure. Determine the maximum length, ℓ, allowed if the friction factor is 0.015 and the flowrate is 0.010 cfs. Neglect minor losses. 8.70 According to fire regulations in a town, the pressure drop in a commercial steel horizontal pipe must not exceed 1.0 psi per 150 ft of pipe for flowrates up to 500 gal/min. If the water temperature is above 50 °F, can a 6-in.-diameter pipe be used? 8.71 As shown in Video V8.15 and Fig. P8.71, water “bubbles up” 3 in. above the exit of the vertical pipe attached to three horizontal pipe segments. The total length of the 0.75-in.diameter galvanized iron pipe between point (1) and the exit is 21 in. Determine the pressure needed at point (1) to produce this flow. 3 2 I.D. = 24 in. = 2 ft · Wp · Wp ■ Figure P8.74 8.75 A motor-driven centrifugal pump delivers 15 °C water at the rate of 10 m3/min from a reservoir, through a 2500-m-long, 30-cm I.D. plastic pipe, to a second reservoir. The water level in the second reservoir is 40 m above the water level in the first reservoir. The pump efficiency is 75%. Find the motor output power. The pipe entrance is square edged. 8.76 An emergency flooding system for a nuclear reactor core is shown in Fig. P8.76. Find the power input required to flood the 475 Problems Main supply line, 4-in. type L copper pipe (actual I.D. is 3.905 in.) core at the rate of 5000 gal/min. Assume a square-edged entrance, 60 °F water, and threaded connections. A patm Q 50,000 gallons water B 70 ft D E 90 ft 140 ft 100 gal/min 750 ft of 3-in. schedule 160 stainless steel pipe (actual I.D. is 2.625 in.), three gate valves, seven 90° regular elbows 500 gal/min C 150 gal/min 100 gal/min ■ Figure P8.79 2000 psig 8.80 Water is pumped through a 60-m-long, 0.3-m-diameter pipe from a lower reservoir to a higher reservoir whose surface is 10 m above the lower one. The sum of the minor loss coefficient for the system is KL = 14.5. When the pump adds 40 kW to the water the flowrate is 0.20 m3/s. Determine the pipe roughness. Core ■ Figure P8.76 8.77 A hydraulic turbine takes water from a lake with the piping system shown in Figure P8.77. Find the head ht available to the turbine for flow rates of 0, 5000, 10,000, 15,000, and 20,000 ft3/min. Equivalent length of 900 ft of 5-ft inside diameter concrete pipe 1100 ft 8.81 Natural gas (ρ = 0.0044 slugs/ft3 and ν = 5.2 × 10−5 ft2/s) is pumped through a horizontal 6-in.-diameter cast-iron pipe at a rate of 800 lb/hr. If the pressure at section (1) is 50 psi (abs), determine the pressure at section (2) 8 mi downstream if the flow is assumed incompressible. Is the incompressible assumption reasonable? Explain. 8.82 As shown in Fig. P8.82, a standard household water meter is incorporated into a lawn irrigation system to measure the volume of water applied to the lawn. Note that these meters measure volume, not volume flowrate. (See Video V8.16.) With an upstream pressure of p1 = 50 psi the meter registered that 120 ft3 of water was delivered to the lawn during an “on” cycle. Estimate the upstream pressure, p1, needed if it is desired to have 150 ft3 delivered during an “on” cycle. List any assumptions needed to arrive at your answer. I.D. = 5 ft Hydraulic turbine ■ Figure P8.77 (1) 8.78 Water flows through a 2-in.-diameter pipe with a velocity of 15 ft/s as shown in Fig. P8.78. The relative roughness of the pipe is 0.004, and the loss coefficient for the exit is 1.0. Determine the height, h, to which the water rises in the piezometer tube. Open Irrigation system: pipes, fittings, nozzles, etc. WATER METER ■ Figure P8.82 8.83 A fan is to produce a constant air speed of 40 m/s throughout the pipe loop shown in Fig. P8.83. The 3-m-diameter pipes are smooth, and each of the four 90° elbows has a loss coefficient of 0.30. Determine the power that the fan adds to the air. 20 m h 15 ft/s 8 ft V = 40 m/s 2 in. 8 ft D=3m 10 m ■ Figure P8.78 8.79 Figure P7.79 shows the 60 °F water flow rates from the branches of a main supply line. Find the total pressure drop (pA − pE) for soldered copper pipe. Assume that the loss coefficient for each tee is based on the threaded type. Fan ■ Figure P8.83 476 Chapter 8 ■ Viscous Flow in Pipes 8.84 Air flows in a horizontal 100-ft-long, 24-in. × 24-in. duct at the rate of 5000 ft3/min. The air then flows through an expansion into 200 ft of 36-in. × 36-in. duct. The expansion has a loss coefficient of 0.80, based on the higher inlet velocity. Calculate the static pressure change across the expansion and the static pressure loss across the entire duct system. The duct is made of galvanized sheet metal. Use a constant air density of 0.0024 slug/ft3. Length of pipe between gages = 60 ft Pipe diameter = 0.1 ft Pipe friction factor = 0.03 Filter loss coefficient = 12 Filter Pump Section 8.5.1 Single Pipes—Determine Flowrate 8.85 The turbine shown in Fig. P8.85 develops 400 kW. Determine the flowrate if (a) head losses are negligible or (b) head loss due to friction in the pipe is considered. Assume f = 0.02. Note: There may be more than one solution or there may be no solution to this problem. 20 m Diffuser T ■ Figure P8.88 8.89 The pump shown in Fig. P8.89 adds 25 kW to the water and causes a flowrate of 0.04 m3/s. Determine the flowrate expected if the pump is removed from the system. Assume f = 0.016 for either case and neglect minor losses. 1m 120 m of 0.30-m-diameter cast-iron pipe 60-mm-diameter, 30-m-long pipe; f = 0.016 40-mm-diameter nozzle ■ Figure P8.85 8.86 Water flows from the nozzle attached to the spray tank shown in Fig. P8.86. Determine the flowrate if the loss coefficient for the nozzle (based on upstream conditions) is 0.75 and the friction factor for the rough hose is 0.11. Nozzle diameter = 7.5 mm p = 150 kPa Pump ■ Figure P8.89 8.90 The vented storage tank shown in Fig. P8.90 is used to refuel race cars at a race track. A total of 40 ft of steel pipe (I.D. = 0.957 in.), two 90° regular elbows, and a globe valve make up the system. Calculate the time needed to put 20 gal of fuel in a car tank. The pressures, p2 and p1, are equal, the connections are threaded, and the fuel has the properties of 68 °F normal octane (ν = 8.31 × 10−6 ft2/s). D = 15 mm 𝓵 = 1.9 m 0.80 m Vent 1 40° patm ■ Figure P8.86 Racing fuel 8.87 Water flows through the pipe shown in Fig. P8.87. Determine the net tension in the bolts if minor losses are neglected and the wheels on which the pipe rests are frictionless. 3.0 m 80 mm Bolts h = 12 ft 2 Galvanized iron p2 = patm ■ Figure P8.90 10 m 20 m †8.91 Gasoline is unloaded from the tanker truck shown in Fig. P8.91 through a 4-in.-diameter rough-surfaced hose. This is a “gravity dump” with no pump to enhance the flowrate. It is 8.88 When the pump shown in Fig. P8.88 adds 0.2 horsepower claimed that the 8800-gallon capacity truck can be unloaded in to the flowing water, the pressures indicated by the two gages are 28 minutes. Do you agree with this claim? Support your answer with appropriate calculations. equal. Determine the flowrate. ■ Figure P8.87 Problems 477 D Midstate Gasoline (1) H (2) L d ■ Figure P8.91 ■ Figure P8.94 8.92 Calculate the water flow rate in the system shown in Fig. P8.92. The piping system includes four gate valves, two half-open globe valves, fourteen 90° regular elbows, and 250 ft of 2-in. schedule 40 commercial steel pipe (with an actual inside diameter of 2.067 in.). Assume threaded connections and a square-edged pipe entrance. 8.95 Water flows through two sections of the vertical pipe shown in Fig. P8.95. The bellows connection cannot support any force in the vertical direction. The 0.4-ft-diameter pipe weights 0.2 lb/ft, and the friction factor is assumed to be 0.02. At what velocity will the force, F, required to hold the pipe be zero? Free jet 2 144 ft 60 ft F f = 0.020 Pipe weighs 0.20 lb/ft 220 ft D = 0.40 ft Piping system Bellows V 20 ft ■ Figure P8.95 30 ft 72 ft2 8.96 Water is circulated from a large tank, through a filter, and back to the tank as shown in Fig. P8.96. The power added to the water by the pump is 200 ft ∙ lb/s. Determine the flowrate through the filter. ■ Figure P8.92 8.93 The pump shown in Fig. P8.93 delivers a head of 250 ft to the water. Determine the power that the pump adds to the water. The difference in elevation of the two ponds is 200 ft. KL exit = 1.0 KL elbow = 1.5 KL valve = 6.0 KL exit Pump KL Filter KL valve = 5.0 KL ent = 0.8 elbow = 1.5 Pipe length = 500 ft Pipe diameter = 0.75 ft Pipe roughness = 0 ■ Figure P8.93 8.94 For the standpipe system shown in Fig. P8.94, calculate the flow rate for Н = 4.0 ft, D = 6.77 in., d = 0.125 in., and L = 48 in. The fluid is 70 °F water. Assume steady flow and neglect the energy loss in the entrance nozzle. The pipe is commercial steel. KL ent = 0.8 KL filter = 12.0 = 1.0 Pump 200 ft. of 0.1-ft-diameter pipe with ε/D = 0.01 ■ Figure P8.96 8.97 A thief siphoned 15 gal of gasoline from a gas tank in the middle of the night. The gas tank is 12 in. wide, 24 in. long, and 18 in. high and was full when the thief started. The siphoning plastic tube has an inside diameter of 0.5 in. and a length of 4.0 ft. Assume that at any instant of time, the steady-state mechanical energy equation is adequate to predict the gasoline flow rate through the tube. As 15 gal is 3465 in3, the gasoline level in the tank will drop 12.0 in. You may use the gasoline level after it has dropped 6.0 in. to estimate the average gasoline flow rate. Use this flow rate to estimate the time needed to siphon the 15 gal of gasoline. Compare your answer with the answer of 190 sec found in problem 3.107, 478 Chapter 8 ■ Viscous Flow in Pipes using Bernoulli’s equation. The siphon discharges at the level of the bottom of the gasoline tank. You may find it useful to use the Blasius equation for smooth pipes found in problem 8.45. 8.98 Estimate the time required for the water depth in the reservoir shown in Fig. P8.98 to drop from a height of 25 m to 5 m. The connections are threaded. 500 m2 T = 10°C 25 m 5m the diameter necessary for drawn copper pipe. The pipe has no elevation change. 8.101 A certain process requires 2.3 cfs of water to be delivered at a pressure of 30 psi. This water comes from a large-diameter supply main in which the pressure remains at 60 psi. If the galvanized iron pipe connecting the two locations is 200 ft long and contains six threaded 90° elbows, determine the pipe diameter. Elevation differences are negligible. 8.102 Water is pumped between two large open reservoirs through 1.5 km of smooth pipe. The water surfaces in the two reservoirs are at the same elevation. When the pump adds 20 kW to the water, the flowrate is 1 m3/s. If minor losses are negligible, determine the pipe diameter. 8.103 Determine the diameter of a steel pipe that is to carry 2000 gal/min of gasoline with a pressure drop of 5 psi per 100 ft of horizontal pipe. 8.104 Water is to be moved from a large, closed tank in which the air pressure is 20 psi into a large, open tank through 2000 ft of smooth pipe at the rate of 3 ft3/s. The fluid level in the open tank is 150 ft below that in the closed tank. Determine the required diameter of the pipe. Neglect minor losses. 30 m 1000 m of 7-cm-inside-diameter, galvanized iron pipe, five regular 90° elbows, one fully open globe valve 8.105 A commercial steel flow channel in a heat exchanger has an equilateral triangle cross section with each side measuring 5.0 in. and a length measuring 96.0 in. Water at 60 °F flowing through the channel has a pressure loss of 0.10 psi. Find the water flow rate. ■ Figure P8.98 8.99 Sheldon and Leonard come home from a long day of studying to discover that their basement is flooded with water. They *8.106 Rainwater flows through the galvanized iron downspout have a submersible pump for such an emergency and connect shown in Fig. P8.106 at a rate of 0.006 m3/s. Determine the size the pump discharge to a 12-ft-long garden hose, as shown in Fig. of the downspout cross section if it is a rectangle with an aspect P8.99. The garden hose may be considered a smooth plastic pipe ratio of 1.7 to 1 and it is completely filled with water. Neglect the with a 1.0-in. indside diameter. If the initial water depth is 6.0 in. velocity of the water in the gutter at the free surface and the head and the pump has an input power of 0.25 hp and an overall effi- loss associated with the elbow. ciency of 0.60, how long will it take to pump out the basement? The pump inlet piping length is negligible, and the garden hose 70 mm outlet is 7.0 ft above the basement floor. g Q 4m h0 = 6 in. 24 ft h = 7 ft 3m ■ Figure P8.106 36 ft *8.107 Repeat Problem 8.106 if the downspout is circular. ■ Figure P8.99 8.108 For a given head loss per unit length, what effect on the flowrate does doubling the pipe diameter have if the flow is (a) laminar, or (b) completely turbulent? Section 8.5.1 Single Pipes—Determine Diameter 8.109 It is necessary to deliver 270 ft3/min of water from reservoir A to reservoir B, as shown in Fig. P8.109. The connecting piping consists of four fully open gate valves, twelve regular 90° elbows, one swing check valve, two fully open globe valves, and 3000 ft of commercial steel pipe. Find the minimum diameter for steel pipe needed to obtain this flow rate. All fittings and valves have flanged connections. Assume a square-edged entrance. 8.100 A company markets ethylene glycol antifreeze in halfgallon bottles. A machine fills and caps the bottles at a rate of 60 per minute. The 68 °F ethylene glycol (ρ = 69.3 lbmft3, v = 1.93 × 10−4 ft2/s) is pumped to the machine from a tank 36 ft away. The pressure at the discharge of the pump is 35 psig, and the pressure at the inlet to the filling machine must be at least 15 psig. Determine Problems Elevation 150.0 ft (with an actual inside diameter of 2.067 in.). Energy losses are important, so consideration is being given to expanding the pipe to a 3.0-in. schedule 40 commercial steel pipe (with an actual inside diameter of 3.068 in.), using a 10° conical expansion. After a length ℓ of the 3.0-in. pipe, it is decreased to a 2.0-in. schedule 40 commercial steel pipe with a gradual contraction (K = 0.10, based on the smaller diameter pipe flow). Find the minimum length ℓ so that the energy loss will be the same as in a 2.0-in. pipe of length ℓ + 2 (5.7 in.) See Fig. P8.114. Reservoir A Elevation 25.0 ft 8.115 The flowrate between tank A and tank B shown in Fig. P8.115 is to be increased by 30% (i.e., from Q to 1.30Q) by the addition of a second pipe (indicated by the dotted lines) running from node C to tank B. If the elevation of the free surface in tank A is 25 ft above that in tank B, determine the diameter, D, of this new pipe. Neglect minor losses and assume that the friction factor for each pipe is 0.02. Reservoir B ■ Figure P8.109 Section 8.5.2 Multiple Pipe Systems 8.110 A 10-m-long, 5.042-cm I.D. copper pipe has two fully open gate valves, a swing check valve, and a sudden enlargement to a 9.919-cm I.D. copper pipe. The 9.919 cm copper pipe is 5.0 m long and then has a sudden contraction to another 5.042-cm copper pipe. Find the head loss for a 20 °C water flow rate of 0.05 m3/s. 8.111 Air, assumed incompressible, flows through the two pipes shown in Fig. P8.111. Determine the flowrate if minor losses are neglected and the friction factor in each pipe is 0.015. Determine the flowrate if the 0.5-in.-diameter pipe were replaced by a 1-in.diameter pipe. Comment on the assumption of incompressibility. p = 0.5 psi T = 150°F 1 in. 0.50 in. 20 ft 479 20 ft ■ Figure P8.111 *8.112 Repeat Problem 8.111 if the pipes are galvanized iron and the friction factors are not known a priori. †8.113 Estimate the power that the human heart must impart to the blood to pump it through the two carotid arteries from the heart to the brain. List all assumptions and show all calculations. 8.114 Normal octane at 68 °F (ν = 8.31 × 10−6 ft2/s) is to be delivered at a flow rate of 3.0 gal/min through a 2.0-in. schedule 40 commercial steel pipe 6-in. diameter; 600 ft long 6-in. diameter; 500 ft long C A B Diameter D, 500 ft long ■ Figure P8.115 8.116 A 250-ft-high building has a 6.065-in.-diameter steel stand9 pipe and a 100-ft-long 216 -in. diameter fire hose on each floor. The nearest fireplug is 100 ft from the standpipe’s ground-level connection. Assume that fire-fighters connect a 6-in.-diameter, 50-ft-long fire hose from the fireplug to the fire truck and a 4-in.-diameter, 50-ft-long fire hose from the fire truck to the standpipe’s groundlevel connection. The National Fire Protection Association (NFPA) requires that a minimum pressure of 65 psig be maintained at the 9 connection of the 216 -in.-diameter hose and the standpipe while maintaining a flow rate of 500 gal/min through the fire hose. What pressure rise must the pump on the fire engine supply to satisfy the NFPA requirement for this building? The fire hydrant water pressure is 80 psig and the water temperature is 60 °F. The connections are threaded. 8.117 With the valve closed, water flows from tank A to tank B as shown in Fig. P8.117. What is the flowrate into tank B when the valve is opened to allow water to flow into tank C also? Neglect all minor losses and assume that the friction factor is 0.02 for all pipes. Elevation = 15 m Diameter of each pipe = 0.10 m A Elevations = 0 2-in. pipe 80 m 5.7 in. 𝓵 40 m B 5.7 in. C 75 m ■ Figure P8.117 Expansion ■ Figure P8.114 3-in. pipe Contraction *8.118 Repeat Problem 8.117 if the friction factors are not known, but the pipes are steel pipes. 480 Chapter 8 ■ Viscous Flow in Pipes 8.119 The three water-filled tanks shown in Fig. P8.119 are connected by pipes as indicated. If minor losses are neglected, determine the flowrate in each pipe. Section 8.6 Pipe Flowrate Measurement 8.122 Water flows through the orifice meter shown in Fig. P8.122 at a rate of 0.10 cfs. If d = 0.1 ft, determine the value of h and the pressure difference associated with this h. Elevation = 60 m Elevation = 20 m Elevation = 0 D = 0.10 m 𝓵 = 200 m f = 0.015 D = 0.08 m 𝓵 = 200 m f = 0.020 D = 0.08 m 𝓵 = 400 m f = 0.020 h d Q 2 in. ■ Figure P8.119 ■ Figure P8.122 8.120 (See The Wide World of Fluids article titled “Deepwater Pipeline,” Section 8.5.2.) Five oil fields, each producing an output of Q barrels per day, are connected to the 28-in.-diameter “mainline pipe” (A–B–C) by 16-in.-diameter “lateral pipes” as shown in Fig. P8.120. The friction factor is the same for each of the pipes and elevation effects are negligible. (a) For section A–B determine the ratio of the pressure drop per mile in the mainline pipe to that in the lateral pipes. (b) Repeat the calculations for section B–C. Q Lateral Q Q A Q B 8.124 Water flows through the orifice meter shown in Fig. P8.122 at a rate of 0.10 cfs. If h = 3.8 ft, determine the value of d. 8.125 Water flows through a 40-mm-diameter nozzle meter in a 75-mm-diameter pipe at a rate of 0.015 m3/s. Determine the pressure difference across the nozzle if the temperature is (a) 10 °C, or (b) 80 °C. 8.126 Gasoline flows through a 35-mm-diameter pipe at a rate of 0.0032 m3/s. Determine the pressure drop across a flow nozzle placed in the line if the nozzle diameter is 20 mm. C Q Main line 8.123 Water flows through the orifice meter shown in Fig P8.122 such that h = 1.6 ft with d = 1.5 in. Determine the flowrate. 8.127 Air at 200 °F and 60 psia flows in a 4-in.-diameter pipe at a rate of 0.52 lb/s. Determine the pressure at the 2-in.-diameter throat of a Venturi meter placed in the pipe. ■ Figure P8.120 †8.121 As shown in Fig. P8.121, cold water (T = 50 °F) flows from the water meter to either the shower or the hot water heater. In the hot water heater it is heated to a temperature of 150 °F. Thus, with equal amounts of hot and cold water, the shower is at a comfortable 100 °F. However, when the dishwasher is turned on, the shower water becomes too cold. Indicate how you would predict this new shower temperature (assume the shower faucet is not adjusted). State any assumptions needed in your analysis. 8.128 A 2.5-in.-diameter flow nozzle meter is installed in a 3.8-in.-diameter pipe that carries water at 160 °F. If the air–water manometer used to measure the pressure difference across the meter indicates a reading of 3.1 ft, determine the flowrate. 8.129 A 0.064-m-diameter nozzle meter is installed in a 0.097-m-diameter pipe that carries water at 60 °C. If the inverted air–water U-tube manometer used to measure the pressure difference across the meter indicates a reading of 1 m, determine the flowrate. 8.130 A 50-mm-diameter nozzle meter is installed at the end of a 80-mm-diameter pipe through which air flows. A manometer attached to the static pressure tap just upstream from the nozzle indicates a pressure of 7.3 mm of water. Determine the flowrate. Hot Dishwasher Shower Cold Water meter 8.131 Water flows through the Venturi meter shown in Fig. P8.131. The specific gravity of the manometer fluid is 1.52. Determine the flowrate. 6 in. Hot water heater 3 in. 2 in. SG = 1.52 ■ Figure P8.121 ■ Figure P8.131 Q Problems 8.132 If the fluid flowing in Problem 8.131 were air, what would the flowrate be? Would compressibility effects be important? Explain. 8.133 The scale reading on the rotameter shown in Fig. P8.133 and Video V8.15 (also see Fig. 8.44) is directly proportional to the volumetric flowrate. With a scale reading of 2.6 the water bubbles up approximately 3 in. How far will it bubble up if the scale reading is 5.0? 6 5 4 3 2 1 0 ■ 481 Lifelong Learning Problems 8.1 LL The field of biomedical engineering has undergone significant growth in recent years. Some universities have undergraduate and graduate programs in this field. Biomedical engineering applies engineering principles to help solve problems in the medical field for human health. Obtain information about biomedical engineering applications in blood flow. Summarize your findings in a brief report. 8.2 LL Data used in the Moody diagram were first published in 1944. Since then, there have been many innovations in pipe material, pipe design, and measurement techniques. Investigate whether there have been any improvements or enhancements to the Moody chart and friction factor relations. Summarize your findings in a brief report. 3 in. Rotameter ■ Figure P8.133 8.3 LL As discussed in Section 8.4.2, flow separation in pipes can lead to losses (we will also see in Chapter 9 that external flow separation is a significant issue). For external flows, many mechanisms have been devised to help mitigate and control flow separation from the surface, for example, from the wing of an airplane. Investigate either passive or active flow control mechanisms that can reduce or eliminate internal flow separation (e.g., flow separation in a diffuser). Summarize your findings in a brief report. 9 Flow over Immersed Bodies Learning Objectives After completing this chapter, you should be able to: ■ identify and discuss the features of external flow. ■ explain the fundamental characteristics of a boundary layer, including laminar, transitional, and turbulent regimes. ■ calculate boundary layer parameters for flow past a flat plate. ■ explain the physical process of boundary layer separation. ■ calculate the lift and drag forces for various objects. Many practical situations involve flow past objects. 482 In this chapter we consider various aspects of the flow over bodies that are completely immersed in a fluid. Examples include the flow of air around airplanes, automobiles, and falling snowflakes or the flow of water around submarines and fish. In these situations the object is completely surrounded by the fluid and the flows are termed external flows. External flows involving air are often termed aerodynamics in response to the important external flows produced when an object such as an airplane flies through the atmosphere. Although this field of external flows is extremely important, there are many other examples that are of equal importance. The fluid force (lift and drag) on surface vehicles (cars, trucks, bicycles) has become a very important topic. By correctly designing cars and trucks, it has become possible to greatly decrease the fuel consumption and improve the handling characteristics of the vehicle. Similar efforts have resulted in improved ships, whether they are surface vessels (surrounded by two fluids, air and water) or submersible vessels (surrounded completely by water). Other applications of external flows involve objects that are not completely surrounded by fluid, although they are placed in some external-type flow. For example, the proper design of a building (whether it is your house or a tall skyscraper) must include consideration of the various wind effects involved. As with other areas of fluid mechanics, various approaches (theoretical, numerical, and experimental) are used to obtain information on the fluid forces developed by external flows. Theoretical (i.e., analytical) techniques can provide some of the needed information about such flows. However, because of the complexities of the governing equations and the complexities of the geometry of the objects involved, the amount of information obtained from purely theoretical methods is limited. Much of the information about external flows comes from experiments carried out, for the most part, on scale models of the actual objects. Such testing includes the obvious wind tunnel testing of model airplanes, buildings, and even entire cities. In some instances the actual device, not a 9.1 General External Flow Characteristics 483 (a) ■ Figure 9.1 (a) Flow past a full-sized (b) streamlined vehicle in the GM aerodynamics laboratory wind tunnel, an 18-ft by 34-ft test section facility driven by a 4000-hp, 43-ft-diameter fan. (Photograph courtesy of General Motors Corporation.) (b) Predicted streamlines for flow past a Formula 1 race car as obtained by using computational fluid dynamics techniques. (Courtesy of ANSYS, Inc.) model, is tested in wind tunnels. Figure 9.1a shows a test of a vehicle in a wind tunnel. Better performance of cars, bikes, skiers, and numerous other objects has resulted from testing in wind tunnels. The use of water tunnels and towing tanks also provides useful information about the flow around ships and other objects. The information gathered from these experiments typically requires dimensional analysis for correlation and/or extrapolation. With advancement in computational fluid dynamics, or CFD, numerical methods are also capable of predicting external flows past objects. Figure 9.1b shows streamlines around a Formula 1 car as predicted by CFD. Appendix A provides an introduction to CFD. In this chapter we consider characteristics of external flow past a variety of objects. We investigate the qualitative aspects of such flows and learn how to determine the various forces on objects surrounded by a moving liquid. 9.1 General External Flow Characteristics For external flows it is usually easiest to use a coordinate system fixed to the object. VIDEO V9.1 Space Shuttle landing A body immersed in a moving fluid experiences a resultant force due to the interaction between the body and the fluid surrounding it. In some instances (such as an airplane flying through still air) the fluid far from the body is stationary and the body moves through the fluid with velocity U. In other instances (such as the wind blowing past a building) the body is stationary and the fluid flows past the body with velocity U. In any case, we can fix the coordinate system in the body and treat the situation as fluid flowing past a stationary body with velocity U, the upstream velocity. For the purposes of this book, we will assume that the upstream velocity is constant in both time and location. That is, there is a uniform, constant velocity fluid flowing past the object. In actual situations this is often not true. For example, the wind blowing past a smokestack is nearly always turbulent and gusty (unsteady) and probably not of uniform velocity from the top to the bottom of the stack. Usually, from an engineering analysis viewpoint, the unsteadiness and nonuniformity are of minor importance. Even with a steady, uniform upstream flow, the flow in the vicinity of an object may be unsteady. Examples of this type of behavior include the flutter that is sometimes found in the flow past airfoils (wings), the regular oscillation of telephone wires that “sing” in a wind, and the irregular turbulent fluctuations in the wake regions behind bodies. The structure of an external flow and the ease with which the flow can be described and analyzed often depend on the nature of the body in the flow (including its orientation with the flow). Three general categories of bodies are shown in Fig. 9.2. They include (a) two-dimensional 484 Chapter 9 ■ Flow over Immersed Bodies y y U y U x x (b) (a) U z x z (c) ■ Figure 9.2 Flow classification: (a) two-dimensional, (b) axisymmetric, (c) three-dimensional. objects (infinitely long perpendicular to the page and of constant cross-sectional size and shape), (b) axisymmetric bodies (formed by rotating their cross-sectional shape about the axis of symmetry), and (c) three-dimensional bodies that may or may not possess a line or plane of symmetry. In practice there can be no truly two-dimensional bodies—nothing extends to infinity. However, many objects are sufficiently long so that the end effects are negligibly small. Another classification of body shape can be made depending on whether the body is streamlined or blunt. The flow characteristics depend strongly on the amount of streamlining present. In general, streamlined bodies (e.g., airfoils, racing cars, etc.) have little effect on the surrounding fluid, compared with the effect that blunt bodies (e.g., parachutes, buildings, etc.) have on the fluid. Usually, but not always, it is easier to force a streamlined body through a fluid than it is to force a similar-sized blunt body at the same velocity. There are important exceptions to this basic rule. 9.1.1 Lift and Drag Concepts A body interacts with the surrounding fluid through pressure and shear stresses. Drag is parallel to upstream velocity and lift is perpendicular. When any body moves through a fluid, an interaction between the body and the fluid occurs; this effect can be given in terms of the forces at the fluid–body interface. These forces can be described in terms of the stresses—wall (surface) shear stresses on the body, τw, due to viscous effects and normal stresses due to the pressure, p. Typical shear stress and pressure distributions are shown in Figs. 9.3a and 9.3b. Both τw and p vary in magnitude and direction along the surface. It is often useful to know the detailed distribution of shear stress and pressure over the surface of the body, although such information is difficult to obtain. Many times, however, only the integrated or resultant effects of these distributions are needed. The resultant force in the direction of the upstream velocity is termed the drag, 𝒟, and the resultant force normal to the upstream velocity is termed the lift, ℒ, as is indicated in Fig. 9.3c. Note that lift may be downward relative to the upstream velocity; in this case lift is termed downforce. For some three-dimensional bodies there may also be a side force that is perpendicular to the plane containing 𝒟 and ℒ. The resultant of the shear stress and pressure distributions can be obtained by integrating the effect of these two quantities on the body surface as is indicated in Fig. 9.4. The x and y components of the fluid force on the small area element dA are dFx = (p dA) cos θ + (τw dA) sin θ p<0 U Pressure distribution ℒ p>0 U (a) τw U Shear stress distribution 𝒟 (c) (b) ■ Figure 9.3 Forces from the surrounding fluid on a two-dimensional object: (a) pressure force, (b) viscous force, and (c) resultant force (lift and drag). 9.1 y U pdA θ General External Flow Characteristics 485 τ w dA dA θ x ■ Figure 9.4 Pressure and shear forces on a small element of the surface of a body. and dFy = −(p dA) sin θ + (τw dA) cos θ Thus, the net x and y components of the force on the object are 𝒟= ∫ dFx = ∫ p cos θ dA + ∫τ ∫ p sin θ dA + ∫ sin θ dA (9.1) τw cos θ dA (9.2) w and ℒ= Lift and drag on a section of a body depend on the orientation of the surface. ∫ dFy = − Of course, to carry out the integrations and determine the lift and drag, we must know the body shape (i.e., θ as a function of location along the body) and the distribution of τw and p along the surface. These distributions are often extremely difficult to obtain, either experimentally or theoretically. The pressure distribution can be obtained experimentally by use of a series of static pressure taps along the body surface. On the other hand, it is usually quite difficult to measure the wall shear stress distribution. THE WIDE WORLD OF FLUIDS Pressure-sensitive paint For many years, the conventional quantitative measure of the surface pressure. One of the biggest method for measuring surface pressure has been to use static advantages of PSP is that it is a global measurement technique, pressure taps consisting of small holes on the surface connected measuring pressure over the entire surface, as opposed to dis- by hoses from the holes to a pressure measuring device. Pressure- crete points at the taps. PSP also has the advantage of being sensitive paint (PSP) is now gaining acceptance as an alternative nonintrusive to the flow field. Although static pressure port holes to the static surface pressure ports. The PSP material is typically are small, they do alter the surface and can slightly alter the flow, a luminescent compound that is sensitive to the pressure on it thus affecting downstream ports. In addition, the use of PSP and can be excited (visualized) by an appropriate light that is cap- eliminates the need for a large number of pressure taps and con- tured by special video imaging equipment. Thus, it provides a necting tubes. This lowers measurement time and cost. It is seen that both the shear stress and pressure force contribute to the lift and drag, since for an arbitrary body θ is neither zero nor 90° along the entire body. The exception is a flat plate aligned either parallel to the upstream flow (θ = 90°) or normal to the upstream flow (θ = 0) as is discussed in Example 9.1. E X AMPLE 9 .1 GIVEN Drag from Pressure and Shear Stress Distributions Air at standard conditions flows past a flat plate as is indicated in Fig. E9.1. In case (a) the plate is parallel to the upstream flow, and in case (b) it is perpendicular to the upstream flow. The pressure and shear stress distributions on the surface are as indicated (obtained either by experiment or theory). FIND Determine the lift and drag on the plate. 486 Chapter 9 ■ Flow over Immersed Bodies y ( y U = 25 ft/s b = width = 10 ft p = p(x) = 0 p = –0.893 lb/ft2 y2 2 p = 0.744 1 – __ 4 lb/ft where y is in feet ) U = 25 ft/s x p = 0 (gage) τw(y) = –τw(–y) τw x p=0 4 ft τw = τw (x) = (1.24 × 10–3)/ x lb/ft2 where x is in feet (a) (b) ■ Figure E9.1 SOLUTION For either orientation of the plate, the lift and drag are obtained from Eqs. 9.1 and 9.2. With the plate parallel to the upstream flow we have θ = 90° on the top surface and θ = 270° on the bottom surface so that the lift and drag are given by ℒ=− ∫ p dA + top ∫ stress is symmetrical about the center of the plate. With the given relatively large pressure on the front of the plate (the center of the plate is a stagnation point) and the negative pressure (less than the upstream pressure) on the back of the plate, we obtain the following drag p dA = 0 𝒟= bottom and 𝒟= ∫ τw dA + top ∫ τw dA = 2 bottom ∫ τw dA ∫ 4 ft ( or top x=0 1.24 × 10−3 x12 lbft2) (10 ft) dx or 𝒟 = 0.0992 lb (Ans) With the plate perpendicular to the upstream flow, we have θ = 0° on the front and θ = 180° on the back. Thus, from Eqs. 9.1 and 9.2 ℒ= ∫ τw dA − front ∫ 𝒟 = 55.6 lb for the drag. On the ultimately streamlined body (a zero-thickness flat plate parallel to the flow) the drag is entirely due to the shear stress at the surface and, in this example, is relatively small. For the ultimately blunted body (a flat plate normal to the upstream flow) the drag is entirely due to the pressure difference between the front and back portions of the object and, in this example, is relatively large. If the flat plate were oriented at an arbitrary angle relative to the upstream flow as indicated in Fig. E9.1c, there would be both a lift and a drag, each of which would be dependent on both the shear stress and the pressure. Both the pressure and shear stress distributions would be different for the top and bottom surfaces. τw dA = 0 Low p U ∫ front p dA − (Ans) COMMENTS Clearly there are two mechanisms responsible back and 𝒟=   −(−0.893) lbft2] (10 ft) dy (1) where we have used the fact that because of symmetry the shear stress distribution is the same on the top and the bottom surfaces, as is the pressure also [whether we use gage (p = 0) or absolute (p = patm ) pressure]. There is no lift generated—the plate does not know up from down. With the given shear stress distribution, Eq. 1 gives 𝒟=2 2 ft y2 0.744 1 − lbft2 [ ( 4) y=−2 ∫ ∫ p dA τw 𝒟≠0 High p τw back Again there is no lift because the pressure forces act parallel to the upstream flow (in the direction of 𝒟 not ℒ) and the shear ℒ≠0 (c) ■ Figure E9.1 (Continued) Although Eqs. 9.1 and 9.2 are valid for any body, the difficulty in their use lies in obtaining the appropriate shear stress and pressure distributions on the body surface. Considerable effort has gone into determining these quantities, but because of the various complexities involved, such information is available only for certain simple situations. Note that even in Example 9.1, the shear stress and pressure distributions were given. 9.1 General External Flow Characteristics 487 Without detailed information concerning the shear stress and pressure distributions on a body, Eqs. 9.1 and 9.2 cannot be used. The widely used alternative is to define dimensionless lift and drag coefficients and determine their approximate values by means of either a simplified analysis, some numerical technique, or an appropriate experiment. The lift coefficient, CL, and drag coefficient, CD, are defined as Lift coefficients and drag coefficients are dimensionless forms of lift and drag. CL = ℒ 1 2 2 ρU A   and CD = 𝒟 1 2 2 ρU A   where A is a characteristic area of the object (see Chapter 7). Typically, A is taken to be frontal area—the projected area seen by a person looking toward the object from a direction parallel to the upstream velocity, U, as indicated by the figure in the margin. It would be the area of the shadow of the object projected onto a screen normal to the upstream velocity as formed by a light shining along the upstream flow. In other situations A is taken to be the planform area—the projected area seen by an observer looking toward the object from a direction normal to the upstream velocity (i.e., from “above” it). Obviously, which characteristic area is used in the definition of the lift and drag coefficients must be clearly stated. Planform area A = c𝓵 Frontal area A = t𝓵 t c U 𝓵 9.1.2 Characteristics of Flow Past an Object 100 Air 10 106 Re 4 10 U, m/s = 1 0.1 102 0.01 1 0.001 0.001 0.01 0.1 𝓵, m 1 10 External flows past objects encompass an extremely wide variety of fluid mechanics phenomena. Clearly the character of the flow field is a function of the shape of the body and its orientation relative to the flow. Flows past relatively simple geometric shapes (i.e., a sphere or circular cylinder) are expected to have less complex flow fields than flows past a complex shape such as an airplane or a tree. However, even the simplest-shaped objects produce rather complex flows. For a given-shaped object, the characteristics of the flow depend very strongly on various parameters such as size, orientation, speed, and fluid properties. As is discussed in Chapter 7, according to dimensional analysis arguments, the character of the flow should depend on the various dimensionless parameters involved. For typical external flows the most important of these parameters are the Reynolds number, Re = ρUℓμ = Uℓν, the Mach number, Ma = Uc, and for flows with a free surface (i.e., flows with an interface between two fluids, such as the flow past a surface ship), the Froude number, Fr = U √gℓ. (Recall that ℓ is some characteristic length of the object and c is the speed of sound.) For the present, we consider how the external flow and its associated lift and drag vary as a function of the Reynolds number. Recall that the Reynolds number represents the ratio of inertial effects to viscous effects. In the absence of all viscous effects (μ = 0), the Reynolds number is infinite. On the other hand, in the absence of all inertial effects (negligible mass or ρ = 0), the Reynolds number is zero. Clearly, any actual flow will have a Reynolds number between (but not including) these two extremes. The nature of the flow past a body depends strongly on whether Re ⪢ 1 or Re ⪡ 1. Most external flows with which we are familiar are associated with moderately sized objects with a characteristic length on the order of 0.01 m < ℓ < 10 m. In addition, typical upstream velocities are on the order of 0.01 ms < U < 100 ms, and the fluids involved are typically water or air. The resulting Reynolds number range for such flows is approximately 10 < Re < 109. This is shown by the figure in the margin for air. As a rule of thumb, flows with Re > 100 are dominated by inertial effects, whereas flows with Re < 1 are dominated by viscous effects. Hence, most familiar external flows are dominated by inertia. On the other hand, there are many external flows in which the Reynolds number is considerably less than 1, indicating in some sense that viscous forces are more important than inertial forces. The gradual settling of small particles of dirt in a lake or stream is governed by low Reynolds number flow principles because of the small diameter of the particles and their small settling speed. Similarly, the Reynolds number for objects moving through large-viscosity oils is small because μ 488 Chapter 9 ■ Flow over Immersed Bodies For low Reynolds number flows, viscous effects are felt far from the object. is large. The general differences between small and large Reynolds number flow past streamlined and blunt objects can be illustrated by considering flows past two objects—one a flat plate parallel to the upstream velocity and the other a circular cylinder. Flows past three flat plates of length ℓ with Re = ρUℓμ = 0.1, 10, and 107 are shown in Fig. 9.5. If the Reynolds number is small, the viscous effects are relatively strong and the plate affects the uniform upstream flow far ahead, above, below, and behind the plate. To reach that portion of the flow field where the velocity has been altered by less than 1% of its undisturbed value (i.e., U − u < 0.01 U), we must travel relatively far from the plate. In low Reynolds number flows, the viscous effects are felt far from the object in all directions. As the Reynolds number is increased (by increasing U, for example), the region in which viscous effects are important becomes smaller in all directions except downstream, as is shown in Fig. 9.5b. One does not need to travel very far ahead, above, or below the plate to reach areas in which the viscous effects of the plate are not felt. The streamlines are displaced from their original Viscous effects important Re = U𝓵/v = 0.1 u < 0.99U y U 𝓵 x Streamlines deflected considerably U (a) Viscous effects important Re = 10 Viscosity not important y u < 0.99U x Streamlines deflected somewhat U U (b) y 7 Re = 10 Viscosity not important Boundary layer Wake region Viscous effects important δ << 𝓵 x Streamlines deflected very slightly U U (c) ■ Figure 9.5 Character of the steady, viscous flow past a flat plate parallel to the upstream velocity: (a) low Reynolds number flow, (b) moderate Reynolds number flow, (c) large Reynolds number flow. 9.1 Thin boundary layers may develop in large Reynolds number flows. © Stanford University, with permission Flow past an array of cylinders. (Photograph courtesy of ONERA, The French Aerospace Lab.) VIDEO V9.2 Streamlined and blunt bodies General External Flow Characteristics 489 uniform upstream conditions, but the displacement is not as great as for the Re = 0.1 situation shown in Fig. 9.5a. If the Reynolds number is large (but not infinite), the flow is dominated by inertial effects and the viscous effects are negligible everywhere except in a region very close to the plate and in the relatively thin wake region behind the plate, as shown in Fig. 9.5c. Since the fluid viscosity is not zero (Re < ∞), it follows that the fluid must stick to the solid surface (the no-slip boundary condition). There is a thin boundary layer region of thickness δ = δ(x) ⪡ ℓ (i.e., thin relative to the length of the plate) next to the plate in which the fluid velocity changes from the upstream value of u = U to zero velocity on the plate. The thickness of this layer increases in the direction of flow, starting from zero at the forward or leading edge of the plate. The flow within the boundary layer may be laminar or turbulent, depending on various parameters involved. The streamlines of the flow outside of the boundary layer are nearly parallel to the plate. As we will see in the next section, the slight displacement of the external streamlines that are outside of the boundary layer is due to the thickening of the boundary layer in the direction of flow. The existence of the plate has very little effect on the streamlines outside of the boundary layer—either ahead, above, or below the plate. On the other hand, the wake region is due entirely to the viscous interaction between the fluid and the plate. One of the great advancements in fluid mechanics occurred in 1904 as a result of the insight of Ludwig Prandtl (1875–1953), a German physicist and aerodynamicist. He conceived of the idea of the boundary layer—a thin region on the surface of a body in which viscous effects are very important and outside of which the fluid behaves essentially as if it were inviscid. Clearly the actual fluid viscosity is the same throughout; only the relative importance of the viscous effects (due to the velocity gradients) is different within or outside of the boundary layer. As is discussed in the next section, by using such a hypothesis it is possible to simplify the analysis of large Reynolds number flows, thereby allowing solution to external flow problems that are otherwise still unsolvable. As with the flow past the flat plate described above, the flow past a blunt object (such as a circular cylinder) also varies with Reynolds number. In general, the larger the Reynolds number, the smaller the region of the flow field in which viscous effects are important. For objects that are not sufficiently streamlined, however, an additional characteristic of the flow is observed. This is termed flow separation and is illustrated by the figure in the margin and in Fig. 9.6. Low Reynolds number flow (Re = UDν < 1) past a circular cylinder is characterized by the fact that the presence of the cylinder and the accompanying viscous effects are felt throughout a relatively large portion of the flow field. As is indicated in Fig. 9.6a, for Re = UDν = 0.1, the viscous effects are important several diameters in any direction from the cylinder. A somewhat surprising characteristic of this flow is that the streamlines are essentially symmetric about the center of the cylinder—the streamline pattern is the same in front of the cylinder as it is behind the cylinder. As the Reynolds number is increased, the region ahead of the cylinder in which viscous effects are important becomes smaller, with the viscous region extending only a short distance ahead of the cylinder. The viscous effects are convected downstream, and the flow loses its upstream to downstream symmetry. Another characteristic of external flows becomes important— the flow separates from the body at the separation location as indicated in Fig. 9.6b. With the increase in Reynolds number, the fluid inertia becomes more important and at some location on the body, denoted the separation location, the fluid’s inertia is such that it cannot follow the curved path around to the rear of the body. (This concept will be further explored in Section 9.2.6.) The result is a separation bubble behind the cylinder in which some of the fluid is actually flowing upstream, against the direction of the upstream flow. (See the photograph in the margin showing laminar flow past an array of cylinders.) At still larger Reynolds numbers, the area affected by the viscous forces is forced closer to the cylinder until it involves only a thin (δ ⪡ D) boundary layer on the front portion of the cylinder and an irregular, unsteady (perhaps turbulent) wake region that extends far downstream of the cylinder. The fluid in the region outside of the boundary layer and wake region flows as if it were inviscid. Of course, the fluid viscosity is the same throughout the entire flow field. Whether or not viscous effects are important depends on which region of the flow field we consider. The velocity gradients within the boundary layer and wake regions are much larger than those in the remainder of the flow field. Since the shear stress (i.e., viscous effect) is the product of the fluid viscosity and 490 Chapter 9 ■ Flow over Immersed Bodies U D x Viscous forces important throughout Re = UD/v = 0.1 (a) Viscous effects important Viscosity not important Separation location U x Separation bubble Re = 50 (b) Viscosity not important Boundary layer U VIDEO Boundary layer separation Viscous effects important Wake region δ << D D x Separated region Re = 105 V9.3 Human aerodynamic wake (c) ■ Figure 9.6 Character of the steady, viscous flow past a circular cylinder: (a) low Reynolds number flow, (b) moderate Reynolds number flow, (c) large Reynolds number flow. Most familiar flows involve large Reynolds numbers. the velocity gradient, it follows that viscous effects are confined to the boundary layer and wake regions. The characteristics described in Figs. 9.5 and 9.6 for flow past a flat plate and a circular cylinder are typical of flows past streamlined and blunt bodies, respectively. The nature of the flow depends strongly on the Reynolds number, and in practice, many bodies exhibit both streamlined and blunt characteristics as illustrated in Example 9.2. (See Ref. 31 for many examples illustrating this behavior.) Most familiar flows are similar to the large Reynolds number flows depicted in Figs. 9.5c and 9.6c, rather than the low Reynolds number flow situations. In the remainder of this chapter we will investigate more thoroughly these ideas and determine how to calculate the forces on immersed bodies. EXAMPLE 9 .2 Characteristics of Flow Past Objects GIVEN It is desired to experimentally determine the various characteristics of flow past a car as shown in Fig E9.2. The following tests could be carried out: (a) U = 20 mms flow of glycerin past a scale model that is 34-mm tall, 100-mm long, and 40-mm wide, (b) U = 20 mms air flow past the same scale model, or (c) U = 25 ms airflow past the actual car, which is 1.7-m tall, 5-m long, and 2-m wide. FIND Would the flow characteristics for these three situations be similar? Explain. ■ Figure E9.2 © Stanford University, with permission. 9.2 Boundary Layer Characteristics 491 SOLUTION The characteristics of flow past an object depend on the Reynolds number. For this instance we could pick the characteristic length to be the height, h, width, b, or length, ℓ, of the car to obtain three possible Reynolds numbers, Reh = Uhν, Reb = Ubν, and Reℓ = Uℓν. These numbers will be different because of the different values of h, b, and ℓ. Once we arbitrarily decide on the length we wish to use as the characteristic length, we must stick with it for all calculations when using comparisons between model and prototype. With the values of kinematic viscosity for air and glycerin obtained from Tables 1.8 and 1.6 as νair = 1.46 × 10−5 m2s and νglycerin = 1.19 × 10−3 m2s, we obtain the following Reynolds numbers for the flows described. Reynolds Number Reh Reb Reℓ (a) Model in Glycerin 0.571 0.672 1.68 (b) Model in Air 46.6 54.8 137.0 (c) Car in Air 2.91 × 106 3.42 × 106 8.56 × 106 Clearly, the Reynolds numbers for the three flows are quite different (regardless of which characteristic length we choose). Based on the previous discussion concerning flow past a flat plate or flow past a circular cylinder, we would expect that the 9.2 flow past the actual car would behave in some way similar to the flows shown in Figs. 9.5c or 9.6c. That is, we would expect some type of boundary layer characteristic in which viscous effects would be confined to relatively thin layers near the surface of the car and the wake region behind it. Whether the car would act more like a flat plate or a cylinder would depend on the amount of streamlining incorporated into the car’s design. Because of the small Reynolds number involved, the flow past the model car in glycerin would be dominated by viscous effects, in some way reminiscent of the flows depicted in Figs. 9.5a or 9.6a. Similarly, with the moderate Reynolds number involved for the airflow past the model, a flow with characteristics similar to those indicated in Figs. 9.5b and 9.6b would be expected. Viscous effects would be important—not as important as with the glycerin flow, but more important than with the full-sized car. It would not be a wise decision to expect the flow past the full-sized car to be similar to the flow past either of the models. The same conclusions result regardless of whether we use Reh, Reb, or Reℓ. As is indicated in Chapter 7, the flows past the model car and the full-sized prototype will not be similar unless the Reynolds numbers for the model and prototype are the same. It is not always an easy task to ensure this condition. One (expensive) solution is to test full-sized prototypes in very large wind tunnels (see Fig. 9.1), or small-scale models in smaller wind tunnels at higher velocity, as long as the Mach number is kept small. Boundary Layer Characteristics Large Reynolds number flow fields may be divided into viscous and inviscid regions. As was discussed in the previous section, it is often possible to treat flow past an object as a combination of viscous flow in the boundary layer and inviscid flow elsewhere. If the Reynolds number is large enough, viscous effects are important only in the boundary layer regions near the object (and in the wake region behind the object). The boundary layer is the result of the no-slip boundary condition that requires the fluid to cling to any solid surface that it flows past. Outside of the boundary layer the velocity gradients normal to the flow are relatively small, and the fluid acts as if it were inviscid, even though the viscosity is not zero. A necessary condition for this structure of the flow is that the Reynolds number be large. 9.2.1 Boundary Layer Structure and Thickness on a Flat Plate VIDEO V9.4 Laminar boundary layer There can be a wide variety in the size of a boundary layer and the structure of the flow within it. Part of this variation is due to the shape of the object on which the boundary layer forms. In this section we consider the simplest situation, one in which the boundary layer is formed on an infinitely long flat plate along which flows a viscous, incompressible fluid as is shown in Fig. 9.7. If the surface were curved or experiences a pressure gradient (i.e., a circular cylinder or an airfoil), the boundary layer structure would be more complex. Such flows are discussed in Section 9.2.6. If the Reynolds number is sufficiently large, only the fluid in a relatively thin boundary layer on the plate will feel the effect of the plate. That is, except in the region next to the plate the flow velocity will be essentially V = U ˆi , the upstream velocity also known as the free-stream velocity. For the infinitely long flat plate extending from x = 0 to x = ∞, it is not obvious how to define the Reynolds number because there is no characteristic length. The plate has no thickness and is not of finite length! 492 Chapter 9 ■ Flow over Immersed Bodies U U U U Edge of boundary layer Fluid particle δ x Leading edge x=0 Laminar boundary layer Turbulent boundary layer ■ Figure 9.7 Distortion of a fluid particle as it flows within the boundary layer. Fluid particles within the boundary layer experience viscous effects. VIDEO V9.5 Laminar/ turbulent transition 100 Re x = 3 × 106 cr xcr, m 10 1 0.1 Re x = 2 × 105 cr 0.01 1 10 U, m/s (air) 100 For a finite length plate, it is clear that the plate length, ℓ, can be used as the characteristic length. For an infinitely long plate we use x, the coordinate distance along the plate from the leading edge, as the characteristic length and define the Reynolds number as Rex = Uxν. Thus, for any fluid or upstream velocity the Reynolds number will be sufficiently large for boundary layer type flow (i.e., Fig. 9.5c) if the plate is long enough. Physically, this means that the flow situations illustrated in Fig. 9.5 could be thought of as occurring on the same plate but should be viewed by looking at longer portions of the plate as we step away from the plate to see the flows in Fig. 9.5a, 9.5b, and 9.5c, respectively. If the plate is sufficiently long, the Reynolds number Re = Uℓν is sufficiently large so that the flow takes on its boundary layer character (except very near the leading edge). The details of the flow field near the leading edge are lost to our eyes because we are standing so far from the plate that we cannot make out these details. On this scale (Fig. 9.5c) the plate has negligible effect on the fluid ahead of the plate. The presence of the plate is felt only in the relatively thin boundary layer and wake regions. As previously noted, Prandtl in 1904 was the first to hypothesize such a concept. It is one of the major turning points in fluid mechanics. A better appreciation of the structure of the boundary layer flow can be obtained by considering what happens to a fluid particle that flows into the boundary layer. As is indicated in Fig. 9.7, a small rectangular particle retains its original shape as it flows in the uniform flow outside of the boundary layer. Once it enters the boundary layer, the particle begins to distort because of the velocity gradient within the boundary layer—the top of the particle has a larger speed than its bottom. The fluid particles do not rotate as they flow along outside the boundary layer, but they begin to rotate once they pass through the fictitious boundary layer surface and enter the world of viscous flow. The flow is said to be irrotational outside the boundary layer and rotational within the boundary layer. (In terms of the kinematics of fluid particles as is discussed in Section 6.1, the flow outside the boundary layer has zero vorticity, and the flow within the boundary layer has nonzero vorticity.) At some distance downstream from the leading edge, the boundary layer flow becomes turbulent and the fluid particles become greatly distorted because of the random, irregular nature of the turbulence as discussed in Sections 8.3.1 and 8.3.2. One of the distinguishing features of turbulent flow is the occurrence of irregular mixing of fluid particles that range in size from the smallest fluid particles up to those comparable in size with the object of interest. For laminar flow, mixing occurs only on the molecular scale. This molecular scale is orders of magnitude smaller in size than typical size scales for turbulent flow mixing. The transition from a laminar boundary layer to a turbulent boundary layer occurs at a critical value of the Reynolds number, Rexcr, on the order of 2 × 105 to 3 × 106, depending on the roughness of the surface and the amount of turbulence in the upstream flow, as is discussed in Section 9.2.4. As shown by the figure in the margin, the location along the plate where the flow becomes turbulent, xcr, moves toward the leading edge as the freestream velocity increases. The purpose of the boundary layer is to allow the fluid to change its velocity from the upstream (or free-stream) value of U to zero on the surface. Thus, V = 0 at y = 0 and V ≈ Uiˆ at the edge of the boundary layer, with the velocity profile, u = u(x, y), bridging the boundary layer thickness, δ. 9.2 y μ=0 u=U a u = 0.99 U U Boundary Layer Characteristics 493 b U U μ ≠0 u = u(y) Equal areas δ U–u δ* a b (a) (b) ■ Figure 9.8 Boundary layer thickness: (a) standard boundary layer thickness, (b) boundary layer displacement thickness. This boundary layer characteristic occurs in a variety of flow situations, not just on flat plates. For example, boundary layers form on the surfaces of cars, in the water running down the gutter of the street, and in the atmosphere as the wind blows across the surface of the Earth (land or water). THE WIDE WORLD OF FLUIDS The albatross: Nature’s aerodynamic solution for long tion, the albatross then makes use of a technique called flights The albatross is a phenomenal seabird that soars just “dynamic soaring” to take advantage of the wind profile over the above ocean waves, taking advantage of the local boundary ocean surface. Based on the boundary layer profile, the alba- layer to travel incredible distances with little to no wing flapping. tross uses the rule of dynamic soaring, which is to climb when This limited physical exertion results in minimal energy con- pointed upwind and dive when pointed downwind, thus con- sumption and, combined with aerodynamic optimization, allows stantly exchanging kinetic and potential energy. Though the the albatross to easily travel 1000 km (620 miles) per day, with albatross loses energy to drag, it can periodically regain energy some tracking data showing almost double that amount. The due to vertical and directional motions within the boundary layer albatross has high aspect ratio wings (up to 11 ft in wingspan) by changing local airspeed and direction. This is not a direct line and a lift-to-drag ratio (ℒ / 𝒟) of approximately 27, both similar to of travel, but it does provide the most efficient method of long- high-performance sailplanes. With this aerodynamic configura- distance flight. VIDEO V9.6 Boundary layer thickness In actuality (both mathematically and physically), there is no sharp “edge” to the boundary layer; that is, u → U as we get farther from the plate. We define the boundary layer thickness, δ, as that distance from the plate at which the fluid velocity is within some arbitrary value of the upstream/free-stream velocity. Typically, as indicated in Fig. 9.8a, δ=y The boundary layer displacement thickness is defined in terms of volumetric flowrate. where u = 0.99U To remove this arbitrariness (i.e., what is so special about 99%; why not 98%?), the following definitions are introduced. Shown in Fig. 9.8b are two velocity profiles for flow past a flat plate— one if there were no viscosity (a uniform profile) and the other if there were viscosity and zero slip at the wall (the boundary layer profile). Because of the velocity deficit, U − u, within the boundary layer, the flowrate across section b–b is less than that across section a–a. However, if we displace the plate at section a–a by an appropriate amount δ*, the boundary layer displacement thickness, the flowrates across each section will be identical. This is true if δ*b U =   ∫ ∞ (U − u)b dy 0 where b is the plate width. Thus, δ* = ∫ 0 ∞ u (1 − U ) dy (9.3) The displacement thickness represents the amount that the thickness of the body must be increased so that the fictitious uniform inviscid flow has the same mass flowrate properties as the actual viscous flow. It represents the outward displacement of the streamlines caused by the viscous 494 Chapter 9 ■ Flow over Immersed Bodies effects on the plate. This idea allows us to simulate the presence that the boundary layer has on the flow outside of the boundary layer by adding the displacement thickness to the actual wall and treating the flow over the thickened body as an inviscid flow. The displacement thickness concept is illustrated in Example 9.3. EXAMPLE 9 .3 Boundary Layer Displacement Thickness GIVEN Air flowing into a 2-ft-square duct with a uniform velocity of 10 fts forms a boundary layer on the walls as shown in Fig. E9.3a. The fluid within the core region (outside the boundary layers) flows as if it were inviscid. From advanced calculations it is determined that for this flow the boundary layer displacement thickness is given by δ* = 0.0070(x) 12 U1 = 10 ft/s Viscous effects important Inviscid core 2-ft square (1) U(x) δ∗ (1) (2) where δ* and x are in feet. x FIND Determine the velocity U = U(x) of the air within the (a) duct but outside of the boundary layer. 12 10 If we assume incompressible flow (a reasonable assumption because of the low velocities involved), it follows that the volume flowrate across any section of the duct is equal to that at the entrance (i.e., Q1 = Q2 ). That is, U1A1 = 10 fts (2 ft) 2 = 40 ft3s = ∫ 4 2 0 u dA = U(2 ft − 2δ*) 2 60 80 100 (b) 1 1 p1 + 2ρU21 = p + 2ρU2 Hence, with ρ = 2.38 × 10−3 slugsft3 and p1 = 0 we obtain 1 ρ (U21 − U2 ) 2 1 = (2.38 × 10−3 slugsft3 ) 2 102 × [ (10 fts) 2 − ft2s2] (1 − 0.0079x12 ) 4 p= 40 ft3s = 4U(1 − 0.0070x12 ) 2 or (1 − 0.0070x12 ) 2 40 layer. It is, however, valid for the inviscid flow outside the boundary layer.) Thus, (2) By combining Eqs. 1 and 2 we obtain 10 20 ■ Figure E9.3 (2) U= 0 x, ft According to the definition of the displacement thickness, δ*, the flowrate across section (2) is the same as that for a uniform flow with velocity U through a duct whose walls have been moved inward by δ*. That is, ∫ 6 u dA (2) 40 ft3s = U, ft/s 8 SOLUTION fts (Ans) COMMENTS Note that U increases in the downstream direc- tion. For example, as shown in Fig. E9.3b, U = 11.6 fts at x = 100 ft. The viscous effects that cause the fluid to stick to the walls of the duct reduce the effective size of the duct, thereby (from conservation of mass principles) causing the fluid to accelerate. (See also Section 8.1.2.) The pressure drop necessary to do this can be obtained by using the Bernoulli equation (Eq. 3.7) along the inviscid streamlines from section (1) to (2). (Recall that this equation is not valid for viscous flows within the boundary or p = 0.119 [1 − 1 lbft2 (1 − 0.0070x12 ) 4 ] For example, p = −0.0401 lbft2 at x = 100 ft. If it were desired to maintain a constant velocity along the centerline of this entrance region of the duct, the walls could be displaced outward by an amount equal to the boundary layer displacement thickness, δ*. 9.2 Boundary Layer Characteristics 495 Another boundary layer thickness definition, the boundary layer momentum thickness, ϴ, is often used when determining the drag on an object. Again because of the velocity deficit, U − u, in the boundary layer, the momentum flux across section b–b in Fig. 9.8 is less than that across section a–a. This deficit in momentum flux for the actual boundary layer flow on a plate of width b is given by ∫ The boundary layer momentum thickness is defined in terms of momentum flux. ρu(U − u) dA = ρb ∫ ∞ u(U − u) dy 0 which by definition is the momentum flux in a layer of uniform speed U and thickness ϴ. That is, ρbU2 ϴ = ρb ∫ ∞ u(U − u) dy 0 or ϴ= ∫ 0 ∞ u u 1 − ) dy ( U U (9.4) All three boundary layer thickness definitions, δ, δ*, and ϴ, are of use in boundary layer analyses. The boundary layer concept is based on the fact that the boundary layer is thin. For the flat plate flow this means that at any location x along the plate, δ ⪡ x. Similarly, δ* ⪡ x and ϴ ⪡ x. Again, this is true if we do not get too close to the leading edge of the plate (i.e., not closer than Rex = Uxν = 1000 or so). The structure and properties of the boundary layer flow depend on whether the flow is laminar or turbulent. As is illustrated in Fig. 9.9 and discussed in Sections 9.2.2 through 9.2.5, both the boundary layer thickness and the wall shear stress are different in these two regimes. 9.2.2 Prandtl / Blasius Boundary Layer Solution In theory, the details of viscous, incompressible flow past any object can be obtained by solving the governing Navier–Stokes equations discussed in Section 6.8.2. For steady, two-dimensional laminar flows with negligible gravitational effects, these equations (Eqs. 6.127a, b, and c) reduce to the following: u ∂u ∂u 1 ∂p ∂ 2u ∂ 2u +υ =− + ν( 2 + 2 ) ρ ∂x ∂x ∂y ∂x ∂y (9.5) u ∂υ ∂υ 1 ∂p ∂ 2υ ∂ 2υ +υ =− + ν( 2 + 2 ) ρ ∂y ∂x ∂y ∂x ∂y (9.6) which express Newton’s second law. In addition, the conservation of mass equation, Eq. 6.31, for steady, two-dimensional incompressible flow is ∂u ∂υ + =0 ∂x ∂y δ Re x cr x τw Laminar Turbulent ■ Figure 9.9 Typical characteristics of boundary layer x thickness and wall shear stress for laminar and turbulent boundary layers. (9.7) 496 Chapter 9 ■ Flow over Immersed Bodies The appropriate boundary conditions are that the fluid velocity (both u and υ) far from the body is the upstream velocity and zero at solid body surfaces (i.e., the fluid sticks to the surface). Although the mathematical problem is well-posed, no one has obtained a general analytical solution to these equations applying to all flows past arbitrary body shapes! Although some analytical solutions exist for specific flows which allow simplifications to the equations, currently much work is being done to obtain numerical solutions to these governing equations for many flow geometries. By using boundary layer concepts introduced in the previous sections, Prandtl was able to impose certain approximations (valid for large Reynolds number flows) and thereby to simplify the governing equations. In 1908, H. Blasius (1883–1970), one of Prandtl’s students, was able to solve these simplified equations for the laminar boundary layer flow past a flat plate parallel to the flow. A brief outline of this technique and the results are presented below. Additional details may be found in the literature (Refs. 1–3). Since the boundary layer is thin, it is expected that the component of velocity normal to the plate is much smaller than that parallel to the plate and that the rate of change of any parameter across the boundary layer should be much greater than that along the flow direction. That is, υ ⪡ u and   ∂ ∂ ⪡ ∂x ∂y Physically, the flow is primarily parallel to the plate surface and any fluid property is convected downstream much more quickly than it is diffused across the streamlines. With these assumptions it can be shown that the governing equations (Eqs. 9.5, 9.6, and 9.7) reduce to the following boundary layer equations: ∂u ∂υ + =0 ∂x ∂y ∂u ∂u 1 ∂p ∂ 2u u +υ = − ( )+ ν 2 ρ ∂x ∂x ∂y ∂y The Navier–Stokes equations can be simplified for boundary layer flow analysis. (9.8) (9.9a) Although both these boundary layer equations and the original Navier–Stokes equations are nonlinear partial differential equations, there is a considerable difference between them. The y momentum equation has been eliminated, leaving only the original, unaltered continuity equation and a modified x momentum equation. For boundary layer flow over a flat plate the pressure is constant throughout the fluid. Thus for flat plate, boundary layer flow, the pressure gradient term in Eq. 9.9a is eliminated, leaving only the x and y components of velocity as unknowns. The flow represents a balance between viscous and inertial effects, with pressure playing no role as indicated in the following flat plate boundary layer equation: ∂u ∂u ∂ 2u u +υ = ν 2 (9.9b) ∂x ∂y ∂y y As shown by the figure in the margin, the boundary conditions for the governing flat plate boundary layer equations (9.8 and 9.9b) are that the fluid sticks to the plate u→U u = υ = 0 on y = 0 (9.10) and that outside of the boundary layer the flow is the uniform free-stream flow u = U. That is, u → U as y → ∞ u=υ =0 u (9.11) Mathematically, the free-stream velocity is approached asymptotically as one moves away from the plate. Physically, the flow velocity is within 1% of the free-stream velocity at a distance of δ from the plate (as defined in the previous section). In mathematical terms, the Navier–Stokes equations (Eqs. 9.5 and 9.6) and the continuity equation (Eq. 9.7) are elliptic equations, whereas the equations for boundary layer flow (Eqs. 9.8 and 9.9a or 9.9b) are parabolic equations. The nature of the solutions to these two sets of equations, therefore, is different. Physically, this fact translates to the idea that what happens downstream of a given location in a boundary layer cannot affect what happens upstream of that point. That is, whether the plate shown in Fig. 9.5c ends with length ℓ or is extended to length 2ℓ, the flow within the first segment of length ℓ will be the same. In addition, the presence of the plate has no 9.2 Boundary Layer Characteristics 497 effect on the flow ahead of the plate. On the other hand, ellipticity allows flow information to propagate in all directions, including upstream. In general, the solutions of nonlinear partial differential equations (such as the boundary layer equations, Eqs. 9.8 and 9.9) are extremely difficult to obtain. However, by applying a clever coordinate transformation and change of variables, Blasius reduced the flat plate partial differential equations to an ordinary differential equation that he was able to solve. A brief description of this process is given below. Additional details can be found in standard books dealing with boundary layer flow (Refs. 1, 2). It can be argued that in dimensionless form the boundary layer velocity profiles on a flat plate should be similar regardless of the location along the plate. That is, y u = g( ) U δ where g(yδ) is an unknown function to be determined. In addition, by applying an order of magnitude analysis of the forces acting on fluid within the boundary layer, it can be shown that the boundary layer thickness grows as the square root of x and inversely proportional to the square root of U. That is, νx 12 δ ∼( ) U Such a conclusion results from a balance between viscous and inertial forces within the boundary layer and from the fact that the velocity varies much more rapidly in the direction across the boundary layer than along it. Thus, we introduce the dimensionless similarity variable η = (Uνx) 12y and the stream function ψ = (ν x U) 12 f (η), where f = f(η) is an unknown function. Recall from Section 6.2.3 that the velocity components for two-dimensional flow are given in terms of the stream function as u = ∂ψ∂y and υ = −∂ψ∂x, which for this flow become The boundary layer equations can be written in terms of a similarity variable. u = Uf ′(η) (9.12) νU 12 υ = ( ) (η f ′ − f ) 4x (9.13) and with the notation ( ) ′ = ddη. We substitute Eqs. 9.12 and 9.13 into the governing equations, Eqs. 9.8 and 9.9b, to obtain (after considerable manipulation—see Problem 9.23) the following nonlinear, third-order ordinary differential equation: 2f ′″ + ff ″ = 0 f f′ → 1 1 As shown by the figure in the margin, the boundary conditions given in Eqs. 9.10 and 9.11 can be written as f = f ′ = 0 at η = 0 and f ′ → 1 as η → ∞ 1 f = f′ = 0 η y ν increasing (x = constant) u (9.14a) (9.14b) The original partial differential equations and boundary conditions have been reduced to an ordinary differential equation by use of the similarity variable η. The two independent variables, x and y, were combined into the similarity variable in a fashion that reduced the partial differential equations (and boundary conditions) to an ordinary differential equation. This type of reduction is not generally possible. For example, this method does not work on the full Navier–Stokes equations, although it does on the flat plate boundary layer equations (Eqs. 9.8 and 9.9b). Although there is no known analytical solution to Eq. 9.14, it is relatively easy to integrate this equation on a computer with the results considered an exact solution. The dimensionless boundary layer profile, uU = f ′(η), obtained by numerical solution of Eq. 9.14 (termed the Blasius solution), is sketched in Fig. 9.10a and is tabulated in Table 9.1. The velocity profiles at different x locations are similar in that there is only one curve necessary to describe the velocity at any point in the boundary layer. Because the similarity variable η contains both x and y, it is seen from Fig. 9.10b that the actual velocity profiles are a function of both x and y. The profile at location x1 is the same as that at x2 except that the y coordinate is stretched by a factor of (x2 x1 ) 12. You can achieve similar changes in the boundary layer velocity profile by altering the kinematic viscosity, ν, as shown by the figure in the margin. 498 Chapter 9 ■ Flow over Immersed Bodies δ ∼ x δ 3 = 3δ1 x 5 = 25 x1 5 u ≈ 0.99 at η ≈ 5 __ U U η = __ vx 1/2 ( )y 4 y δ 2 = 2δ1 x 4 = 16 x1 3 x 3 = 9 x1 2 δ1 1 0 x 2 = 4 x1 0 0.2 0.4 0.6 0.8 1.0 0 x = x1 U 0 u f ′ (η ) = __ U u (b) ( a) ■ Figure 9.10 Blasius boundary layer profile: (a) boundary layer profile in dimensionless form using the similarity variable η, (b) similar boundary layer profiles at different locations along the flat plate. From the solution it is found that uU ≈ 0.99 when η = 5.0. Thus, δ=5 νx AU (9.15) or 5 δ = x √Rex where Rex = Uxν. It can also be shown that the displacement and momentum thicknesses are given by For large Reynolds numbers the boundary layer is relatively thin. 1.721 δ* = x √Rex (9.16) 0.664 ϴ = x √Rex (9.17) and As postulated, the boundary layer is thin provided that Rex is large (i.e., δx → 0 as Rex → ∞). ■ Table 9.1 Laminar Flow along a Flat Plate (the Blasius Solution) η = y(Uνx) 12 0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 f ′(η) = uU η f ′(η) 0 0.1328 0.2647 0.3938 0.5168 0.6298 0.7290 0.8115 0.8761 3.6 4.0 4.4 4.8 5.0 5.2 5.6 6.0 ∞ 0.9233 0.9555 0.9759 0.9878 0.9916 0.9943 0.9975 0.9990 1.0000 9.2 Boundary Layer Characteristics 499 With the velocity profile known, it is an easy matter to determine the wall shear stress, τw = μ (∂u∂y) y=0, where the velocity gradient is evaluated at the plate. The value of ∂u∂y at y = 0 can be obtained from the Blasius solution to give τw = 0.332U 32 τw τw ~ 1 √x x ρμ B x (9.18) As indicated by Eq. 9.18 and illustrated in the figure in the margin, the shear stress decreases with increasing x because of the increasing thickness of the boundary layer—the velocity gradient at the wall decreases with increasing x. Also, τw varies as U32, not as U as it does for fully developed laminar pipe flow. These variations are discussed in Section 9.2.3. 9.2.3 Momentum Integral Boundary Layer Equation for a Flat Plate One of the important aspects of boundary layer theory is the determination of the drag caused by shear forces on a body. As was discussed in the previous section, such results can be obtained from the governing differential equations for laminar boundary layer flow. Since exact solutions are extremely difficult (or impossible) to obtain, it is of interest to have an alternative approximate method. The momentum integral method described in this section provides such an alternative. We consider the uniform flow past a flat plate and the fixed control volume as shown in Fig. 9.11. The fluid adjacent to the plate makes up the lower portion of the control surface. The upper surface coincides with the streamline just outside the edge of the boundary layer at section (2). It need not (in fact, does not) coincide with the edge of the boundary layer except at section (2). In agreement with advanced theory and experiment, we assume that the pressure is constant throughout the flow field for a flat plate. The flow entering the control volume at the leading edge of the plate [section (1)] is uniform, while the velocity of the flow exiting the control volume [section (2)] varies from the upstream velocity at the edge of the boundary layer to zero velocity on the plate. If we apply the x component of the momentum equation (Eq. 5.22) to the steady flow of fluid within the control volume we obtain ∑ Fx = ρ ∫ uV ∙ nˆ dA + ρ (1) ∫ uV ∙ nˆ dA (2) where for a plate of width b ∑ Fx = −𝒟 = − ∫ ∫ τw dA = −b plate τw dx (9.19) plate and 𝒟 is the drag that the plate exerts on the fluid. Note that the net force caused by the uniform pressure distribution does not contribute to this flow. Since the plate is solid and the upper surface of the control volume is a streamline, there is no flow through these areas. Thus, −𝒟 = ρ ∫ U(−U) dA + ρ (1) The drag on a flat plate depends on the velocity profile within the boundary layer. ∫ u2 dA (2) or 𝒟 = ρU2bh − ρb ∫ δ u2 dy (9.20) 0 y U Control surface Streamline U h Boundary layer edge (1) τ w(x) δ (x) u x (2) ■ Figure 9.11 Control volume used in the derivation of the momentum integral equation for boundary layer flow. 500 Chapter 9 ■ Flow over Immersed Bodies Although the height h is not known, it is known that for conservation of mass the flowrate through section (1) must equal that through section (2), or Uh = ∫ δ u dy 0 Drag on a flat plate is related to momentum deficit within the boundary layer. which can be written as ρU2bh = ρb δ Uu dy (9.21) 0 Thus, by combining Eqs. 9.20 and 9.21 we obtain the drag in terms of the deficit of momentum flux across the outlet of the control volume as 𝒟 = ρb y ∫ ∫ δ u(U − u) dy (9.22) 0 δ y U–u δ u(U – u) u The idea of a momentum deficit is illustrated in the figure in the margin. If the flow were inviscid, the drag would be zero, since we would have u ≡ U and the right-hand side of Eq. 9.22 would be zero. (This is consistent with the fact that τw = 0 if μ = 0.) Equation 9.22 points out the important fact that boundary layer flow on a flat plate is governed by a balance between shear drag (the left-hand side of Eq. 9.22) and a decrease in the momentum of the fluid (the right-hand side of Eq. 9.22). As x increases, δ increases and the drag increases. The thickening of the boundary layer is necessary to overcome the drag of the viscous shear stress on the plate. This is contrary to horizontal fully developed pipe flow in which the momentum of the fluid remains constant and the shear force is overcome by the pressure gradient along the pipe. The development of Eq. 9.22 and its use was first put forth in 1921 by T. von Kármán (1881– 1963), a Hungarian/German-American aerodynamicist. By comparing Eqs. 9.22 and 9.4 we see that the drag can be written in terms of the momentum thickness, ϴ, as 𝒟 = ρbU2 ϴ (9.23) Note that this equation is valid for laminar or turbulent flows. The shear stress distribution can be obtained from Eq. 9.23 by differentiating both sides with respect to x to obtain d𝒟 dϴ = ρbU2 dx dx (9.24) The increase in drag per length of the plate, d𝒟dx, occurs at the expense of an increase of the momentum boundary layer thickness, which represents a decrease in the momentum of the fluid. Since d𝒟 = τwbdx (see Eq. 9.19) it follows that d𝒟 = bτw dx (9.25) Hence, by combining Eqs. 9.24 and 9.25 we obtain the momentum integral equation for the boundary layer flow on a flat plate τw = ρU2 dϴ dx (9.26) Although no integral is visible in this form of the equation, recall that there is one “hiding” in the ϴ term (see Eq. 9.4). The usefulness of this relationship lies in the ability to obtain approximate results easily by using rather crude assumptions. For example, if we knew the detailed velocity profile in the boundary layer (i.e., the Blasius solution discussed in the previous section), we could evaluate either the right-hand side of Eq. 9.23 to obtain the drag or the right-hand side of Eq. 9.26 to obtain the shear stress. Fortunately, even a rather crude guess at the velocity profile will allow us to obtain reasonable drag and shear stress results from Eq. 9.26. This method is illustrated in Example 9.4. 9.2 E X AMPLE 9 .4 Boundary Layer Characteristics 501 Momentum Integral Boundary Layer Equation Consider the laminar flow of an incompressible fluid past a flat plate at y = 0. The boundary layer velocity profile is approximated as u = Uyδ for 0 ≤ y ≤ δ and u = U for y > δ, as is shown in Fig. E9.4. y FIND Determine the shear stress by using the momentum inte- δ GIVEN u=U gral equation. Compare these results with the Blasius results given by Eq. 9.18. SOLUTION From Eq. 9.26 the shear stress is given by τw = ρU 2 dϴ 0 (1) dx u = Uy/δ ■ Figure E9.4 u U while for laminar flow we know that τw = μ( ∂u∂y) y=0. For the assumed profile we have τw = μ U δ (2) This can be integrated from the leading edge of the plate, x = 0 (where δ = 0) to an arbitrary location x where the boundary layer thickness is δ. The result is and from Eq. 9.4 ϴ= ∫ ∞ 0 = ∫ 0 δ u u 1 − ) dy = U( U ∫ 0 δ 6μ δ2 = x 2 ρU u u 1 − ) dy U( U or y y ( δ )(1 − δ ) dy δ = 3.46 or δ ϴ= 6 (3) Note that as yet we do not know the value of δ (but suspect that it should be a function of x). By combining Eqs. 1, 2, and 3 we obtain the following differential equation for δ: μU ρU2 dδ = δ 6 dx or δ dδ = νx BU (4) Note that this approximate result (i.e., the velocity profile is not actually the simple straight line we assumed) compares favorably with the (much more laborious to obtain) Blasius result given by Eq. 9.15. The wall shear stress can also be obtained by combining Eqs. 1, 3, and 4 to give τw = 0.289U32 ρμ B x (Ans) Again this approximate result is fairly close (within 13%) to the Blasius value of τw given by Eq. 9.18. 6μ dx ρU As is illustrated in Example 9.4, the momentum integral equation, Eq. 9.26, can be used along with an assumed velocity profile to obtain reasonable, approximate boundary layer results. The accuracy of these results depends on how closely the shape of the assumed velocity profile approximates the actual profile. Thus, we consider a general velocity profile u = g(Y) for 0 ≤ Y ≤ 1 U Approximate velocity profiles are used in the momentum integral equation. and u = 1 for Y > 1 U where the dimensionless coordinate Y = yδ varies from 0 to 1 across the boundary layer. The dimensionless function g(Y) can be any shape we choose, although it should be a reasonable 502 Chapter 9 ■ Flow over Immersed Bodies approximation to the boundary layer profile, as shown by the figure in the margin. In particular, it should certainly satisfy the boundary conditions u = 0 at y = 0 and u = U at y = δ. That is, 1 u = g(Y ) U Y 0 0 u U g(0) = 0 and g(1) = 1 1 The linear function g(Y) = Y used in Example 9.4 is one such possible profile. Other conditions, such as dgdY = 0 at Y = 1 (i.e., ∂u∂y = 0 at y = δ), and d 2g/dY 2 = 0 at Y = 0 (i.e., ∂2u/∂y2 = 0 at y = 0) could also be incorporated into the function g(Y) to more closely approximate the actual profile. For a given g(Y), the drag can be determined from Eq. 9.22 as 𝒟 = ρb ∫ δ u(U − u) dy = ρbU2δ 0 ∫ 1 g(Y)[1 − g(Y)] dY 0 or 𝒟 = ρbU2δC1 (9.27) where the dimensionless constant C1 has the value ∫ C1 = 1 g(Y)[1 − g(Y)] dY 0 Also, the wall shear stress can be written as τw = μ ∂u ∂y | μU dg δ dY = y=0 where the dimensionless constant C2 has the value C2 = dg dY By combining Eqs. 9.25, 9.27, and 9.28 we obtain δ dδ = | | = Y=0 μU C δ 2 (9.28) Y=0 μC2 dx ρUC1 which can be integrated from δ = 0 at x = 0 to give δ= 2νC2x B UC1 or √2C2 C1 δ = x √Rex (9.29) By substituting this expression back into Eqs. 9.28 we obtain τw = Approximate boundary layer results are obtained from the momentum integral equation. C1C2 32 ρμ U B 2 B x (9.30) To use Eqs. 9.29 and 9.30 we must determine the values of C1 and C2. Several assumed velocity profiles and the resulting values of δ are given in Fig. 9.12 and Table 9.2. The more closely the assumed shape approximates the actual (i.e., Blasius) profile, the more accurate the final results. For any assumed profile shape, the functional dependence of δ and τw on the physical parameters ρ, μ, U, and x is the same. Only the constants are different. That is, δ ∼ (μxρU) 12 or δRex12 x = constant, and τw ∼ (ρμU3 x) 12, where Rex = ρUxμ. It is often convenient to use the dimensionless local friction coefficient, cf , defined as cf = τw 1 2 2 ρU (9.31) 9.2 Boundary Layer Characteristics 503 1.0 Linear Cubic Sine wave y __ δ 0.5 Blasius Parabolic 0 0 0.5 ■ Figure 9.12 Typical approximate boundary layer profiles used in the momentum integral equation. 1.0 u __ U to express the wall shear stress. From Eq. 9.30 we obtain the approximate value cf = √2C1C2 √2C1C2 μ = A ρUx √Rex while the Blasius solution result is given by cf = 0.664 (9.32) √Rex These results are also indicated in Table 9.2. For a flat plate of length ℓ and width b, the net friction drag, 𝒟f , can be expressed in terms of the friction drag coefficient, CDf , as CDf = The friction drag coefficient is an integral of the local friction coefficient. b 𝒟f 1 2 2 ρU bℓ ∫ ℓ τw dx 0 = 1 2 2 ρU bℓ or CDf = 1 ℓ ∫ ℓ (9.33) cf dx 0 ■ Table 9.2 Flat Plate Momentum Integral Results for Various Assumed Laminar Flow Velocity Profiles Profile Character a. Blasius solution b. Linear uU = yδ c. Parabolic uU = 2yδ − (yδ) 2 d. Cubic uU = 3(yδ)2 − (yδ) 32 e. Sine wave uU = sin[π(yδ)2] δR e12 x x cf Re12 x CDf Re12 ℓ 5.00 0.664 1.328 3.46 0.578 1.156 5.48 0.730 1.460 4.64 0.646 1.292 4.79 0.655 1.310 504 Chapter 9 ■ Flow over Immersed Bodies We use the above approximate value of cf = (2C1C2μρUx) 12 to obtain CDf = 0.04 CDf 0.03 Laminar boundary layer 0.02 √8C1C2 where Reℓ = Uℓν is the Reynolds number based on the plate length. The corresponding value obtained from the Blasius solution (Eq. 9.32) and shown by the figure in the margin gives CDf = 0.01 0.00 Re𝓵 √Reℓ 1.328 √Reℓ These results are also indicated in Table 9.2. The momentum integral boundary layer method provides a relatively simple technique to obtain useful boundary layer results. As is discussed in Sections 9.2.5 and 9.2.6, this technique can be extended to boundary layer flows on curved surfaces (where the pressure and fluid velocity at the edge of the boundary layer are not constant) and to turbulent flows. 9.2.4 Transition from Laminar to Turbulent Flow The boundary layer on a flat plate will become turbulent if the plate is long enough. VIDEO The analytical results given in Table 9.2 are restricted to laminar boundary layer flows along a flat plate with zero pressure gradient. They agree quite well with experimental results up to the point where the boundary layer flow becomes turbulent, which will occur for any free-stream velocity and any fluid provided the plate is long enough. This is true because the parameter that predicts the transition to turbulent flow is the Reynolds number—in this case the Reynolds number based on the distance from the leading edge of the plate, Rex = Uxν. The value of the Reynolds number at the transition location is a rather complex function of various parameters involved, including the roughness of the surface, the curvature of the surface (for example, a flat plate or a sphere), and some measure of the disturbances in the flow outside the boundary layer. On a flat plate with a sharp leading edge in a typical airstream, the transition takes place at a distance x from the leading edge given by Rexcr = 2 × 105 to 3 × 106. Unless otherwise stated, we will use Rexcr = 5 × 105 in our calculations. The actual transition from laminar to turbulent boundary layer flow may occur over a region of the plate, not at a specific single location. This occurs, in part, because of the spottiness of the transition. Typically, the transition begins at random locations on the plate in the vicinity of Rex = Rexcr. These spots grow rapidly as they are convected downstream until the entire width of the plate is covered with turbulent flow. The photo shown in Fig. 9.13 illustrates this transition process. The complex process of transition from laminar to turbulent flow involves the instability of the flow field. Small disturbances imposed on the boundary layer flow (i.e., from a vibration of the plate, a roughness of the surface, or a “wiggle” in the flow past the plate) will either grow (instability) or decay (stability), depending on where the disturbance is introduced into the flow. If these disturbances occur at a location with Rex < Rexcr they will die out, and the boundary layer will return to laminar flow at that location. Disturbances imposed at a location with Rex > Rexcr will grow and transform the boundary layer flow downstream of this location into turbulence. The study of the initiation, growth, and structure of these turbulent bursts or spots is an active area of fluid mechanics research. V9.7 Transition on flat plate ■ Figure 9.13 Turbulent spots and the transition from laminar to turbulent boundary layer flow on a flat plate. Flow from left to right. (Photograph courtesy of Brian Cantwell, Stanford University.) 9.2 Boundary Layer Characteristics 505 0.06 x = 5.25 ft x = 6.76 ft x = 8.00 ft 0.05 U = 89 ft/s; airflow y, ft 0.04 Transitional 0.03 Turbulent 0.02 0.01 Laminar 0 0 0.2 0.4 0.6 0.8 1 u __ U ■ Figure 9.14 Typical boundary layer profiles on a flat plate for laminar, transitional, and turbulent flow (Ref. 1). Transition from laminar to turbulent flow also involves a noticeable change in the shape of the boundary layer velocity profile. Typical profiles obtained in the neighborhood of the transition location are indicated in Fig. 9.14. The turbulent profiles are flatter, have a larger velocity gradient at the wall, and produce a larger boundary layer thickness than do the laminar profiles. E X AMPLE 9 .5 Boundary Layer Transition GIVEN A fluid flows steadily past a flat plate with a velocity of U = 10 fts. point if the fluid is (a) water at 60 °F, (b) standard air, or (c) glycerin at 68 °F? FIND At approximately what location will the boundary layer become turbulent, and how thick is the boundary layer at that SOLUTION For any fluid, the laminar boundary layer thickness on a flat plate is found from Eq. 9.15 as νx δ=5 AU The boundary layer remains laminar up to νRexcr xcr = U Thus, if we assume Rexcr = 5 × 105 we obtain xcr = 5 × 105 ν = 5 × 104 ν 10 fts and from Eq. (1) δcr ≡ δ x=xcr = 5[ 12 ν (5 × 104 ν) ] = 354 ν 10 (1) ■ Table E9.5 Fluid a. Water b. Air c. Glycerin ν (ft2s) xcr (ft) δcr (ft) 1.21 × 10−5 1.57 × 10−4 1.28 × 10−2 0.605 7.85 640.0 0.00428 0.0556 4.53 (Ans) where ν is in ft2s and xcr and δcr are in feet. The values of the kinematic viscosity obtained from Tables 1.5 and 1.7 are listed in Table E9.5 along with the corresponding xcr and δcr. COMMENT Laminar flow can be maintained on a longer portion of the plate if the viscosity is increased. However, the boundary layer flow eventually becomes turbulent, provided the plate is long enough. Similarly, the boundary layer thickness is greater if the viscosity is increased. 506 Chapter 9 ■ Flow over Immersed Bodies 9.2.5 Turbulent Boundary Layer Flow Plate © J. P. Crimaldi, M. B. Wiley, & J. R. Koseff Random transport of finite-sized fluid masses occurs within turbulent boundary layers. VIDEO V9.8 Random flow but not turbulent The structure of turbulent boundary layer flow is very complex, seemingly random, and irregular. It shares many of the characteristics described for turbulent pipe flow in Section 8.3. In particular, the velocity at any given location in the flow is unsteady in a random fashion. The flow can be thought of as a jumbled mix of intertwined eddies (or swirls) of different sizes (diameters and angular velocities). The figure in the margin shows a laser-induced fluorescence visualization of a turbulent boundary layer on a flat plate (side view). The various fluid quantities involved (i.e., mass, momentum, energy) are convected downstream in the free-stream direction as in a laminar boundary layer. For turbulent flow they are also convected across the boundary layer (in the direction perpendicular to the plate) by the random transport of relatively small, finite-sized fluid masses associated with the turbulent eddies. There is considerable mixing involved with these finite-sized eddies—considerably more than is associated with the mixing found in laminar flow where it is confined to the molecular scale. Although there is considerable random motion of fluid particles perpendicular to the plate, there is very little net transfer of mass across the boundary layer—the largest flowrate by far is parallel to the plate. There is, however, a considerable net transfer of x component of momentum perpendicular to the plate because of the random motion. Fluid masses moving toward the plate (in the negative y direction) have some of their excess momentum (they come from areas of higher velocity) removed by the plate. Conversely, masses moving away from the plate (in the positive y direction) gain momentum from the fluid (they come from areas of lower velocity). The net result is that the plate acts as a momentum sink, continually extracting momentum from the fluid. For laminar flows, such cross-stream transfer of these properties takes place solely on the molecular scale. For turbulent flow the randomness is associated with fluid mixing. Consequently, the shear force for turbulent boundary layer flow is considerably greater than it is for laminar boundary layer flow (see Section 8.3.2). There are no “exact” solutions for turbulent boundary layer flow. As is discussed in Section 9.2.2, it is possible to solve the Prandtl boundary layer equations for laminar flow past a flat plate to obtain the Blasius solution (which is “exact” within the framework of the assumptions involved in the boundary layer equations). Since there is no precise expression for the shear stress in turbulent flow (see Section 8.3), solutions are not available for turbulent flow. However, as introduced in Appendix A considerable headway has been made in obtaining numerical (computer) solutions for turbulent flow by using approximate shear stress models. Also, progress is being made in the area of direct, full numerical integration of the basic governing equations, the Navier–Stokes equations. Approximate turbulent boundary layer results can also be obtained by use of the momentum integral equation, Eq. 9.26, which is valid for either laminar or turbulent flat plate flow. What is needed for the use of this equation are reasonable approximations to the velocity profile u = U g(Y), where Y = yδ and u is the time-averaged velocity (the overbar notation, u, of Section 8.3.2 has been dropped for convenience), and a functional relationship describing the wall shear stress. For laminar flow the wall shear stress was used as τw = μ(∂u∂y) y=0. In theory, such a technique should work for turbulent boundary layers also. However, as is discussed in Section 8.3, the details of the velocity gradient at the wall are not well understood for turbulent flow. Thus, it is necessary to use some empirical relationship for the wall shear stress. This is illustrated in Example 9.6. EXAMPLE 9 .6 Turbulent Boundary Layer Properties GIVEN Consider turbulent flow of an incompressible fluid past a flat plate. The boundary layer velocity profile is assumed to be uU = (yδ) 17 = Y 17 for Y = yδ ≤ 1 and u = U for Y > 1 as shown in Fig. E9.6. This is a reasonable approximation of experimentally observed profiles, except very near the plate where this formula gives ∂u∂y = ∞ at y = 0. Note the differences between the assumed turbulent profile and the laminar profile. Also assume that the shear stress agrees with the experimentally determined formula: τw = 0.0225ρU2( ν 14 Uδ ) (1) FIND Determine the boundary layer thicknesses δ, δ*, and ϴ and the wall shear stress, τw, as a function of x. Determine the friction drag coefficient, CDf . 9.2 Boundary Layer Characteristics 507 SOLUTION Whether the flow is laminar or turbulent, it is true that the drag force is accounted for by a reduction in the momentum of the fluid flowing past the plate. The shear is obtained from Eq. 9.26 in terms of the rate at which the momentum boundary layer thickness, ϴ, increases with distance along the plate as 0.8 0.6 dϴ dx (1) ( ) 0.4 For the assumed velocity profile, the boundary layer momentum thickness is obtained from Eq. 9.4 as ϴ= ∫ ∞ 0 u u 1 − ) dy = δ U( U 1 ∫ 0 Laminar 0.2 u u 1 − ) dY U( U 0 or by integration ϴ=δ ∫ 1 Y 17 (1 − Y 17 ) dY = 0 0 0.5 u __ (2) ν 14 7 2 dδ 0.0225ρU2 ( Uδ ) = 72 ρU dx or ν 14 δ14 dδ = 0.231( ) dx U ■ Figure E9.6 You may have noticed that the assumed velocity profile is the one-seventh power-law introduced in Section 8.3.3. This assumed profile not only fails at the plate surface, but also fails at the outer edge of the boundary layer because the velocity gradient is not zero. The given shear stress formula is actually obtained by transforming an experimentally determined formula for the friction factor in pipe flow (due to Blasius) into a form suitable for boundary layers (for example, by replacing pipe radius with boundary layer thickness). Similarly, from Eq. 2, This can be integrated from δ = 0 at x = 0 to obtain ν 15 45 7 ϴ = 72 δ = 0.0360 (U) x 15 45 x (3) (Ans) or in dimensionless form δ 0.370 = x Re15 x Strictly speaking, the boundary layer near the leading edge of the plate is laminar, not turbulent, and the precise boundary condition should be the matching of the initial turbulent boundary layer thickness (at the transition location) with the thickness of the laminar boundary layer at that point. In practice, however, the laminar boundary layer often exists over a relatively short portion of the plate, and the error associated with starting the turbulent boundary layer with δ = 0 at x = 0 can be negligible. The displacement thickness, δ*, and the momentum thickness, ϴ, can be obtained from Eqs. 9.3 and 9.4 by integrating as follows: δ* = ∫ ∞ 0 =δ ∫ 0 u (1 − U ) dy = δ 1 ∫ 0 1 (4) (Ans) The functional dependence for δ, δ*, and ϴ is the same; only the constants of proportionality are different. Typically, ϴ < δ* < δ. By combining Eqs. 1 and 3, we obtain the following result for the wall shear stress τw = 0.0225ρU2 [ = 0.0288ρU 14 ν U(0.370) (νU) 15x45 ] 2 (Ans) Re15 x This can be integrated over the length of the plate to obtain the friction drag on one side of the plate, 𝒟f , as 𝒟f = ∫ ℓ bτw dx = b(0.0288ρU2 ) 0 ∫ ℓ 0 ν 15 ( Ux ) dx or 𝒟f = 0.0360ρU2 u (1 − U )dY A Re15 ℓ where A = bℓ is the area of the plate. (This result can also be obtained by combining Eq. 9.23 and the expression for the momentum thickness given in Eq. 4.) The corresponding friction drag coefficient, CDf, is δ (1 − Y17 ) dY = 8 Thus, by combining this with Eq. 3 we obtain ν 15 δ* = 0.0463( ) x45 U 1.0 U 7 δ 72 where δ is an unknown function of x. By combining the assumed shear force dependence (Eq. 1) with Eq. 2, we obtain the following differential equation for δ: ν δ = 0.370( ) U 1 __ y 7 u = __ __ U δ Turbulent y Y = __ δ τw = ρU2 1.0 (Ans) CDf = 𝒟f 1 2 2 ρU A = 0.0720 Re15 ℓ (Ans) 508 Chapter 9 ■ Flow over Immersed Bodies COMMENT Note that for the turbulent boundary layer flow the boundary layer thickness increases with x as δ ∼ x45 and the shear stress decreases as τw ∼ x−15. For laminar flow these dependencies are x12 and x−12, respectively. The random character of the turbulent flow causes a different structure of the flow. In general, the drag coefficient for a flat plate of length ℓ is a function of the Reynolds number, Reℓ, and the relative roughness, εℓ. The results of numerous experiments covering a wide range of the parameters of interest are shown in Fig. 9.15. For laminar boundary layer flow the drag coefficient is a function of only the Reynolds number—surface roughness is not important. This is similar to laminar flow in a pipe. However, for turbulent flow, the surface roughness does affect the shear stress and, hence, the drag coefficient. This is similar to turbulent pipe flow in which the surface roughness may protrude into or through the viscous sublayer next to the wall and alter the flow in this thin, but very important, layer (see Section 8.4.1). Values of the roughness, ε, for different materials can be obtained from Table 8.1. The drag coefficient diagram of Fig. 9.15 (boundary layer flow) shares many characteristics in common with the familiar Moody diagram (pipe flow) of Fig. 8.20, even though the mechanisms governing the mean flow are quite different. Fully developed horizontal pipe flow is governed by a balance between pressure forces and viscous forces. The fluid inertia remains constant throughout the flow. Boundary layer flow on a horizontal flat plate is governed by a balance between inertia effects and viscous forces. The pressure remains constant throughout the flow. (As is discussed in Section 9.2.6, for boundary layer flow on curved surfaces, the pressure is not constant.) On the other hand, the structure of the turbulence in both flows can be quite similar. It is often convenient to have an equation for the drag coefficient as a function of the Reynolds number and relative roughness rather than the graphical representation given in Fig. 9.15. Although there is not one equation valid for the entire Reℓ − εℓ range, the equations presented in Table 9.3 do work well for the conditions indicated. 0.014 5 × 10– 3 ε –3 𝓵 = 3 × 10 0.012 Completely turbulent 2 × 10–3 0.010 1 × 10–3 Turbulent 0.008 5 × 10– 4 CDf The flat plate drag coefficient is a function of relative roughness and Reynolds number. Obviously the results presented in this example are valid only in the range of validity of the original data—the assumed velocity profile and shear stress. This range covers smooth flat plates with 5 × 105 < Reℓ < 107. 2 × 10– 4 0.006 1 × 10– 4 5 × 10–5 2 × 10–5 0.004 5 × 10– 6 Transitional 1 × 10– 6 0.002 Laminar 0 105 Turbulent smooth plate 106 107 108 109 Re𝓵 ■ Figure 9.15 Friction drag coefficient for a flat plate parallel to the upstream flow (Ref. 18, with permission). 9.2 Boundary Layer Characteristics 509 ■ Table 9.3 Equations for the Flat Plate Drag Coefficient (Ref. 1) Equation CDf CDf CDf CDf Flow Conditions = 1.328(Reℓ ) = 0.455(log Reℓ ) 2.58 − 1700Reℓ = 0.455(log Reℓ ) 2.58 = [1.89 − 1.62 log(εℓ)]−2.5 0.5 E X AMPLE 9 .7 Laminar flow Transitional with Rexcr = 5 × 105 Turbulent, smooth plate Completely turbulent Drag on a Flat Plate GIVEN The water ski shown in Fig. E9.7a moves through FIND Estimate the drag caused by the shear stress on the 70 °F water with a velocity U. bottom of the ski for 0 < U < 30 fts. SOLUTION Clearly the ski is not a flat plate, and it is not aligned exactly parallel to the upstream flow. However, we can obtain a reasonable approximation to the shear force by using the flat plate results. That is, the friction drag, 𝒟f, caused by the shear stress on the bottom of the ski (the wall shear stress) can be determined as 1 𝒟f = 2ρU2ℓbCDf 𝒟f = 1.94(10) 2 (0.00308) = 0.598 lb By covering the range of upstream velocities of interest we obtain the results shown in Fig. E9.7b. (Ans) COMMENTS If Re ≲ 1000, the results of boundary layer With A = ℓb = 4 ft × 0.5 ft = 2 ft , ρ = 1.94 slugsft , and μ = 2.04 × 10−5 lb ˙ sft2 (see Table B.1) we obtain 2 3 1 𝒟f = 2 (1.94 slugsft3 ) (2.0 ft2 )U2CDf = 1.94 U2CDf (1) where 𝒟f and U are in pounds and fts, respectively. The friction coefficient, CDf, can be obtained from Fig. 9.15 or from the appropriate equations given in Table 9.3. As we will see, for this problem, much of the flow lies within the transition regime where both the laminar and turbulent portions of the boundary layer flow occupy comparable lengths of the plate. We choose to use the values of CDf from the table. For the given conditions we obtain (1.94 slugsft3 ) (4 ft)U ρUℓ = 3.80 × 105 U = μ 2.04 × 10−5 lb ∙ sft2 where U is in fts. With U = 10 fts, or Reℓ = 3.80 × 106, we obtain from Table 9.3 CDf = 0.455(log Reℓ ) 2.58 − 1700Reℓ = theory are not valid—inertia effects are not dominant enough and the boundary layer is not thin compared with the length of the plate. For our problem this corresponds to U = 2.63 × 10−3 fts. For all practical purposes U is greater than this value, and the flow past the ski is of the boundary layer type. The approximate location of the transition from laminar to turbulent boundary layer flow as defined by Recr = ρUxcr μ = 5 × 105 is indicated in Fig. E9.7b. Up to U = 1.31 fts the entire boundary layer is laminar. The fraction of the boundary layer that is laminar decreases as U increases until only the front 0.18 ft is laminar when U = 30 fts. For anyone who has water skied, it is clear that it can require considerably more force to be pulled along at 30 fts than the 2 × 4.88 lb = 9.76 lb (two skis) indicated in Fig. E9.7b. As is discussed in Section 9.3, the total drag on an object such as a water ski consists of more than just the friction drag. Other components, including pressure drag and wave-making drag, add considerably to the total resistance. 5 5 Entire boundary layer laminar 4 4 𝒟f 𝒟f , lb b = width = 0.5 ft x U 3 3 xcr , ft Reℓ = 0.00308. From Eq. 1 the corresponding drag is xcr 2 2 1 1 x = 4 ft = 𝓵 x=0 (a) 0 0 5 10 15 U, ft /s (b) ■ Figure E9.7 20 25 30 510 Chapter 9 ■ Flow over Immersed Bodies 9.2.6 Effects of Pressure Gradient Ufs ≠ Ufs 1 (2) 2 (1) p1 ≠ p2 The free-stream velocity on a curved surface is not constant. Ufs = 2U Ufs = U U Ufs = 0 The boundary layer discussions in the previous parts of Section 9.2 have dealt with flow along a flat plate in which the pressure is constant throughout the fluid. In general, when a fluid flows past an object other than a flat plate, the pressure field is not uniform. As shown in Fig. 9.6, if the Reynolds number is large, relatively thin boundary layers will develop along the surfaces. Within these layers the component of the pressure gradient in the streamwise direction (i.e., along the body surface) is not zero, although the pressure gradient normal to the surface is negligibly small. That is, if we were to measure the pressure while moving across the boundary layer from the body to the boundary layer edge, we would find that the pressure is essentially constant. However, the pressure does vary in the direction along the body surface if the body is curved, as shown by the figure in the margin. The variation in the free-stream velocity, Ufs, the fluid velocity at the edge of the boundary layer, is the cause of the pressure gradient in this direction. The characteristics of the entire flow (both within and outside of the boundary layer) are often highly dependent on the pressure gradient effects on the fluid within the boundary layer. For a flat plate parallel to the upstream flow, the upstream velocity (that far ahead of the plate) and the free-stream velocity (that at the edge of the boundary layer) are equal—U = Ufs. This is a consequence of the negligible thickness of the plate. For bodies of nonzero thickness, these two velocities are different. This can be seen in the flow past a circular cylinder of diameter D. The upstream velocity and pressure are U and p0, respectively. If the fluid were completely inviscid (μ = 0), the Reynolds number would be infinite (Re = ρUDμ = ∞) and the streamlines would be symmetrical, as are shown in Fig. 9.16a. The fluid velocity along the surface would vary from Ufs = 0 at the very front and rear of the cylinder (points A and F are stagnation points) to a maximum of Ufs = 2U at the top and bottom of the cylinder (point C). This is also indicated in the figure in the margin. The pressure on the surface of the cylinder would be symmetrical about the vertical midplane of the cylinder, reaching a maximum value of p0 + ρU2 2 (the stagnation pressure) at both the front and back of the cylinder, and a minimum of p0 − 3ρU2 2 at the top and bottom of the cylinder. The pressure and free-stream velocity distributions are shown in Figs. 9.16b and 9.16c. These characteristics can be obtained from potential flow analysis of Section 6.6.3. Because of the absence of viscosity (therefore, τw = 0) and the symmetry of the pressure distribution for inviscid flow past a circular cylinder, it is clear that the drag on the cylinder is zero. The pressure distribution is symmetrical, so the “push” experienced in the flow direction on the front of the cylinder is balanced by the equal push opposite to the flow direction on the back. The predication Ufs C U, p0 A θ F (a) p0 + A 1 ρU 2 2 F p p0 – Ufs 1 ρU 2 2 U p0 – ρ U 2 p0 – 3 ρU 2 2 0 C 2U p0 C 90 180 0 A 0 F 90 θ , degrees θ , degrees (b) (c) 180 ■ Figure 9.16 Inviscid flow past a circular cylinder: (a) streamlines for the flow if there were no viscous effects, (b) pressure distribution on the cylinder’s surface, (c) free-stream velocity on the cylinder’s surface. 9.2 The pressure gradient in the external flow is imposed throughout the boundary layer fluid. Separation location © Stanford University, with permission Boundary Layer Characteristics 511 of zero drag for an inviscid fluid even extends to a plate normal to the flow. Based on common experience, however, we know that there must be a net drag. Clearly, since there is no purely inviscid fluid, the reason for the observed drag must lie on the shoulders of the viscous effects. To test this hypothesis, we could conduct an experiment by measuring the drag on an object (such as a circular cylinder) in a series of fluids with decreasing values of viscosity. To our initial surprise we would find that no matter how small we make the viscosity (provided it is not precisely zero) we would measure a finite drag, essentially independent of the value of μ. As was noted in Section 6.6.3, this leads to what has been termed d’Alembert’s paradox—the drag on an object in an inviscid fluid is zero, but the drag on an object in a fluid with vanishingly small (but nonzero) viscosity is not zero. The reason for the above paradox can be described in terms of the effect of the pressure gradient on boundary layer flow. Consider large Reynolds number flow of a real (viscous) fluid past a circular cylinder. As was discussed in Section 9.1.2, we expect the viscous effects to be confined to thin boundary layers near the surface at the base of which the fluid sticks to the surface (V = 0)—a necessary condition for any real fluid (μ ≠ 0). The basic idea of boundary layer theory is that the boundary layer is thin enough so that it does not greatly disturb the flow outside the boundary layer. Based on this reasoning, for large Reynolds numbers the flow throughout most of the flow field would be expected to be as is indicated in Fig. 9.16a, the inviscid flow field. The pressure distribution indicated in Fig. 9.16b is imposed on the boundary layer flow along the surface of the cylinder. In fact, there is negligible pressure variation across the thin boundary layer so that the pressure within the boundary layer is that given by the inviscid flow field. This pressure distribution along the cylinder is such that the stationary fluid at the nose of the cylinder (Ufs = 0 at θ = 0) is accelerated to its maximum velocity (Ufs = 2U at θ = 90°) and then is decelerated back to zero velocity at the rear of the cylinder (Ufs = 0 at θ = 180°). This is accomplished by a balance between pressure and inertia effects; viscous effects are absent for the inviscid flow outside the boundary layer. Physically, in the absence of viscous effects, a fluid particle traveling from the front to the back of the cylinder coasts down the “pressure hill” from θ = 0 to θ = 90° (from point A to C in Fig. 9.16b) and then back up the hill to θ = 180° (from point C to F) without any loss of energy. There is an exchange between kinetic and pressure energy, but there are no energy losses. The same pressure distribution is imposed on the viscous fluid within the boundary layer. The decrease in pressure in the direction of flow along the front half of the cylinder is termed a favorable pressure gradient. The increase in pressure in the direction of flow along the rear half of the cylinder is termed an adverse pressure gradient. Consider a fluid particle within the boundary layer indicated in Fig. 9.17a. In its attempt to flow from A to F, it experiences the same pressure distribution as the particles in the free stream immediately outside the boundary layer—the inviscid flow field pressure. However, because of the viscous effects involved, the particle in the boundary layer experiences a loss of energy as it flows along. This loss means that the particle does not have enough energy to coast all of the way up the pressure hill (from C to F) and to reach point F at the rear of the cylinder. This kinetic energy deficit is seen in the velocity profile detail at point C, shown in Fig. 9.17a. Because of friction, the boundary layer fluid cannot travel from the front to the rear of the cylinder. (This conclusion can also be obtained from the concept that due to viscous effects the particle at C does not have enough momentum to allow it to coast up the pressure hill to F.) The situation is similar to a bicyclist coasting down a hill and up the other side of the valley. If there were no friction, the rider starting with zero speed could reach the same height from which he or she started. Clearly friction (rolling resistance, aerodynamic drag, etc.) causes a loss of energy (and momentum), making it impossible for the rider to reach the height from which he or she started without supplying additional energy (i.e., pedaling). The fluid within the boundary layer does not have such an energy supply. Thus, the fluid flows against the increasing pressure as far as it can, at which point it stops, and ultimately, reverses its direction; the boundary layer separates from (lifts off) the surface. This boundary layer separation is indicated in Fig. 9.17a as well as the figures in the margin. Typical velocity profiles at representative locations along the surface are shown in Fig. 9.17b. At the separation location (profile D), the velocity gradient at the wall and the 512 Chapter 9 ■ Flow over Immersed Bodies Boundary layer thickens rapidly in rising pressure Boundary layer remains thin in falling pressure δ C B Boundary layer fluid has kinetic energy deficit C D Boundary layer separation location Front stagnation point E Backflow A F Falling pressure Wake Constant pressure Rising pressure Edge of boundary layer (not a streamline) Outer flow streamlines (a) 1.0 2.0 y __ δ E 0.5 D C Separated flow B p – p0 Cp = _______ 1 _ U2 2ρ Locations indicated in Figs. 9.17a and 9.19 A 1.0 Inviscid theory Cp = 1 – 4 sin2 θ 0.0 Turbulent Laminar –1.0 –2.0 –0.2 0 0.2 0.4 u ___ Ufs ∂u For profile D, ___ = 0 at y = 0 ∂y (b) 0.6 0.8 1.0 –3.0 0 45° 90° 135° 180° θ (c) ■ Figure 9.17 Boundary layer characteristics on a circular cylinder: (a) boundary layer separation location, (b) typical boundary layer velocity profiles at various locations on the cylinder, (c) surface pressure distributions for inviscid flow and boundary layer flow. VIDEO V9.9 Snow drifts wall shear stress are zero. Beyond that location (from D to E) there is reverse flow in the boundary layer. A wake forms behind the cylinder where the pressure is low and approximately constant. As is indicated in Fig. 9.17c, because of the boundary layer separation, the average pressure on the rear half of the cylinder is considerably less than that on the front half. Thus, a large pressure drag is developed, even though (because of small viscosity) the viscous shear drag may be quite small. D’Alembert’s paradox is explained. No matter how small the viscosity, provided it is not zero, there will be a boundary layer that separates from the surface, giving a drag that is, for the most part, independent of the value of μ. 9.2 Boundary Layer Characteristics 513 THE WIDE WORLD OF FLUIDS Increasing truck mpg A large portion of the aerodynamic drag shape, thereby making the originally flat rear of the trailer a on semis (tractor-trailer rigs) is a result of the low pressure on somewhat more “aerodynamic” shape. Based on thorough wind the flat back end of the trailer. Researchers have recently devel- tunnel testing and actual tests conducted with a prototype oped a drag-reducing attachment that could reduce fuel costs design used in a series of cross-country runs, it is estimated on these big rigs by 10%. The device consists of a set of flat that trucks using the device could save approximately $6000 a plates (attached to the rear of the trailer) that fold out into a box year in fuel costs. Viscous effects within the boundary layer cause boundary layer separation. VIDEO V9.10 Effect of body shape on pressure gradient The location of separation, the width of the wake region behind the object, and the pressure distribution on the surface depend on the nature of the boundary layer flow. Compared with a laminar boundary layer, a turbulent boundary layer flow has more kinetic energy and momentum associated with it because: (1) as is indicated in Fig. E9.6, the velocity profile is fuller, more nearly like the ideal uniform profile, and (2) there can be considerable energy associated with the swirling, random components of the velocity that do not appear in the time-averaged x component of velocity. Thus, as is indicated in Fig. 9.17c, the turbulent boundary layer can flow farther around the cylinder (farther up the pressure hill) before it separates than can the laminar boundary layer. These effects of turbulence delaying separation are clearly visible in Figure 9.18. The two spheres are identical, except one has a patch of sandpaper on its nose. This sandpaper causes the boundary layer to become turbulent, which delays separation. Note that the wake is turbulent in both cases; there is no visible difference between the laminar and turbulent boundary layers on the spheres, but the wake is much different in size. The effect is lower drag. Note that because a flat plate parallel to the flow has no pressure gradient and thus no separation, a turbulent boundary layer would increase drag on a flat plate. The structure of the flow field past a circular cylinder is completely different for a zero viscosity fluid than it is for a viscous fluid, no matter how small the viscosity is, provided it is not zero. This is due to boundary layer separation. Similar concepts hold for other shaped bodies as well. The flow past an airfoil at zero angle of attack (the angle between the upstream flow and the axis of the object) is shown in Fig. 9.19a; flow past the same airfoil at a 5° angle of attack is shown in Fig. 9.19b. Over the front portion of the airfoil the pressure decreases in the direction of flow—a favorable pressure gradient. Over the rear portion the pressure increases in the direction of flow—an adverse pressure gradient. The boundary layer velocity profiles at representative locations are similar to those indicated in Fig. 9.17b for flow past a circular cylinder. If the adverse pressure gradient is not too great (because the body is not too “thick” in some sense), the boundary layer (a) (b) ■ Figure 9.18 Effect of turbulence in boundary layer on delaying separation. An 8.5 inch bowling ball is dropped into water at a speed of 25 ft/s. (a) The ball is smooth; (b) the ball has a patch of sandpaper on the nose that causes transition. (U.S. Navy photographs.) 514 Chapter 9 ■ Flow over Immersed Bodies (a) (b) ■ Figure 9.19 Flow visualization photographs of flow past an airfoil (the boundary layer velocity profiles for the points indicated are similar to those indicated in Fig. 9.17b): (a) zero angle of attack, no separation, (b) 5° angle of attack, flow separation. Dye in water. (Photographs courtesy of ONERA, The French Aerospace Lab.) Streamlined bodies generally have no separated flow. fluid can flow into the slightly increasing pressure region (i.e., from C to the trailing edge in Fig. 9.19a) without separating from the surface. However, if the pressure gradient is too adverse (because the angle of attack is too large), the boundary layer will separate from the surface as indicated in Fig. 9.19b. Such situations can lead to the catastrophic loss of lift called stall, which is discussed in Section 9.4. Streamlined bodies are generally those designed to eliminate (or at least to reduce) the effects of separation, whereas nonstreamlined bodies generally have relatively large drag due to the low pressure in the separated regions (the wake). Although the boundary layer may be quite thin, it can appreciably alter the entire flow field because of boundary layer separation. These ideas are discussed in Section 9.3. 9.2.7 Momentum Integral Boundary Layer Equation with Nonzero Pressure Gradient The boundary layer results discussed in Sections 9.2.2 and 9.2.3 are valid only for boundary layers with zero pressure gradients. They correspond to the velocity profile labeled C in Fig. 9.17b. Boundary layer characteristics for flows with nonzero pressure gradients can be obtained from the nonlinear, partial differential boundary layer equations, Eqs. 9.8 and 9.9a, provided the pressure gradient is appropriately accounted for. Such an approach is beyond the scope of this book (Refs. 1, 2). An alternative approach is extend the momentum integral boundary layer equation technique (Section 9.2.3) so that it is applicable for flows with nonzero pressure gradients. The momentum integral equation for boundary layer flows with zero pressure gradient, Eq. 9.26, is a statement of the balance between the shear force on the plate (represented by τw ) and rate of change of momentum of the fluid within the boundary layer [represented by ρU2 (dϴdx)]. For such flows the freestream velocity is constant (Ufs = U). If the free-stream velocity is not constant [Ufs = Ufs (x), where x is the distance measured along the curved body], the pressure will not be constant. This 9.3 Drag 515 follows from the Bernoulli equation with negligible gravitational effects, since p + ρU2fs2 is constant along the streamlines outside the boundary layer. Thus, dUfs dp = −ρUfs dx dx (9.34) For a given body the free-stream velocity and the corresponding pressure gradient on the surface can be approximated from inviscid flow techniques (potential flow) discussed in Section 6.7. (This is how the circular cylinder results of Fig. 9.16 were obtained.) Flow in a boundary layer with nonzero pressure gradient is very similar to that shown in Fig. 9.11, except that the upstream velocity, U, is replaced by the free-stream velocity, Ufs (x), and the pressures at sections (1) and (2) are not necessarily equal. By using the x component of the momentum equation (Eq. 5.22) with the appropriate shear forces and pressure forces acting on the control surface indicated in Fig. 9.11, the following integral momentum equation for boundary layer flows is obtained: Pressure gradient effects can be included in the momentum integral equation. 9.3 τw = ρ dUfs d (U2fs ϴ ) + ρδ* Ufs dx dx (9.35) The derivation of this equation is similar to that of the corresponding equation for constant-pressure boundary layer flow, Eq. 9.26, although the inclusion of the pressure gradient effect brings in additional terms (Refs. 1, 2, 3). For example, both the boundary layer momentum thickness, ϴ, and the displacement thickness, δ*, are involved. Equation 9.35, the general momentum integral equation for two-dimensional boundary layer flow, represents a balance between viscous forces (represented by τw ), pressure forces (represented by ρUfs dUfsdx = −dpdx), and the fluid momentum (represented by ϴ, the boundary layer momentum thickness). In the special case of a flat plate, Ufs = U = constant, and Eq. 9.35 reduces to Eq. 9.26. Equation 9.35 can be used to obtain boundary layer information in a manner similar to that done for the flat plate boundary layer (Section 9.2.3). That is, for a given body shape the free-stream velocity, Ufs, is determined, and a family of approximate boundary layer profiles is assumed. Equation 9.35 is then used to provide information about the boundary layer thickness, wall shear stress, and other properties of interest. The details of this technique are not within the scope of this book (Refs. 1, 3). Drag As was discussed in Section 9.1, any object moving through a fluid will experience a drag, 𝒟—a net force in the direction of flow due to the pressure and shear forces on the surface of the object. This net force, a combination of flow direction components of the normal and tangential forces on the body, can be determined by use of Eqs. 9.1 and 9.2, provided the distributions of pressure, p, and wall shear stress, τw, are known. Only in very rare instances can these distributions be determined analytically. The boundary layer flow past a flat plate parallel to the upstream flow as is discussed in Section 9.2 is one such case. Current advances in computational fluid dynamics, CFD (i.e., the use of computers to solve the governing equations of the flow field), have provided encouraging results for more complex shapes. However, much work in this area remains. Most of the information pertaining to drag on objects is a result of numerous experiments with wind tunnels, water tunnels, towing tanks, and other ingenious devices that are used to measure the drag on scale models. As was discussed in Chapter 7, these data can be put into dimensionless form, and the results can be appropriately ratioed for prototype calculations. Typically, the result for a given-shaped object is a drag coefficient, CD, where CD = 𝒟 1 2 2 ρU A (9.36) 516 Chapter 9 ■ Flow over Immersed Bodies and CD is a function of other dimensionless parameters such as Reynolds number, Re, Mach number, Ma, Froude number, Fr, and relative roughness of the surface, εℓ. That is, CD = ϕ(shape, Re, Ma, Fr, εℓ) The character of CD as a function of these parameters is discussed in this section. 9.3.1 Friction Drag Friction drag, 𝒟f, is that part of the drag that is due directly to the shear stress, τw, on the object. It is a function of not only the magnitude of the wall shear stress but also of the orientation of the surface on which it acts. This is indicated by the factor τw sin θ in Eq. 9.1. If the surface is parallel to the upstream velocity, the entire shear force contributes directly to the drag. This is true for the flat plate parallel to the flow as was discussed in Section 9.2. If the surface is perpendicular to the upstream velocity, the shear stress contributes nothing to the drag. Such is the case for a flat plate normal to the upstream velocity as was discussed in Section 9.1. In general, the surface of a body will contain portions parallel to and normal to the upstream flow, as well as any direction in between. A circular cylinder is such a body. Because the viscosity of most common fluids is small, the contribution of the shear force to the overall drag on a body is often quite small. Such a statement should be worded in dimensionless terms. That is, because the Reynolds number of most familiar flows is quite large, the percent of the drag caused directly by the shear stress is often quite small. For highly streamlined bodies or for low Reynolds number flow, however, most of the drag may be due to friction drag. The friction drag on a flat plate of width b and length ℓ oriented parallel to the upstream flow can be calculated from 1 𝒟f = 2ρU2bℓCDf Friction (viscous) drag is the drag produced by viscous shear stresses. where CDf is the friction drag coefficient. The value of CDf , given as a function of Reynolds number, Reℓ = ρUℓμ, and relative surface roughness, εℓ, in Fig. 9.15 and Table 9.3, is a result of boundary layer analysis and experiments (see Section 9.2). Typical values of roughness, ε, for various surfaces are given in Table 8.1. As with the pipe flow discussed in Chapter 8, the flow is divided into two distinct categories—laminar or turbulent, with a transitional regime connecting them. The drag coefficient (and, hence, the drag) is not a function of the plate roughness if the flow is laminar. However, for turbulent flow the roughness does considerably affect the value of CDf. As with pipe flow, this dependence is a result of the surface roughness elements protruding into or through the laminar sublayer (see Section 8.3). Most objects are not flat plates parallel to the flow; instead, they are curved or variably angled surfaces along which the pressure varies. As was discussed in Section 9.2.6, this means that the boundary layer character, including the velocity gradient at the wall, is different for most objects from that for a flat plate. This can be seen in the change of shape of the boundary layer profile along the cylinder in Fig. 9.17b. The precise determination of the shear stress along the surface of a curved body is quite difficult to obtain. Although approximate results can be obtained by a variety of techniques (Refs. 1, 2), these are outside the scope of this text. As is shown by the following example, if the shear stress is known, its contribution to the drag can be determined. EXAMPLE 9 .8 Drag Coefficient Based on Friction Drag GIVEN A viscous, incompressible fluid flows past the circular cylinder shown in Fig. E9.8a. According to a more advanced theory of boundary layer flow, the boundary layer remains attached to the cylinder up to the separation location at θ ≈ 108.8°, with the dimensionless wall shear stress as is indicated in Fig. E9.8b (Ref. 1). The shear stress on the cylinder in the wake region, 108.8 < θ < 180°, is negligible. FIND Determine CDf, the drag coefficient for the cylinder based on the friction drag only. 9.3 517 Drag SOLUTION The friction drag, 𝒟f, can be determined from Eq. 9.1 as 𝒟f = ∫τ w D sin θ dA = 2( ) b 2 π ∫ τw sin θ dθ where b is the length of the cylinder. Note that θ is in radians (not degrees) to ensure the proper dimensions of dA = 2 (D2) b dθ. Thus, 0 CDf = p dA U θ θ 1 2 2 ρU bD = 2 ρU2 ∫ π τw sin θ dθ 0 This can be put into dimensionless form by using the dimensionless shear stress parameter, F(θ) = τw √Re(ρU22), given in Fig. E9.8b as follows: τ wdA D ρ UD Re = _____ μ 𝒟f Ddθ dA = b __ CDf = 2 (a) ∫ 0 π τw 1 2 2 ρU sin θ dθ = 1 ∫ √Re 0 π τw √Re 1 2 2 ρU sin θ dθ where Re = ρUDμ. Thus, 2 τ w Re F(θ ) ≡ _______ 1 _ ρU2 8 CDf = F = 0 for θ > 108 6 0 Separation 0 20 40 60 80 100 120 θ, degrees CDf = 6 F(θ ) sin θ π F(θ) sin θ dθ (1) The function F(θ) sin θ, obtained from Fig. E9.8b, is plotted in Fig. E9.8c. The necessary integration to obtain CDf from Eq. 1 can be done by an appropriate numerical technique or by an approximate graphical method to determine the area under the given curve. The result is ʃ0π F(θ) sin θ dθ = 5.93, or (b) 5.93 √Re (Ans) COMMENTS Note that the total drag must include both the 4 2 0 ∫ √Re 0 4 2 1 π __ 4 0 π __ 2 θ , rad π _3__ 4 (c ) ■ Figure E9.8 π shear stress (friction) drag and the pressure drag. As we will see in Example 9.9, for the circular cylinder most of the drag is due to the pressure force. The above friction drag result is valid only if the boundary layer flow on the cylinder is laminar. As is discussed in Section 9.3.3, for a smooth cylinder this means that Re = ρUDμ < 3 × 105. It is also valid only for flows that have a Reynolds number sufficiently large to ensure the boundary layer structure to the flow. For the cylinder, this means Re > 100. 9.3.2 Pressure Drag Pressure (form) drag is the drag produced by normal stresses. pdA θ U dA Pressure drag, 𝒟p, is that part of the drag that is due directly to the pressure, p, on an object. It is often referred to as form drag because of its strong dependency on the shape or form of the object. Pressure drag is a function of the magnitude of the pressure and the orientation of the surface element on which the pressure force acts. For example, the pressure force on either side of a flat plate parallel to the flow may be very large, but it does not contribute to the drag because it acts in the direction normal to the upstream velocity. On the other hand, the pressure force on a flat plate normal to the flow provides the entire drag. As previously noted, for most bodies, there are portions of the surface that are parallel to the upstream velocity, others normal to the upstream velocity, and the majority of which are at some angle in between, as shown by the figure in the margin. The pressure drag can be obtained from Eq. 9.1 provided a detailed description of the pressure distribution and the body shape is given. That is, 𝒟p = ∫ p cos θ dA 518 Chapter 9 ■ Flow over Immersed Bodies which can be rewritten in terms of the pressure drag coefficient, CDp, as CDp = 𝒟p 1 2 2 ρU A = ∫ p cos θ dA 1 2 2 ρU A ∫ C cos θ dA p = (9.37) A Here Cp = (p − p0 )(ρU2 2) is the pressure coefficient, where p0 is a reference pressure. The magnitude of the reference pressure does not influence the drag directly because the net pressure force on a body is zero if the pressure is constant (i.e., p0 ) on the entire surface. For flows in which inertial effects are large relative to viscous effects (i.e., large Reynolds number flows), the pressure difference, p − p0, scales directly with the dynamic pressure, ρU2 2, and the pressure coefficient is independent of Reynolds number. In such situations we expect the drag coefficient to be relatively independent of Reynolds number. For flows in which viscous effects are large relative to inertial effects (i.e., very small Reynolds number flows), it is found that both the pressure difference and wall shear stress scale with the characteristic viscous stress, μUℓ, where ℓ is a characteristic length. In such situations we expect the drag coefficient to be proportional to 1Re. That is, CD ∼ 𝒟(ρU2 2) ∼ (μUℓ)(ρU2 2) ∼ μρUℓ = 1Re. These characteristics are similar to the friction factor dependence of f ∼ 1 Re for laminar pipe flow and f ∼ constant for large Reynolds number flow (see Section 8.4). As is discussed in Section 9.2.6 if the viscosity were zero, the pressure drag on any shaped object (symmetrical or not) in a steady flow would be zero. There perhaps would be large pressure forces on the front portion of the object, but there would be equally large (and oppositely directed) pressure forces on the rear portion. If the viscosity is not zero, the net pressure drag may be nonzero because of boundary layer separation. Example 9.9 illustrates this concept. The pressure coefficient is a dimensionless form of the pressure. EXAMPLE 9 .9 Drag Coefficient Based on Pressure Drag GIVEN A viscous, incompressible fluid flows past the circular FIND Determine the pressure drag coefficient for this flow. cylinder shown in Fig. E9.8a. The pressure coefficient on the surface of the cylinder (as determined from experimental measurements) is as indicated in Fig. E9.9a. Combine the results of Examples 9.8 and 9.9 to determine the drag coefficient for a circular cylinder. Compare your results with those given in Fig. 9.23. SOLUTION The pressure (form) drag coefficient, CDp, can be determined from Eq. 9.37 as CDp = 1 A ∫ C cos θ dA = bD1 ∫ p 0 2π D Cp cos θ b( ) dθ 2 or because of symmetry CDp = ∫ π Cp cos θ dθ by pulling on the cylinder in the upstream direction. The positive area under the Cp cos θ curve is greater than the negative area— there is a net pressure drag. In the absence of viscosity, these two contributions would be equal—there would be no pressure (or friction) drag. The net drag on the cylinder is the sum of friction and pressure drag. Thus, from Eq. 1 of Example 9.8 and Eq. 1 of this example, we obtain the drag coefficient 0 where b and D are the length and diameter of the cylinder. To obtain CDp, we must integrate the Cp cos θ function from θ = 0 to θ = π radians. Again, this can be done by some numerical integration scheme or by determining the area under the curve shown in Fig. E9.9b. The result is CDp = 1.17 (1) (Ans) Note that the positive pressure on the front portion of the cylinder (0 ≤ θ ≤ 30°) and the negative pressure (less than the upstream value) on the rear portion (90 ≤ θ ≤ 180°) produce positive contributions to the drag. The negative pressure on the front portion of the cylinder (30 < θ < 90°) reduces the drag CD = CDf + CDp = 5.93 √Re + 1.17 (2) (Ans) This result is compared with the standard experimental value (obtained from Fig. 9.23) in Fig. E9.9c. The agreement is very good over a wide range of Reynolds numbers. For Re < 10 the curves diverge because the flow is not a boundary layer type flow—the shear stress and pressure distributions used to obtain Eq. 2 are not valid in this range. The drastic divergence in the curves for Re > 3 × 105 is due to the change from a laminar to turbulent boundary layer, with the corresponding change in the pressure distribution. This phenomenon was discussed in Section 9.2.6 and will be elaborated upon in Section 9.3.3. 9.3 Drag 519 1.0 0.5 Cp cosθ 2 p-p Cp = _____o 1 _ ρ U2 1.0 0 -0.5 0.5 0 -1.0 -0.5 -1.5 0 30 60 90 120 150 180 0 θ , degrees π __ 2 ___ 3π 4 π θ , rad ( a) 100 π __ 4 ( b) Experimental value Eq. 2 10 CD 1.0 0.1 100 101 102 103 104 105 106 ___ Re = UD v (c) ■ Figure E9.9 COMMENT It is of interest to compare the friction drag to the total drag on the cylinder. That is, 𝒟f 𝒟 = CDf CD = 5.93 √Re (5.93 √Re) + 1.17 = 1 1 + 0.197√Re For Re = 103, 104, and 105 this ratio is 0.138, 0.0483, and 0.0158, respectively. In other words, friction drag accounts for less than 15% of the total drag and can even be as low as 1%. Thus, most of the drag on the blunt cylinder is pressure drag—a result of the boundary layer separation. 9.3.3 Drag Coefficient Data and Examples VIDEO V9.11 Skydiving practice As was discussed in previous sections, the net drag is produced by both pressure and shear stress effects. In most instances these two effects are considered together, and an overall drag coefficient, CD, as defined in Eq. 9.36 is used. There is an abundance of such drag coefficient data available in the literature. This information covers incompressible and compressible viscous flows past objects of almost any shape of interest—both man-made and natural objects. In this section we consider a small portion of this information for representative situations. Additional data can be obtained from various sources (Refs. 5, 6). Shape Dependence. Clearly the drag coefficient for an object depends on the shape of the object, with shapes ranging from those that are streamlined to those that are blunt. The drag on an ellipse with aspect ratio ℓD, where D and ℓ are the thickness and length parallel to the flow, illustrates this dependence. The drag coefficient CD = 𝒟(ρU2 bD2), based on the frontal area, A = bD, where b is the length normal to the flow, is as shown in Fig. 9.20. The blunter the body, the larger the drag coefficient. With ℓD = 0 (i.e., a flat plate normal to the flow) we obtain the flat plate value of CD = 1.9. With ℓD = 1 the corresponding value for a circular cylinder is obtained. As ℓD becomes larger, the value of CD decreases. 520 Chapter 9 ■ Flow over Immersed Bodies 2.5 • Flat plate normal to flow 2.0 U UD Re = v = 105 b = length D 1.5 CD 𝓵 1.0 CD = 𝒟 1 ρ U 2 bD 2 Flat plate parallel to flow 0.5 CD = 0 0 𝒟 1 ρ U 2 b𝓵 2 1 2 3 𝓵 D The drag coefficient may be based on the frontal area or the planform area. 4 5 6 ■ Figure 9.20 Drag coefficient for an ellipse with the characteristic area either the frontal area, A = bD, or the planform area, A = bℓ (Ref. 5). For very large aspect ratios (ℓD → ∞) the ellipse behaves as a flat plate parallel to the flow. For such cases, the friction drag is greater than the pressure drag, and the value of CD based on the frontal area, A = bD, would increase with increasing ℓD. (This occurs for larger ℓD values than those shown in the figure.) For such extremely thin bodies (i.e., an ellipse with ℓD → ∞, a flat plate, or very thin airfoils) it is customary to use the planform area, A = bℓ, in defining the drag coefficient. After all, it is the planform area on which the shear stress acts, rather than the much smaller (for thin bodies) frontal area. The ellipse drag coefficient based on the planform area, CD = 𝒟(ρU2bℓ2), is also shown in Fig. 9.20. Clearly the drag obtained by using either of these drag coefficients would be the same. They merely represent two different ways to package the same information. From a design perspective, we often focus on minimizing drag, but this is not always the goal. Boundary layer separation and its subsequent wake promote mixing. This is desirable in heat exchange devices or fluid mixing vessels. If, on the other hand, you are designing low-drag shapes, keep in mind two important principles. ■ ■ If the body is long and thin (i.e., streamlined such as a parallel flat plate or airfoil), the drag is primarily caused by friction. This drag can be reduced by keeping the flow laminar as much as possible. This condition implies smooth surfaces and, if the body has thickness (like an airfoil), means selecting the shape that delays transition to turbulence as long as possible. If the body is blunt, the (high Reynolds number) drag is primarily pressure drag. This drag can be reduced by delaying separation as long as possible. One way to delay boundary layer transition is to promote transition to a turbulent boundary layer (Fig. 9.17 and Fig. 9.18). This will be discussed more in a few pages in the subsection titled “Surface Roughness.” A better method is streamlining, that is, elongating the rear portion of the body. Incredibly the drag on the two two-dimensional objects drawn to scale in Fig. 9.21 is the same. The width of the wake for the streamlined strut is very thin, on the order of that for the much smaller diameter circular cylinder. Consider the additional structural support offered with option (b) over option (a) with no added drag! If the nose of the body is flat, nose rounding can also reduce the drag. Figure 9.22 illustrates the benefits of nose rounding and rearward taper. The round nose-rear tapered body would have higher drag if it were reversed, with the tapered section pointing forward. The forward portion of a body can be quite blunt with a large (negative) pressure gradient, because separation does not occur in falling pressure. It is rearward (rising) pressure change that must be accomplished gradually. Be cautious not to add too much length due to taper; more surface area adds more friction drag, which could negate the benefits of reducing pressure drag. 9.3 Drag 521 U, ρ U Diameter = D CD ≈ 2 (a) U 𝒟 a = 𝒟b U, ρ 10 D U (b) CD ≈ 1.1 CD ≈ 0.1 ■ Figure 9.22 Effect of streamlining ■ Figure 9.21 Two objects of considerably different size that have the same drag force: (a) circular cylinder CD = 1.2; (b) streamlined strut CD = 0.12. on drag reduction. Drag coefficients are based on frontal area. All bodies are two-dimensional. Reynolds Number Dependence. Another parameter on which the drag coefficient can be very dependent is the Reynolds number. The main categories of Reynolds number dependence are (1) very low Reynolds number flow, (2) moderate Reynolds number flow (laminar boundary layer), and (3) very large Reynolds number flow (turbulent boundary layer). Examples of these three situations are discussed below. Low Reynolds number flows (Re < 1), also known as creeping flows, result from small body sizes (e.g., dust or pulverized coal particles), large fluid viscosities (e.g., oil or grease), or very low density (e.g., high altitude flight). The mechanisms of drag are fundamentally different in low and high Reynolds number flow. The boundary layer concept, so valuable in high Reynolds number flow, is invalid at low Reynolds number. At low Reynolds number, viscous forces are not confined to a thin region near the body but are significant everywhere. That is, the boundary layer is so thick that it distorts the stream line pattern everywhere (see Fig. 9.6a). Low Reynolds number flows are governed by a balance of viscous and pressure forces, and inertia effects are negligibly small. In such instances the drag on a three-dimensional body is expected to be a function of the upstream velocity, U, the body size, ℓ, and the viscosity, μ. Thus, for a small grain of sand settling in a lake (see figure in the margin) 𝒟 = f (U, ℓ, μ) From dimensional considerations (see Section 7.7.1) Re < 1 𝓵 μ U 𝒟 = f(U, 𝓵, μ) For very small Reynolds number flows, the drag coefficient varies inversely with the Reynolds number. 𝒟 = CμℓU (9.38) where the value of the constant C depends on the shape of the body. If we put Eq. 9.38 into dimensionless form using the standard definition of the drag coefficient, we obtain CD = 𝒟 1 2 2 2 ρU ℓ = 2CμℓU 2C = Re ρU2ℓ2 where Re = ρUℓμ. The use of the dynamic pressure, ρU2 2, in the definition of the drag coefficient is somewhat misleading in the case of creeping flows (Re < 1) because it introduces inertia, which as mentioned is negligibly small compared to pressure and viscous effects. Nonetheless, use of this standard drag coefficient definition gives the 1Re dependence for small Re drag coefficients. Typical values of CD for low Reynolds number flows past a variety of objects are given in Table 9.4. It is of interest that the drag on a disk normal to the flow is only 1.5 times greater than that on a disk parallel to the flow. For large Reynolds number flows this ratio is considerably larger (see Example 9.1). Streamlining (i.e., making the body slender) can produce a considerable drag reduction for large Reynolds number flows; for very small Reynolds number flows it can actually increase the drag because of an increase in the area on which shear forces act. For most objects, the low Reynolds number flow results are valid up to a Reynolds number of about 1. 522 Chapter 9 ■ Flow over Immersed Bodies ■ Table 9.4 Low Reynolds Number Drag Coefficients (Ref. 7) (Re = ρUDμ, A = πD24) CD = 𝒟( ρU2A2) (for Re ≲ 1) Object a. Circular disk normal to flow Object 20.4/Re d. Circular disk parallel to flow 24.0/Re c. Sphere D U D U CD d. Hemisphere 13.6/Re 22.2/Re U U D D EXAMPLE 9 .10 Low Reynolds Number Flow Drag GIVEN While workers spray paint on to the ceiling of a room, FIND Determine the time it would take this particle to fall 8 ft numerous small paint aerosols are dispersed into the air. Eventually these particles will settle out and fall to the floor or other surfaces. Consider a small spherical paint particle of diameter D = 1 × 10−5 m (or 10 μm) and specific gravity SG = 1.2. from near the ceiling to the floor. Assume that the air within the room is motionless. SOLUTION A free-body diagram of the particle (relative to the moving particle) is shown in Fig. E9.10a. The particle moves downward with a constant velocity U that is governed by a balance between the weight of the particle, 𝒲, the buoyancy force of the surrounding air, FB, and the drag, 𝒟, of the air on the particle. From the free-body diagram, we obtain We assume (because of the smallness of the object) that the flow will be creeping flow (Re < 1) with CD = 24Re so that 𝒟= π 1 π 24 1 ρ U2 D2CD = ρairU 2 D2( 2 air 4 2 4 ρairUDμair ) or 𝒲 = 𝒟 + FB 𝒟 = 3πμairUD where, if V is the particle volume π 𝒲 = γpoint V = SG γH2O D3 6 (1) and FB = γair V = γair π 3 D 6 (3) We must eventually check to determine if the assumption that Re < 1 is valid. Equation 3 is called Stokes’s law in honor of G. G. Stokes, a British mathematician and a physicist. By combining Eqs. 1, 2, and 3, we obtain SG γH2O (2) π 3 π D = 3πμairUD + γair D3 6 6 Then, solving for U, U= FB 18μair (4) From Tables 1.5 and 1.7 we obtain γH2O = 9800 N/m3 , γair = 12.0 N/m3, and μair = 1.79 × 10−5 N ∙ s/m2. Thus, from Eq. 4 we obtain 𝒟 𝒲 U D2 (SGγH2O − γair ) U= (10−5 m) 2[(1.2) (9800 N/m3 ) − (12.0 N/m3 )] 18(1.79 × 10−5 N ∙ s/m2 ) or ■ Figure E9.10a U = 0.00365 m/s 9.3 Let tfall denote the time it takes for the particle to fall 8 ft. Thus, since 8 ft is approximately 2.44 m, tfall = 2.44 m = 668 s 0.00365 m/s Therefore, it will take over 11 minutes for the paint particle to fall to the floor. Since air, proper engineering controls are needed to reduce inhalation exposures (e.g., appropriate ventilation) or, in extreme cases, personal protective equipment (PPE, e.g., respirator) may be required. PPE is considered a last resort because in practice it is the least effective means to protect workers. It is more effective and usually less expensive in the long run to design out or control a problem with engineering ingenuity—prevention through design. (1.23 kg/m3 ) (1 × 10−5 m) (0.00365 m/s) ρDU = μ 1.79 × 10−5 N ∙ s/m2 = 0.00251 Re = COMMENTS By repeating the calculations for various par- ticle diameters, the results shown in Fig. E9.10b are obtained. Note that very small particles fall extremely slowly. In fact, particles in the general size range that can be inhaled deeply into the lungs (on the order of 5 μm or less) will remain airborne most of the working day, thus continually exposing workers to potential hazards. It is often the smaller particles that cause the greatest health concerns, particularly to individuals exposed day after day for many years. In work environments that generate particle-laden Flow past a cylinder can take on a variety of different structures. VIDEO V9.12 Karman vortex street 5 4.5 4 3.5 tfall, hours we see that Re < 1, and the form of the drag coefficient used is valid. 523 Drag 3 2.5 2 1.5 1 10 μ m, 0.18 hour 0.5 0 0 2 4 6 8 10 12 14 16 18 20 D, μm ■ Figure E9.10b Moderate Reynolds number flows tend to take on a boundary layer flow structure. For such flows past streamlined bodies, the drag coefficient tends to decrease slightly with Reynolds number. The CD ∼ Re−12 dependence for a laminar boundary layer on a flat plate (see Table 9.3) is such an example. Moderate Reynolds number flows past blunt bodies generally produce drag coefficients that are relatively constant. The CD values for the spheres and circular cylinders shown in Fig. 9.23a indicate this character in the range 103 < Re < 105. The structure of the flow field at selected Reynolds numbers indicated in Fig. 9.23a is shown in Fig. 9.23b. For a given object there is a wide variety of flow situations, depending on the Reynolds number involved. The curious reader is strongly encouraged to study the many beautiful photographs and videos of these (and other) flow situations found in Refs. 8 and 31. For many shapes there is a sudden change in the character of the drag coefficient when the boundary layer becomes turbulent. This is illustrated in Fig. 9.15 for the flat plate and in Fig. 9.23 for the sphere and the circular cylinder. The Reynolds number at which this transition takes place is a function of the shape of the body. For streamlined bodies, the drag coefficient increases when the boundary layer becomes turbulent because most of the drag is due to the shear force, which is greater for turbulent flow than for laminar flow. On the other hand, the drag coefficient for a relatively blunt object, such as a cylinder or sphere, actually decreases when the boundary layer becomes turbulent. As is discussed in Section 9.2.6, a turbulent boundary layer can travel further along the surface into the adverse pressure gradient on the rear portion of the cylinder before separation occurs. The result is a thinner wake and smaller pressure drag for turbulent boundary layer flow. This is indicated in Fig. 9.23 by the sudden decrease in CD for 105 < Re < 106. In a portion of this range the actual drag (not just the drag coefficient) decreases with increasing speed. It would be very difficult to control the steady flight of such an object in this range—an increase in velocity requires a decrease in thrust (drag). In all other Reynolds number ranges the drag increases with an increase in the upstream velocity; 524 Chapter 9 ■ Flow over Immersed Bodies 400 200 100 60 40 20 CD A 10 6 4 2 24 CD = ___ B Re Smooth cylinder C D 1 0.6 0.4 Smooth sphere 0.2 E 0.1 0.06 –1 10 100 101 102 103 104 105 106 107 UD Re = ρ____ μ (a) VIDEO VIDEO V9.13 Flow past cylinder No separation Steady separation bubble (A) (B) V9.14 Oscillating sign Oscillating Karman vortex street wake (C) Laminar boundary layer, wide turbulent wake Turbulent boundary layer, narrow turbulent wake (D) (E) (b) VIDEO V9.15 Flow past a flat plate ■ Figure 9.23 (a) Drag coefficient as a function of Reynolds number for a smooth circular cylinder and a smooth sphere. (b) Typical flow patterns for flow past a circular cylinder at various Reynolds numbers as indicated in (a). even though CD may decrease with Re, remember that drag calculation requires drag coefficient times velocity squared. For extremely blunt bodies, like a flat plate perpendicular to the flow, the flow separates at the edge of the plate regardless of the nature of the boundary layer flow. Thus, the drag coefficient shows very little dependence on the Reynolds number. The drag coefficient for a flat plate perpendicular to the flow is about 1.9 for any Reynolds number. The drag coefficients for a series of two-dimensional bodies of varying bluntness are given as a function of Reynolds number in Fig. 9.24. The characteristics described above are evident. 9.3 Drag 525 1.0 VIDEO Ellipse V9.16 Flow past an ellipse D 0.5 D 0.1 Airfoil CD Flat plate 0.18 D D 0.01 𝒟 CD = _________ 1 ρ U2 bD __ 2 D b = length 104 105 106 107 ___ Re = UD v ■ Figure 9.24 Character of the drag coefficient as a function of Reynolds number for objects with various degrees of streamlining, (two-dimensional flow) (Ref. 5). E X AMPLE 9 .11 Terminal Velocity of a Falling Object GIVEN Hail is produced by the repeated rising and falling of FIND Estimate the velocity, U, of the updraft needed to make ice particles in the updraft of a thunderstorm, as is indicated in Fig. E9.11a. When the hail becomes large enough, the aerodynamic drag from the updraft can no longer support the weight of the hail, and it falls from the storm cloud. D = 1.5-in.-diameter (i.e., “golf ball-sized”) hail. 40,000 to 50,000 ft SOLUTION As is discussed in Example 9.10, for steady-state conditions a force balance on an object falling through a fluid at its terminal velocity, U, gives Anvil-shaped cloud Hail Updraft Down draft Storm movement Ground Rain ■ Figure E9.11a 𝒲 = 𝒟 + FB where FB = γair V is the buoyant force of the air on the particle, 𝒲 = γice V is the particle weight, and 𝒟 is the aerodynamic drag. This equation can be rewritten as 1 2 2 ρairU π 2 D C D = 𝒲 − FB 4 (1) With V = πD36 and since γice ⪢ γair (i.e., 𝒲 ⪢ FB ), Eq. 1 can be simplified to U= 4 ρice gD 12 U =( 3 ρair CD ) (2) By using ρice = 1.84 slugsft3, ρair = 2.38 × 10−3 slugsft3, and D = 1.5 in. = 0.125 ft, Eq. 2 becomes U=[ 4(1.84 slugsft3 ) (32.2 fts2 ) (0.125 ft) 3(2.38 × 10−3 slugsft3 )CD 12 ] 64.5 √CD 64.5 √0.5 = 91.2 fts The corresponding Reynolds number (assuming v = 1.57 × 10−4 ft2s) is Re = 91.2 fts (0.125 ft) UD = 7.26 × 104 = ν 1.57 × 10−4 ft2s For this value of Re we obtain from Fig. 9.23, CD = 0.5. Thus, our assumed value of CD = 0.5 was correct. The corresponding value of U is or U= where U is in fts. To determine U, we must know CD. Unfortunately, CD is a function of the Reynolds number (see Fig. 9.23), which is not known unless U is known. Thus, we must use an iterative technique similar to that done with the Moody chart for certain types of pipe flow problems (see Section 8.5). From Fig. 9.23 we expect that CD is on the order of 0.5. Thus, we assume CD = 0.5 and from Eq. 3 obtain (3) U = 91.2 fts = 62.2 mph (Ans) 526 Chapter 9 ■ Flow over Immersed Bodies COMMENTS By repeating the calculations for various alti- Hailstones greater than 6 in. in diameter have been reported. In reality, a hailstone is seldom spherical and often not smooth. The hailstone shown in Fig. E9.11c is larger than golf-ball sized and is clearly ellipsoidal in shape with additional smaller-scale surface roughness features. However, the calculated updraft velocities are in agreement with measured values. tudes, z, above sea level (using the properties of the U.S. Standard Atmosphere given in Appendix C), the results shown in Fig. E9.11b are obtained. Because of the decrease in density with altitude, the hail falls even faster through the upper portions of the storm than when it hits the ground. Clearly, an airplane flying through such an updraft would feel its effects (even if it were able to dodge the hail). As seen from Eq. 2, the larger the hail, the stronger the necessary updraft. 140 120 U, mph 100 80 60 (0, 62.2 mph) 40 20 0 0 10,000 20,000 z, ft 30,000 40,000 ■ Figure E9.11b Compressibility Effects. The above discussion is restricted to incompressible flows. If the velocity of the object is sufficiently large, compressibility effects become important and the drag coefficient becomes a function of the Mach number, Ma = Uc, where c is the speed of sound in the fluid. The introduction of Mach number effects complicates matters because the drag coefficient for a given object is then a function of both Reynolds number and Mach number—CD = ϕ(Re, Ma). The Mach number and Reynolds number effects are often closely connected because both are directly proportional to the upstream velocity. For example, both Re and Ma increase with increasing flight speed of an airplane. The changes in CD due to a change in U are due to changes in both Re and Ma. The precise dependence of the drag coefficient on Re and Ma is generally quite complex (Ref. 13). However, the following simplifications are often justified. For low Mach numbers, the drag coefficient is essentially independent of Ma as is indicated in Fig. 9.25. For this situation, if Ma < 0.5 or so, compressibility effects are unimportant. On the other hand, for larger Mach number flows, the drag coefficient can be strongly dependent on Ma, with only secondary Reynolds number effects. For most objects, values of CD increase dramatically in the vicinity of Ma = 1 (i.e., sonic flow). This change in character, indicated by Fig. 9.26, is due to the existence of shock waves as indicated by the figure in the margin on the next page. Shock waves are extremely narrow regions in the flow field across which the flow parameters change in a nearly discontinuous manner, which are discussed in Chapter 11. Shock waves, which cannot exist in subsonic flows, provide a mechanism for the generation of drag that is not present in the relatively low-speed subsonic flows. (See Fig. E11.3b in Example 11.3.) 1.0 2 𝒟 CD = __________ 1 __ ρU2 bD The drag coefficient is usually independent of Mach number for Mach numbers up to approximately 0.5. ■ Figure E9.11c (Photograph courtesy of NOAA.) 4D U D 0.5 0 0 b = length 0.5 Ma ■ Figure 9.25 Drag coefficient as a function of Mach number for 1.0 two-dimensional object in subsonic flow (Ref. 5). 9.3 Drag 527 1.8 Circular cylinder D 1.6 U 1.4 1.2 CD 1.0 D Sphere U 0.8 0.6 Sharp pointed ogive 0.4 D 4 2 0 U 𝒟 CD = ___________ π D2 1 ρU2__ __ 0.2 0 1 2 3 Ma U Ma = 1.5 Shock wave U Ma = 3 Depending on the body shape, an increase in surface roughness may increase or decrease drag. Surface roughness can cause the boundary layer to become turbulent. 4 5 6 ■ Figure 9.26 Drag coefficient as a function of Mach number for supersonic flow (adapted from Ref. 19). The character of the drag coefficient as a function of Mach number is different for blunt bodies than for sharp bodies. As is shown in Fig. 9.26, sharp-pointed bodies develop their maximum drag coefficient in the vicinity of Ma = 1 (sonic flow), whereas the drag coefficient for blunt bodies increases with Ma far above Ma = 1. This behavior is due to the nature of the shock wave structure and the accompanying flow separation. The leading edges of wings for subsonic aircraft are usually quite rounded and blunt, while those of supersonic aircraft tend to be quite pointed and sharp. More information on these important topics can be found in standard texts about compressible flow and aerodynamics (Refs. 9, 10, and 29). Surface Roughness. As is indicated in Fig. 9.15, the drag on a flat plate parallel to the flow is quite dependent on the surface roughness, provided the boundary layer flow is turbulent. In such cases the surface roughness protrudes through the laminar sublayer adjacent to the surface (see Section 8.4) and alters the wall shear stress. In addition to the increased turbulent shear stress, surface roughness can alter the Reynolds number at which the boundary layer flow becomes turbulent. Thus, a rough flat plate may have a larger portion of its length covered by a turbulent boundary layer than does the corresponding smooth plate. This also acts to increase the net drag on the plate. In general, for streamlined bodies, the drag increases with increasing surface roughness. Great care is taken to design the surfaces of airplane wings to be as smooth as possible, since protruding rivets or screw heads can cause a considerable increase in the drag. On the other hand, for an extremely blunt body, such as a flat plate normal to the flow, the drag is independent of the surface roughness, since the shear stress is not in the upstream flow direction and contributes nothing to the drag. For blunt bodies like a circular cylinder or sphere, an increase in surface roughness can actually cause a decrease in the drag. This is illustrated for a sphere in Fig. 9.27. As is discussed in Section 9.2.6, when the Reynolds number reaches the critical value (Re = 3 × 105 for a smooth sphere), the boundary layer becomes turbulent and the wake region behind the sphere becomes considerably narrower than if it were laminar (see Figs. 9.18 and 9.23b). The result is a considerable drop in pressure drag with a slight increase in friction drag, combining to give a smaller overall drag (and CD). The boundary layer can be tripped into turbulence at a smaller Reynolds number by using a rough-surfaced sphere. For example, the critical Reynolds number for a golf ball is approximately Re = 4 × 104. In the range 4 × 104 < Re < 4 × 105, the drag on the standard rough (i.e., 528 Chapter 9 ■ Flow over Immersed Bodies 0.6 ε __ = relative roughness D 0.5 Golf ball 2 4 𝒟 CD = ___________ π D2 1 ρU2__ __ 0.4 0.3 0.2 ε __ = 1.25 × 10–2 ε __ = 0 (smooth) D D 0.1 ε __ = 1.5 × 10–3 D 0 4 × 104 105 4 × 105 106 4 × 106 ___ Re = UD v ■ Figure 9.27 The effect of surface roughness on the drag coefficient of a sphere in the Reynolds number range for which the laminar boundary layer becomes turbulent (Ref. 5). dimpled) golf ball is considerably less (CDrough CDsmooth ≈ 0.250.5 = 0.5) than for the smooth ball. As is shown in Example 9.12, this is precisely the Reynolds number range for well-hit golf balls—hence, a reason for dimples on golf balls. The Reynolds number range for well-hit table tennis balls is less than Re = 4 × 104. Thus, table tennis balls are smooth. EXAM PLE 9 .12 Effect of Surface Roughness GIVEN A well-hit golf ball (diameter D = 1.69 in., weight 𝒲 = 0.0992 lb) can travel at U = 200 fts as it leaves the tee. A well-hit table tennis ball (diameter D = 1.50 in., weight 𝒲 = 0.00551 lb) can travel at U = 60 fts as it leaves the paddle. FIND Determine the drag on a standard golf ball, a smooth VIDEO V9.17 Wake of golf ball and smooth ball golf ball, and a table tennis ball for the conditions given. Also determine the deceleration of each ball for these conditions. SOLUTION CD = 0.50 for the table tennis ball. Hence, from Eq. 1 for the standard golf ball For either ball, the drag can be obtained from 𝒟= π 1 ρU 2 D2CD 2 4 (1) where the drag coefficient, CD, is given in Fig. 9.27 as a function of the Reynolds number and surface roughness. For the golf ball in standard air (200 fts) (1.6912 ft) UD Re = = 1.79 × 105 = ν 1.57 × 10−4 ft2s while for the table tennis ball Re = UD = ν 1 π 1.69 2 (0.00238 slugsft3 ) (200 fts) 2 ( ft (0.25) 2 4 12 ) = 0.185 lb (Ans) 𝒟= for the smooth golf ball 1 π 1.69 2 (0.00238 slugsft3 ) (200 fts) 2 ( ft (0.51) 2 4 12 ) = 0.378 lb (Ans) 𝒟= and for the table tennis ball (60 fts) (1.5012 ft) −4 1.57 × 10 ft s 2 = 4.78 × 104 The corresponding drag coefficients are CD = 0.25 for the standard golf ball, CD = 0.51 for the smooth golf ball, and 1 π 1.50 2 (0.00238 slugsft3 ) (60 fts) 2 ( ft (0.50) 2 4 12 ) = 0.0263 lb (Ans) 𝒟= 9.3 The corresponding decelerations are a = 𝒟m = g𝒟𝒲, where m is the mass of the ball. Thus, the deceleration relative to the acceleration of gravity, ag (i.e., the number of g’s deceleration) is ag = 𝒟𝒲 or a 0.185 lb = 1.86 for the standard golf ball = g 0.0992 lb 0.378 lb a = 3.81 for the smooth golf ball = g 0.0992 lb a 0.0263 lb = 4.77 for the table tennis ball = g 0.00551 lb (Ans) (Ans) (Ans) 529 By repeating the above calculations, the drag as a function of speed for both a standard golf ball and a smooth golf ball is shown in Fig. E9.12. The Reynolds number range for which a rough golf ball has smaller drag than a smooth one (i.e., 4 × 104 to 3.6 × 105) corresponds to a flight velocity range of 45 < U < 400 ft/s. This is comfortably within the range of most golfers. (The fastest tee shot by top professional golfers is approximately 280 ft/s.) As discussed in Section 9.4.2, the dimples (roughness) on a golf ball also help produce a lift (due to the spin of the ball) that allows the ball to travel farther than a smooth ball. 1.2 COMMENTS When considering drag force, don’t be fooled by 1 Smooth golf ball 0.8 𝒟, lb the magnitude of CD alone. Notice that the drag force on the table tennis ball is considerably lower than on the standard golf ball, even though the drag coefficient is larger. This is a result of diameter difference and especially squared velocity difference. Note that there is a considerably smaller deceleration for the rough golf ball than for the smooth one. Because of its much larger drag-tomass ratio, the table tennis ball slows down relatively quickly and does not travel as far as the golf ball. Note that with U = 60 fts the standard golf ball has a drag of 𝒟 = 0.0200 lb and a deceleration of ag = 0.202, considerably less than the ag = 4.77 of the table tennis ball. Conversely, a table tennis ball hit from a tee at 200 fts would decelerate at a rate of a = 1740 fts2, or ag = 54.1. It would not travel nearly as far as the golf ball. Drag 0.6 0.4 Standard golf ball 0.2 0 0 100 200 300 400 500 U, ft/s ■ Figure E9.12 THE WIDE WORD OF FLUIDS Dimpled baseball bats For many years it has been known that pondered that question, performed experiments with dimpled dimples on golf balls can create a turbulent boundary layer and bats to determine the answer, and received a patent for his reduce the aerodynamic drag, allowing longer drives than with dimpled bat invention. The result is that a batter can swing a smooth balls. Thus, why not put dimples on baseball bats so that dimpled bat approximately 3 to 5% faster than a smooth bat. tomorrow’s baseball sluggers can swing the bat faster and, Theoretically, this extra speed will translate to an extra 10 to therefore, hit the ball farther? MIT instructor Jeffery De Tullio 15 ft distance on a long hit. (See Problem 9.98.) V9.18 Jet ski VIDEO The drag coefficient for surface ships is a function of the Froude number. Froude Number Effects. Another parameter on which the drag coefficient may be strongly dependent is the Froude number, Fr = U √gℓ. As is discussed in Chapter 10, the Froude number is a ratio of the free-stream speed to a typical wave speed on the interface of two fluids, such as the surface of the ocean. An object moving on the surface, such as a ship, often produces waves that require a source of energy to generate. This energy comes from the ship and is manifest as a drag. [Recall that the rate of energy production (power) equals speed times force.] The nature of the waves produced often depends on the Froude number of the flow and the shape of the object—the waves generated by a water skier “plowing” through the water at a low speed (low Fr) are different than those generated by the skier “planing” along the surface at high speed (large Fr). Thus, the drag coefficient for surface ships is a function of Reynolds number (viscous effects) and Froude number (wave-making effects); CD = ϕ(Re, Fr). As was discussed in Chapter 7, it is often quite difficult to run model tests under conditions similar to those of the prototype (i.e., same Re and Fr for surface ships). Fortunately, the viscous and wave effects can often be separated, with the total drag being the sum of the drag of these individual effects. A detailed account of this important topic can be found in standard texts (Ref. 11). As is indicated in Fig. 9.28, the wave-making drag, 𝒟w, can be a complex function of the Froude number and the body shape. The rather “wiggly” dependence of the wave drag coefficient, 530 Chapter 9 ■ Flow over Immersed Bodies 0.0015 𝓵 U Hull with no bow bulb 0.0010 U 𝒟w CD = __________ 1 __ ρ U2 𝓵2 CDw Hull with bow bulb 2 0.0005 Design speed, Fr = 0.267 0 0.1 0.2 0.3 0.4 U Fr = ____ √𝓵g ■ Figure 9.28 Typical drag coefficient data as a function of Froude number and hull characteristics for that portion of the drag due to the generation of waves (adapted from Ref. 25). 𝒟pole 𝒟flag The drag on a complex body can be approximated as the sum of the drag on its parts. CDw = 𝒟w (ρU2ℓ2 2), on the Froude number shown is typical. It results from the fact that the structure of the waves produced by the hull is a strong function of the ship speed or, in dimensionless form, the Froude number. This wave structure is also a function of the body shape. For example, the bow wave, which is often the major contributor to the wave drag, can be reduced by use of an appropriately designed bulb on the bow, as is indicated in Fig. 9.28. In this instance the streamlined body (hull without a bulb) has more drag than the less streamlined one. Composite Body Drag. Approximate drag calculations for a complex body can often be obtained by treating the body as a composite collection of its various parts. For example, the total force on a flag pole because of the wind (see the figure in the margin) can be approximated by adding the aerodynamic drag produced by the various components involved—the drag on the flag and the drag on the pole. In some cases considerable care must be taken in such an approach because of the interactions between the various parts. It may not be correct to merely add the drag of the components to obtain the drag of the entire object, although such approximations are often reasonable. EXAMPLE 9 .13 Drag on a Composite Body GIVEN A 60-mph (i.e., 88-fps) wind blows past the water tower shown in Fig. E9.13a. FIND Estimate the moment (torque), M, needed at the base to keep the tower from tipping over. U = 60 mph = 88 fps Ds = 40 ft Dc = 15 ft b = 50 ft 𝒲 b + Ds /2 Dc b 𝒟c b/2 Rx SOLUTION We treat the water tower as a sphere resting on a circular cylinder and assume that the total drag is the sum of the drag from these parts. The free-body diagram of the tower is shown in Fig. 𝒟s Ds M (a) ■ Figure E9.13 (b) Ry 9.3 E9.13b. By summing moments about the base of the tower, we obtain Ds b M = 𝒟s b + + 𝒟c ( ) ( 2) 2 (1) 𝒟s = 0.5(0.00238 slugsft3 ) (88 fts) 2 1 ρU 2bDcCDc 2 (2) (3) UDs (88 fts) (40 ft) = 2.24 × 107 = ν 1.57 × 10−4 ft2s and From Eq. 1 the corresponding moment needed to prevent the tower from tipping is 40 50 M = 3470 lb 50 ft + ft + 4840 lb( ft) ( 2 ) 2 = 3.64 × 105 ft ˙ lb (Ans) COMMENT The above result is only an estimate because (a) the Rec = UDc (88 fts) (15 ft) = 8.41 × 106 = ν 1.57 × 10−4 ft2s The corresponding drag coefficients, CDs and CDc, can be approximated from Fig. 9.23 as CDs ≈ 0.3 and CDc ≈ 0.7 VIDEO and 𝒟c = 0.5(0.00238 slugsft3 ) (88 fts) 2 (50 ft × 15 ft) (0.7) = 4840 lb are the drag on the sphere and cylinder, respectively. For standard atmospheric conditions, the Reynolds numbers are Res = π (40 ft) 2 (0.3) 4 = 3470 lb and 𝒟c = 531 Note that the value of CDs was obtained by an extrapolation of the given data to Reynolds numbers beyond those given (a potentially dangerous practice!). From Eqs. 2 and 3 we obtain where 1 π 𝒟s = ρU 2 D2s CDs 2 4 Drag V9.19 Drag on a truck wind is probably not uniform from the top of the tower to the ground, (b) the tower is not exactly a combination of a smooth sphere and a circular cylinder, (c) the cylinder is not of infinite length, (d) there will be some interaction between the flow past the cylinder and that past the sphere so that the net drag is not exactly the sum of the two, and (e) a drag coefficient value was obtained by extrapolation of the given data. However, such approximate results are often quite accurate. The aerodynamic drag on automobiles provides an example of the use of adding component drag forces. The power required to move a car along a level street is used to overcome the rolling resistance and the aerodynamic drag. For speeds above approximately 30 mph, the aerodynamic drag becomes a significant contribution to the net propulsive force needed. The contributions of the drag due to various portions of cars (i.e., front end, windshield, roof, rear end, windshield peak, rear roof trunk, and cowl) has been determined by numerous model and full-sized tests as well as by numerical calculations. As a result, it is possible to predict the aerodynamic drag on cars of a wide variety of body styles. As is indicated in Fig. 9.29, the drag coefficient for cars has decreased rather continuously over the years. This reduction is a result of careful design of the shape and the details (such as window molding, rear view mirrors, etc.). An additional reduction in drag has been accomplished by a reduction of the projected area. The net result is a considerable increase in the gas mileage, especially at highway speeds. Considerable additional information about the aerodynamics of road vehicles can be found in the literature (Ref. 30). THE WIDE WORLD OF FLUIDS At 12,600 mpg it doesn’t cost much to “fill ’er up” Typical gas considerations, this amazing fuel efficiency was obtained. The consumption for a Formula 1 racer, a sports car, and a sedan is type (and number) of tires, the appropriate engine power and approximately 2 mpg, 15 mpg, and 30 mpg, respectively. Thus, weight, the specific chassis design, and the design of the body just how did the winning entry in the 2005 Shell Eco-Marathon shell are all important and interrelated considerations. To reduce achieve an incredible 12,600 mpg? To be sure, this vehicle is not drag, the aerodynamic shape of the high-efficiency vehicle was as fast as a Formula 1 racer (although the rules require it to aver- given special attention through theoretical considerations and age at least 15 mph), and it can’t carry as large a load as your wind tunnel model testing. The result is an amazing vehicle that family sedan can (the vehicle has barely enough room for the can travel a long distance without hearing the usual “fill ’er up.” driver). However, by using a number of clever engineering design (See Problem 9.99.) 532 Chapter 9 ■ Flow over Immersed Bodies 1920s, CD ≈ 0.8 VIDEO V9.20 Automobile streamlining 1970s, CD ≈ 0.4 2000s, CD ≈ 0.3 ■ Figure 9.29 The historical trend of streamlining automobiles to reduce their aerodynamic drag and increase their miles per gallon (adapted from Ref. 5). The effect of several important parameters (shape, Re, Ma, Fr, and roughness) on the drag coefficient for various objects has been discussed in this section. As stated previously, drag coefficient information for a very wide range of objects is available in the literature. Some of this information is given in Figs. 9.30, 9.31, and 9.32 below for a variety of two- and three-dimensional, natural and Reference area A Shape (b = length) Drag coefficient 𝒟 CD = ________ 1 ρU2A __ 2 D R/D CD R 0 0.02 0.17 0.33 2.2 2.0 1.2 1.0 Square rod with rounded corners A = bD R/D R D Rounded equilateral triangle A = bD Reynolds number Re = ρ UD/μ 0 0.02 0.08 0.25 Re = 105 CD 1.4 1.2 1.3 1.1 2.1 2.0 1.9 1.3 Re = 105 Semicircular shell A = bD 2.3 1.1 Re = 2 × 104 D Semicircular cylinder A = bD 2.15 1.15 Re > 104 D T-beam A = bD 1.80 1.65 Re > 104 D I-beam A = bD 2.05 Re > 104 D Angle A = bD 1.98 1.82 Re > 104 Hexagon A = bD 1.0 Re > 104 D D 𝓵 D Rectangle A = bD 𝓵/D CD < 0.1 0.5 0.65 1.0 2.0 3.0 1.9 2.5 2.9 2.2 1.6 1.3 Re = 105 ■ Figure 9.30 Typical drag coefficients for regular two-dimensional objects (Refs. 5, 6). 9.4 Reference area Shape A CD 4 1.17 0.42 Re > 104 Hollow hemisphere π 2 A = __ D 4 1.42 0.38 Re > 104 Thin disk π 2 A = __ D 1.1 Re > 103 Solid hemisphere D D 4 𝓵 D D D Circular rod parallel to flow π 2 A = __ D 4 𝓵/D CD 0.5 1.0 2.0 4.0 1.1 0.93 0.83 0.85 θ , degrees CD 10 30 60 90 0.30 0.55 0.80 1.15 Re > 105 Cone π 2 A = __ D Cube A = D2 1.05 Re > 104 Cube A = D2 0.80 Re > 104 π 2 A = __ D 0.04 Re > 105 4 533 Reynolds number Re = ρ UD/μ π 2 A = __ D D θ Drag coefficient Lift Re > 104 D D 4 Streamlined body ■ Figure 9.31 Typical drag coefficients for regular three-dimensional objects (Ref. 5). man-made objects. Recall that a drag coefficient of unity is equivalent to the drag produced by the 1 1 dynamic pressure acting on an area of size A. That is, 𝒟 = 2ρU2ACD = 2ρU2A if CD = 1. Typical nonstreamlined objects have drag coefficients on this order. 9.4 Lift As is indicated in Section 9.1, any object moving through a fluid will experience a net force of the fluid on the object. All objects experience a drag force, 𝒟, which is parallel to the upstream flow. A force normal to the flow, lift ℒ, may be present only if the object is not symmetrical or if it does not produce a symmetrical flow field (such as flow around a rotating sphere). Considerable effort has been put forth to understand the various properties of the generation of lift. Some objects, such as an airfoil on an airplane wing, a propellor, or turbine and compressor blades are designed to 534 Chapter 9 ■ Flow over Immersed Bodies Shape Reference area D Parachute Frontal area π 2 A = __ D 4 Porous parabolic dish Frontal area π 2 A = __ D 4 Drag coefficient CD 1.4 0 0.2 0.5 1.42 1.20 0.82 0.95 0.90 0.80 Porosity D Porosity = open area/total area Average person D Fluttering flag Standing CD A = 9 ft2 Sitting CD A = 6 ft2 Crouching CD A = 2.5 ft2 A = 𝓵D l Empire State Building 𝓵/D CD 1 2 3 0.07 0.12 0.15 Frontal area 1.4 Frontal area 1.8 Upright commuter A = 5.5 ft2 1.1 Racing A = 3.9 ft2 0.88 Drafting A = 3.9 ft2 0.50 Streamlined A = 5.0 ft2 0.12 Standard Frontal area 0.96 With fairing Frontal area 0.76 With fairing and gap seal Frontal area 0.70 U = 10 m/s U = 20 m/s U = 30 m/s Frontal area 0.43 0.26 0.20 Dolphin Wetted area 0.0036 at Re = 6 × 106 (flat plate has CDf = 0.0031) Large birds Frontal area 0.40 Six-car passenger train Bikes Tractor-trailer trucks Fairing Gap seal Tree U ■ Figure 9.32 Typical drag coefficients for objects of interest (Refs. 5, 6, 15, 20). 9.4 Lift 535 Denotes p > p0 Denotes p < p0 U, p0 ■ Figure 9.33 Pressure distribution on the surface of an automobile. generate lift. Other objects are designed to reduce the lift generated. For example, the lift on a car tends to reduce the contact force between the wheels and the ground, causing reduction in traction and cornering ability. It is desirable to reduce this lift on a passenger car, or for the case of a racecar, reverse the lift to the downward direction (known as downforce). 9.4.1 Surface Pressure Distribution The lift can be determined from Eq. 9.2 if the distributions of pressure and wall shear stress around the entire body are known. As is indicated in Section 9.1, such data are usually not known. Typically, the lift is given in terms of the lift coefficient, The lift coefficient is a dimensionless form of the lift. CL = ℒ 1 2 2 ρU A (9.39) which is obtained from experiments, advanced analysis, or numerical considerations. The lift coefficient is a function of the appropriate dimensionless parameters and, as the drag coefficient, can be written as The lift coefficient is a function of other dimensionless parameters. Usually most lift comes from pressure forces, not viscous forces. CL = ϕ(shape, Re, Ma, Fr, εℓ) The Froude number, Fr, is important only if there is a free surface present, as with an underwater “wing” used to support a high-speed hydrofoil surface ship. Often the surface roughness, ε, is relatively unimportant in terms of lift—it has more of an effect on the drag. The Mach number, Ma, is of importance for relatively high-speed subsonic and supersonic flows (i.e., Ma > 0.8), and the Reynolds number effect is often not great. The most important parameter that affects the lift coefficient is the shape of the object. Considerable effort has gone into designing optimally shaped lift-producing devices. We will emphasize the effect of the shape on lift—the effects of the other dimensionless parameters can be found in the literature (Refs. 13, 14, and 29). Most common lift-generating devices (i.e., airfoils, fans, spoilers on cars, etc.) operate in the large Reynolds number range in which the flow has a boundary layer character, with viscous effects confined to the boundary layers and wake regions. For such cases the wall shear stress, τw, contributes little to the lift. Most of the lift comes from the surface pressure distribution. A typical pressure distribution on a moving car is shown in Fig. 9.33. The distribution, for the most part, is consistent with simple Bernoulli equation analysis. Locations with high-speed flow (i.e., over the roof and hood) have low pressure, while locations with low-speed flow (i.e., on the grill and windshield) have high pressure. It is easy to believe that the integrated effect of this pressure distribution would provide a net upward force. For objects operating in very low Reynolds number regimes (i.e., Re < 1), viscous effects are important, and the contribution of the shear stress to the lift may be as important as that of the pressure. Such situations include the flight of minute insects and the swimming of microscopic organisms. The relative importance of τw and p in the generation of lift in a typical large Reynolds number flow is shown in Example 9.14. 536 Chapter 9 ■ Flow over Immersed Bodies EXAMPLE 9 .14 Lift from Pressure and Shear Stress Distributions GIVEN When a uniform wind of velocity U blows past the FIND If the pressure in the building is atmospheric (i.e., the semicircular building shown in Fig. E9.14a,b, the wall shear stress and pressure distributions on the outside of the building are as given previously in Figs. E9.8b and E9.9a, respectively. value, p0, far from the building), determine the lift coefficient and the lift on the roof. SOLUTION dimensionless form by using the dynamic pressure, ρU22, planform area, A = bD, and dimensionless shear stress From Eq. 9.2 we obtain the lift as ℒ=− ∫ p sin θ dA + ∫ τ w (1) cos θ dA As is indicated in Fig. E9.14b, we assume that on the inside of the building the pressure is uniform, p = p0, and that there is no shear stress. Thus, Eq. 1 can be written as ℒ=− π ∫ ∫ 0 bD − 2 [ ∫ ℒ= π D τw cos θ b( ) dθ 2 π (p − p0 ) sin θ dθ + ∫ ∫ 1 π ∫ 2√Re 0 π 0 0 1 1 2 ρU A [− 2 2 + or ℒ= to give D (p − p0 ) sin θ b( ) dθ 2 0 + F(θ) = τw (Re) 12(ρU22) τw cos θ dθ] (2) (p − p0 ) 1 2 2 ρU 0 sin θ dθ F(θ) cos θ dθ] ∫ π (p − p0 ) 1 2 2 ρU 0 U p0 pdA sin θ dθ = −1.76 τ w dA p0 dA θ D/2 dθ (a) (b) 1.0 0 6 –0.5 4 F(θ ) cos θ 2 p – p0 Cp sin θ = _______ sin θ 1 ρU2 __ 0.5 –1.0 –1.5 0 π __ 4 π __ 2 θ, rad (c) ■ Figure E9.14 3π ___ 4 (3) From the data in Figs. E9.8b and E9.9a, the values of the two integrals in Eq. 3 can be obtained by determining the area under the curves of [(p − p0 )(ρU22)] sin θ versus θ and F(θ) cos θ versus θ plotted in Figs. E9.14c and E9.14d. The results are where b and D are the length and diameter of the building, respectively, and dA = b(D2)dθ. Equation 2 can be put into U p0 π π 2 0 0 π __ 4 π __ 2 θ, rad (d ) 3π ___ 4 π 9.4 and 537 Hence, the lift coefficient is ∫ π F(θ) cos θ dθ = 3.92 CL = 0.88 + 0 Thus, the lift is ℒ= 1 1 1 ρU2A [(− ) (−1.76) + (3.92) ] 2 2 2√Re or ℒ = (0.88 + 1.96 1 √Re )( 2 ρU2A) (Ans) and 1.96 (3.82 × 106 ) 12 = 0.88 + 0.001 = 0.881 Note that the pressure contribution to the lift coefficient is 0.88, whereas that due to the wall shear stress is only 1.96(Re12 ) = 0.001. The Reynolds number dependency of CL is quite minor. The lift is pressure dominated. Recall from Example 9.9 that this is also true for the drag on a similar shape. From Eq. 4 with A = 20 ft × 50 ft = 1000 ft2 , we obtain the lift for the assumed conditions as 1 1 ℒ = 2ρU2ACL = 2 (0.00238 slugsft3 ) (30 fts) 2 (1000 ft2 ) (0.881) CL = ℒ 1 2 2 ρU A = 0.88 + 1.96 √Re (4) (Ans) U = 30 fts, b = 50 ft, and standard atmospheric conditions (ρ = 2.38 × 10−3 slugsft3 and ν = 1.57 × 10−4 ft2s), which gives a Reynolds number of Re = ℒ ~ U2 U or ℒ = 944 lb COMMENTS Consider a typical situation with D = 20 ft, ℒ Lift (30 fts) (20 ft) UD = 3.82 × 106 = ν 1.57 × 10−4 ft2s There is a considerable tendency for the building to lift off the ground. Clearly this is due to the object being nonsymmetrical. The lift force on a complete circular cylinder is zero, although the opposing fluid forces do tend to pull the upper and lower halves apart. A typical device designed to produce lift does so by generating a pressure distribution that is different on the top and bottom surfaces. For large Reynolds number flows these pressure distributions are usually directly proportional to the dynamic pressure, ρU2 2, with viscous effects being of secondary importance. Hence, as indicated by the figure in the margin, for a given airfoil the lift is proportional to the square of the airspeed. Two airfoils used to produce lift are indicated in Fig. 9.34. Clearly the symmetrical one cannot produce lift unless the angle of attack, α, is nonzero. Because of the asymmetry of the nonsymmetric airfoil, the pressure distributions on the upper and lower surfaces are different, and a lift is produced even with α = 0. Of course, there will be a certain value of α (less than zero for this case) for which the lift is zero. For this situation, the pressure distributions on the upper and lower surfaces are different, but their resultant (integrated) pressure forces will be equal and opposite. Since most airfoils are thin, it is customary to use the planform area, A = bc, in the definition of the lift coefficient. Here b is the length of the airfoil, or span, and c is the chord length—the length from the leading edge to the trailing edge as indicated in Fig. 9.34. Typical lift coefficients so defined are on the order of unity. That is, the lift force is on the order of the dynamic pressure times the planform area of the wing, ℒ ≈ (ρU2 2)A. The wing loading, defined as the average lift per unit area of the wing, ℒA, therefore, increases with speed. For example, the wing loading of the 1903 Wright Flyer aircraft was 1.5 lbft2, while for the present-day Boeing 747 aircraft it is 150 lbft2. The wing loading for a bumble bee is approximately 1 lbft2 (Ref. 15). U U α α Symmetrical c Nonsymmetrical ■ Figure 9.34 Symmetrical and nonsymmetrical airfoils. 538 Chapter 9 ■ Flow over Immersed Bodies 120 100 NACA 64(1) – 412 airfoil Re = 7 × 105 Stall 1.5 80 1.0 60 α = 8° α = 6° C __L CD 40 α = 4° CL 0.5 VIDEO V9.21 Stalled airfoil α = 2° 20 α = 0° α = –2° 0 0 α = –4° –20 –40 –8 –0.5 –4 0 4 8 0 α , degrees (a) 0.005 0.010 α = –6° 0.015 0.02 CD (b) ■ Figure 9.35 Two representations of the same lift and drag data for a typical airfoil: (a) lift-to-drag ratio as a function of angle of attack, with the onset of boundary layer separation on the upper surface indicated by the occurrence of stall, (b) the lift and drag polar diagram with the angle of attack indicated (Ref. 27). At large angles of attack the boundary layer separates and the wing stalls. Not stalled Stalled Although viscous effects and the wall shear stress contribute little to the direct generation of lift, they play an extremely important role in the design and use of lifting devices. This is because of the viscosity-induced boundary layer separation that can occur on nonstreamlined bodies such as airfoils that have too large an angle of attack (see Fig. 9.19). As is indicated in Fig. 9.35, 9.36, and 9.39, up to a certain point, the lift coefficient increases rather steadily with the angle of attack, α. If α is too large, the boundary layer on the upper surface separates, the flow over the wing develops a wide, turbulent wake region, the lift decreases, and the drag increases. This condition, as indicated by the figures in the margin, is termed stall. Such conditions are extremely dangerous if they occur while the airplane is flying at a low altitude where there is not sufficient time and altitude to recover from the stall. In many lift-generating devices the important quantity is the ratio of the lift to drag developed, ℒ𝒟 = CLCD. Such information is often presented in terms of CLCD versus α, as is shown in Fig. 9.35a, or in a lift-drag polar of CL versus CD with α as a parameter, as is shown in Fig. 9.35b. The most efficient angle of attack (i.e., largest CLCD) can be found by drawing a line tangent to the CL − CD curve from the origin, as is shown in Fig. 9.35b. High-performance airfoils generate lift that is perhaps 100 or more times greater than their drag. This translates into the fact that in still air they can glide a horizontal distance of 100 m for each 1 m drop in altitude. THE WIDE WORLD OF FLUIDS Courtesy of NASA VIDEO V9.22 Bat flying Bats feel turbulence Researchers have discov- when maneuvering to avoid obstacles, their eleva- ered that at certain locations on the wings of bats, tion control is erratic. When the hairs grow back, there are special touch-sensing cells with a tiny the bats regain their complete flying skills. It is hair poking out of the center of the cell. These proposed that these touch-sensing cells are used cells, which are very sensitive to air flowing across to detect turbulence on the wing surface and the wing surface, can apparently detect turbu- thereby tell bats when to adjust the angle of lence in the flow over the wing. If these hairs are attack and curvature of their wings in order to removed, the bats fly well in a straight line, but avoid stalling out in midair. 9.4 VIDEO V9.23 Trailing edge flap Lift 539 As is indicated above, the lift and drag on an airfoil can be altered by changing the angle of attack. This actually represents a change in the shape of the object. Other shape changes can be used to alter the lift and drag when desirable. In modern airplanes it is common to utilize leading edge and trailing edge flaps as is shown in Fig. 9.36. To generate the necessary lift during the relatively low-speed landing and takeoff procedures, the airfoil shape is altered by extending special flaps on the front andor rear portions of the wing. Use of the flaps considerably enhances the lift, although it is at the expense of an increase in the drag (the airfoil is in a “dirty” configuration). This increase in drag is not of much concern during landing and takeoff operations—the decrease in landing or takeoff speed is more important than is a temporary increase in drag. During normal flight with the flaps retracted (the “clean” configuration), the drag is relatively small, and the needed lift force is achieved with the smaller lift coefficient and the larger dynamic pressure (higher speed). THE WIDE WORLD OF FLUIDS Learning from nature For hundreds of years humans looked has no technological parallel. Birds can control the airflow toward nature, particularly birds, for insight about flying. over their wings by moving the feathers on their wingtips and However, all early airplanes that closely mimicked birds proved to the leading edges of their wings, providing designs that are be unsuccessful. Only after much experimenting with rigid (or at more efficient than the flaps and rigid, pivoting tail surfaces least nonflapping) wings did human flight become possible. of current aircraft (Ref. 15). On a smaller scale, understanding Recently, however, engineers have been turning to living systems— the mechanism by which insects dynamically manage unsta- birds, insects, and other biological models—in an attempt to pro- ble flow to generate lift may provide insight into the develop- duce breakthroughs in aircraft design. Perhaps it is possible that ment of microscale air vehicles. With new hi-tech materials, nature’s basic design concepts can be applied to airplane sys- computers, and automatic controls, aircraft of the future may tems. For example, by morphing and rotating their wings in three mimic nature more than was once thought possible. (See dimensions, birds have remarkable maneuverability that to date Problem 9.122.) A wide variety of lift and drag information for airfoils can be found in standard aerodynamics books (Ref. 13, 14, 29). 3.0 VIDEO V9.24 Leading edge flap 2.0 CL No flaps 1.0 Trailing edge slotted flap Double slotted trailing edge flaps (Data not Leading edge flap shown) 0 0 0.1 0.2 CD 0.3 ■ Figure 9.36 Typical lift and drag alterations possible with the use of various types of flap designs (Data from Ref. 21). 540 Chapter 9 ■ Flow over Immersed Bodies EXAMPLE 9 .15 Lift and Power for Human-Powered Flight GIVEN In 1977 the Gossamer Condor, shown in Fig. E9.15a, FIND Determine won the Kremer prize by being the first human-powered aircraft to complete a prescribed figure-of-eight course around two turning points 0.5 mi apart (Ref. 22). The following data pertain to this aircraft: (a) the lift coefficient, CL, and (b) the power, 𝒫, required by the pilot. flight speed = U = 15 fts wing size = b = 96 ft, c = 7.5 ft (average) U weight (including pilot) = 𝒲 = 210 lb drag coefficient = CD = 0.046 (based on planform area) power train efficiency = η ■ Figure E9.15a = power to overcome dragpilot power = 0.8 (Photograph © Don Monroe.) SOLUTION (a) For steady flight conditions the lift must be exactly balanced by the weight, or 1 𝒲 = ℒ = 2 ρU2ACL Thus, CL = 2𝒲 ρU2A where A = bc = 96 ft × 7.5 ft = 720 ft2, 𝒲 = 210 lb, and ρ = 2.38 × 10−3 slugsft3 for standard air. This gives CL = 2(210 lb) −3 (2.38 × 10 = 1.09 COMMENT This power level is obtainable by a well-conditioned athlete (as is indicated by the fact that the flight was successfully completed). Note that only 80% of the pilot’s power (i.e., 0.8 × 0.302 = 0.242 hp, which corresponds to a drag of 𝒟 = 8.86 lb) is needed to force the aircraft through the air. The other 20% is lost because of the power train inefficiency. By repeating the calculations for various flight speeds, the results shown in Fig. E9.15b are obtained. Note from Eq. 1 that for a constant drag coefficient, the power required increases as U3—a doubling of the speed to 30 ft/s would require an eightfold increase in power (i.e., 2.42 hp, well beyond the range of any human). slugsft3 ) (15 fts) 2 (720 ft2 ) (Ans) a reasonable number. The overall lift-to-drag ratio for the aircraft is CL CD = 1.090.046 = 23.7. 2.5 (b) The product of the power that the pilot supplies and the power train efficiency equals the useful power needed to overcome the drag, 𝒟. That is, 2.0 1.5 𝒫, hp η𝒫 = 𝒟U where 1.0 1 𝒟 = 2ρU2ACD 0.5 Thus, 0 1 2 ρACDU3 𝒟U 2 ρU ACDU 𝒫= = = η η 2η 0 5 (2.38 × 10−3 slugsft3 ) (720 ft2 ) (0.046) (15 fts) 3 2(0.8) 1 hp 𝒫 = 166 ft ∙ lbs (Ans) ( 550 ft ∙ lbs ) = 0.302 hp 10 15 U, ft/s (1) or 𝒫= (15, 0.302) ■ Figure E9.15b 20 25 30 9.4 Lift 541 9.4.2 Circulation Inviscid flow analysis can be used to obtain ideal flow past airfoils. Since viscous effects are of minor importance in the generation of lift, it should be possible to calculate the lift force on an airfoil by integrating the pressure distribution obtained from the equations governing inviscid flow past the airfoil. That is, the potential flow theory discussed in Chapter 6 should provide a method to determine the lift. Although the details are beyond the scope of this book, the following is found from such calculations (Ref. 4). The calculation of the inviscid flow past a two-dimensional airfoil gives a flow field as indicated in Fig. 9.37. The predicted flow field past an airfoil with no lift (i.e., a symmetrical airfoil at zero angle of attack, Fig. 9.37a) appears to be quite accurate (except for the absence of thin boundary layer regions). However, as is indicated in Fig. 9.37b, the calculated flow past the same airfoil at a nonzero angle of attack (but one small enough so that boundary layer separation would not occur) is not proper near the trailing edge. In addition, the calculated lift for a nonzero angle of attack is zero—in conflict with the known fact that such airfoils produce lift. In reality, the flow should pass smoothly over the top surface as is indicated in Fig. 9.37c, without the strange behavior indicated near the trailing edge in Fig. 9.37b. As is shown in Fig. 9.37d, the unrealistic flow situation can be corrected by adding an appropriate clockwise swirling flow around the airfoil. The results are twofold: (1) The unrealistic behavior near the trailing edge is eliminated (i.e., the flow pattern of Fig. 9.37b is changed to that of Fig. 9.37c), and (2) the average velocity on the upper surface of the airfoil is increased while that on the lower surface is decreased. From the Bernoulli equation concepts (i.e., pγ + V2 2g + z = constant), the average pressure on the upper surface is decreased and that on the lower surface is increased. The net effect is to change the original zero lift condition to that of a lift-producing airfoil. The addition of the clockwise swirl is termed the addition of circulation. The amount of swirl (circulation) needed to have the flow leave the trailing edge smoothly is a function of the airfoil size and shape and can be calculated from potential flow (inviscid) theory (see Section 6.6.3 and Ref. 29). Although the addition of circulation to make the flow field physically realistic may seem artificial, it has well-founded mathematical and physical grounds. For example, consider the flow past a α=0 ℒ=0 (a) α>0 ℒ=0 (b) α>0 ℒ>0 (c) + = “(b) + circulation = (c)” (d) ■ Figure 9.37 Inviscid flow past an airfoil: (a) symmetrical flow past the symmetrical airfoil at a zero angle of attack; (b) same airfoil at a nonzero angle of attack—no lift, flow near trailing edge not realistic; (c) same conditions as for (b) except circulation has been added to the flow—nonzero lift, realistic flow; (d) superposition of flows to produce the final flow past the airfoil. 542 Chapter 9 ■ Flow over Immersed Bodies U B Bound vortex A Trailing vortex (a) Bound vortex Low pressure A B High pressure Trailing vortex (b) ■ Figure 9.38 Flow past a finite-length wing: (a) the horseshoe vortex system produced by the bound vortex and the trailing vortices; (b) the leakage of air around the wing tips produces the trailing vortices. VIDEO V9.25 Wing tip vortices (Photograph courtesy of NASA.) finite-length airfoil, as is indicated in Fig. 9.38. For lift-generating conditions the average pressure on the lower surface is greater than that on the upper surface. Near the tips of the wing this pressure difference will cause some of the fluid to attempt to migrate from the lower to the upper surface, as is indicated in Fig. 9.38b. At the same time, this fluid is swept downstream, forming a trailing vortex (swirl) from each wing tip (see Fig. 4.3). It is speculated that the reason some birds migrate in vee-formation is to take advantage of the updraft produced by the trailing vortex of the preceding bird. [It is calculated that for a given expenditure of energy, a flock of 25 birds flying in vee-formation could travel 70% farther than if each bird were to fly separately (Ref. 15).] The trailing vortices from the right and left wing tips are connected by the bound vortex along the length of the wing. It is this vortex that generates the circulation that produces the lift. The combined vortex system (the bound vortex and the trailing vortices) is termed a horseshoe vortex. The strength of the trailing vortices (which is equal to the strength of the bound vortex) is proportional to the lift generated. Large aircraft (for example, a Boeing 747) can generate very strong trailing vortices that persist for a long time before viscous effects and instability mechanisms finally cause them to die out. Such vortices are strong enough to flip smaller aircraft out of control if they follow too closely behind the large aircraft. The figure in the margin clearly shows a trailing vortex produced during a wake vortex study in which an airplane flew through a column of smoke. The wing tip vortices are detrimental to wing performance, causing a decrease in lift and increase in drag, termed induced drag (Ref. 13). Increasing the aspect ratio, 𝒜, of the wing reduces these effects. The aspect ratio is defined as the ratio of the square of the wing length to the planform area, 𝒜 = b2/A. If the chord length, c, is constant along the length of the wing (a rectangular planform wing), this reduces to 𝒜 = b/c. Figure 9.39 shows the effect of aspect ratio on the lift and drag coefficient for a typical finite wing. In general, the lift coefficient increases and the drag coefficient decreases with an increase in aspect ratio. Thus, long wings are more efficient because the wing tip losses are lower than those for short wings. High-performance soaring airplanes and highly efficient soaring birds (e.g., the albatross and sea gull) have long, narrow wings. Such wings, however, have considerable inertia that inhibits rapid maneuvers. Thus, highly maneuverable fighter or acrobatic airplanes and birds (e.g., the falcon) have small-aspect-ratio wings. Other methods of reducing the 9.4 543 Lift 1.4 1.2 𝒜=7 1.0 𝒜=3 0.03 0.8 CL 0.6 𝒜=1 𝒜=1 0.4 𝒜=3 0.02 CD 0.2 𝒜=7 0.01 0 –0.2 –0.4 –10 0 10 20 α , degrees 0 –10 0 10 20 α , degrees (a) (b) ■ Figure 9.39 Typical lift and drag coefficient data as a function of angle of attack and the aspect ratio of a typical finite wing with the same airfoil shape: (a) lift coefficient, (b) drag coefficient. effects of induced drag include tapering a wing from the root to the tip or the addition of winglets (see The Wide World of Fluids). THE WIDE WORLD OF FLUIDS Why winglets? Winglets, those upward turning ends of air- Boeing 747-400 has a conventional, vertical upper winglet; and plane wings, boost the performance by reducing induced drag. the Boeing Business Jet (a derivative of the Boeing 737) has an This is accomplished by reducing the strength of the wing tip 8-ft winglet with a curving transition from wing to winglet. Since vortices formed by the difference between the high pressure the airflow around the winglets is quite complicated, the wing- on the lower surface of the wing and the low pressure on the lets must be carefully designed and tested for each aircraft. In upper surface of the wing. In essence, the winglet provides an the past, winglets were more likely to be retrofitted to existing effective increase in the aspect ratio of the wing without wings, but new airplanes are being designed with winglets from extending the wingspan. Winglets come in a variety of styles— the start. Unlike tailfins on cars, winglets really do work. (See the Airbus A320 has a very small upper and lower winglet; the Problem 9.123.) A spinning sphere or cylinder can generate lift. As is indicated above, the generation of lift is directly related to the production of a swirl or vortex flow around the object. A nonsymmetric airfoil, by design, generates its own prescribed amount of swirl and lift. A symmetric object like a circular cylinder or sphere, which normally provides no lift, can generate swirl and lift if it rotates. As is discussed in Section 6.6.3, the inviscid flow past a circular cylinder has the symmetrical flow pattern indicated in Fig. 9.40a. By symmetry the lift and drag are zero. However, if the cylinder is rotated about its axis in a stationary real (μ ≠ 0) fluid, the rotation will drag some of the fluid around, producing circulation about the cylinder as in Fig. 9.40b. When this circulation is combined with an ideal, uniform upstream flow, the flow pattern indicated in Fig. 9.40c is obtained. The flow is no longer symmetrical about the horizontal plane through the center of the cylinder; the average pressure is greater on the lower half of the cylinder than on the upper half, and a lift is generated. This effect is called the Magnus effect, after Heinrich Magnus (1802–1870), a German chemist and physicist who first investigated this phenomenon. A similar lift is generated on a rotating sphere. It accounts for the various types of pitches in baseball (e.g., curve ball, “rising” fastball, sinker, etc.), the ability of a soccer player to hook the ball, and the hook or slice of a golf ball. Typical lift and drag coefficients for a smooth, spinning sphere are shown in Fig. 9.41. Although the drag coefficient is fairly independent of the rate of rotation, the lift coefficient is strongly dependent on it. In addition (although not indicated in the figure), both CL and CD are 544 Chapter 9 ■ Flow over Immersed Bodies ω ω S S S S S = stagnation point (highest pressure) “(a) + (b) = (c)” (a) (b) (c) ■ Figure 9.40 Inviscid flow past a circular cylinder: (a) uniform upstream flow without circulation, (b) free vortex at the center of the cylinder, (c) combination of free vortex and uniform flow past a circular cylinder giving nonsymmetric flow and a lift. 0.8 𝒟 CD = ____________ π 2 1 ρU2 __ __ D 2 4 0.6 ω CD, CL U D Smooth sphere 0.4 ℒ CL = ____________ π 2 1 ρU2 __ __ D 2 4 0.2 ___ = 6 × 104 Re = UD v 0 A dimpled golf ball has less drag and more lift than a smooth one. 0 1 2 ω D/2U 3 4 5 ■ Figure 9.41 Lift and drag coefficients for a spinning smooth sphere (Ref. 23). dependent on the roughness of the surface. As was discussed in Section 9.3, in a certain Reynolds number range an increase in surface roughness actually decreases the drag coefficient. Similarly, an increase in surface roughness can increase the lift coefficient because the roughness helps drag more fluid around the sphere, increasing the circulation for a given angular velocity. Thus, a rotating, rough golf ball travels farther than a smooth one because the drag is less and the lift is greater. However, do not expect a severely roughed up (cut) ball to work better—extensive testing has gone into obtaining the optimum surface roughness for golf balls. EXAMPLE 9 .16 Lift on a Rotating Sphere GIVEN A table tennis ball weighing 2.45 × 10−2 N with diameter D = 3.8 × 10−2 m is hit at a velocity of U = 12 ms with a back spin of angular velocity ω as is shown in Fig. E9.16. FIND What is the value of ω if the ball is to travel on a horizontal path, not dropping due to the acceleration of gravity? 9.5 Chapter Summary and Study Guide 545 SOLUTION For horizontal flight, the lift generated by the spinning of the ball must exactly balance the weight, 𝒲, of the ball so that 1 𝒲 = ℒ = 2ρU2ACL or CL = ω 2𝒲 ρU (π4)D2 U Horizontal path with backspin 2 where the lift coefficient, CL, can be obtained from Fig. 9.39. For standard atmospheric conditions with ρ = 1.23 kgm3 we obtain Path without spin 2(2.45 × 10−2 N) CL = (1.23 kgm3 ) (12 ms) 2 (π4) (3.8 × 10−2 m) 2 = 0.244 ■ Figure E9.16 which, according to Fig. 9.39, can be achieved if ωD = 0.9 2U or ω= 2(12 ms) (0.9) 2U(0.9) = = 568 rads D 3.8 × 10−2 m Thus, ω = (568 rads) (60 smin) (1 rev2π rad) = 5420 rpm (Ans) COMMENT Is it possible to impart this angular velocity to the ball? With larger angular velocities the ball will rise and follow an upward curved path. Similar trajectories can be produced by a well-hit golf ball—rather than falling like a rock, the golf ball trajectory is actually curved up and the spinning ball travels a greater distance than one without spin. However, if topspin is imparted to the ball (as in an improper tee shot) the ball will curve downward more quickly than under the action of gravity alone—the ball is “topped” and a negative lift is generated. Similarly, rotation about a vertical axis will cause the ball to hook or slice to one side or the other. Recall the deceleration of a table tennis ball in Example 9.12. As the velocity drops, the spin on the ball will no longer impart a horizontal path. The fact that CD is higher than CL as shown in Fig. 9.41 makes it difficult to impart lift. It is, in fact, impossible for a baseball pitcher to impart enough angular and translational velocity on a thrown baseball to cause it to rise. A “rising” fastball is an illusion to the hitter because it drops less than every other pitch witnessed. (See Problem 9.120.) 9.5 Chapter Summary and Study Guide drag lift lift coefficient drag coefficient wake region boundary layer laminar boundary layer turbulent boundary layer boundary layer thickness transition free-stream velocity favorable pressure gradient adverse pressure gradient boundary layer separation friction drag pressure drag stall circulation Magnus effect In this chapter the flow past objects is discussed. It is shown how the pressure and shear stress distributions on the surface of an object produce the net lift and drag forces on the object. The character of flow past an object is a function of the Reynolds number. For large Reynolds number flows, a thin boundary layer forms on the surface. Properties of this boundary layer flow are discussed. These include the boundary layer thickness, whether the flow is laminar or turbulent, and the wall shear stress exerted on the object. In addition, boundary layer separation and its relationship to the pressure gradient are considered. The drag, which contains portions due to friction (viscous) effects and pressure effects, is written in terms of the dimensionless drag coefficient. It is shown how the drag coefficient is a function of shape, with objects ranging from very blunt to very streamlined. Other parameters affecting the drag coefficient include the Reynolds number, Froude number, Mach number, and surface roughness. The lift is written in terms of the dimensionless lift coefficient, which is strongly dependent on the shape of the object. Variation of the lift coefficient with shape is illustrated by the variation of an airfoil’s lift coefficient with angle of attack. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. 546 Chapter 9 ■ Flow over Immersed Bodies ■ ■ ■ ■ ■ ■ ■ determine the lift and drag on an object from the given pressure and shear stress distributions on the object. for flow past a flat plate, calculate the boundary layer thickness, the wall shear stress, and the friction drag, and determine whether the flow is laminar or turbulent. explain the concept of the pressure gradient and its relationship to boundary layer separation. explain methods to reduce drag for both blunt and streamlined objects. for a given object, obtain the drag coefficient from appropriate tables, figures, or equations and calculate the drag on the object. explain why golf balls have dimples. for a given object, obtain the lift coefficient from appropriate figures and calculate the lift on the object. Some of the important equations in this chapter are: 𝒟= Drag force Lift force ℒ= ∫ ∫ dFx = ∫ p cos θ dA + ∫τ ∫ p sin θ dA + ∫ dFy = − Lift coefficient and drag coefficient CL = Boundary layer displacement thickness δ* = ℒ 1 2 2 ρU A ∞ ∫ 0 ϴ= Boundary layer momentum thickness Blasius boundary layer thickness, displacement thickness, and momentum thickness for flat plate ∫ 0 ∞ , CD = 𝒟 (9.39), (9.36) u u 1 − ) dy ( U U (9.4) Drag on flat plate 𝒟 = ρbU2 ϴ √Rex (9.2) (9.3) τw = 0.332U32 0.664 τw cos θ dA 1 2 2 ρU A 5 1.721 ϴ 0.664 δ δ* = = , , = x x x √Rex √Rex √Rex cf = (9.1) u (1 − U ) dy Blasius wall shear stress for flat plate Blasius wall friction coefficient and friction drag coefficient for flat plate sin θ dA w , CDf = ρμ B x (9.15), (9.16), (9.17) (9.18) (9.23) 1.328 √Reℓ (9.32) References 1. 2. 3. 4. 5. 6. 7. 8. 9. Schlichting, H., Boundary Layer Theory, 8th Ed., McGraw-Hill, New York, 2000. Rosenhead, L., Laminar Boundary Layers, Oxford University Press, London, 1963. White, F. M., Viscous Fluid Flow, 3rd Ed., McGraw-Hill, New York, 2005. Currie, I. G., Fundamental Mechanics of Fluids, McGraw-Hill, New York, 1974. Blevins, R. D., Applied Fluid Dynamics Handbook, Van Nostrand Reinhold, New York, 1984. Hoerner, S. F., Fluid-Dynamic Drag, published by the author, Library of Congress No. 64,19666, 1965. Happel, J., Low Reynolds Number Hydrodynamics, Prentice-Hall, Englewood Cliffs, N.J., 1965. Van Dyke, M., An Album of Fluid Motion, Parabolic Press, Stanford, Calif., 1982. Thompson, P. A., Compressible-Fluid Dynamics, McGraw-Hill, New York, 1972. Problems 547 10. Zucrow, M. J., and Hoffman, J. D., Gas Dynamics, Vol. I, Wiley, New York, 1976. 11. Clayton, B. R., and Bishop, R. E. D., Mechanics of Marine Vehicles, Gulf Publishing Co., Houston, Tex., 1982. 12. CRC Handbook of Tables for Applied Engineering Science, 2nd Ed., CRC Press, Boca Raton, Fla., 1973. 13. Shevell, R. S., Fundamentals of Flight, 2nd Ed., Prentice-Hall, Englewood Cliffs, N.J., 1989. 14. Kuethe, A. M., and Chow, C. Y., Foundations of Aerodynamics, Bases of Aerodynamics Design, 4th Ed., Wiley, New York, 1986. 15. Vogel, J., Life in Moving Fluids, 2nd Ed., Willard Grant Press, Boston, 1994. 16. Kreider, J. F., Principles of Fluid Mechanics, Allyn and Bacon, Newton, Mass., 1985. 17. Dobrodzicki, G. A., Flow Visualization in the National Aeronautical Establishment’s Water Tunnel, National Research Council of Canada, Aeronautical Report LR-557, 1972. 18. White, F. M., Fluid Mechanics, 6th Ed., McGraw-Hill, New York, 2008. 19. Vennard, J. K., and Street, R. L., Elementary Fluid Mechanics, 7th Ed., Wiley, New York, 1995. 20. Gross, A. C., Kyle, C. R., and Malewicki, D. J., The Aerodynamics of Human Powered Land Vehicles, Scientific American, Vol. 249, No. 6, 1983. 21. Abbott, I. H., and Von Doenhoff, A. E., Theory of Wing Sections, Dover Publications, New York, 1959. 22. MacReady, P. B., “Flight on 0.33 Horsepower: The Gossamer Condor,” Proc. AIAA 14th Annual Meeting (Paper No. 78-308), Washington, D.C., 1978. 23. Goldstein, S., Modern Developments in Fluid Dynamics, Oxford Press, London, 1938. 24. Achenbach, E., Distribution of Local Pressure and Skin Friction around a Circular Cylinder in CrossFlow Up to Re = 5 × 106, Journal of Fluid Mechanics, Vol. 34, Pt. 4, 1968. 25. Inui, T., Wave-Making Resistance of Ships, Transactions of the Society of Naval Architects and Marine Engineers, Vol. 70, 1962. 26. Sovran, G., et al., eds., Aerodynamic Drag Mechanisms of Bluff Bodies and Road Vehicles, Plenum Press, New York, 1978. 27. Abbott, I. H., von Doenhoff, A. E., and Stivers, L. S., Summary of Airfoil Data, NACA Report No. 824, Langley Field, Va., 1945. 28. Society of Automotive Engineers Report HSJ1566, “Aerodynamic Flow Visualization Techniques and Procedures,” 1986. 29. Anderson, J. D., Fundamentals of Aerodynamics, 5th Ed., McGraw-Hill, New York, 2011. 30. Hucho, W. H., Aerodynamics of Road Vehicles, Butterworth–Heinemann, 1987. 31. Homsy, G. M., et al., Multimedia Fluid Mechanics, 2nd Ed., CD-ROM, Cambridge University Press, New York, 2008. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are no unique answers to these problems. Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. Section 9.1 General External Flow Characteristics 9.1 Assume that water flowing past the equilateral triangular bar shown in Fig. P9.1 produces the pressure distributions indicated. Determine the lift and drag on the bar and the corresponding lift and drag coefficients (based on frontal area). Neglect shear forces. Linear distribution U = 5 ft/s 0.1 ft p = 0.5 ρ U2 b = length = 4 ft ■ Figure P9.1 p = –0.25 ρ U2 548 Chapter 9 ■ Flow over Immersed Bodies 9.2 Fluid flows past the two-dimensional bar shown in Fig. P9.2. The pressures on the ends of the bar are as shown, and the average shear stress on the top and bottom of the bar is τavg. Assume that the drag due to pressure is equal to the drag due to viscous effects. (a) Determine τavg in terms of the dynamic pressure, ρU2/2. (b) Determine the drag coefficient for this object. 9.8 A 0.10-m-diameter circular cylinder moves through air with a speed U. The pressure distribution on the cylinder’s surface is approximated by the three straight-line segments shown in Fig. P9.8. Determine the drag coefficient on the cylinder. Neglect shear forces. 3 p = 1 ρU2 2 2 1 h Width = b τ avg 10h 0 p, N /m2 U p = –0.2 1 ρU2 2 τ avg ■ Figure P9.2 60 α = 7° –1 –2 –4 –5 –6 ■ Figure P9.8 Animal 2 pave = 2.3 kN/m τ ave = 7.6 × 10–2 kN/m2 ■ Figure P9.4 *9.5 The pressure distribution on the l-m-diameter circular disk in Fig. P9.5 is given in the table. Determine the drag on the disk. r (m) p (kNm2 ) 0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 4.34 4.28 4.06 3.72 3.10 2.78 2.37 1.89 1.41 0.74 0.0 p = –5 kN/m2 (a) (b) (c) (d) (e) large whale flying duck large dragonfly invertebrate larva bacterium Speed 10 ms 20 ms 7 ms 1 mms 0.01 mms Re 300,000,000 300,000 30,000 0.3 0.00003 †9.10 Estimate the Reynolds numbers associated with the following objects moving through water. (a) a kayak, (b) a minnow, (c) a submarine, (d) a grain of sand settling to the bottom, (e) you swimming. Are viscous or inertial effects more dominant for each case? 9.11 Approximately how fast can the wind blow past a 0.25-in.diameter twig if viscous effects are to be of importance throughout the entire flow field (i.e., Re < 1)? Explain. Repeat for a 0.004-in.diameter hair and a 6-ft-diameter smokestack. p = p(r) U θ , deg 80 100 120 140 160 180 9.9 Typical values of the Reynolds number for various animals moving through air or water are listed below. For which cases is inertia of the fluid important? For which cases do viscous effects dominate? For which cases would the flow be laminar; turbulent? Explain. pave = –1.2 kN/m2 τ ave = 5.8 × 10–2 kN/m2 U 40 –3 9.3 Repeat Problem 9.1 if the object is a cone (made by rotating the equilateral triangle about the horizontal axis through its tip) rather than a triangular bar. 9.4 The average pressure and shear stress acting on the surface of the 1-m-square flat plate are as indicated in Fig. P9.4. Determine the lift and drag generated. Determine the lift and drag if the shear stress is neglected. Compare these two sets of results. 20 r D = 1m ■ Figure P9.5 9.6 A nonspinning ball having a mass of 3 oz. is thrown vertically upward with a velocity of 100 mph and has zero velocity at a height 250 ft above the release point. Assume that the air drag on the ball is constant and find this constant “average” air drag. Neglect the buoyant force of air on the ball. 9.7 When you walk through still air at a rate of 1 m/s, would you expect the character of the airflow around you to be most like that depicted in Fig. 9.6a, b, or c? Explain. (See Video 9.3.) 9.12 Consider the following cases. (a) A small 0.6-in.-long fish swims with a speed of 0.8 in/s. Would a boundary layer type flow be developed along the sides of the fish? Explain. (b) A 12-ft-long kayak moves with a speed of 5 ft/s. Would a boundary layer type flow be developed along the sides of the boat? Explain. Section 9.2 Boundary Layer Characteristics 9.13 Water flows past a flat plate that is oriented parallel to the flow with an upstream velocity of 0.5 m/s. Determine the approximate location downstream from the leading edge where the boundary layer becomes turbulent. What is the boundary layer thickness at this location? 9.14 A viscous fluid flows past a flat plate such that the boundary layer thickness at a distance 1.3 m from the leading edge is 12 mm. Determine the boundary layer thickness at distances of 0.20, 2.0, and 20 m from the leading edge. Assume laminar flow. Problems 549 9.15 If the upstream velocity of the flow in Problem 9.14 is U = 1.5 m/s, determine the kinematic viscosity of the fluid. at the center and the trailing edge of the plate. Assume a laminar boundary layer. 9.16 Water flows past a flat plate with an upstream velocity of U = 0.02 m/s. Determine the water velocity a distance of 10 mm from the plate at distances of x = 1.5 m and x = 15 m from the leading edge. 9.21 An atmospheric boundary layer is formed when the wind blows over the Earth’s surface. Typically, such velocity profiles can be written as a power law: u = ayn, where the constants a and n depend on the roughness of the terrain. As is indicated in Fig. P9.21, typical values are n = 0.40 for urban areas, n = 0.28 for woodland or suburban areas, and n = 0.16 for flat open country (Ref. 23). (a) If the velocity is 20 ft/s at the bottom of the sail on your boat (y = 4 ft), what is the velocity at the top of the mast (y = 30 ft)? (b) If the average velocity is 10 mph on the tenth floor of an urban building, what is the average velocity on the sixtieth floor? 9.17 The typical shape of small cumulus clouds is as indicated in Fig. P9.17. Based on boundary layer ideas, explain why it is clear that the wind is blowing from right to left as indicated. u ~ y0.40 y, m 450 U 300 u ~ y0.28 u ~ y0.16 150 0 ■ Figure P9.17 ■ Figure P9.21 9.18 Because of the velocity deficit, U − u, in the boundary layer, the streamlines for flow past a flat plate are not exactly parallel to the plate. This deviation can be determined by use of the displacement thickness, δ*. For air blowing past the flat plate shown in Fig. P9.18, plot the streamline A−B that passes through the edge of the boundary layer (y = δB at x = ℓ) at point B. That is, plot y = y(x) for streamline A−B. Assume laminar boundary layer flow. U= y 1 m/s Streamline A–B B A Edge of boundary layer δB x 𝓵=4m 9.22 A 30-story office building (each story is 12 ft tall) is built in a suburban industrial park. Plot the dynamic pressure, ρu22, as a function of elevation if the wind blows at hurricane strength (75 mph) at the top of the building. Use the atmospheric boundary layer information of Problem 9.21. 9.23 Show that by writing the velocity in terms of the similarity variable η and the function f(η), the momentum equation for boundary layer flow on a flat plate (Eq. 9.9b) can be written as the ordinary differential equation given by Eq. 9.14. *9.24 Integrate the Blasius equation (Eq, 9.14) numerically to determine the boundary layer profile for laminar flow past a flat plate. Compare your results with those of Table 9.1. 9.25 An airplane flies at a speed of 400 mph at an altitude of 10,000 ft. If the boundary layers on the wing surfaces behave as those on a flat plate, estimate the extent of laminar boundary layer flow along the wing. Assume a transitional Reynolds number of Rexcr = 5 × 105. If the airplane maintains its 400-mph speed but descends to sea-level elevation, will the portion of the wing covered by a laminar boundary layer increase or decrease compared with its value at 10,000 ft? Explain. ■ Figure P9.18 9.19 Air enters a square duct through a 1-ft opening as is shown in Fig. P9.19. Because the boundary layer displacement thickness increases in the direction of flow, it is necessary to increase the cross-sectional size of the duct if a constant U = 2 ft/s velocity is to be maintained outside the boundary layer. Plot a graph of the duct size, d, as a function of x for 0 ≤ x ≤ 10 ft if U is to remain †9.26 If the boundary layer on the hood of your car behaves as one constant. Assume laminar flow. on a flat plate, estimate how far from the front edge of the hood the boundary layer becomes turbulent. How thick is the boundary U= layer at this location? 2 ft/s d(x) 1 ft x 2 ft/s 9.27 A laminar boundary layer velocity profile is approximated by u/U = [2 − (y/δ)] (y/δ) for y ≤ δ, and u = U for y > δ. (a) Show that this parabolic profile satisfies the appropriate boundary conditions. (b) Use the momentum integral equation to determine the boundary layer thickness, δ = δ(x). Compare the result with the exact Blasius solution. *9.28 Choose two of the velocity profiles and corresponding boundary layer parameter formulas presented in Table 9.2. Plot the thickness of the boundary layer and the wall shear stress versus x for 9.20 A smooth, flat plate of length ℓ = 6 m and width b = 4 m values of x from zero to the laminar–turbulent transition point. The is placed in water with an upstream velocity of U = 0.5 m/s. fluid is 20 °C water and the free-stream velocity is 10 m/s. Ask your Determine the boundary layer thickness and the wall shear stress instructor to specify which velocity profile models you should use. ■ Figure P9.19 550 Chapter 9 ■ Flow over Immersed Bodies *9.29 Air at 20 psia and 90 °F flows over a flat plate at 100 ft/s. Find the value of x at which the Reynolds number is 104. Choose three velocity profile models from Table 9.2 and plot the velocity profiles (u versus y) on the same axes. Ask your instructor to specify which velocity profiles to choose. 9.30 A laminar boundary layer velocity profile is approximated by the two straight-line segments indicated in Fig. P9.30. Use the momentum integral equation to determine the boundary layer thickness, δ = δ(x), and wall shear stress, τw = τw(x). Compare these results with those in Table 9.2. δ u 2U ___ 3 U ■ Figure P9.30 9.31 A laminar boundary layer velocity profile is approximated by u/U = 2(y/δ) − 2(y/δ)3 + (y/δ)4 for y ≤ δ, and u = U for y > δ. (a) Show that this profile satisfies the appropriate boundary conditions. (b) Use the momentum integral equation to determine the boundary layer thickness, δ = δ(x). Compare the result with the exact Blasius solution. 9.32 Consider 20 °C water flowing over a thin, wide, smooth flat plate aligned with the flow. The approach velocity is 60 km/hr and the Reynolds number based on the plate length is 2 × 105. Calculate the drag per unit width of the plate for each of the velocity profiles listed in Table 9.2. Next calculate the drag coefficient CD as used in Fig. 9.15. Which velocity profile(s) provide the best estimation of CD? Why? *9.33 For a fluid of specific gravity SG = 0.86 flowing past a flat plate with an upstream velocity of U = 5 m/s, the wall shear stress on the flat plate was determined to be as indicated in the table below. Use the momentum integral equation to determine the boundary layer momentum thickness, ϴ = ϴ(x). Assume ϴ = 0 at the leading edge, x = 0. 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 9.35 Two different fluids flow over two identical flat plates with the same laminar free-stream velocity. Both fluids have the same viscosity, but one is twice as dense as the other. What is the relationship between the drag forces for these two plates? 9.36 Fluid flows past a flat plate with a drag force 𝒟1. If the freestream velocity is doubled, will the new drag force, 𝒟2, be larger or smaller than 𝒟1 and by what amount? 9.38 The drag coefficient for a newly designed hybrid car is predicted to be 0.21. The cross-sectional area of the car is 30 ft2. Determine the aerodynamic drag on the car when it is driven through still air at 55 mph. δ /2 x (m) 9.34 Should a canoe paddle be made rough to get a “better grip on the water” for paddling purposes? Explain. 9.37 A model is placed in an airflow at standard conditions with a given velocity and then placed in water flow at standard conditions with the same velocity. If the drag coefficients are the same between these two cases, how do the drag forces compare between the two fluids? y 0 Section 9.3 Drag 9.39 A 5-m-diameter parachute of a new design is to be used to transport a load from flight altitude to the ground with an average vertical speed of 3 m/s. The total weight of the load and parachute is 200 N. Determine the approximate drag coefficient for the parachute. 9.40 A 180-1b man parachutes from a plane using a hemispherical parachute in air at 20 °F and 14.60 psia. Calculate the parachute diameter required for the man’s terminal velocity not to exceed 20 ft/s. Neglect the weight of the parachute. 9.41 A 70-kg soldier on a secret mission has to parachute from an airplane over the desert. It takes five seconds for him to release his parachute after he jumps. Taking the mass of the parachute as 5 kg, estimate the force acting on the soldier as he releases the parachute. The diameter of the parachute is 2.0 m. Neglect drag on the soldier’s body prior to the opening of the parachute. The air is at 20 °C. 9.42 The aerodynamic drag on a car depends on the “shape” of the car. For example, the car shown in Fig. P9.42 has a drag coefficient of 0.35 with the windows and roof closed. With the windows and roof open, the drag coefficient increases to 0.45. With the windows and roof open, at what speed is the amount of power needed to overcome aerodynamic drag the same as it is at 65 mph with the windows and roof closed? Assume the frontal area remains the same. Recall that power is force times velocity. τw (Nm2 ) — 13.4 9.25 7.68 6.51 5.89 6.57 6.75 6.23 5.92 5.26 Windows and roof closed: CD = 0.35 Windows open; roof open: CD = 0.45 ■ Figure P9.42 9.43 An automobile engine has a maximum power output of 70 hp, which occurs at an engine speed of 2200 rpm. A 10% power loss occurs through the transmission and differential. The rear wheels have a radius of 15.0 in. and the automobile is to have a maximum speed of 75 mph along a level road. At this speed, the power absorbed by the tires because of their continuous deformation is 27 hp. Find the maximum permissible drag coefficient for the automobile at this speed. The car’s frontal area is 24.0 ft2. 551 Problems 9.44 A woman is riding a bicycle down an 18% slope. Her velocity is 25 km/hr into an oncoming 25-km/hr wind. The air is at 15 °C and 101 kPa. Assuming that the only forces affecting the speed are the weight and drag, calculate the drag coefficient if the frontal area is 0.6 m2 and the combined mass is 54 kg. Speculate whether the rider is in the upright or racing position. 9.51 For small Reynolds number flows, the drag coefficient of an object is given by a constant divided by the Reynolds number (see Table 9.4). Thus, as the Reynolds number tends to zero, the drag coefficient becomes infinitely large. Does this mean that for small velocities (hence, small Reynolds numbers) the drag is very large? Explain. 9.45 A baseball is thrown by a pitcher at 95 mph through standard air. The diameter of the baseball is 2.82 in. Estimate the drag force on the baseball. Explain how the actual drag force may be different than your estimate. 9.52 A rectangular cartop carrier of 1.6-ft height, 4.0-ft length (front to back), and 4.2-ft width is attached to the top of a car. Estimate the additional power required to drive the car with the carrier at 65 mph through still air compared with the power required to drive only the car at 65 mph. 9.46 A logging boat is towing a log at 1.5 m/s against a river current that is 1 m/s. The log is 0.5 m in diameter and 2 m long. (a) Estimate the power required if the axis of the log is parallel to the flow. (b) Estimate the power required if the axis of the log is perpendicular to the flow. (c) Explain any reason(s) why the actual power required may be different than your estimates. 9.53 As shown in Video V9.2 and Fig. P9.53a, a kayak is a relatively streamlined object. As a first approximation in calculating the drag on a kayak, assume that the kayak acts as if it were a smooth, flat plate 17 ft long and 2 ft wide. Determine the drag as a function of speed and compare your results with the measured values given in Fig. P9.50b. Comment on reasons why the two sets of values may differ. 9.47 How fast do small water droplets of 0.06-μm (6 × 10−8 m) diameter fall through the air under standard sea-level conditions? Assume the drops do not evaporate. Repeat the problem for standard conditions at 5000-m altitude. 9.48 Determine the drag on a small circular disk of 0.01-ft diameter moving 0.01 ft/s through oil with a specific gravity of 0.87 and a viscosity 10,000 times that of water. The disk is oriented normal to the upstream velocity. By what percent is the drag reduced if the disk is oriented parallel to the flow? U 𝓵 𝓵 Measured drag 𝒟, lb 9.49 The square, flat plate shown in Fig. P9.49a is cut into four equal-sized pieces and arranged as shown in Fig. P9.45b. Determine the ratio of the drag on the original plate [case (a)] to the drag on the plates in the configuration shown in (b). Assume laminar boundary flow. Explain your answer physically. (a) 8 6 4 2 0 (a) U 𝓵/4 4𝓵 ( b) 2 4 6 Kayak speed U, ft/s 8 (b) ■ Figure P9.53 ■ Figure P9.49 9.54 A coal barge 1000 ft long and 100 ft wide is submerged a depth of 12 ft in 60 °F water. It is being towed at a speed of 12 mph. Estimate the friction drag on the barge. 9.50 Water flows past a triangular flat plate oriented parallel to the free stream as shown in Fig. P9.50. Integrate the wall shear stress over the plate to determine the friction drag on one side of the plate. Assume laminar boundary layer flow. 9.55 A three-bladed helicopter blade rotates at 200 rpm. If each blade is 12 ft long and 1.5 ft wide, estimate the torque needed to overcome the friction on the blades if they act as flat plates. 9.56 A thin, smooth sign is attached to the side of a truck as is indicated in Fig. P9.56. Estimate the friction drag on the sign when the truck is driven at 55 mph. 5 ft 45° U = 0.2 m/s Eli Van Lines 1.0 m 45° ■ Figure P9.50 20 ft ■ Figure P9.56 4 ft 552 Chapter 9 ■ Flow over Immersed Bodies 9.57 A 38.1-mm-diameter, 0.0245-N table tennis ball is released from the bottom of a 4-m-deep swimming pool. Assuming that the ball has reached its terminal velocity within 1 m, estimate the time required to rise the remaining distance to the surface. 9.66 A 20-m/s wind blows against a 20-m-tall, 0.12-m-diameter flag pole. (a) Determine the anchoring moment at the base of the pole. (b) Determine the anchoring moment if a 2-m by 2.5-m flag is attached to the top of the pole. See Fig. 9.32 for drag coefficient data for flags. 9.58 A hot-air balloon roughly spherical in shape has a volume of 70,000 ft3 and a weight of 500 lb (including passengers, basket, †9.67 During a flash flood, water rushes over a road as shown in balloon fabric, etc.). If the outside air temperature is 80 °F and Fig. P9.67 with a speed of 12 mph. Estimate the maximum water the temperature within the balloon is 165 °F, estimate the rate at depth, h, that would allow a car to pass without being swept away. which it will rise under steady-state conditions if the atmospheric List all assumptions and show all calculations. pressure is 14.7 psi. 9.59 It is often assumed that “sharp objects can cut through the air better than blunt ones.” Based on this assumption, the drag on the object shown in Fig. P9.59 should be less when the wind blows from right to left than when it blows from left to right. Experiments show that the opposite is true. Explain. U? U = 12 mph h U? ■ Figure P9.59 ■ Figure P9.67 9.60 An object falls at a rate of 100 ft/s immediately prior to the time that the parachute attached to it opens. The final descent rate with the chute open is 10 ft/s. Calculate and plot the speed of falling as a function of time from when the chute opens. Assume that the chute opens instantly, that the drag coefficient and air density remain constant, and that the flow is quasisteady. 9.68 With the rider in the racing position, how much more power is required to pedal a bicycle at 15 mph into a 20-mph head-wind than at 15 mph through still air? Pedaling at 15 mph in still air, how much more power is required for an upright position instead of a racing position? †9.61 Estimate the velocity with which you would contact the †9.69 Estimate the wind velocity necessary to knock over a 20-lb ground if you jumped from an airplane at an altitude of 5,000 ft garbage can that is 3 ft tall and 2 ft in diameter. List your assumpand (a) air resistance is negligible, (b) air resistance is important, tions. but you forgot your parachute, or (c) you use a 25-ft-diameter 9.70 On a day without any wind, your car consumes x gallons of parachute. gasoline when you drive at a constant speed, U, from point A to 9.62 As is discussed in Section 9.3, the drag on a rough golf ball point B and back to point A. Assume that you repeat the journey, may be less than that on an equal-sized smooth ball. Does it follow driving at the same speed, on another day when there is a steady that a 10-m-diameter spherical water tank resting on a 20-m-tall wind blowing from B to A. Would you expect your fuel consumpsupport should have a rough surface so as to reduce the moment tion to be less than, equal to, or greater than x gallons for this windy round-trip? Support your answer with appropriate analysis. needed at the base of the support when a wind blows? Explain. 9.63 A 12-mm-diameter cable is strung between a series of poles that are 50 m apart. Determine the horizontal force this cable puts on each pole if the wind velocity is 30 m/s. 9.64 A strong wind can blow a golf ball off the tee by pivoting it about point 1 as shown in Fig. P9.64. Determine the wind speed necessary to do this. Radius = 0.845 in. 9.71 The structure shown in Fig. P9.71 consists of three cylindrical support posts to which an elliptical flat plate sign is attached. Estimate the drag on the structure when a 50-mph wind blows against it. 16 ft Adrianna’s Petting Zoo 5 ft U Weight = 0.0992 lb (1) 0.6 ft 15 ft 0.20 in. 0.8 ft 15 ft 1 ft 15 ft ■ Figure P9.64 9.65 A 22 in. by 34 in. speed limit sign is supported on a 3-in.-wide, 5-ft-long pole. Estimate the bending moment in the pole at ground level when a 30-mph wind blows against the sign. (See Video V9.14.) List any assumptions used in your calculations. ■ Figure P9.71 Problems 9.72 A 25-ton (50,000-lb) truck coasts down a steep 7% mountain grade without brakes, as shown in Fig. P9.72. The truck’s ultimate steady-state speed, V, is determined by a balance between weight, rolling resistance, and aerodynamic drag. Determine V if the rolling resistance for a truck on concrete is 1.2% of the weight and the drag coefficient based on frontal area is 0.76. 553 b = width = 10 ft 12 ft Isabelle 2001 12 ft (a) CD = 0.70 7 V 100 Truck width = 10 ft Isabelle 2001 ■ Figure P9.72 9.73 Phil’s Pizza Parlor decides to place a thin, rectangular, plastic sign on top of its delivery van as shown in Fig. P9.73. The sign measures 2 ft by 5 ft. (a) Estimate the extra power required to drive the van in standard still air at 35 mph if the sign faces forward rather than sideways. (b) The supports for the sign consist of two 5 steel pipes 10 in. long and of 116 in. outside diameter. Estimate the power required to overcome air drag on these two supports with the sign facing forward. 5 ft 35 mph L’S PHI PIZ ZA 2 ft (b) CD = 0.96 ■ Figure P9.74 9.76 Estimate the energy required for an average person (see Fig. 9.32) to run a mile in 4 minutes in still standard air. Compare your estimate if you instead modeled the person as a cylinder 6 ft tall and 2 ft in diameter. 9.77 As shown in Video V9.11 and Fig. P9.77, a vertical wind tunnel can be used for skydiving practice. Estimate the vertical wind speed needed if a 160-lb person is to be able to “float” motionless when the person (a) curls up as in a crouching position or (b) lies flat. See Fig. 9.30 for appropriate drag coefficient data. U ■ Figure 9.73 Phil’s pizza van. 9.74 As shown in Video V9.19 and Fig. P9.74, the aerodynamic drag on a truck can be reduced by the use of appropriate air deflectors. A reduction in drag coefficient from CD = 0.96 to CD = 0.70 corresponds to a reduction of how many horsepower needed at a highway speed of 65 mph? *9.75 A full-sized automobile has a frontal area of 24 ft2, and a compact car has a frontal area of 13 ft2. Both have a drag coefficient of 0.5 based on the frontal area. Find the horsepower required to move each automobile along a level road in still air at 55 mph. Assume that the power required to deform the tires continuously at this speed (called rolling resistance) is equal to the power to overcome the air resistance. Estimate the gas mileage of both automobiles if the energy supplied to the drive wheels is 1 8 4 that available in the fuel. A gallon of fuel has 1.0 × 10 ft ∙ 1b available energy. ■ Figure P9.77 1 9.78 Compare the rise velocity of an 8 -in.-diameter air bubble 1 in water to the fall velocity of an 8 -in.-diameter water drop in air. Assume each to behave as a solid sphere. 9.79 A 50-lb box shaped like a 1-ft cube falls from the cargo hold of an airplane at an altitude of 30,000 ft. If the drag coefficient of the falling box is 1.2, determine the time it takes for the box to hit the ocean. Assume that it falls at the terminal velocity corresponding to its current altitude and use a standard atmosphere (see Table C.1). 9.80 A 500-N cube of specific gravity SG = 1.8 falls through water at a constant speed U. Determine U if the cube falls (a) as oriented in Fig. P9.80a, (b) as oriented in Fig. P9.80b. 554 Chapter 9 ■ Flow over Immersed Bodies 10-mm diameter 1 m long 0.6 m 20-mm diameter 1.5 m long g (a) (b) ■ Figure P9.80 40-mm diameter 5 m long 0.25 m *9.81 The helium-filled balloon shown in Fig P9.81 is to be used as a wind-speed indicator. The specific weight of the helium is y = 0.011 lb/ft3, the weight of the balloon material is 0.20 lb, and the weight of the anchoring cable is negligible. Plot a graph of θ as a function of U for 1 ≤ U ≤ 50 mph. Would this be an effective device over the range of U indicated? Explain. U 0.5 m ■ Figure P9.83 9.86 A 2-mm-diameter meteor of specific gravity 2.9 has a speed of 6 km/s at an altitude of 50,000 m where the air density is 1.03 × 10−3 kg/m3. If the drag coefficient at this large Mach number condition is 1.5, determine the deceleration of the meteor. 2-ft diameter 9.87 Air flows past two equal sized spheres (one rough, one smooth) that are attached to the arm of a balance as is indicated in Fig. P9.87. With U = 0 the beam is balanced. What is the minimum air velocity for which the balance arm will rotate clockwise? D = 0.1 m Rough sphere ε /D = 1.25 × 10–2 Smooth sphere θ 0.3 m ■ Figure P9.81 9.82 A 0.30-m-diameter cork ball (SG = 0.21) is tied to an object on the bottom of a river as is shown in Fig. P9.82. Estimate the speed of the river current. Neglect the weight of the cable and the drag on it. 0.5 m U ■ Figure P9.87 9.88 A 2-in.-diameter sphere weighing 0.14 lb is suspended by the jet of air shown in Fig. P9.88 and Video V3.2. The drag coefficient for the sphere is 0.5. Determine the reading on the pressure gage if friction and gravity effects can be neglected for the flow between the pressure gage and the nozzle exit. U 30° Area = 0.3 ft2 ■ Figure P9.82 9.83 A shortwave radio antenna is constructed from circular tubing, as is illustrated in Fig. P9.83. Estimate the wind force on the antenna in a 100-km/hr wind. Area = 0.6 ft2 Air Pressure gage †9.84 Estimate the wind force on your hand when you hold it out of your car window while driving 55 mph. Repeat your calculations if you were to hold your hand out of the window of an airplane flying 550 mph. ■ Figure P9.88 †9.85 Estimate the energy that a runner expends to overcome aerodynamic drag while running a complete marathon race. This expenditure of energy is equivalent to climbing a hill of what height? List all assumptions and show all calculations. 9.89 A smooth orange ball weighs 64 lb (at sea level) and has a diameter of 1.5 in. The discharge of a vacuum cleaner is directed 1 upward and supports the ball 2 in. above the hose outlet similar to 1 Problems 0.6 9.91 A marine location marker is a smoke-producing device usually dropped from an airplane and used to mark a reference point in the ocean. One is being tested in a wind tunnel to determine the drag force when it is carried by an airplane at a velocity of 200 mph. The marker is a cylinder with flat ends, has a diameter of 12 in., and is 48 in. long. Calculate the drag force on the marker in 10 °F air if the air flow is parallel to the cylinder’s axis. 9.92 The United Nations Building in New York is approximately 87.5 m wide and 154 m tall. (a) Determine the drag on this building if the drag coefficient is 1.3 and the wind speed is a uniform 20 m/s. (b) Repeat your calculations if the velocity profile against the building is a typical profile for an urban area (see Problem 9.21) and the wind speed halfway up the building is 20 m/s. 9.93 An airplane flies at 150 km/hr. (a) The airplane is towing a banner that is b = 0.8 m tall and ℓ = 25 m long. If the drag coefficient based on area bℓ is CD = 0.06, estimate the power required to tow the banner. (b) For comparison, determine the power required if the airplane was instead able to tow a rigid flat plate of the same size. (c) Explain why one had a larger power requirement (and larger drag) than the other. (d) Finally, determine the power required if the airplane was towing a smooth spherical balloon with a diameter of 2 m. 9.94 The paint stirrer shown in Fig. P9.94 consists of two circular disks attached to the end of a thin rod that rotates at 80 rpm. The specific gravity of the paint is SG = 1.1 and its viscosity is μ = 2 × 10−2 lb ∙ s/ft2. Estimate the power required to drive the mixer if the induced motion of the liquid is neglected. 80 rpm _7_ in. 8 L 0.5 U Calm wind 0.4 Strong wind U H W CD that shown in Video V3.2. If the hose inside diameter is 4.0 in., estimate the volume flow rate through the vacuum cleaner. The air temperature is 100 °F. 9.90 A 60 mph wind blows against a football stadium scoreboard that is 36 ft tall, 80 ft wide, and 8 ft thick (parallel to the wind). Estimate the wind force on the scoreboard. See Fig. 9.30 for drag coefficient data. 555 0.3 0.2 0.1 0 10,000 100,000 Re = ρUL /μ 1,000,000 ■ Figure P9.97 9.98 (See The Wide World of Fluids article “Dimpled Baseball Bats,” Section 9.3.3.) How fast must a 3.5-in.-diameter, dimpled baseball bat move through the air in order to take advantage of drag reduction produced by the dimples on the bat? Although there are differences, assume the bat (a cylinder) acts the same as a golf ball in terms of how the dimples affect the transition from a laminar to a turbulent boundary layer. 9.99 (See The Wide World of Fluids article “At 12,600 mpg It Doesn’t Cost Much to ‘Fill ‘er Up,’” Section 9.3.3.) (a) Determine the power it takes to overcome aerodynamic drag on a small (6 ft2 cross section), streamlined (CD = 0.12) vehicle traveling 15 mph. (b) Compare the power calculated in part (a) with that for a large (36 ft2 cross-sectional area), nonstreamlined (CD = 0.48) SUV traveling 65 mph on the interstate. Section 9.4 Lift 9.100 A rectangular wing with an aspect ratio of 6 is to generate 1000 lb of lift when it flies at a speed of 200 ft/s. Determine the length of the wing if its lift coefficient is 1.0. 9.101 A 1.2-lb kite with an area of 6 ft2 flies in a 20-ft/s wind such that the weightless string makes an angle of 55° relative to the horizontal. If the pull on the string is 1.5 lb, determine the lift and drag coefficients based on the kite area. 9.102 A Piper Cub airplane has a gross weight of 1750 lb, a cruising speed of 115 mph, and a wing area of 179 ft2. Determine the lift coefficient of this airplane for these conditions. 9.103 A light aircraft with a wing area of 200 ft2 and a weight of 2000 lb has a lift coefficient of 0.40 and a drag coefficient of 0.05. Determine the power required to maintain level flight. 1.5 in. ■ Figure P9.94 †9.95 If the wind becomes strong enough, it is “impossible” to paddle a canoe into the wind. Estimate the wind speed at which this will happen. List all assumptions and show all calculations. 9.96 By appropriate streamlining, the drag coefficient for an airplane is reduced by 12% while the frontal area remains the same. For the same power output, by what percentage is the flight speed increased? 9.97 As indicated in Fig. P9.97, the orientation of leaves on a tree is a function of the wind speed, with the tree becoming “more streamlined” as the wind increases. The resulting drag coefficient for the tree (based on the frontal area of the tree, HW) as a function of Reynolds number (based on the leaf length, L) is approximated as shown. Consider a tree with leaves of length L = 0.3 ft. What wind speed will produce a drag on the tree that is 6 times greater than the drag on the tree in a 15 ft/s wind? 9.104 An airplane weighs 320,000 lb, has a wing area of 2800 ft2, and has a wing length of 140 ft. The atmospheric pressure and temperature are 14.67 psia and 60 °F, respectively. The airplane is moving along a runway at 200 mph, and the wing has the lift and drag coefficients given in Fig. 9.39. Find the lift and drag on the airplane wings for an angle of attack of 10°. 9.105 As shown in V9.25 and Fig. P9.105, a spoiler (i.e., an upside-down airfoil) is mounted above the rear wheels of a race car to produce negative lift (i.e., downforce), thereby improving tractive force. The spoiler’s airfoil is angled 10º with the race track and has lift and drag coefficient characteristics of Fig. 9.39. The race car is traveling at 200 mph in air at 90 ºF and 14.72 psia. (a) If the coefficient of friction between the wheels and pavement is 0.6, by how much would use of the spoiler increase the maximum tractive force that would be generated between the wheels and the track? Assume the airspeed past the spoiler equals the car speed and that the airfoil acts directly over the drive wheels. (b) How much drag is added to the car with the use of the spoiler’s airfoil? 556 Chapter 9 ■ Flow over Immersed Bodies b = spoiler length = 4 ft Spoiler sfglfbkjxfdbaerg 200 mph 33 11⁄3 1 ℒairplane = 2 CL ρAwingV 2 ft akkjbgfkgbsgboiabkv and GOOD YEAR The Frugal Hoosier 1 𝒟airplane = 2 CD ρAwingV 2, where V is the airplane velocity. Which wing design would you recommend? Support your recommendation. *9.109 Air blows over the flat-bottomed, two-dimensional object shown in Fig. P9.109. The shape of the object, y = y(x), and the fluid speed along the surface, u = u(x), are given in the table. 9.106 The wings of old airplanes are often strengthened by the use Determine the lift coefficient for this object. of wires that provided cross-bracing as shown in Fig. P9.106. If the drag coefficient for the wings was 0.020 (based on the planform y area), determine the ratio of the drag from the wire bracing to that u = u(x) from the wings. ■ Figure P9.105 U u=U x c ■ Figure P9.109 Speed: 70 mph Wing area: 148 ft 2 Wire: length = 160 ft diameter = 0.05 in. ■ Figure P9.106 9.107 A wing generates a lift ℒ when moving through sea-level air with a velocity U. How fast must the wing move through the air at an altitude of 10,000 m with the same lift coefficient if it is to generate the same lift? 9.108 A design group has two possible wing designs (A and B) for an airplane wing. The planform area of either wing is 130 m2 and each must provide a lift of 1,550,000 N. 0.020 2.0 0.015 1.5 CL CDA CDB CLA 0.5 0.010 0.005 y(% c) u/U 0 2.5 5.0 7.5 10 20 30 40 50 60 70 80 90 100 0 3.72 5.30 6.48 7.43 9.92 11.14 11.49 10.45 9.11 6.46 3.62 1.26 0 0 0.971 1.232 1.273 1.271 1.276 1.295 1.307 1.308 1.195 1.065 0.945 0.856 0.807 *9.110 When air flows past the airfoil shown in Fig. P9.110, the velocity just outside the boundary layer, u, is as indicated. Estimate the lift coefficient for these conditions. CLB 1.0 x(% c) CD 1.6 Upper surface 0 0 1.2 –0.005 –0.5 ( Uu ) Lower surface 2 –1.0 –12 –8 –4 0 4 8 12 –0.010 16 0.8 Angle of attack α (degrees) ■ Figure P9.108 NACA 632−015 u 0.4 U The airplane is to fly at 700 km/hr at an altitude of 10,000 m in the Standard Atmosphere. The lift and drag coefficients are shown in Fig. P9.108. Both sets of lift and drag coefficients give the total lift ℒ and the total drag 𝒟 on the airplane per unit area of the wing so that 0 0 0.2 ■ Figure P9.110 0.4 x c 0.6 0.8 1.0 Problems 557 9.111 A Boeing 747 aircraft weighing 580,000 lb when loaded *9.121 A baseball leaves the pitcher’s hand with horizontal velocwith fuel and 100 passengers takes off with an airspeed of 140 mph. ity of 90 mph and travels a distance of 45 ft. Neglect air drag and With the same configuration (i.e., angle of attack, flap settings, gravity, so the ball moves in a horizontal plane. The ball has a mass etc.), what is its takeoff speed if it is loaded with 372 passengers? of 5 oz, a circumference of 9 in., a rotational speed of 1600 rev/min, and a baseball lift coefficient of 0.75. How far does the baseball Assume each passenger with luggage weighs 200 lb. “break’’ in the horizontal plane in 60 °F, still air? 9.112 Show that for unpowered flight (for which the lift, drag, and weight forces are in equilibrium) the glide slope angle, θ, is given 9.122 (See The Wide World of Fluids article “Learning from Nature,” Section 9.4.1.) As indicated in Fig. P9.122, birds can by tan θ = CDCL. significantly alter their body shape and increase their planform 9.113 A sail plane with a lift-to-drag ratio of 25 flies with a speed area, A, by spreading their wing and tail feathers, thereby reducing of 50 mph. It maintains or increases its altitude by flying in ther- their flight speed. If during landing the planform area is increased mals, columns of vertically rising air produced by buoyancy effects by 50% and the lift coefficient increased by 30% while all other of nonuniformly heated air. What vertical airspeed is needed if the parameters are held constant, by what percent is the flight speed sail plane is to maintain a constant altitude? reduced? 9.114 If the lift coefficient for a Boeing 777 aircraft is 15 times greater than its drag coefficient, can it glide from an altitude of 30,000 ft to an airport 80 mi away if it loses power from its engines? Explain. (See Problem 9.112.) 9.115 Over the years there has been a dramatic increase in the flight speed (U), altitude (h), weight (𝒲), and wing loading (𝒲/A = weight divided by wing area) of aircraft. Use the data given in the table below to determine the lift coefficient for each of the aircraft listed. Aircraft Year 𝒲, lb Wright Flyer Douglas DC-3 Douglas DC-6 Boeing 747 1903 1935 1947 1970 750 25,000 105,000 800,000 U, mph 𝒲/A, lb/ft2 35 180 315 570 1.5 25.0 72.0 150.0 ■ Figure P9.122 h, ft 0 10,000 15,000 30,000 9.116 If the required takeoff speed of a particular airplane is 120 mi/hr at sea level, what will be required at Denver (elevation 5000 ft)? Use properties of the U.S. Standard Atmosphere. 9.117 The landing speed of a winged aircraft such as the Space Shuttle is dependent on the air density. (See Video V9.1.) By what percent must the landing speed be increased on a day when the temperature is 110 °F compared to a day when it is 50 °F? Assume that the atmospheric pressure remains constant. 9.118 Commercial airliners normally cruise at relatively high altitudes (30,000 to 35,000 ft). Discuss how flying at this high altitude (rather than 10,000 ft, for example) can save fuel costs. 9.119 A pitcher can pitch a “curve ball” by putting sufficient spin on the ball when it is thrown. A ball that has absolutely no spin will follow a “straight” path. A ball that is pitched with a very small amount of spin (on the order of one revolution during its flight between the pitcher’s mound and home plate) is termed a knuckle ball. A ball pitched this way tends to “jump around” and “zig-zag” back and forth. Explain this phenomenon. Note: A baseball has seams. 9.120 For many years, hitters have claimed that some baseball pitchers have the ability to actually throw a rising fastball. Assuming that a top major leaguer pitcher can throw a 95-mph pitch and impart an 1800-rpm spin to the ball, is it possible for the ball to actually rise? Assume the baseball diameter is 2.9 in. and its weight is 5.25 oz. 9.123 (See The Wide World of Fluids article “Why Winglets?,” Section 9.4.2.) It is estimated that by installing appropriately designed winglets on a certain airplane the drag coefficient will be reduced by 5%. For the same engine thrust, by what percent will the aircraft speed be increased by use of the winglets? ■ Lifelong Learning Problems 9.1 LL One of the “The Wide World of Fluids” articles in this chapter discusses pressure-sensitive paint—a new technique of measuring surface pressure. There have been other advances in fluid measurement techniques, particularly in velocity measurements. One such technique is particle image velocimetry, or PIV. Obtain information about PIV and its advantages. Summarize your findings in a brief report. 9.2 LL For typical aircraft flying at cruise conditions, it is advantageous to have as much laminar flow over the wing as possible because there is an increase in friction drag once the flow becomes turbulent. Various techniques have been developed to help promote laminar flow over the wing, both in airfoil geometry configurations as well as active flow control mechanisms. Obtain information on one of these techniques. Summarize your findings in a brief report. 9.3 LL We have seen in this chapter that streamlining an automobile can help to reduce the drag coefficient. One of the methods of reducing the drag has been to reduce the projected area. However, it is difficult for some road vehicles, such as a tractor-trailer, to reduce this projected area due to the storage volume needed to haul the required load. Over the years, work has been done to help minimize some of the drag on this type of vehicle. Obtain information on a method that has been developed to reduce drag on a tractortrailer. Summarize your findings in a brief report. 10 Open-Channel Flow Learning Objectives After completing this chapter, you should be able to: ■ discuss the general characteristics of open-channel flow. ■ use a specific energy diagram. ■ apply appropriate equations to analyze uniform flow in an open channel. ■ calculate key properties of a hydraulic jump. ■ determine flowrates based on open-channel flow-measuring devices. VIDEO V10.1 Offshore oil drilling platform. 10.1 Open-channel flow can have a variety of characteristics. Open-channel flow involves the flow of a liquid in a channel or conduit that is not completely filled. A free surface exists between the flowing fluid (usually water) and fluid above it (usually the atmosphere). The main driving force for such flows is the fluid weight—gravity forces the fluid to flow downhill. Most open-channel flow results are based on correlations obtained from model and full-scale experiments. Additional information can be gained from various analytical and numerical efforts. Open-channel flows are essential to the world as we know it. The natural drainage of water through the numerous creek and river systems is an example of open-channel flow. Although the flow geometry for these systems is extremely complex, the resulting flow properties are of considerable economic, ecological, and recreational importance. Other examples of open-channel flows include the flow of rainwater in the gutters of our houses; the flow in canals, drainage ditches, sewers, and gutters along roads; the flow of small rivulets and sheets of water on a road during a rain; and the flow in the chutes of water rides in amusement parks. Open-channel flow involves a free surface which can deform into various shapes. Thus, a brief introduction into the properties and characteristics of surface waves is included. The purpose of this chapter is to investigate the basic features of open-channel flow. Because of the amount and variety of material available, only a brief introduction to the topic can be presented. Further information can be obtained from the references indicated. General Characteristics of Open-Channel Flow In our study of pipe flow (Chapter 8), we found that there are many ways to classify a flow— developing, fully developed, laminar, turbulent, and so on. For open-channel flow, the existence of a free surface allows additional types of flow. The extra freedom that allows the fluid to select its free-surface location and configuration (because it does not completely fill a pipe or conduit) allows 559 560 Chapter 10 ■ Open-Channel Flow Uniform flow Rapidly varying flow (Photo by Marty Melchior) important phenomena in open-channel flow that cannot occur in pipe flow. Some of the classifications of the flows are described below. The manner in which the fluid depth, y, varies with time, t, and distance along the channel, x, is used to partially classify a flow. For example, the flow is unsteady or steady depending on whether the depth at a given location does or does not change with time. Some unsteady flows can be viewed as steady flows if the reference frame of the observer is changed. For example, a tidal bore (difference in water level) moving up a river is unsteady to an observer standing on the bank, but steady to an observer moving along the bank with the speed of the wave front of the bore. Other flows are unsteady regardless of the reference frame used. The complex, time-dependent, windgenerated waves on a lake are in this category. In this book we will consider only steady openchannel flows. An open-channel flow is classified as uniform flow (UF) if the depth of flow does not vary along the channel (dy/dx = 0). Conversely, it is nonuniform flow or varied flow if the depth varies with distance (dy/dx ≠ 0). Nonuniform flows are further classified as rapidly varying flow (RVF) if the flow depth changes considerably over a relatively short distance; dy/dx ∼ 1. Gradually varying flows (GVF) are those in which the flow depth changes slowly with distance along the channel; dy/dx ⪡ 1. Examples of these types of flow are illustrated in Fig. 10.1 and the photographs in the margin. The relative importance of the various types of forces involved (pressure, weight, shear, inertia) is different for the different types of flows. As for any flow geometry, open-channel flow may be laminar, transitional, or turbulent, depending on various conditions involved. We can determine which type of flow exists by examining the Reynolds number, Re = ρVRh /μ, where V is the average velocity of the fluid and Rh is the hydraulic radius of the channel (see Section 10.4). A general rule is that open-channel flow is laminar if Re < 500, turbulent if Re > 12,500, and transitional otherwise. The values of these dividing Reynolds numbers are only approximate—a precise knowledge of the channel geometry is necessary to obtain specific values. Since most open-channel flows involve water (which has a fairly small viscosity) and have relatively large characteristic lengths, it is rare to have laminar openchannel flows. For example, flow of 50 °F water (ν = 1.41 × 10 −5 ft 2/s) with an average velocity of V = 1 ft/s in a stream with a hydraulic radius of Rh = 10 ft has Re = VRh /ν = 7.1 × 105. The flow is turbulent. However, flow of a thin sheet of water down a driveway with an average velocity of V = 0.25 ft/s such that Rh = 0.02 ft (in such cases the hydraulic radius is approximately equal to the fluid depth; see Section 10.4) has Re = 355. The flow is laminar. In some cases stratified flows are important. In such situations layers of two or more fluids of different densities flow in a channel. A layer of oil on water is one example of this type of flow. All of the open-channel flows considered in this book are homogeneous flows. That is, the fluid has uniform properties throughout. Open-channel flows involve a free surface that can deform from its undisturbed, relatively flat configuration to form waves. Such waves move across the surface at speeds that depend on their size (height, length) and properties of the channel (depth, fluid velocity, etc.). The character of an open-channel flow may depend strongly on how fast the fluid is flowing relative to how fast UF uniform flow GVF gradually varying flow RVF rapidly varying flow y RVF UF RVF UF RVF GVF ■ Figure 10.1 Classification of open-channel flow. RVF UF 10.2 Fr = V √gy 1 Supercritical Critical Subcritical Surface Waves 561 a typical wave moves relative to the fluid. The dimensionless parameter that describes this behavior is termed the Froude number, Fr = V/(gℓ) 1/2, where ℓ is an appropriate characteristic length of the flow. This dimensionless parameter was introduced in Chapter 7 and is discussed more fully in Section 10.2. As shown by the figure in the margin, the special case of a flow with a Froude number of unity, Fr = 1, is termed a critical flow. If the Froude number is less than 1, the flow is subcritical (or tranquil). A flow with the Froude number greater than 1 is termed supercritical (or rapid). 0 10.2 Surface Waves The distinguishing feature of flows involving a free surface (as in open-channel flows) is the opportunity for the free surface to deform into various shapes. The surface of a lake or the ocean is seldom “smooth as a mirror.” It is usually distorted into ever-changing patterns associated with surface waves as shown in the photos in the margin. Some of these waves are very high, some barely ripple the surface; some waves are very long (the distance between wave crests), some are short; some are breaking waves that form whitecaps, others are quite smooth. Although a general study of this wave motion is beyond the scope of this book, an understanding of certain fundamental properties of simple waves is necessary for open-channel flow considerations. The interested reader is encouraged to use some of the excellent references available for further study about wave motion (Refs. 1, 2, 3). THE WIDE WORLD OF FLUIDS Rogue waves There is a long history of stories concerning shore oil platforms indicate that such rogue waves can reach giant rogue ocean waves that come out of nowhere and cap- a height of 30 m. Although researchers still do not under- size ships. The movie The Poseidon Adventure (1972, 2006) is stand the formation of these large rogue waves, there are based on such an event. Although these giant, freakish several suggestions as to how ordinary smaller waves can be waves were long considered fictional, recent satellite obser- focused into one spot to produce a giant wave. Additional vations and computer simulations prove that, although rare, theoretical calculations and wave tank experiments are need- they are real. Such waves are single, sharply peaked mounds ed to adequately grasp the nature of such waves. Perhaps it of water that travel rapidly across an otherwise relatively calm will eventually be possible to predict the occurrence of these ocean. Although most ships are designed to withstand waves destructive waves, thereby reducing the loss of ships and life up to 15 m high, satellite measurements and data from off- because of them. 10.2.1 Wave Speed VIDEO V10.2 Filling your car’s gas tank. Consider the situation illustrated in Fig. 10.2a in which a single elementary wave of small height, δy, is produced on the surface of a channel by suddenly moving the initially stationary end wall with speed δV. The water in the channel was stationary at the initial time, t = 0. A stationary observer will observe a single wave move down the channel with a wave speed c, with no fluid motion ahead of the wave and a fluid velocity of δV behind the wave. The motion is unsteady for such an observer. For an observer moving along the channel with speed c, the flow will appear steady as shown in Fig. 10.2b. To this observer, the fluid velocity will be V = −ciˆ on the observer’s right and V = (−c + δV)iˆ to the left of the observer. The relationship between the various parameters involved for this flow can be obtained by application of the continuity and momentum equations to the control volume shown in Fig. 10.2b as follows. With the assumption of uniform one-dimensional flow, the continuity equation (Eq. 5.12) becomes −cyb = (−c + δV)(y + δy)b where b is the channel width. This simplifies to c= (y + δy)δV δy 562 Chapter 10 ■ Open-Channel Flow δy δV Moving end wall c = wave speed δV Stationary fluid y x (a) Control surface Channel width = b Stationary wave ^ V = (– c + δ V )i ^ y +δy V = – ci (1) (b) y x (2) ■ Figure 10.2 (a) Production of a single elementary wave in a channel as seen by a stationary observer. (b) Wave as seen by an observer moving with a speed equal to the wave speed. or in the limit of small-amplitude waves with δy ⪡ y c=y δV δy (10.1) Similarly, the momentum equation (Eq. 5.22) is 1 2 1 γy2b − 2 γ(y + δy) 2b = ρbcy[(c − δV) − c] where we have written the mass flowrate as m˙ = ρbcy and have assumed that the pressure variation is hydrostatic within the fluid. That is, the pressure forces on the channel cross sections (1) and (2) are F1 = γyc1A1 = γ(y + δy) 2b/2 and F2 = γyc2 A2 = γy2 b/2, respectively. If we again impose the assumption of small-amplitude waves [i.e., (δy) 2 ⪡ y δy], the momentum equation reduces to g δV = c δy (10.2) Combination of Eqs. 10.1 and 10.2 gives the wave speed c = √gy 10 c, m/s 8 6 c = √gy 4 2 0 0 2 4 6 8 10 y, m The wave speed can be obtained from the continuity and momentum equations or the continuity and energy equations. (10.3) as indicated by the figure in the margin. The speed of a small-amplitude solitary wave as is indicated in Fig. 10.2 is proportional to the square root of the fluid depth, y, and independent of the wave amplitude, δy. The fluid density is not involved, although the acceleration of gravity is. This is a result of the fact that such wave motion is a balance between inertial effects (proportional to ρ) and weight or hydrostatic pressure effects (proportional to γ = ρg). A ratio of these forces eliminates the common factor ρ but retains g. For extremely small waves (like those produced by insects on water as shown in Video V10.3), Eq. 10.3 is not valid because the effects of surface tension are significant. The wave speed can also be calculated by using the energy and continuity equations rather than the momentum and continuity equations as is done above. A simple wave on the surface is shown in Fig. 10.3. As seen by an observer moving with the wave speed, c, the flow is steady. Since the pressure is constant at any point on the free surface, the Bernoulli equation for this frictionless flow is simply V2 + y = constant 2g 10.2 Stationary wave V=c Surface Waves 563 δy V = c + δV Moving fluid y ■ Figure 10.3 Stationary simple wave in a flowing fluid. or by differentiating V δV + δy = 0 g V10.3 Water strider VIDEO Also, by differentiating the continuity equation, Vy = constant, we obtain y δV + V δy = 0 We combine these two equations to eliminate δV and δy and use the fact that V = c for this situation (the observer moves with speed c) to obtain the wave speed given by Eq. 10.3. The above results are restricted to waves of small amplitude because we have assumed onedimensional flow. That is, δy/y ⪡ 1. More advanced analysis and experiments show that the wave speed for finite-sized solitary waves exceeds that given by Eq. 10.3. To a first approximation, one obtains (Ref. 4) 10 c, m/s 6 δy y = 0.2 4 δy y =0 2 0 δy 1/2 c ≈ √gy (1 + ) y δy y = 0.4 8 0 2 4 6 y, m VIDEO V10.4 Sinusoidal waves 8 10 As indicated by the figure in the margin, the larger the amplitude, the faster the wave travels. A more general description of wave motion can be obtained by considering continuous (not solitary) waves of sinusoidal shape as is shown in Fig. 10.4. By combining waves of various wavelengths, λ, and amplitudes, δy, it is possible to describe very complex surface patterns found in nature, such as the wind-driven waves on a lake. Mathematically, such a process consists of using a Fourier series (each term of the series represented by a wave of different wavelength and amplitude) to represent an arbitrary function (the free-surface shape). A more advanced analysis of such sinusoidal surface waves of small amplitude shows that the wave speed varies with both the wavelength and fluid depth as (Ref. 1) c=[ gλ 2πy 12 tanh( 2π λ )] (10.4) where tanh(2πy/λ) is the hyperbolic tangent of the argument 2πy/λ. The result is plotted in Fig. 10.5. For conditions for which the water depth is much greater than the wavelength (y ⪢ λ, as in the ocean), the wave speed is independent of y and given by c= gλ A 2π Surface at time t δy = amplitude c λ = length y = mean depth ■ Figure 10.4 Sinusoidal surface wave. Surface at time t + δ t c δt 564 Chapter 10 ■ Open-Channel Flow c2 __ gy c = √gy Shallow layer 1.0 g c = __λ √ 2π y << λ Deep layer Eq. 10.4 y >> λ 0 c, m/s 150 y >> λ 100 50 0 0 2 4 6 8 10 λ , km __ λ 0 y ■ Figure 10.5 Wave speed as a function of wavelength. This result, shown in the figure in the margin, follows from Eq. 10.4, since tanh(2πy/λ) → 1 as yλ → ∞. Note that waves with very long wavelengths [e.g., waves created by a tsunami (“tidal wave”) with wavelengths on the order of several kilometers] travel very rapidly. On the other hand, if the fluid layer is shallow (y ⪢ λ, as often happens in open channels), the wave speed is given by c = (gy) 12, as derived for the solitary wave in Fig. 10.2. This result also follows from Eq. 10.4, since tanh(2πy/λ) → 2πy/λ as y/λ → 0. These two limiting cases are shown in Fig. 10.5. For moderate depth layers (y ∼ λ), the results are given by the complete Eq. 10.4. Note that for a given fluid depth, the long wave travels the fastest. Hence, for our purposes we will consider the wave speed to be this limiting situation, c = (gy) 12. THE WIDE WORLD OF FLUIDS Tsunami, the nonstorm wave A tsunami, often miscalled a waves approached shore and moved into shallower water, they “tidal wave,” is a wave produced by a disturbance (for example, slowed down considerably and reached heights up to 30m. an earthquake, volcanic eruption, or meteorite impact) that verti- Because the rate at which a wave loses its energy is inversely cally displaces the water column. Tsunamis are characterized as related to its wavelength, tsunamis, with their wavelengths on shallow-water waves, with long periods, very long wavelengths, the order of 100km, not only travel at high speeds, but they also and extremely large wave speeds. For example, the waves of the travel great distances with minimal energy loss. The furthest great December 2005, Indian Ocean tsunami traveled with reported death from the Indian Ocean tsunami occurred approx- speeds of 500–1000 m/s. Typically, these waves were of small imately 8000km from the epicenter of the earthquake that pro- amplitude in deep water far from land. Satellite radar measured duced it. A more recent example is the tsunami that struck the wave height as less than 1m in these areas. However, as the Japan in 2011. V–c 10.2.2 Froude Number Effects V>c Stationary V=c c–V V c the wave will be washed to the left with speed V − c. The above ideas can be expressed in dimensionless form by use of the Froude number, Fr = V(gy) 12, where we take the characteristic length to be the fluid depth, y. Thus, the Froude number, Fr = V/(gy) 12 = V/c, is the ratio of the fluid velocity to the wave speed. The following characteristics are observed when a wave is produced on the surface of a moving stream, as happens when a rock is thrown into a river. If the stream is not flowing, the wave spreads equally in all directions. If the stream is nearly stationary or moving in a tranquil manner (i.e., V < c), the wave can move upstream. Upstream locations are said to be in hydraulic communication with the downstream locations. That is, an observer upstream of a disturbance can tell that there has been a disturbance on the surface because that disturbance can propagate upstream to 10.2 Surface Waves 565 the observer. Viscous effects, which have been neglected in this discussion, will eventually damp out such waves far upstream. Such flow conditions, V < c, or Fr < 1, are termed subcritical. On the other hand, if the stream is moving rapidly so that the flow velocity is greater than the wave speed (i.e., V > c), no upstream communication with downstream locations is possible. Any disturbance on the surface downstream from the observer will be washed farther downstream. Such conditions, V > c or Fr > 1, are termed supercritical. For the special case of V = c or Fr = 1, the upstream propagating wave remains stationary and the flow is termed critical. E X AMPLE 10 .1 Surface Waves GIVEN At a certain location along the Rock River shown in Fig. E10.1a, the velocity, V, of the flow is a function of the depth, y, of the river as indicated in Fig. E10.1b. A reasonable approximation to these experimental results is V = 5 y2/3 (1) where V is in ft/s and y is in ft. FIND For what range of water depth will a surface wave on the river be able to travel upstream? 14 12 V 10 V, ft/s ■ Figure E10.1a SOLUTION While the river travels to the left with speed V, the surface wave travels upstream (to the right) with speed c = (g y) 1/2 relative to the water (not relative to the ground). Hence relative to the stationary ground, the wave travels to the right with speed c − V = (g y) 1/2 − 5 y2/3 = (32.2 ft/s2 y) 1/2 − 5 y2/3 (2) V = 5y2/3 8 6 4 Measured values 2 0 0 1 2 3 4 3 4 y, ft ■ Figure E10.1b For the wave to travel upstream, c − V > 0 so that from Eq. 2, (32.2 y) 1/2 > 5 y2/3 or y < 2.14 ft 1.2 (Ans) 1 COMMENT As shown above, if the river depth is less than Fr = V/c = (5 y2/3 )(g y) 1/2 = 5 y1/6(32.2 ft/s2 ) 1/2 = 0.881 y1/6 This result is plotted in Fig. E10.1c. Note that in agreement with the above answer, for y < 2.14 the flow is subcritical; the wave can travel upstream. (2.14,1) 0.8 Fr 2.14 ft, its velocity is less than the wave speed and the wave can travel upstream against the current. This is consistent with the fact that if a wave is to travel upstream, the flow must be subcritical (i.e., Fr = V/c < 1). For this flow 0.6 0.4 0.2 0 0 1 2 y, ft ■ Figure E10.1c The character of an open-channel flow may depend strongly on whether the flow is subcritical or supercritical. The characteristics of the flow may be completely opposite for subcritical flow than for supercritical flow. For example, as is discussed in Section 10.3, a “bump” on the bottom of a river (such 566 VIDEO Chapter 10 ■ Open-Channel Flow V10.5 Bicycle through a puddle 10.3 as a submerged log) may cause the surface of the river to dip below the level it would have had if the log were not there, or it may cause the surface level to rise above its undisturbed level. Which situation will happen depends on the value of Fr. Similarly, for supercritical flows it is possible to produce steplike discontinuities in the fluid depth (called a hydraulic jump; see Section 10.6.1). For subcritical flows, however, changes in depth must be smooth and continuous. Certain open-channel flows, such as the broad-crested weir (Section 10.6.3), depend on the existence of critical flow conditions for their operation. As strange as it may seem, there exist many similarities between the open-channel flow of a liquid and the compressible flow of a gas. The governing dimensionless parameter in each case is the fluid velocity, V, divided by a wave speed, the surface wave speed for open-channel flow or sound wave speed for compressible flow. Many of the differences between subcritical (Fr < 1) and supercritical (Fr > 1) open-channel flows have analogs in subsonic (Ma < 1) and supersonic (Ma > 1) compressible gas flow, where Ma is the Mach number. Some of these similarities are discussed in this chapter and in Chapter 11. Energy Considerations 10.3.1 Energy Balance The slope of the bottom of most open channels is very small; the bottom is nearly horizontal. A typical segment of an open-channel flow is shown in Fig. 10.6. The slope of the channel bottom (or bottom slope), S0 = (z1 − z2 )ℓ, is assumed constant over the segment shown. The fluid depths and velocities are y1 , y2 , V1 , and V2 as indicated. Note that the fluid depth is measured in the vertical direction and the distance x is horizontal. For most open-channel flows the value of S0 is very small (the bottom is nearly horizontal). For example, the Mississippi River drops a distance of 1470 ft in its 2350-mi length to give an average value of S0 = 0.000118. In such circumstances the values of x and y are often taken as the distance along the channel bottom and the depth normal to the bottom, with negligibly small differences introduced by the two coordinate schemes. With the assumption of a uniform velocity profile across any section of the channel, the onedimensional energy equation for this flow (Eq. 5.84) becomes p1 p2 V 21 V 22 + z1 = + z2 + hL + + γ γ 2g 2g (10.5) where hL is the head loss due to viscous effects between sections (1) and (2) and z1 − z2 = S0ℓ. Since the pressure is essentially hydrostatic at any cross section, we find that p1 γ = y1 and p2 γ = y2 so that Eq. 10.5 becomes y1 + V 21 V 22 + S0ℓ = y2 + + hL 2g 2g (10.6) One of the difficulties of analyzing open-channel flow, similar to that discussed in Chapter 8 for pipe flow, is associated with the determination of the head loss in terms of other physical parameters. Without getting into such details at present, we write the head loss in terms of the slope of the energy line, Sf = hL/ℓ (often termed the friction slope), as indicated in Fig. 10.6. Recall from 1 Sf 2 V1 ___ 2g y1 S0 hL Energy lin e V1 y2 (1) (2) 1 Slope = Sf 2 V2 ___ 2g V2 Slope = S0 𝓵 z1 z2 Horizontal datum ■ Figure 10.6 Typical open-channel geometry. x 10.3 Energy Considerations 567 Chapter 3 that the energy line is located at a distance z (the elevation from some datum to the channel bottom) plus the pressure head (p/γ) plus the velocity head (V 2/2g) above the datum. Therefore, Eq. 10.6 can be written as y1 − y2 = (V 22 − V 21 ) + (Sf − S0 )ℓ 2g (10.7) If there is no head loss, the energy line is horizontal (Sf = 0), and the total energy of the flow is free to shift between kinetic energy and potential energy in a conservative fashion. In the specific instance of a horizontal channel bottom (S0 = 0) and negligible head loss (Sf = 0), Eq. 10.7 simply becomes y1 − y2 = (V22 − V21 ) 2g 10.3.2 Specific Energy The specific energy is the sum of potential energy and kinetic energy (per unit weight). The specific energy or specific head, E, defined as E=y+ V2 2g (10.8) is often useful in open-channel flow considerations. The energy equation, Eq. 10.7, can be written in terms of E as E1 = E2 + (Sf − S0 )ℓ (10.9) If head losses are negligible, then Sf = 0 so that (Sf − S0 )ℓ = −S0ℓ = z2 − z1 and the sum of the specific energy and the elevation of the channel bottom remains constant (i.e., E1 + z1 = E2 + z2, a statement of the Bernoulli equation). If we consider a simple channel whose cross-sectional shape is a rectangle of width b, the specific energy can be written in terms of the flowrate per unit width, q = Qb = Vybb = Vy, as E=y+ q2 2gy2 (10.10) For a given channel of constant width, the value of q remains constant along the channel, although the depth, y, may vary. To gain insight into the flow processes involved, we consider the specific energy diagram, a graph of y and E, with q fixed, as shown in Fig. 10.7. The relationship between the flow depth, y, and the velocity head, V 2/2g, as given by Eq. 10.8, is indicated in the figure. y V2 2g y=E ysub y q = q" > q yc ysup q = q' Emin yneg E ■ Figure 10.7 Specific energy diagram. 568 Chapter 10 ■ Open-Channel Flow For a given value of specific energy, a flow may have two alternate depths. For given q and E, Eq. 10.10 can be rearranged into a cubic equation [y3 − Ey2 + (q2/2g) = 0] with three solutions, ysup, ysub, and yneg. If the specific energy is large enough (i.e., E > Emin, where Emin is a function of q), two of the solutions are positive and the other, yneg, is negative. The negative root, represented by the curved dashed line in Fig. 10.7, has no physical meaning and can be ignored. Thus, for a given flowrate and specific energy there are two possible depths, unless the vertical line from the E axis does not intersect the specific energy curve corresponding to the value of q given (i.e., E < Emin ). These two depths are termed alternate depths. For large values of E the upper and lower branches of the specific energy diagram (ysub and ysup ) approach y = E and y = 0, respectively. These limits correspond to a very deep channel flowing very slowly (E = y + V2/2g → y as y → ∞ with q = Vy fixed), or a very highspeed flow in a shallow channel (E = y + V 2/2g → V 2/2g as y → 0). As is indicated in Fig. 10.7, ysup < ysub. Thus, since q = Vy is constant along the curve, it follows that Vsup > Vsub, where the subscripts “sub” and “sup” on the velocities correspond to the depths so labeled. The specific energy diagram consists of two portions divided by the Emin “nose” of the curve. We will show that the flow conditions at this location correspond to critical conditions (Fr = 1), those on the upper portion of the curve correspond to subcritical conditions (hence, the “sub” subscript), and those on the lower portion of the curve correspond to supercritical conditions (hence, the “sup” subscript). To determine the value of Emin, we use Eq. 10.10 and set dE/dy = 0 to obtain q2 1/3 yc = (g) (10.11) where the subscript “c” denotes conditions at Emin. By substituting this back into Eq. 10.10, we obtain Emin = 3yc 2 By combining Eq. 10.11 and Vc = q/yc , we obtain Vc = y Fr < 1 Fr = 1 Fr > 1 E 1/2 q (y3/2 c g ) = = √gyc yc yc or Frc ≡ Vc /(gyc ) 1/2 = 1. Thus, critical conditions (Fr = 1) occur at the location of Emin. Since the fluid is deeper and the velocity smaller for the upper part of the specific energy diagram (compared with the conditions at Emin ), such flows are subcritical (Fr < 1). Conversely, flows for the lower part of the diagram are supercritical. This is shown by the figure in the margin. Thus, for a given flowrate, q, if E > Emin there are two possible depths of flow, one subcritical and the other supercritical. It is often possible to determine various characteristics of a flow by considering the specific energy diagram. Example 10.2 illustrates this for a situation in which the channel bottom elevation is not constant. EXAMPLE 10 .2 Specific Energy Diagram—Quantitative GIVEN Water flows up a 0.5-ft-tall ramp in a constant width, rectangular channel at a rate q = 5.75 ft2/s as is shown in Fig. E10.2a. (For now disregard the “bump.”) The upstream depth is 2.3 ft, and viscous effects are negligible. FIND Determine the elevation of the water surface downstream of the ramp, y2 + z2. 10.3 Energy Considerations 569 Free surface with ramp SOLUTION V1 = With S0ℓ = z1 − z2 and hL = 0, conservation of energy (Eq. 10.6 which, under these conditions, is actually the Bernoulli equation) requires that V 21 V 22 y1 + + z1 = y2 + + z2 2g 2g 2.5 ft/s V 22 64.4 0.89 ft (a) 4 3 0.5 y1 = 2.30 y, ft (1) 2 yc = 1.01 (2) Equations 1 and 2 can be combined to give − z2 = 0.5 ft (1) or 1.90y22 Free surface with bump Ramp y2V2 = y1V1 y32 y2 0.5 ft z1 = 0 where V2 and y2 are in ft/s and ft, respectively. The continuity equation provides the second equation y2V2 = 5.75 ft2/s Bump (c) 2.3 ft For the conditions given (z1 = 0, z2 = 0.5 ft, y1 = 2.3 ft, and V1 = qy1 = 2.5 ft/s), this becomes 1.90 = y2 + y2 y1 = V2 (2) (c) 1 + 0.513 = 0 which has solutions y2 = 1.72 ft,  y2 = 0.638 ft, or y2 = −0.466 ft Note that two of these solutions are physically realistic, but the negative solution is meaningless. This is consistent with the previous discussions concerning the specific energy (recall the three roots indicated in Fig. 10.7). The corresponding elevations of the free surface are either q= 5.75 ft2/s (2') 0 0 1 2 Emin = 1.51 E1 = 2.40 E2 = 1.90 3 4 E, ft (b) ■ Figure E10.2 y2 + z2 = 1.72 ft + 0.50 ft = 2.22 ft or y2 + z2 = 0.638 ft + 0.50 ft = 1.14 ft Which of these two flows is to be expected? This question can be answered by use of the specific energy diagram obtained from Eq. 10.10, which for this problem is E=y+ 0.513 y2 where E and y are in feet. The diagram is shown in Fig. E10.2b. The upstream condition corresponds to subcritical flow; the downstream condition is either subcritical or supercritical, corresponding to points 2 or 2′. Note that since E1 = E2 + (z2 − z1 ) = E2 + 0.5 ft, it follows that the downstream conditions are located 0.5 ft to the left of the upstream conditions on the diagram. With a constant-width channel, the value of q remains the same for any location along the channel. That is, all points for the flow from (1) to (2) or (2′ ) must lie along the q = 5.75 ft2/s curve shown. Any deviation from this curve would imply either a change in q or a relaxation of the one-dimensional flow assumption. To stay on the curve and go from (1) around the critical point (point c) to point (2′ ) would require a reduction in specific energy to Emin. As is seen from Fig. E10.2a, this would require a specified elevation (bump) in the channel bottom so that critical conditions would occur above this bump. The height of this bump can be obtained from the energy equation (Eq. 10.9) written between points (1) and (c) with Sf = 0 (no viscous effects) and S0ℓ = z1 − zc . That is, E1 = Emin − z1 + zc . In particular, since E1 = y1 + 0.513y21 = 2.40 ft and Emin = 3yc2 = 3(q2g) 132 = 1.51 ft, the top of this bump would need to be zc − z1 = E1 − Emin = 2.40ft − 1.51 ft = 0.89 ft above the channel bottom at section (1). The flow could then accelerate to supercritical conditions (Fr2′ > 1) as is shown by the free surface represented by the dashed line in Fig. E10.2a. Since the actual elevation change (a ramp) shown in Fig. E10.2a does not contain a bump, the downstream conditions will correspond to the subcritical flow denoted by (2), not the supercritical condition (2′ ). Without a bump on the channel bottom, the state (2′ ) is inaccessible from the upstream condition state (1). Such considerations are often termed the accessibility of flow regimes. Thus, the surface elevation is y2 + z2 = 2.22 ft (Ans) Note that since y1 + z1 = 2.30 ft and y2 + z2 = 2.22 ft, the elevation of the free surface decreases as it goes across the ramp. 570 Chapter 10 ■ Open-Channel Flow COMMENT If the flow conditions upstream of the ramp were supercritical, the free-surface elevation and fluid depth would increase as the fluid flows up the ramp. This is indicated in Fig. E10.2c along with the corresponding specific energy diagram, as is shown in Fig. E10.2d. For this case the flow starts at (1) on the lower (supercritical) branch of the specific energy curve and ends at (2) on the same branch with y2 > y1. Since both y and z increase from (1) to (2), the surface elevation, y + z, also increases. Thus, flow up a ramp is different for subcritical than it is for supercritical conditions. y y2 > y1 V1 > c1 y1 V2 > c2 (2) y2 y1 0.5 ft (1) 0.5 ft (c) E (d ) ■ Figure E10.2 (Continued) 10.4 VIDEO Uniform Flow V10.6 Merging channels Many channels are designed to carry fluid at a uniform depth and velocity all along their length. Irrigation canals are frequently of uniform depth and cross section for considerable lengths. Natural channels such as rivers and creeks are seldom of uniform shape, although a reasonable approximation to the flowrate in such channels can often be obtained by assuming uniform flow. Uniform flow is the open channel analog to fully developed pipe flow. In this section we will discuss various aspects of such flows. Uniform flow (dy/dx = 0) can be accomplished by adjusting the bottom slope, S0, so that it precisely equals the slope of the energy line, Sf . That is, S0 = Sf . From an energy point of view, uniform flow is achieved by a balance between the potential energy lost by the fluid as it coasts downhill and the energy that is dissipated by viscous effects (head loss) associated with shear stresses throughout the fluid. Similar conclusions can be reached from a force balance analysis as discussed in the following section. 10.4.1 Uniform Flow Approximations We consider fluid flowing in an open channel of constant cross-sectional size and shape such that the depth remains constant as is indicated in Fig. 10.8. The area of the section is A, and the wetted perimeter (i.e., the length of the perimeter of the cross section in contact with the fluid) is P. The interaction between the fluid and the atmosphere at the free surface is assumed negligible so that this portion of the perimeter is not included in the definition of the wetted perimeter. Free surface a A = flow area Q P = wetted perimeter a Section a – a (a) (b) ■ Figure 10.8 Uniform flow in an open channel. 10.4 Lines of constant velocity y Centerline velocity profiles Uniform Flow 571 y u u (a) Actual Uniform τ w = shear stress distribution τw ■ Figure 10.9 Typical velocity and shear stress distributions in an open channel: (a) velocity distribution throughout the cross section, (b) shear stress distribution on the wetted perimeter. (b) The wall shear stress acts on the wetted perimeter of the channel. Since the fluid must adhere to the solid surfaces, the actual velocity distribution in an open channel is not uniform. Some typical velocity profiles measured in channels of various shapes are indicated in Fig. 10.9a. The maximum velocity is often found somewhat below the free surface, and the fluid velocity is zero on the wetted perimeter, where a wall shear stress, τw, is developed. This shear stress is seldom uniform along the wetted perimeter, with typical variations as are indicated in Fig. 10.9b. Fortunately, reasonable analytical results can be obtained by assuming a uniform velocity profile, V, and a constant wall shear stress, τw. Similar assumptions were made for pipe flow situations (Chapter 8), with the friction factor being used to obtain the head loss. 10.4.2 The Chezy and Manning Equations The basic equations used to determine the uniform flowrate in open channels were derived many years ago. Continual refinements have taken place to obtain better values of the empirical coefficients involved. The result is a semiempirical equation that provides reasonable engineering results. A more refined analysis is perhaps not warranted because of the complexity and uncertainty of the flow geometry (i.e., channel shape and the irregular makeup of the wetted perimeter, particularly for natural channels). Under the assumptions of steady uniform flow, the x component of the momentum equation (Eq. 5.22) applied to the control volume indicated in Fig. 10.10 simply reduces to ∑Fx = ρQ(V2 − V1 ) = 0 𝓵 Control surface τ =0 V1 F1 y1 = y2 θ (1) τ w P𝓵 V2 = V1 y2 F2 θ 𝒲 = γ A𝓵 (2) x ■ Figure 10.10 Control volume for 𝒲 sin θ uniform flow in an open channel. 572 Chapter 10 ■ Open-Channel Flow For steady, uniform depth flow in an open channel there is no fluid acceleration. y1 y2 = y1 p(y) y since V1 = V2. There is no acceleration of the fluid, and the momentum flux across section (1) is the same as that across section (2). The flow is governed by a simple balance between the forces in the direction of the flow. Thus, ∑Fx = 0, or F1 − F2 − τwPℓ + 𝒲 sin θ = 0 where F1 and F2 are the hydrostatic pressure forces across either end of the control volume, as shown by the figure in the margin. Because the flow is at a uniform depth (y1 = y2 ), it follows that F1 = F2 so that these two forces do not contribute to the force balance. The term 𝒲sin θ is the component of the fluid weight that acts down the slope, and τw Pℓ is the shear force on the fluid, acting up the slope as a result of the interaction of the water and the channel’s wetted perimeter. Thus, Eq. 10.12 becomes f Wholly turbulent 𝒲 S0 𝒲 sin θ = Pℓ Pℓ τw = Equal pressure distributions For uniform depth, channel flow is governed by a balance between friction and weight. where we have used the approximation that sin θ ≈ tan θ = S0, since the bottom slope is typically very small (i.e., S0 ⪡ 1). Since 𝒲 = γAℓ, the force balance equation becomes τw = γAℓS0 = γRhS0 Pℓ Re (10.13) where the hydraulic radius is defined as Rh = A/P. Note that the hydraulic radius is one-fourth the hydraulic diameter. Most open-channel flows are turbulent. In fact, typical Reynolds numbers are quite large, well above the transitional value and into the wholly turbulent regime. As was discussed in Chapter 8, and shown by the figure in the margin, for very large Reynolds number pipe flows (wholly turbulent flows), the friction factor, f, is found to be independent of Reynolds number, and dependent only on the relative roughness, ε/D, of the pipe surface. For such cases, the wall shear stress is proportional to the dynamic pressure, ρV2/2, and independent of the viscosity. That is, τw = Kρ ε f = f ( D) (10.12) V2 2 where K is a constant dependent upon the roughness of the pipe. It is not unreasonable that similar shear stress dependencies occur for the large Reynolds number open-channel flows. In such situations, Eq. 10.13 becomes Kρ V2 = γRhS0 2 or V = C √RhS0 (10.14) where the constant C is termed the Chezy coefficient and Eq. 10.14 is termed the Chezy equation. This equation, one of the oldest in fluid mechanics, was developed in 1768 by A. Chezy (1718– 1798), a French engineer who designed a canal for the Paris water supply. The value of the Chezy coefficient, which must be determined by experiments, is not dimensionless but has the dimensions of (length) 1/2 per time (i.e., the square root of the units of acceleration). From a series of experiments it was found that the slope dependence of Eq. 10.14 (V ∼ S1/2 0 ) is reasonable but that the dependence on the hydraulic radius is more nearly V ∼ R2/3 h rather than V ∼ R1/2 h . In 1889, R. Manning (1816–1897), an Irish engineer, developed the following somewhat modified equation for open-channel flow to more accurately describe the Rh dependence: V= 1/2 R2/3 h S0 n (10.15) Equation 10.15 is termed the Manning equation, and the parameter n is the Manning resistance coefficient (usually shortened to “Manning’s n”). Its value is dependent on the surface material of 10.4 Uniform Flow 573 ■ Table 10.1 Values of the Manning Coefficient, n (Ref. 6) Wetted Perimeter The Manning equation is used to obtain the velocity or flowrate in an open channel. n A. Natural channels Clean and straight Sluggish with deep pools Major rivers 0.030 0.040 0.035 B. Floodplains Pasture, farmland Light brush Heavy brush Trees 0.035 0.050 0.075 0.15 C. Excavated earth channels Clean Gravelly Weedy Stony, cobbles 0.022 0.025 0.030 0.035 Wetted Perimeter D. Artificially lined channels Glass Brass Steel, smooth Steel, painted Steel, riveted Cast iron Concrete, finished Concrete, unfinished Planed wood Clay tile Brickwork Asphalt Corrugated metal Rubble masonry n 0.010 0.011 0.012 0.014 0.015 0.013 0.012 0.014 0.012 0.014 0.015 0.016 0.022 0.025 the channel’s wetted perimeter and is obtained from experiments. It is not dimensionless, having the units of s/m1/3 or s/ft1/3. As is discussed in Chapter 7, any correlation should be expressed in dimensionless form, with the coefficients that appear being dimensionless coefficients, such as the friction factor for pipe flow or the drag coefficient for flow past objects. Thus, Eq. 10.15 should be expressed in dimensionless form. Unfortunately, the Manning equation is so widely used and has been used for so long that it will continue to be used in its dimensional form with a coefficient, n, that is not dimensionless. The values of n found in the literature (such as Table 10.1) were developed for SI units. Standard practice is to use the same value of n when using the BG system of units and to insert a conversion factor into the equation. Thus, the velocity in uniform flow in an open channel is obtained from the Manning equation written as V= κ 2/3 1/2 R S n h 0 (10.16) Q= κ 2/3 1/2 AR S n h 0 (10.17) 0 0.05 n Trees Heavy brush Pasture, farmland Q Light brush and 0.1 0.15 where κ = 1 if SI units are used, and κ = 1.49 if BG units are used. The value 1.49 is the cube root of the number of feet per meter: (3.281 ft/m) 1/3 = 1.49. Thus, by using Rh in meters, A in m2, and κ = 1, the average velocity is m/s and the flowrate m3/s. By using Rh in feet, A in ft2, and κ = 1.49, the average velocity is ft/s and the flowrate ft3/s. Typical values of the Manning coefficient are indicated in Table 10.1. As expected, the rougher the wetted perimeter, the larger the value of n. For example, the roughness of floodplain surfaces increases from pasture to brush to tree conditions. So does the corresponding value of the Manning coefficient. Thus, for a given depth of flooding, the flowrate varies with floodplain roughness as indicated by the figure in the margin. Precise values of n are often difficult to obtain. Except for artificially lined channel surfaces like those found in new canals or flumes, the channel surface structure may be quite complex and variable. There are various methods used to obtain a reasonable estimate of the value of n for a 574 VIDEO Chapter 10 ■ Open-Channel Flow V10.7 Uniform channel flow given situation (Ref. 5). For the purpose of this book, the values from Table 10.1 are sufficient. Note that the error in Q is directly proportional to the error in n. A 10% error in the value of n produces a 10% error in the flowrate. Considerable effort has been put forth to obtain the best estimate of n, with extensive tables of values covering a wide variety of surfaces (Ref. 7). It should be noted that the values of n given in Table 10.1 are valid only for water as the flowing fluid. Both the friction factor for pipe flow and the Manning coefficient for channel flow are parameters that relate the wall shear stress to the makeup of the bounding surface. Thus, various results are available that describe n in terms of the equivalent pipe friction factor, f, and the surface roughness, ε (Ref. 8). For our purposes we will use the values of n from Table 10.1. 10.4.3 Uniform Flow Examples A variety of interesting and useful results can be obtained from the Manning equation. The following examples illustrate some of the typical considerations. The main parameters involved in uniform flow are the size and shape of the channel cross section (A, Rh ), the slope of the channel bottom (S0 ), the character of the material lining the channel bottom and walls (n), and the average velocity or flowrate (V or Q). Although the Manning equation is a rather simple equation, the ease of using it depends in part on which variables are known and which are to be determined. Determination of the flowrate of a given channel with flow at a given depth (often termed the normal flowrate for normal depth, sometimes denoted yn ) is obtained from a straightforward calculation as is shown in Example 10.3. EXAMPLE 10 .3 Uniform Flow, Determine Flowrate GIVEN Water flows in the canal of trapezoidal cross section shown in Fig. E10.3a. The bottom drops 1.4 ft per 1000 ft of length. The canal is lined with new finished concrete. y = 5 ft 40° 40° FIND Determine (a) 12 ft the flowrate and (b) the Froude number for this flow. ■ Figure E10.3a Assume uniform flow SOLUTION (a) From Table 10.1, we obtain n = 0.012 for the finished concrete. Thus, From Eq. 10.17, Q= 1.49 2/3 1/2 ARh S0 n (1) where we have used κ = 1.49, since the dimensions are given in BG units. For a depth of y = 5 ft, the flow area is A = 12 ft (5 ft) + 5 ft ( 5 ft = 89.8 ft2 tan 40° ) so that with a wetted perimeter of P =12 ft + 2(5/sin 40° ft) = 27.6 ft, the hydraulic radius is determined to be Rh = A/P = 3.25 ft. Note that even though the channel is quite wide (the freesurface width is 23.9 ft), the hydraulic radius is only 3.25 ft, which is less than the depth. Thus, with S0 = 1.4 ft/1000 ft = 0.0014, Eq. 1 becomes Q= 10.98 1.49 (89.8 ft2 ) (3.25 ft) 2/3 (0.0014) 1/2 = n n where Q is in ft3/s. Q= 10.98 = 915 cfs 0.012 (Ans) (b) The Froude number based on the maximum depth for the flow can be determined from Fr = V/(gy) 1/2. For the finished concrete case, Fr = 10.2 ft/s = 0.804 (32.2 ft/s2 × 5 ft) 1/2 (Ans) The flow is subcritical. COMMENTS The corresponding average velocity, V = QA, is 10.2 fts. It does not take a very steep slope [S0 = 0.0014 or θ = tan−1 (0.0014) = 0.080°) for this velocity. By repeating the calculations for various surface types (i.e., various Manning coefficient values), the results shown in Fig. E10.3b are obtained. Note that the increased roughness causes 10.4 1200 1000 Q, cfs a decrease in the flowrate. This is an indication that for the turbulent flows involved, the wall shear stress increases with surface roughness. [For water at 50 °F, the Reynolds number based on the 3.25-ft hydraulic radius of the channel and a smooth concrete surface is Re = RhV/ν = 3.25 ft (10.2 ft/s)/ (1.41 × 10−5 ft2/s) = 2.35 × 106, well into the turbulent regime.] The same results would be obtained for the channel if its size were given in meters. We would use the same value of n but set κ = 1 for this SI units situation. Finished concrete 800 Brickwork Asphalt 600 400 Rubble masonry 200 0 0 0.005 0.01 0.015 0.02 n 0.025 575 Uniform Flow 0.03 ■ Figure E10.3b In some instances a trial-and-error or iteration method must be used to solve for the dependent variable. This is often encountered when the flowrate, channel slope, and channel material are known, and the flow depth is to be determined as illustrated in the following examples. E X AMPLE 10 .4 Uniform Flow, Determine Normal Depth GIVEN Water flows in the channel shown in Fig. E10.3a at a rate of Q = 10.0 m3/s. The canal lining is weedy. FIND Determine the depth of the flow. SOLUTION In this instance neither the flow area nor the hydraulic radius is known, although they can be written in terms of the depth, y. Since the flowrate is given in m3/s, we will solve this problem using SI units. Hence, the bottom width is (12 ft) (1 m/3.281 ft) = 3.66 m and the area is and-error methods. The only physically meaningful root of Eq. 1 (i.e., a positive, real number) gives the solution for the normal flow depth at this flowrate as y A = y( + 3.66y = 1.19y2 + 3.66y tan 40° ) COMMENT By repeating the calculations for various flowrates, the results shown in Fig. E10.4 are obtained. Note that the water depth is not linearly related to the flowrate. That is, if the flowrate is doubled, the depth is not doubled. where A and y are in square meters and meters, respectively. Also, the wetted perimeter is y = 1.50 m y P = 3.66 + 2 ( sin 40° ) = 3.11y + 3.66 m (Ans) 3.0 so that 2.5 1.19y2 + 3.66y A Rh = = m P 3.11y + 3.66 2.0 Q = 10 = = y, m where Rh and y are in meters. Thus, with n = 0.030 (from Table 10.1), Eq. 10.17 can be written as κ 1/2 3 AR2/3 h S0 m /s n 1.19y2 + 3.66y 2/3 1.0 (1.19y2 + 3.66y)( 0.030 3.11y + 3.66 ) 1/2 × (0.0014) 1.5 (10, 1.50) 1.0 0.5 which can be rearranged into the form (1.19y2 + 3.66y) 5 − 515(3.11y + 3.66) 2 = 0 (1) where y is in meters. The solution of Eq. 1 can be easily obtained by use of a simple rootfinding numerical technique or by trial- 0 0 5 10 15 Q, m3/s ■ Figure E10.4 20 25 30 576 Chapter 10 ■ Open-Channel Flow In Example 10.4 we found the uniform depth for a given flowrate. Since the equation for this depth is a nonlinear equation (usually a polynomial), it may be that there is more than one solution to the problem. For a given channel there may be two or more depths that carry the same flowrate. Although this is not normally so, it can and does happen, as is illustrated by Example 10.5. EXAMPLE 10 .5 Uniform Flow, Maximum Flowrate GIVEN Water flows in a round pipe of diameter D at a depth of 0 ≤ y ≤ D, as is shown in Fig. E10.5a. The pipe is laid on a constant slope of S0, and the Manning coefficient is n. FIND (a) At what depth does the maximum flowrate occur? D (b) Show that for certain flowrates there are two depths possible with the same flowrate. Explain this behavior. y θ SOLUTION (a) (a) According to the Manning equation (Eq. 10.17), the flowrate is Q= κ 23 1/2 AR S n h 0 1.0 (1) where S0, n, and κ are constants for this problem. From geometry it can be shown that A= D2 (θ − sin θ) 8 Q _____ Qmax Qmax 0.5 Q full = 0.929 Q max 0 0 0.5 y __ 1.0 y = 0.938D D where θ, the angle indicated in Fig. E10.5a, is in radians. Similarly, the wetted perimeter is P= Dθ 2 so that the hydraulic radius is Rh = D(θ − sin θ) A = P 4θ Therefore, Eq. 1 becomes Q= κ 12 D8/3 (θ − sin θ) 5/3 S ] n 0 8(4) 2/3 [ θ2/3 This can be written in terms of the flow depth by using y = (D/2)[1 − cos(θ/2)]. A graph of flowrate versus flow depth, Q = Q(y), has the characteristic indicated in Fig. E10.5b. In particular, the maximum flowrate, Qmax, does not occur when the pipe is full; (b) ■ Figure E10.5 Qfull = 0.929Qmax. It occurs when y = 0.938D, or θ = 5.28 rad = 303°. Thus, Q = Qmax when y = 0.938D (Ans) (b) For any 0.929 < Q/Qmax < 1 there are two possible depths that give the same Q. The reason for this behavior can be seen by considering the gain in flow area, A, compared to the increase in wetted perimeter, P, for y ≈ D. The flow area increase for an increase in y is very slight in this region, whereas the increase in wetted perimeter, and hence the increase in shear force holding back the fluid, is relatively large. The net result is a decrease in flowrate as the depth increases. COMMENT For most practical problems, the slight difference between the maximum flowrate and full pipe flowrates is negligible, particularly in light of the usual inaccuracy of the value of n. Analysis of a circular conduit (pipe) running partially full is a somewhat unique problem. First, one must be careful to use the correct geometric formulas for A and P depending on whether the conduit is more than or less than half full. Second, it has been shown that Manning’s n is not constant but varies with depth (Ref. 10). 10.4 Uniform Flow 577 THE WIDE WORLD OF FLUIDS Done without GPS or lasers Two thousand years before the France. This aqueduct is part of a circuitous, 50-km-long open invention of such tools as the Global Positioning System (GPS) or channel that transported water to the Roman Colony at Nîmes laser surveying equipment, Roman engineers were able to from a spring located 20 km away. The spring is only 14.6 m design and construct structures that made a lasting contribu- above the point of delivery, giving an average bottom slope of tion to Western civilization. For example, one of the best surviv- only 3 × 10−4. It is obvious that to carry out such a project, the ing examples of Roman aqueduct construction is the Pont du Roman understanding of hydraulics, surveying, and construction Gard, an aqueduct that spans the Gardon River near Nîmes, was well advanced. (See Problem 10.50.) For many open channels, the surface roughness varies across the channel. In many constructed channels and in most natural channels, the surface roughness (and, hence, the Manning coefficient) varies along the wetted perimeter of the channel. A drainage ditch, for example, may have a rocky bottom surface with concrete side walls to prevent erosion. Thus, the effective n will be different for shallow depths than for deep depths. Similarly, a river channel may have one value of n appropriate for its normal channel and another very different value of n during its flood stage when a portion of the flow occurs across fields or through floodplain woods. An ice-covered channel usually has a different value of n for the ice than for the remainder of the wetted perimeter (Ref. 7). (Strictly speaking, such ice-covered channels are not “open” channels, although analysis of their flow is often based on open-channel flow equations. This is acceptable, since the ice cover is often thin enough so that it represents a fixed boundary in terms of the shear stress resistance, but it cannot support a significant pressure differential as in pipe flow situations.) A variety of methods have been used to determine an appropriate value of the effective roughness of channels that contain subsections with different values of n. Which method gives the most accurate, easy-to-use results is not firmly established, since the results are nearly the same for each method (Ref. 5). A reasonable approximation is to divide the channel cross section into N subsections, each with its own wetted perimeter, Pi, area, Ai, and Manning coefficient, ni. The Pi values do not include the imaginary boundaries between the different subsections. The total flowrate is assumed to be the sum of the flowrates through each section. This technique is illustrated by Example 10.6. E X AMPLE 10 .6 Uniform Flow, Variable Roughness GIVEN Water flows along the drainage canal having the proper- 3 ft ties shown in Fig. E10.6a. The bottom slope is S0 = 1 ft/500 ft = 0.002. 2 ft 3 ft n3 = 0.030 n1 = 0.020 (1) 0.6 ft y FIND Estimate the flowrate when the depth in the center channel 0.8 ft (2) is y = 0.8 ft + 0.6 ft = 1.4 ft. (3) n2 = 0.015 ■ Figure E10.6a SOLUTION We divide the cross section into three subsections as is indicated in Fig. E10.6a and write the flowrate as Q = Q1 + Q2 + Q3, where for each section 1.49 Qi = A R2/3 S1/2 ni i hi 0 The appropriate values of Ai, Pi, Rhi, and ni are listed in Table E10.6. Note that the imaginary portions of the perimeters between sections (denoted by the vertical dashed lines in Fig. E10.6a) are not included in the Pi. That is, for section (2) A2 = 2 ft (0.8 + 0.6) ft = 2.8 ft2 and P2 = 2 ft + 2(0.8 ft) = 3.6 ft ■ Table E10.6 i Ai (ft2) Pi (ft) Rhi (ft) ni 1 2 3 1.8 2.8 1.8 3.6 3.6 3.6 0.500 0.778 0.500 0.020 0.015 0.030 so that Rh2 = A2 2.8 ft2 = = 0.778 ft P2 3.6 ft 578 Chapter 10 ■ Open-Channel Flow 40 Thus, the total flowrate is Q = Q1 + Q2 + Q3 = 1.49(0.002) + 30 (1.8 ft2 ) (0.500 ft) 2/3 (2.8 ft2 ) (0.778 ft) 2/3 + 0.020 0.015 Q, ft3/s ×[ 1/2 (1.8 ft2 ) (0.500 ft) 2/3 ] 0.030 or (1.4 ft, 16.8 ft3/s) 10 Q = 16.8 ft3/s (Ans) 0 COMMENTS If the entire channel cross section were con- sidered as one flow area, then A = A1 + A2 + A3 = 6.4 ft and P = P1 + P2 + P3 = 10.8 ft, or Rh = A/P = 6.4 ft2/10.8 ft = 0.593 ft. The flow rate is given by Eq. 10.17, which can be written as 2 Q= 20 0 0.5 1 1.5 2 y, ft ■ Figure E10.6b 1.49 1/2 AR2/3 h S0 neff As expected, the effective roughness (Manning n) is between the minimum (n2 = 0.015) and maximum (n3 = 0.030) values for the individual subsections. By repeating the calculations for various depths, y, the results shown in Fig. E10.6b are obtained. Note that there are two distinct portions of the graph—one when the water is contained entirely within the main, center channel (y < 0.8 ft) and the other when the water overflows into the side portions of the channel (y > 0.8 ft). where neff is the effective value of n for this channel. With Q = 16.8 ft3/s as determined above, the value of neff is found to be 1/2 1.49AR2/3 h S0 Q 1.49(6.4) (0.593) 2/3 (0.002) 1/2 = = 0.0179 16.8 neff = One type of problem often encountered in open-channel flows is that of determining the best hydraulic cross section, defined as the section of the minimum area for a given flowrate, Q, slope, S0, and roughness coefficient, n. By using Rh = A/P we can write Eq. 10.17 as For a given flowrate, the channel of minimum area is denoted as the best hydraulic cross section. Q= κ A 2/3 κ A5/3S1/2 0 A ( ) S1/2 0 = n n P2/3 P which can be rearranged as A =( nQ 3/5 2/5 ) P κS1/2 0 where the quantity in the parentheses is a constant for a given application. Thus, a channel with minimum A is one with a minimum P, so that both the amount of excavation needed and the amount of material to line the surface are minimized by the best hydraulic cross section. The best hydraulic cross section possible is that of a semicircular channel. No other shape has as small a wetted perimeter for a given area. It is often desired to determine the best shape for a class of cross sections. The results for rectangular, trapezoidal (with 60° sides), and triangular shapes are shown in Fig. 10.11. For example, the best hydraulic cross section for a rectangle is one whose depth is half its width; for a triangle it is a 90° triangle. b y = b/2 b y= √ 3b/2 90° 60° b ■ Figure 10.11 Best hydraulic cross sections for a rectangle, a 60º trapezoid, and a triangle. 10.5 E X AMPLE 10 .7 Gradually Varied Flow 579 Uniform Flow—Best Hydraulic Cross Section GIVEN Water flows uniformly in a rectangular channel of width b and depth y. FIND Determine the aspect ratio, b/y, for the best hydraulic cross section. SOLUTION The uniform flow is given by Eq. 10.17 as Q= κ 1/2 AR 2/3 h S0 n That is, the rectangle with the best hydraulic cross section is twice as wide as it is deep, or (1) where A = by and P = b + 2y, so that Rh = A/P = by/(b + 2y). We rewrite the hydraulic radius in terms of A as Ay A A Rh = = = (2y + b) (2y + A/y) (2y2 + A) so that Eq. 1 becomes Q= 2/3 Ay κ S1/2 A( 2 0 ) n 2y + A This can be rearranged to give A5/2y = K(2y2 + A) (2) where K = (nQ/kS0 ) is a constant. The best hydraulic section is the one that gives the minimum A for all y. That is, dA/dy = 0. By differentiating Eq. 2 with respect to y, we obtain 1/2 3/2 5 3/2 dA dA A y + A5/2 = K (4y + 2 dy dy ) which, with dA/dy = 0, reduces to A5/2 = 4Ky With K = A the form 5/2 b/y = 2 COMMENTS A rectangular channel with b/y = 2 will give the smallest area (and smallest wetted perimeter) for a given flowrate. Conversely, for a given area, the largest flowrate in a rectangular channel will occur when b/y=2. For A = by = constant, if y → 0 then b → ∞, and the flowrate is small because of the large wetted perimeter P = b + 2y → ∞. The maximum Q occurs when y = b/2. However, as seen in Fig. E10.7, the maximum represented by this optimal configuration is a rather weak one. For example, for aspect ratios between 1 and 4, the flowrate is within 96% of the maximum flowrate obtained with the same area and b/y = 2. An alternate but equivalent method to obtain the aforementioned answer is to use the fact that dRhdy = 0, which follows from Eq. 1 using dQdy = 0 (constant flowrate) and dAdy = 0 (best hydraulic cross section has minimum area). Differentiation of Rh = Ay(2y2 + A) with constant A gives b/y = 2 when dRh dy = 0. The best hydraulic cross section can be calculated for other shapes in a similar fashion. The results (given here without proof) for rectangular, trapezoidal (with 60° sides), and triangular shapes are shown in Fig. 10.11. (3) 1.00 y(2y + A) from Eq. 2, Eq. 3 can be written in 2 A5/2 = 4A5/2y2 (2y + A) 2 which simplifies to y = (A/2) . Thus, because A = by, the best hydraulic cross section for a rectangular shape has a width b and a depth 1/2 by 1/2 A 1/2 y =( = (2) 2) 0.95 Q _____ y Qmax 0.90 b A = by = constant 0.85 0 or 1 2 b __ 3 4 5 y 2y2 = by 10.5 (Ans) ■ Figure E10.7 Gradually Varied Flow In many situations the flow in an open channel is not uniform. This can occur for several reasons: The bottom slope is not constant, the cross-sectional shape and area vary in the flow direction, or there is some obstruction across a portion of the channel. Physically, the difference between the component of weight and the shear forces in the direction of flow produces a change in the fluid momentum that requires a change in velocity and, from continuity considerations, a change in depth. Free ebooks ==> www.Ebook777.com 580 Chapter 10 ■ Open-Channel Flow In this section we consider gradually varying flows. For such flows, dydx ⪡ 1 and it is reasonable to impose the one-dimensional velocity assumption. At any section the total head is H = V 2 2g + y + z and the energy equation (Eq. 10.5) becomes H1 = H2 + hL where hL is the head loss between sections (1) and (2). As shown in Fig. 10.6, the slope of the energy line is dH/dx = dhL /dx = Sf and the slope of the channel bottom is dz/dx = S0. Thus, since dy dH d V2 V dV dz = ( + y + z) = + + g dx dx 2g dx dx dx we obtain dy dhL V dV = + + S0 g dx dx dx or dy V dV + = Sf − S0 g dx dx (10.18) For a given flowrate per unit width, q, in a rectangular channel of constant width b, we have V = q/y or by differentiation q dy dV V dy =− 2 =− y dx dx y dx so that the kinetic energy term in Eq. 10.18 becomes dy V dV V2 dy =− = −Fr2 g dx gy dx dx (10.19) where Fr = V/(gy) 1/2 is the local Froude number of the flow. Substituting Eq. 10.19 into Eq. 10.18 and simplifying gives (Sf − S0 ) dy = dx (1 − Fr2 ) dy dx 0 Sf = So 0.5 Sf > So Sf < So 1.5 2 Fr (10.20) We see that the rate of change of fluid depth, dy/dx, depends on the local slope of the channel bottom, S0, the slope of the energy line, Sf , and the Froude number, Fr. As shown by the figure in the margin, the value of dy/dx can be either negative, zero, or positive, depending on the values of these three parameters. That is, the channel flow depth may be constant or it may increase or decrease in the flow direction, depending on the values of S0, Sf , and Fr. The behavior of subcritical flow may be the opposite of that for supercritical flow, as shown by the denominator, 1 − Fr2, of Eq. 10.20. Although in the derivation of Eq. 10.20 we assumed q is constant (i.e., a rectangular channel), Eq. 10.20 is valid for channels of any constant cross-sectional shape, provided the Froude number is interpreted properly (Ref. 3). In this book we will consider only rectangular cross-sectional channels when using this equation. In order to proceed further with Eq. 10.20, we need to relate Sf (= hL/ℓ) to other problem variables such as V, Rh, and roughness (n). An exact relationship would be very complicated because the flow is most likely turbulent and the shear stress is not uniformly distributed. In the absence of more accurate information, the following model is used: we assume that the energy loss at any location along the channel is the same as the energy loss that occurs in a uniform flow with the same values of the flow variables (i.e., V, Rh, n). The fluid velocity would then be given by the Manning equation (Eq. 10.15). The friction slope is found by noting that for uniform flow, the friction slope and the channel slope are equal, S0 = Sf , so we can write the “ Manning equation” as V= κ 2/3 1/2 R S n h f Sf = n2 V 2 ) κ 2( R 4/3 h Solving for the friction slope gives www.Ebook777.com (10.21) 10.6 Rapidly Varied Flow 581 If we substitute the equations for Sf, V, Rh, and F into Eq. 10.20, we get −4/3 by n2 Q2 − S0 dy κ2 ( b2y2 )( b + 2y ) = dx Q 2 1 −( gby ) (10.22) where Q (flowrate), g (acceleration of gravity), and κ (1 for SI, 1.49 for BG) are constants and b (channel width), n (roughness coefficient), and S0 (channel slope) are known functions of x (distance along the channel). A solution of this differential equation gives the fluid depth (or surface profile) as a function of position along the channel. It is usually not possible to integrate the differential equation analytically; however, a numerical solution is quite feasible. Using a computer, the channel can be divided into many small steps for accurate computation. A pencil-and-paper calculation using only a few reaches, called the unit step method, can be used to calculate the profile with acceptable accuracy. An interesting feature of either method is that for supercritical flow, the calculations begin at a known point and proceed downstream, while for subcritical flow, calculations begin at a point and proceed upstream. See Refs. 3, 5, and 6 for details. 10.6 Rapidly Varied Flow In many cases the flow depth may change significantly in a short distance. In many open channels, flow depth changes occur over a relatively short distance so that dy/dx ∼ 1. Such rapidly varied flow conditions are often quite complex and difficult to analyze precisely. Fortunately, many useful approximate results can be obtained by using a simple one-dimensional model along with appropriate experimentally determined coefficients when necessary. In this section we discuss several of these flows. Some rapidly varied flows occur in constant area channels for reasons that are not immediately obvious. The hydraulic jump is one such case. As is indicated in Fig. 10.12, the flow may change from a relatively shallow, high-speed condition into a relatively deep, low-speed condition within a horizontal distance of just a few flow depths. Other rapidly varied flows may be due to a sudden change in the channel geometry such as the flow in an expansion or contraction section of a channel as is indicated in Fig. 10.13. ■ Figure 10.12 Hydraulic jump. VIDEO V10.9 Bridge pier scouring w Flo ■ Figure 10.13 Rapidly varied flow may occur in a channel transition section. 582 Chapter 10 ■ Open-Channel Flow V10.10 Big Sioux River Bridge collapse VIDEO Upstream velocity profile Bridge pier Horseshoe vortex Scouring of channel bottom ■ Figure 10.14 The complex three-dimensional flow structure around a bridge pier. V10.8 Erosion in a channel VIDEO In such situations the flow field is often two- or three-dimensional. There may be regions of flow separation, flow reversal, or unsteady oscillations of the free surface. For the purpose of some analyses, these complexities can be neglected and a simplified analysis can be undertaken. In other cases, however, it is the complex details of the flow that are the most important property of the flow; any analysis must include their effects. The scouring of a river bottom in the neighborhood of a bridge pier, as is indicated in Fig. 10.14, is such an example. A one- or two-dimensional model of this flow would not be sufficient to describe the complex structure of the flow that is responsible for the erosion near the foot of the bridge pier. Many open-channel flow-measuring devices are based on principles associated with rapidly varied flows. Among these devices are broad-crested weirs, sharp-crested weirs, critical flow flumes, and sluice gates. The operation of such devices is discussed in the following sections. 10.6.1 The Hydraulic Jump Under certain conditions it is possible that the fluid depth will change very rapidly over a short length of the channel without any change in the channel configuration. Such changes in depth can be approximated as a discontinuity in the free-surface elevation (dy/dx → ∞). For reasons discussed below, this step change in depth is always from a shallow to a deeper depth—always a step up, never a step down. Physically, this near discontinuity, called a hydraulic jump, may result when there is a conflict between the upstream and downstream influences that control a particular section (or reach) of a channel. For example, a sluice gate may require that the conditions immediately downstream of the gate be supercritical flow, while obstructions in the channel on the downstream end of the reach may require that the flow be subcritical. The hydraulic jump provides the mechanism (a nearly discontinuous one) to make the transition between the two types of flow. The simplest type of hydraulic jump occurs in a horizontal, rectangular channel as is indicated in Fig. 10.15. Although the flow within the jump itself is extremely complex and agitated, it is reasonable to assume that the flow at sections (1) and (2) is nearly uniform, steady, and onedimensional. In addition, we neglect any force from wall shear stresses, τw, within the relatively short segment between these two sections. Under these conditions the x component of the momentum equation (Eq. 5.22) for the control volume indicated can be written as A hydraulic jump is a steplike increase in fluid depth in an open channel. p y1/2 c F1 − F2 = ρQ(V2 − V1 ) = ρV1y1b(V2 − V1 ) y1 where, as indicated by the figure in the margin, the pressure force at either section is hydrostatic. That is, F1 = pc1A1 = γy21b2 and F2 = pc2A2 = γy22b2, where pc1 = γy1 2 and pc2 = γy2 2 are the pressures at the centroids of the channel cross sections and b is the channel width (see Section 2.8). Thus, the momentum equation becomes y21 y22 V1y1 − = (V2 − V1 ) g 2 2 (10.23) 10.6 Energy line hL (2) Control volume Q (1) y1 V2 V1 583 Rapidly Varied Flow F2 y2 F1 x τw = 0 ■ Figure 10.15 Hydraulic jump geometry. In addition to the momentum equation, we have the conservation of mass equation (Eq. 5.12) y1bV1 = y2bV2 = Q VIDEO V10.11 Hydraulic jump in a river (10.24) and the energy equation (Eq. 5.84) y1 + V 21 V 22 = y2 + + hL 2g 2g (10.25) The head loss, hL, in Eq. 10.25 is due to the violent turbulent mixing and dissipation that occur within the jump itself. We have neglected any head loss due to wall shear stresses, which is minor compared to loss from turbulence. Clearly Eqs. 10.23, 10.24, and 10.25 have a solution y1 = y2, V1 = V2, and hL = 0. This represents the case of no jump. Since these are nonlinear equations, it may be possible that more than one solution exists. The other solutions can be obtained as follows. By combining Eqs. 10.23 and 10.24 to eliminate V2 we obtain y21 y22 V1y1 V1y1 V 21y1 − = − V = (y − y2 ) 1) g ( y2 gy2 1 2 2 which can be simplified by cancelling a common nonzero factor y1 − y2 from each side to give y2 2 y2 2 ( y1 ) + ( y1 ) − 2 Fr1 = 0 where Fr1 = V1  √gy1 is the upstream Froude number. By using the quadratic formula we obtain y2 1 = (−1 ± √1 + 8Fr21 ) y1 2 The depth ratio across a hydraulic jump depends on the Froude number only. Clearly the solution with the minus sign is not possible (it would give a negative y2 y1 ). Thus, y2 1 = (−1 + √1 + 8Fr21 ) y1 2 (10.26) This depth ratio, y2 /y1, across the hydraulic jump is shown as a function of the upstream Froude number in Fig. 10.16. The portion of the curve for Fr1 < 1 is dashed in recognition of the fact that to have a hydraulic jump the flow must be supercritical. That is, the solution as given by Eq. 10.26 must be restricted to Fr1 ≥ 1, for which y2 /y1 ≥ 1. This can be shown by consideration of the energy equation, Eq. 10.25, as follows. The dimensionless head loss, hL /y1, can be obtained from Eq. 10.25 as y2 y1 2 hL Fr21 =1− + 1 − ( y2 ) ] y1 y1 2 [ (10.27) 584 Chapter 10 ■ Open-Channel Flow 4 3 y2 __ y2 __ y1 y1 2 hL __ y1 hL __ No jump possible or y1 1 0 –1 (Photograph courtesy of U.S. Army Corps of Engineers.) Hydraulic jumps dissipate energy. 0 1 2 V1 Fr1 = ______ √gy1 3 4 ■ Figure 10.16 Depth ratio and dimensionless head loss across a hydraulic jump as a function of upstream Froude number. where, for given values of Fr1, the values of y2 y1 are obtained from Eq. 10.26. As is indicated in Fig. 10.16, the head loss is negative if Fr1 < 1. Since negative head losses violate the second law of thermodynamics (viscous effects dissipate energy, they cannot create energy; see Section 5.3), it is not possible to produce a hydraulic jump with Fr1 < 1. The head loss across the jump is indicated by the lowering of the energy line shown in Fig. 10.15. A flow must be supercritical (Froude number > 1) to produce a hydraulic jump. This is analogous to the compressible flow discussed in Chapter 11 in which it is shown that the flow of a gas must be supersonic (Mach number > 1) to produce the discontinuity called a shock wave. However, the fact that a flow is supercritical (or supersonic) does not guarantee the production of a hydraulic jump (or shock wave). The mathematically trivial solution y1 = y2 and V1 = V2 is also possible. The fact that there is an energy loss across a hydraulic jump is useful in many situations. For example, the relatively large amount of energy contained in the fluid flowing down the spillway of a dam like that shown in the figure in the margin could cause damage to the channel below the dam. By placing suitable flow control objects in the channel downstream of the spillway, it is possible (if the flow is supercritical) to produce a hydraulic jump on the apron of the spillway and thereby dissipate a considerable portion of the energy of the flow. That is, the dam spillway produces supercritical flow, and the channel downstream of the dam requires subcritical flow. The resulting hydraulic jump provides the means to change the character of the flow. THE WIDE WORLD OF FLUIDS Grand Canyon rapids building Virtually all of the rapids in the change over the years as debris is added to or removed from the Grand Canyon were formed by rock debris carried into the rapids. For example, Crystal Rapid, one of the notorious rafting Colorado River from side canyons. Severe storms wash large stretches of the river, changed very little between the first pho- amounts of sediment into the river, building debris fans that nar- tos of 1890 and those of 1966. However, a debris flow from a row the river. This debris forms crude dams that back up the river severe winter storm in 1966 greatly constricted the river. Within to form quiet pools above the rapids. Water exiting the pool a few minutes the configuration of Crystal Rapid was completely through the narrowed channel can reach supercritical conditions changed. The new, immature rapid was again drastically changed and produce hydraulic jumps downstream. Since the configura- by a flood in 1983. While Crystal Rapid is now considered full tion of the jumps is a function of the flowrate, the difficulty in grown, it will undoubtedly change again, perhaps in 100 or 1000 running the rapids can change from day to day. Also, rapids years. (See Problem 10.80.) 10.6 E X AMPLE 10 .8 585 Rapidly Varied Flow Hydraulic Jump GIVEN Water on the horizontal apron of the 100-ft-wide spillway shown in Fig. E10.8a has a depth of 0.60 ft and a velocity of 18 ft/s. FIND Determine the depth, y2, after the jump, the Froude numbers before and after the jump, Fr1 and Fr2, and the power dissipated, 𝒫d, within the jump. SOLUTION Conditions across the jump are determined by the upstream Froude number V1 18 ft/s = = 4.10 Fr1 = [(32.2 ft/s2 ) (0.60 ft)]1/2 √gy1 b = width = 100 ft Spillway apron (Ans) y2 V1 = 18 ft/s Thus, the upstream flow is supercritical, and it is possible to generate a hydraulic jump as sketched. From Eq. 10.26 we obtain the depth ratio across the jump as V2 y1 = 0.60 ft Downstream obstacles (a) 1000 y2 1 = (−1 + √1 + 8 Fr21 ) y1 2 1 = [−1 + √1 + 8(4.10) 2] = 5.32 2 800 or (Ans) Since Q1 = Q2, or V2 = (y1V1 )y2 = 0.60 ft (18 ft/s)3.19 ft = 3.39 ft/s, it follows that V2 𝒫d, hp 600 y2 = 5.32 (0.60 ft) = 3.19 ft 400 (0.60 ft, 277 hp) 3.39 ft/s Fr2 = = = 0.334 (Ans) 2 [(32.2 ft/s ) (3.19 ft)]1/2 √gy2 200 As is true for any hydraulic jump, the flow changes from supercritical to subcritical flow across the jump. The power (energy per unit time) dissipated, 𝒫d, by viscous effects within the jump can be determined from the head loss as (see Eq. 5.85) 0 𝒫d = γQhL = γby1V1hL (1.54 ft, 0 hp) 0 0.2 0.4 0.6 1 1.2 1.4 1.6 (b) 4 q = 10.8 ft2/s (1) (2) 3 y, ft where hL is obtained from Eq. 10.25 or 10.27 as V 21 V 22 (18.0 ft/s) 2 − y + = 0.60 ft + hL = (y1 + 2 2g ) ( 2g ) [ 2(32.2 ft/s2 ) ] 2 (3.39 ft/s) − [3.19 ft + 2(32.2 ft/s2 ) ] 0.8 y1, ft 2 (2') 1 (1) hL = 2.26 ft or 0 hL = 2.26 ft 0 1 Thus, from Eq. 1, 3 E2 = 3.37 4 5 6 E1 = 5.63 E, ft 𝒫d = (62.4 lb/ft ) (100 ft) (0.60 ft) (18.0 ft/s) (2.26 ft) = 1.52 × 105 ft ∙ lb/s (c) 3 or 𝒫d = 2 1.52 × 10 ft ∙ lb/s = 277 hp 550[(ft ∙ lb/s)/hp] 5 (Ans) COMMENTS This power, which is dissipated within the highly turbulent motion of the jump, is converted into an increase in water temperature, T. That is, T2 > T1. Although the power ■ Figure E10.8 dissipated is considerable, the difference in temperature is not great because the flowrate is quite large. By repeating the calculations for the given flowrate Q1 = A1V1 = b1y1V1 = 100 ft (0.6 ft) (18 ft/s) = 1080 ft3/s but with various upstream depths, y1, the results shown in Fig. E10.8b are obtained. Note that a slight change in water depth can produce a considerable change in energy dissipated. Also, if y1 > 1.54 ft the flow is subcritical (Fr1 < 1) and no hydraulic jump can occur. 586 Chapter 10 ■ Open-Channel Flow The hydraulic jump flow process can be illustrated by use of the specific energy concept introduced in Section 10.3. Equation 10.23 can be written in terms of the specific energy, E = y + V 22g, as E1 = E2 + hL, where E1 = y1 + V 212g = 5.63 ft and E2 = y2 + V 222g = 3.37 ft. As is discussed in Section 10.3, the specific energy diagram for this flow can be obtained by using V = q/y, where q = q1 = q2 = Q = y1V1 = 0.60 ft (18.0 ft/s) b = 10.8 ft2/s where y and E are in feet. The resulting specific energy diagram is shown in Fig. E10.8c. Because of the head loss across the jump, the upstream and downstream values of E are different. In going from state (1) to state (2), the fluid does not proceed along the specific energy curve and pass through the critical condition at state 2′. Rather, it jumps from (1) to (2), as is represented by the dashed line in the figure. From a one-dimensional consideration, the jump is a discontinuity. In actuality, the jump is a complex three-dimensional flow incapable of being represented as a discontinuous change on the one-dimensional specific energy diagram. Thus, E=y+ The actual structure of a hydraulic jump depends on the Froude number. q2 2 2gy =y+ (10.8 ft2/s) 2 2 2 2(32.2 ft/s )y =y+ 1.81 y2 The actual structure of a hydraulic jump is a complex function of Fr1, even though the depth ratio and head loss are given quite accurately by a simple one-dimensional flow analysis (Eqs. 10.26 and 10.27). A detailed investigation of the flow indicates that there are essentially five types of surface and jump conditions. The classification of these jumps is indicated in Table 10.2, along with sketches of the structure of the jump. For flows that are barely supercritical, the jump is more like a standing wave, without a nearly step change in depth. In some Froude number ranges the jump is unsteady, with regular periodic oscillations traveling downstream. (Recall that the wave cannot travel upstream against the supercritical flow.) The length of a hydraulic jump (the distance between the nearly uniform upstream and downstream flows) may be of importance in the design of channels. Although its value cannot be determined theoretically, experimental results indicate that over a wide range of Froude numbers, the jump is approximately seven downstream depths long (Ref. 5). ■ Table 10.2 Classification of Hydraulic Jumps (Ref. 12) Fr1 y2y1 <1 1 1 to 1.7 1 to 2.0 Classification Jump impossible Standing wave or undulant jump 1.7 to 2.5 2.0 to 3.1 Weak jump 2.5 to 4.5 3.1 to 5.9 Oscillating jump 4.5 to 9.0 5.9 to 12 Stable, well-balanced steady jump; insensitive to downstream conditions >9.0 >12 Rough, somewhat intermittent strong jump Sketch V1 y1 V2 = V1 y2 10.6 Rapidly Varied Flow 587 Q V1 V2 S0 > S0c S0 < S0c (a) Submerged jump Q Free jump ■ Figure 10.17 Hydraulic jump variations: (a) jump caused by a change in channel slope, (b) submerged jump. (b) VIDEO V10.12 Hydraulic jump in a sink Hydraulic jumps can occur in a variety of channel flow configurations, not just in horizontal, rectangular channels as discussed above. Jumps in nonrectangular channels (i.e., circular pipes, trapezoidal canals) behave in a manner quite like those in rectangular channels, although the details of the depth ratio and head loss are somewhat different from jumps in rectangular channels. Other common types of hydraulic jumps include those that occur in sloping channels, as is indicated in Fig. 10.17a and the submerged hydraulic jumps that can occur just downstream of a sluice gate, as is indicated in Fig. 10.17b. Details of these and other jumps can be found in standard open-channel flow references (Refs. 3 and 5). 10.6.2 Sharp-Crested Weirs A sharp-crested weir can be used to determine the flowrate. A weir is an obstruction on a channel bottom over which the fluid must flow. It provides a convenient method of determining the flowrate in an open channel by making of a single depth measurement. A sharp-crested weir is essentially a vertical sharp-edged flat plate placed across the channel in a way such that the fluid must flow across the sharp edge and drop into the pool downstream of the weir plate, as is shown in Fig. 10.18. The specific shape of the flow area in the plane of the weir plate is used to designate the type of weir. Typical shapes include the rectangular weir, the triangular weir, and the trapezoidal weir, as indicated in Fig. 10.19. The complex nature of the flow over a weir makes it impossible to obtain precise analytical expressions for the flow as a function of other parameters, such as the weir height, Pw, weir head, H, the fluid depth upstream, and the geometry of the weir plate (angle θ for triangular weirs or aspect ratio, b/H, for rectangular weirs). The flow structure is far from one-dimensional, with a variety of interesting flow phenomena. Draw down Nappe H Q Pw Weir plate ■ Figure 10.18 Sharp-crested weir geometry. 588 Chapter 10 ■ Open-Channel Flow Channel walls b θ (a) Weir plate (b) (c) ■ Figure 10.19 Sharp-crested weir plate geometry: (a) rectangular, (b) triangular, (c) trapezoidal. The main mechanisms governing flow over a weir are gravity and inertia. From a highly simplified point of view, gravity accelerates the fluid from its free-surface elevation upstream of the weir to larger velocity as it flows down the hill formed by the nappe. Although viscous and surface tension effects are usually of secondary importance, such effects cannot be entirely neglected. Generally, appropriate experimentally determined coefficients are used to account for these effects. As a first approximation, we assume that the velocity profile upstream of the weir plate is uniform and that the pressure within the nappe is atmospheric. In addition, we assume that the fluid flows horizontally over the weir plate with a nonuniform velocity profile, as indicated in Fig. 10.20. With pB = 0 the Bernoulli equation for flow along the arbitrary streamline A – B indicated can be written as pA u22 V21 + zA = (H + Pw − h) + + γ 2g 2g pA′/γ A′ VA′ where h is the distance that point B is below the free surface. We do not know the location of point A from which came the fluid that passes over the weir at point B. However, since the total head for any particle along the vertical section (1) is the same, zA + pAγ + V 21 2g = H + Pw + V 21 2g, the specific location of A (i.e., A or A′ shown in the figure in the margin) is not needed, and the velocity of the fluid over the weir plate is obtained from Eq. 10.28 as pA/γ u2 = A zA VA zA′ (10.28) V 21 2g(h + ) B 2g The flowrate can be calculated from Q= ∫ u2 dA = (2) z V 12 ___ h=H h=0 u2ℓ dh Energy line Free surface and hydraulic grade line 2g u2(h) h B A ∫ V1 h H dh B Pw zA x (a) ■ Figure 10.20 Assumed flow structure over a weir. (b) 𝓵 (10.29) 10.6 Rapidly Varied Flow 589 where ℓ(h) is the cross-channel width of a strip of the weir area, as is indicated in Fig. 10.20b. For a rectangular weir, ℓ is constant. For other weirs, such as triangular or circular weirs, the value of ℓ is known as a function of h. For a rectangular weir, ℓ = b, and the flowrate becomes Q = √2g b ∫ H 0 V 21 1/2 h + dh ( 2g ) or Q= V 21 3/2 V 21 3/2 2 √2g b [(H + ) − ( ) ] 3 2g 2g (10.30) Equation 10.30 is a rather cumbersome expression that can be simplified by using the fact that with Pw ⪢ H (as often happens in practical situations) the upstream velocity is negligibly small. That is, V 21 2g ⪡ H and Eq. 10.30 simplifies to the basic rectangular weir equation Q= 2 3 √2g b H3/2 (10.31) Note that the weir head, H, is the height of the upstream free surface above the crest of the weir. As is indicated in Fig. 10.18, because of the drawdown effect, H is not the distance of the free surface above the weir crest as measured directly above the weir plate. Because of the numerous approximations made to obtain Eq. 10.31, an experimentally determined correction factor must be used to obtain the actual flowrate as a function of weir head. Thus, the final form is A weir coefficient is used to account for nonideal conditions excluded in the simplified analysis. 2 Q = Cwr 3 √2g b H3/2 where Cwr is the rectangular weir coefficient. From dimensional analysis it is expected that Cwr is a function of Reynolds number (viscous effects), Weber number (surface tension effects), and H/Pw (geometry). In most practical situations, the Reynolds and Weber number effects are negligible, and the following correlation, shown in the figure in the margin, can be used (Refs. 4, 7): 1 Cwr 0 (10.32) 0 VIDEO H/Pw V10.13 Triangular weir 1 Cwr = 0.611 + 0.075( H Pw ) (10.33) More precise values of Cwr can be found in the literature, if needed (Refs. 3 and 15). The triangular sharp-crested weir is often used for flow measurements, particularly for measuring flowrates over a wide range of values. For small flowrates, the head, H, for a rectangular weir would be very small and the flowrate could not be measured accurately. However, with the triangular weir, the flow width decreases as H decreases so that even for small flowrates, reasonable heads are developed. Accurate results can be obtained over a wide range of Q. The triangular weir equation can be obtained from Eq. 10.29 by using θ ℓ = 2(H − h) tan ( ) 2 where θ is the angle of the V-notch (see Figs. 10.19 and 10.20). After carrying out the integration and again neglecting the upstream velocity (V 12 2g ⪡ H), we obtain Q= 8 θ tan ( ) √2g H52 15 2 An experimentally determined triangular weir coefficient, Cwt, is used to account for the real-world effects neglected in the analysis so that Q = Cwt 8 θ tan( ) √2g H5/2 15 2 (10.34) 590 Chapter 10 ■ Open-Channel Flow 0.66 θ = 20° 0.64 45° 0.62 60° Cwt 90° 0.60 0.58 Minimum Cwt for all θ 0.56 ■ Figure 10.21 Weir coefficient for 0 0.2 0.4 0.6 0.8 1.0 triangular sharp-crested weirs (Ref. 10, with permission from ASCE). H, ft VIDEO V10.14 Low-head dam Flowrate over a weir depends on whether the nappe is free or submerged. Typical values of Cwt for triangular weirs are in the range of 0.58 to 0.62, as is shown in Fig. 10.21. Note that although Cwt and θ are dimensionless, the value of Cwt is given as a function of the weir head, H, which is a dimensional quantity. Although using dimensional parameters is not recommended (see the dimensional analysis discussion in Chapter 7), tradition dictates that such parameters are often used for open-channel flow. The above results for sharp-crested weirs are valid provided the area under the nappe is ventilated to atmospheric pressure. Although this is not a problem for triangular weirs, for rectangular weirs it is sometimes necessary to provide ventilation tubes to ensure atmospheric pressure in this region. In addition, depending on downstream conditions, it is possible to obtain the submerged weir operation, as is indicated in Fig. 10.22. Clearly the flowrate will be different for these situations from that given by Eqs. 10.30 and 10.32. 10.6.3 Broad-Crested Weirs H/Lw = 0.08 H Lw H/Lw = 0.50 A broad-crested weir is a structure in an open channel that has a horizontal crest above which the fluid pressure may be considered hydrostatic. A typical configuration is shown in Fig. 10.23. Generally, to ensure proper operation, these weirs are restricted to the range 0.08 < H/Lw < 0.50. These conditions are drawn to scale in the figure in the margin. For long weir blocks (H/Lw less than 0.08), head losses across the weir cannot be neglected. On the other hand, for short weir blocks (H/Lw greater than 0.50), the streamlines of the flow over the weir block are not horizontal. Although broad-crested weirs can be used in channels of any cross-sectional shape, we restrict our attention to rectangular channels. The operation of a broad-crested weir is based on the fact that nearly uniform critical flow is achieved in the short reach above the weir block. (If H/Lw < 0.08, viscous effects are important, (a) (b) ■ Figure 10.22 Flow conditions over a weir without a free nappe: (a) plunging nappe, (b) submerged nappe. 10.6 Rapidly Varied Flow 591 (1) (2) y1 V1 H Pw V2 = Vc y2 = yc Weir block ■ Figure 10.23 Broad-crested weir Lw geometry. and the flow is subcritical over the weir.) If the kinetic energy of the upstream flow is negligible, then V21 2g ⪡ y1 and the upstream specific energy is E1 = V21 2g + y1 ≈ y1. Observations show that as the flow passes over the weir block, it accelerates and reaches critical conditions, y2 = yc and Fr2 = 1 (i.e., V2 = c2 ), corresponding to the nose of the specific energy curve (see Fig. 10.7). The flow does not accelerate to supercritical conditions. To do so would require the ability of the downstream fluid to communicate with the upstream fluid to let it know that there is an end of the weir block. Since waves cannot propagate upstream against a critical flow, this information cannot be transmitted. The flow remains critical, not supercritical, across the weir block. The Bernoulli equation can be applied between point (1) upstream of the weir and point (2) over the weir where the flow is critical to obtain H + Pw + V 2c V 21 = yc + Pw + 2g 2g or, if the upstream velocity head is negligible, H − yc = (V 2c − V 21 ) V 2c = 2g 2g However, since V2 = Vc = (gyc ) 1/2, we find that V 2c = gyc so that we obtain The broad-crested weir is governed by critical flow across the weir block. H − yc = yc 2 or yc = 2H 3 Thus, the flowrate is Q = by2V2 = bycVc = byc (gyc ) 1/2 = b √g y3/2 c or 2 3/2 3/2 Q = b √g ( H 3) Again an empirical weir coefficient is used to account for the various real-world effects not included in the above simplified analysis. That is, 2 3/2 Q = Cwb b √g ( ) H 3/2 3 1 (10.35) where approximate values of Cwb, the broad-crested weir coefficient shown in the figure in the margin, can be obtained from the equation (Ref. 6) Cwb 0 0 1 H/Pw Cwb = 1.125 ( 1 + HPw 1/2 2 + HPw ) ) (10.36) 592 Chapter 10 ■ Open-Channel Flow EXAMPLE 10 .9 Sharp-Crested and Broad-Crested Weirs GIVEN Water flows in a rectangular channel of width b = 2 m with flowrates between Qmin = 0.02 m3/s and Qmax = 0.60 m3/s. This flowrate is to be measured by using either (a) a rectangular sharp-crested weir, (b) a triangular sharp-crested weir with θ = 90°, or (c) a broad-crested weir. In all cases the bottom of the flow area over the weir is a distance Pw = 1 m above the channel bottom. FIND Plot a graph of Q = Q(H) for each weir and comment on which weir would be best for this application. SOLUTION (a) For the rectangular weir with Pw = 1 m, Eqs. 10.32 and 10.33 give Q = Cwr 0.6 Qmax = 0.60 Rectangular 2 √2g bH 3/2 3 H 2 √2g bH 3/2 Pw ) 3 0.4 Broad-crested Q, m3/s = (0.611 + 0.075 Thus, Q = (0.611 + 0.075H) 2 √2(9.81 m/s2 ) (2 m) H 3/2 3 0.2 or Triangular Q = 5.91(0.611 + 0.075H)H 3/2 (1) where H and Q are in meters and m3/s, respectively. The results from Eq. 1 are plotted in Fig. E10.9. (b) Similarly, for the triangular weir, Eq. 10.34 gives θ 8 tan ( ) √2g H 5/2 15 2 8 = Cwt tan(45°) √2(9.81 m/s2 ) H 5/2 15 (2) For the broad-crested weir, Eqs. 10.30 and 10.31 give 2 3/2 Q = Cwbb√g ( ) H3/2 3 1 + H/Pw 1/2 2 3/2 3/2 = 1.125 ( b√g (3) H 2 + H/Pw ) Thus, with Pw=1 m 2 3/2 1 + H 1/2 (2 m) √9.81 m/s2 ( ) H3/2 ) 2+H 3 or Q = 3.84 ( 1 + H 1/2 3/2 H 2 + H) 0.4 0.6 0.8 COMMENTS Although it appears as though any of the three where H and Q are in meters and m3/s and Cwt is obtained from Fig. 10.21. For example, with H = 0.20 m = 0.656 ft, we find Cwt = 0.584, or Q = 2.36 (0.584) (0.20) 5/2 = 0.0247 m3/s. The triangular weir results are also plotted in Fig. E10.9. Q = 1.125 ( 0.2 ■ Figure E10.9 or (c) 0 H, m Q = Cwt Q = 2.36Cwt H5/2 Qmin = 0.02 0 (3) where, again, H and Q are in meters and m3/s. This result is also plotted in Fig. E10.9. weirs would work well for the upper portion of the flowrate range, neither the rectangular nor the broad-crested weir would be very accurate for small flowrates near Q = Qmin because of the small head, H, at these conditions. The triangular weir, however, would allow reasonably large values of H at the lowest flowrates. The corresponding heads with Q = Qmin = 0.02 m3/s for rectangular, triangular, and broad-crested weirs are 0.0312, 0.182, and 0.0375 m, respectively. In addition, as discussed in this section, for proper operation the broad-crested weir geometry is restricted to 0.08 < H/Lw < 0.50, where Lw is the weir block length. From Eq.3 with Qmax = 0.60 m3/s, we obtain Hmax = 0.349. Thus, we must have Lw > Hmax0.5 = 0.698m to maintain proper critical flow conditions at the largest flowrate in the channel. However, with Q = Qmin = 0.02 m3/s, we obtain Hmin = 0.0375 m. Thus, we must have Lw < Hmin0.08 = 0.469 m to ensure that frictional effects are not important. Clearly, these two constraints on the geometry of the weir block, Lw, are incompatible. A broad-crested weir will not function properly under the wide range of flowrates considered in this example. The sharpcrested triangular weir would be the best of the three types considered, provided the channel can handle the Hmax=0.719-m head. 10.6 Rapidly Varied Flow 593 y2 y1 a (a) (b) (c) ■ Figure 10.24 Three variations of underflow gates: (a) vertical gate, (b) radial gate, (c) drum gate. 10.6.4 Underflow (Sluice) Gates A variety of underflow gate (or sluice gate) structures is available for flowrate control at the crest of an overflow spillway (as shown by the figure in the margin), or at the entrance of an irrigation canal or river from a lake. Three types are illustrated in Fig. 10.24. Each has certain advantages and disadvantages in terms of costs of construction, ease of use, and the like, although the basic fluid mechanics involved is the same in all instances. The flow under a gate is said to be free outflow when the fluid issues as a jet of (Photograph courtesy of Pend supercritical flow with a free surface open to the atmosphere as shown in Fig. 10.24. In Oreille Public Utility District.) such cases it is customary to write this flowrate as the product of the distance, a, between the channel bottom and the bottom of the gate times the convenient reference velocity (2gy1 ) 1/2. That is, q = Cd a√2gy1 VIDEO V10.15 Spillway gate The flowrate from an underflow gate depends on whether the outlet is free or drowned. (10.37) where q is the flowrate per unit width. The discharge coefficient, Cd, is a function of the contraction coefficient, Cc = y2/a, and the depth ratio y1/a. Typical values of the discharge coefficient for free outflow (or free discharge) from a vertical sluice gate are on the order of 0.55 to 0.60 as indicated by the top line in Fig. 10.25 (Ref. 3). As indicated in Fig. 10.26, in certain situations the depth downstream of the gate is controlled by some downstream obstacle and the jet of water issuing from under the gate is overlaid by a mass of water that is quite turbulent. The flowrate for a submerged (or drowned) gate can be obtained from the same equation that is used for free outflow (Eq. 10.37), provided the discharge coefficient is modified appropriately. Typical values of Cd for drowned outflow cases are indicated as the series of lower curves in Fig. 10.25. Consider flow for a given gate and upstream conditions (i.e., given y1 a) corresponding to 0.6 Free outflow 0.5 0.4 Cd 0.3 Drowned outflow 0.2 0.1 y3 __ a = 2 0 2 3 4 4 5 6 6 7 8 8 10 12 14 16 y1 __ a ■ Figure 10.25 Typical discharge coefficients for underflow gates (Ref. 3). 594 Chapter 10 ■ Open-Channel Flow y1 y3 q a ■ Figure 10.26 Drowned outflow from a sluice gate. VIDEO V10.16 Unsteady under and over a vertical line in the figure. With y3 a = y1 a (i.e., y3 = y1 ) there is no head to drive the flow so that Cd = 0 and the fluid is stationary. For a given upstream depth (y1 a fixed), the value of Cd increases with decreasing y3 a until the maximum value of Cd is reached. This maximum corresponds to the free discharge conditions and is represented by the free outflow line so labeled in Fig. 10.25. For values of y3 a that give Cd values between zero and its maximum, the jet from the gate is overlaid (drowned) by the downstream water and the flowrate is therefore reduced when compared with a free discharge situation. Similar results are obtained for the radial gate and drum gate. EXAMPLE 10 .10 Sluice Gate GIVEN Water flows under the sluice gate shown in Fig. E10.10. The channel width is b = 20 ft, the upstream depth is y1 = 6 ft, and the gate is a = 1.0 ft off the channel bottom. FIND Plot a graph of flowrate, Q, as a function of y3. 250 Free outflow SOLUTION Submerged outflow 200 From Eq. 10.37 we have Q = bq = baCd √2gy1 Q, cfs 150 = 20 ft (1.0 ft) Cd √2(32.2 ft/s2 ) (6.0 ft) 100 or Q = 393Cd cfs The value of Cd is obtained from Fig. 10.25 along the vertical line y1/a = 6 ft/1 ft = 6. For y3 = 6 ft (i.e., y3 /a = 6 = y1/a) we obtain Cd = 0, indicating that there is no flow when there is no head difference across the gate. The value of Cd increases as y3 /a decreases, reaching a maximum of Cd = 0.56 when y3 /a = 3.2. Thus, with y3 = 3.2a = 3.2 ft y1 = 6 ft Q (1) 50 y3 a = 1 ft 0 0 1 2 3 4 5 6 y3, ft ■ Figure E10.10 Q = 393 (0.56) cfs = 220 cfs The flowrate for 3.2 ft ≤ y3 ≤ 6 ft is obtained from Eq. 1 and the Cd values of Fig. 10.24 with the results as indicated in Fig. E10.10. COMMENT For y3 < 3.2 ft the flowrate is independent of y3, and the outflow is a free (not submerged) outflow. For such 10.7 cases the inertia of the water flowing under the gate is sufficient to produce free outflow even with y3 > a. Chapter Summary and Study Guide This chapter discussed various aspects of flows in an open channel. A typical open-channel flow is driven by the component of gravity in the direction of flow. The character of such flows can be a strong function of the Froude number, which is the ratio of the fluid speed to the free-surface wave speed. The specific energy diagram is used to provide insight into the flow processes involved in open-channel flow. 10.7 open-channel flow Froude number critical flow subcritical flow supercritical flow wave speed specific energy specific energy diagram uniform flow wetted perimeter hydraulic radius Chezy equation Manning equation Manning coefficient gradually varied flow rapidly varied flow hydraulic jump sharp-crested weir weir head broad-crested weir underflow gate sluice gate Chapter Summary and Study Guide 595 Uniform flow is achieved by a balance between the potential energy lost by the fluid as it coasts downhill and the energy dissipated by viscous effects. Alternately, it represents a balance between weight and friction forces. The relationship among the flowrate, the slope of the channel, the geometry of the channel, and the roughness of the channel surfaces is given by the Manning equation. Values of the empirical Manning coefficient used in the Manning equation are dependent on the surface material roughness. The Manning equation is not dimensionally homogeneous. For gradually varied flow the fluid depth changes slowly along the length of the channel. The fluid depth is described by a first order ordinary differential equation, which can be solved numerically. The hydraulic jump is an example of nonuniform depth open-channel flow. If the Froude number of a flow is greater than 1, the flow is supercritical, and a hydraulic jump may occur. The momentum and mass equations are used to obtain the relationship between the upstream Froude number and the depth ratio across the jump. The energy dissipated in the jump and the head loss can then be determined by use of the energy equation. The use of weirs to measure the flowrate in an open channel is discussed. The relationships between the flowrate and the weir head are given for both sharp-crested and broad-crested weirs. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ ■ ■ ■ ■ ■ write out meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. determine the Froude number for a given flow and explain the concepts of subcritical, critical, and supercritical flows. plot and interpret the specific energy diagram for a given flow. use the Manning equation to analyze uniform flow in an open channel. calculate properties such as the depth ratio and the head loss for a hydraulic jump. determine the flowrates over sharp-crested weirs, broad-crested weirs, and under underflow gates. Some of the important equations in this chapter are: Froude number Fr = V(gy) 1/2 Wave speed c = √gy Specific energy E=y+ Manning equation V= Hydraulic jump depth ratio y2 1 = (−1 + √1 + 8Fr21 ) y1 2 (10.26) Hydraulic jump head loss y2 y1 2 hL Fr21 =1− + 1 − ( y2 ) ] y1 y1 2 [ (10.27) Rectangular sharp-crested weir Q = Cwr 2 √2g b H3/2 3 (10.32) Triangular sharp-crested weir Q = Cwt 8 θ tan √2g H 5/2 15 ( 2 ) (10.34) Broad-crested weir 2 3/2 Q = Cwb b √g( ) H 3/2 3 (10.35) Underflow (sluice) gate q = Cd a√2gy1 (10.37) (10.3) V2 2g κ 2/3 1/2 R S n h 0 (10.8) (10.16) 596 Chapter 10 ■ Open-Channel Flow References 1. Currie, C. G., and Currie, I. G., Fundamental Mechanics of Fluids, 3rd Ed., Marcel Dekker, New York, 2003. 2. Stoker, J. J., Water Waves, Interscience, New York, 1957. 3. Henderson, F. M., Open Channel Flow, Macmillan, New York, 1966. 4. Rouse, H., Elementary Fluid Mechanics, Wiley, New York, 1946. 5. French, R. H., Open Channel Hydraulics, McGraw-Hill, New York, 1992. 6. Chow, V. T., Open Channel Hydraulics, McGraw-Hill, New York, 1959. 7. Blevins, R. D., Applied Fluid Dynamics Handbook, Van Nostrand Reinhold, New York, 1984. 8. Daugherty, R. L., and Franzini, J. B., Fluid Mechanics with Engineering Applications, McGraw-Hill, New York, 1977. 9. Vennard, J. K., and Street, R. L., Elementary Fluid Mechanics, 7th Ed., Wiley, New York, 1995. 10. Camp, T. R., “Design of Sewers to Facilitate Flow,” Sewage Works Journal, 18(3), 1946. 11. Lenz, A. T., “Viscosity and Surface Tension Effects on V-Notch Weir Coefficients,” Transactions of the American Society of Civil Engineers, Vol. 108, 759–802, 1943. 12. White, F. M., Fluid Mechanics, 5th Ed., McGraw-Hill, New York, 2003. 13. U.S. Bureau of Reclamation, Research Studies on Stilling Basins, Energy Dissipators, and Associated Appurtenances, Hydraulic Lab Report Hyd.-399, June 1, 1955. 14. Wallet, A., and Ruellan, F., Houille Blanche, Vol. 5, 1950. 15. Spitzer, D. W., ed., Flow Measurement: Practical Guides for Measurement and Control, Instrument Society of America, Research Triangle Park, N.C., 1991. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are no unique answers to these problems. Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. Section 10.2 Surface Waves 10.1 On a distant planet small-amplitude waves travel across a 1-m-deep pond with a speed of 5 m/s. Determine the acceleration of gravity on the surface of that planet. 10.2 The flowrate per unit width in a wide channel is q = 2.3 m2/s. Is the flow subcritical or supercritical if the depth is (a) 0.2 m, (b) 0.8 m, or (c) 2.5 m? 10.3 A rectangular channel 3 m wide carries 10 m3/s at a depth of 2 m. Is the flow subcritical or supercritical? For the same flowrate, what depth will give critical flow? 10.4 Do shallow waves propagate at the same speed in all fluids? Explain why or why not. 10.5 Waves on the surface of a tank are observed to travel at a speed of 2 m/s. How fast would these waves travel if (a) the tank were in an elevator accelerating downward at a rate of 4 m/s2, (b) the tank accelerates horizontally at a rate of 9.81 m/s2, (c) the tank were aboard the orbiting Space Shuttle? Explain. 10.6 In flowing from section (1) to section (2) along an open channel, the water depth decreases by a factor of 2 and the Froude number changes from a subcritical value of 0.5 to a supercritical value of 3.0. Determine the channel width at (2) if it is 12 ft wide at (1). 10.7 Water flows with an average velocity of 1.0 m/s and a normal depth of 0.5 m in a wide rectangular channel. Is the flow subcritical or supercritical? 10.8 A trout jumps, producing waves on the surface of a 0.8-m-deep mountain stream. If it is observed that the waves do not travel upstream, what is the minimum velocity of the current? Problems 10.9 Observations at a shallow sandy beach show that even though the waves several hundred yards out from the shore are not parallel to the beach, the waves often “break” on the beach nearly parallel to the shore as indicated in Fig. P10.9. Explain this behavior based on the wave speed c = (gy)1/2. 597 where zB and x are in meters. The water depth far upstream of the bump is y1 = 2 m. Plot a graph of the water depth, y = y(x), and the surface elevation, z = z(x), for −4 m ≤ x ≤ 4 m. Assume onedimensional flow. z V1 Wave crest y1 y(x) z(x) c 0 Ocean x zB = 0.2e–x 2 ■ Figure P10.16 Beach 10.17 Water in a rectangular channel flows into a gradual contraction section as is indicated in Fig. P10.17. If the flowrate is Q = 25 ft3/s and the upstream depth is y1 = 2 ft, determine the downstream depth, y2. ■ Figure P10.9 10.10 (See The Wide World of Fluids article titled “Tsunami, the Nonstorm Wave,” Section 10.2.1.) Often when an earthquake shifts a segment of the ocean floor, a relatively small-amplitude wave of very long wavelength is produced. Such waves go unnoticed as they move across the open ocean; only when they approach the shore do they become dangerous (e.g., a tsunami). Determine the wave speed if the wavelength, λ, is 6000 ft and the ocean depth is 15,000 ft. 10.11 What is the minimum water depth necessary for a 40-ftwide stream to handle 4000 ft3/s if the flow is not supercritical? V1 b1 = 4 ft b2 = 3 ft V2 y2 V2 Top view V1 y1 Section 10.3 Energy Considerations 10.12 Water flows in a 10-m-wide open channel with a flowrate of 5 m3/s. Determine the two possible depths if the specific energy of the flow is E = 0.6 m. 10.13 Water flows in a 10-ft-wide rectangular channel with a flowrate of 200 ft3/s. Plot the specific energy diagram for this flow. Determine the two possible flowrates when the specific energy is 6 ft. 10.14 Water flows in a rectangular channel at a rate of q = 20 cfs/ft. When a Pitot tube is placed in the stream, water in the tube rises to a level of 4.5 ft above the channel bottom. Determine the two possible flow depths in the channel. Illustrate this flow on a specific energy diagram. 10.15 Water flows in a 5-ft-wide rectangular channel with a flowrate of Q = 30 ft3/s and an upstream depth of y1 = 2.5 ft as is shown in Fig. P10.15. Determine the flow depth and the surface elevation at section (2). V1 y1 Q y2 V2 (1) Side view (2) ■ Figure P10.17 10.18 A channel has a rectangular cross section, a width of 40 m, and a flow rate of 4000 m3/s. The normal water depth is 20 m. The flow then encounters a 4.0-m-high dam. Find the water depth directly above the dam if the flow is critical. Assume frictionless flow. 10.19 Repeat Problem 10.17 if the upstream depth is y1 = 0.5 ft. Assume that there are no losses between sections (1) and (2). 10.20 A rectangular channel has a gradual contraction in width from 59 ft to 30 ft and a bed level drop of 6 in. below the upstream channel bed, which the increased velocity caused by the contraction scoured. If the upstream and downstream flow depths are both 10.1 ft, compute the discharge in the channel. 10.21 Water flows in a rectangular channel with a flowrate per unit width of q = 1.5 m2/s and a depth of 0.5 m at section (1). The head loss between sections (1) and (2) is 0.03 m. Plot the specific energy diagram for this flow and locate states (1) and (2) on this diagram. Is it possible to have a head loss of 0.06 m? Explain. ■ Figure P10.15 10.22 Water flows in a horizontal rectangular channel with a flowrate per unit width of q = 10 ft2/s and a depth of 1.0 ft at the downstream section (2). The head loss between section (1) upstream and section (2) is 0.2 ft. Plot the specific energy diagram for this flow and locate states (1) and (2) on this diagram. 10.16 Water flows over the bump in the bottom of the rectangular channel shown in Fig. P10.16 with a flowrate per unit width of2 q = 4 m2/s. The channel bottom contour is given by zB = 0.2e−x , 10.23 Water flows in a horizontal, rectangular channel with an initial depth of 1 m and an initial velocity of 4 m/s. Determine the depth downstream if losses are negligible. Note that there may be more than one solution. 0.2 ft (2) (1) 598 Chapter 10 ■ Open-Channel Flow 10.24 A smooth transition section connects two rectangular channels as shown in Fig. P10.24. The channel width increases from 6.0 to 7.0 ft, and the water surface elevation is the same in each channel. If the upstream depth of flow is 3.0 ft, determine h, the amount the channel bed needs to be raised across the transition section to maintain the same surface elevation. 10.28 Repeat Problem 10.27 if the Froude number is 2.75. 10.29 Supercritical, uniform flow of water occurs in a 5.0-m-wide, rectangular, horizontal channel. The flow has a depth of 1.5 m and a flow rate of 45.0 m3/s. The water flow encounters a 0.25-m rise in the channel bottom. Find the normal depth after the rise in the channel bottom. Is the flow after the rise subcritical, critical, or supercritical? Assume frictionless flow. Section 10.4.2 The Manning Equation Q 6 ft 7 ft 10.30 Water flows in a 5-m-wide channel with a speed of 2 m/s and a depth of 1 m. The channel bottom slopes at a rate of 1 m per 1000 m. Determine the Manning coefficient for this channel. 10.31 The following data are taken from measurements on Indian Fork Creek: A = 26 m2, P = 16 m, and S0 = 0.02 m/62 m. Determine the average shear stress on the wetted perimeter of this channel. Top view Q 3 ft 10.32 Consider laminar flow down a wide rectangular channel making an angle θ with the horizontal. The fluid has kinematic viscosity v and the volume flow rate per unit width is given by h Side view q= ■ Figure P10.24 10.25 Water flows over a bump of height h = h(x) on the bottom of a wide rectangular channel as is indicated in Fig. P10.25. If energy losses are negligible, show that the slope of the water surface is given by dy/dx = –(dh/dx)[1 − (V 2/gy)], where V = V(x) and y = y(x) are the local velocity and depth of flow. Comment on the sign (i.e., <0, = 0, or >0) of dydx relative to the sign of dhdx. y y1 where y is the fluid depth perpendicular to the channel bottom. Find Manning’s n for the flow. 10.33 The following data are obtained for a particular reach of the Provo River in Utah: A = 183 ft2, free-surface width = 55 ft, average depth = 3.3 ft, Rh = 3.32 ft, V = 6.56 ft/s, length of reach = 116 ft, and elevation drop of reach = 1.04 ft. Determine (a) the average shear stress on the wetted perimeter, (b) the Manning coefficient, n, and (c) the Froude number of the flow. 10.34 At a particular location, the cross section of the Columbia River is as indicated in Fig. P10.34. If on a day without wind it takes 5 min to float 0.5 mile along the river, which drops 0.46 ft in that distance, determine the value of the Manning coefficient, n. 𝓵(x) V1 gy3 sin θ , 3v V(x) h(x) ■ Figure P10.25 10.26 Consider 100 ft3/s of water flowing down a rectangular channel measuring 10 ft wide. The normal depth is 3.00 ft. A 4.0-ft-diameter pier is located in the channel. Find the water depth as it flows past the pier. Assume frictionless flow. 10.27 Water flows in the river shown in Fig. P10.27 with a uniform bottom slope. The total head at each section is measured by using Pitot tubes as indicated. Determine the value of dydx at the location where the Froude number is 0.357. (3) Depth, ft 60 x 40 20 0 0 400 800 1200 Width, ft 1600 2000 ■ Figure P10.34 10.35 A uniform flow of 110,000 ft3/s is measured in a natural channel that is approximately rectangular in shape with a 2650-ft width and 17.5-ft depth. The water-surface elevation drops 0.37 ft per mile. Based on the computed Manning n, characterize the type of natural channel observed. Also compute the Froude number and determine whether the flow is subcritical or supercritical. (4) Q Section 10.4.3 Uniform Flow—Determine Flowrate (1) z1 = 620.1 ft z3 = 628.3 ft x2 – x1 = 4100 ft ■ Figure P10.27 z (2) z2 = 618.7 ft z4 = 625.0 ft 10.36 A 2-m-diameter pipe made of finished concrete lies on a slope of 1-m elevation change per 1000-m horizontal distance. Determine the flowrate when the pipe is half full. 10.37 By what percent is the flowrate reduced in the rectangular channel shown in Fig. P10.37 because of the addition of the thin center board? All surfaces are of the same material. Problems Center board 599 10.46 The smooth, concrete-lined channel shown in Fig. P10.46 is built on a slope of 2 m/km. Determine the flowrate if the depth is y = 1.5 m. b/2 6m Concrete b/2 1.0 m 3m 0.5 m b ■ Figure P10.37 ■ Figure P10.46 10.38 A large trapezoidal channel cut through stone has side slopes of 1:1 and a bed width of 291 ft. Find the uniform flow in the channel when the flow depth is 30.4 ft and the bed slopes 1 ft per mile. 10.47 At a given location, under normal conditions a river flows with a Manning coefficient of 0.030, and a cross section as indicated in Fig. P10.47a. During flood conditions at this location, the river has a Manning coefficient of 0.040 (because of trees and brush in the floodplain) and a cross section as shown in Fig. P10.47b. Determine the ratio of the flowrate during flood conditions to that during normal conditions. 10.39 The great Kings River flume in Fresno County, California, was used from 1890 to 1923 to carry logs from an elevation of 4500 ft where trees were cut to an elevation of 300 ft at the railhead. The flume was 54 miles long, constructed of wood, and had a V-cross section as indicated in Fig. P10.39. It is claimed that logs would travel the length of the flume in 15 hours. Do you agree with this claim? Provide appropriate calculations to support your answer. 12 ft 800 ft (a) 20 ft 8 ft 800 ft 1 ft 1000 ft (b) 1 ft ■ Figure P10.47 ■ Figure P10.39 10.40 An unfinished concrete rectangular channel is 5 m wide and has a slope of 0.50°. The water is 0.5 m deep. Find the discharge rate for uniform flow. 10.41 A trapezoidal channel with a bottom width of 3.0 m and sides with a slope of 2:1 (horizontal:vertical) is lined with fine gravel (n = 0.020) and is to carry 10 m3/s. Can this channel be built with a slope of S0 = 0.00010 if it is necessary to keep the velocity below 0.75 m/s to prevent scouring of the bottom? Explain. 10.42 Water flows in a 2-m-diameter finished concrete pipe so that it is completely full and the pressure is constant all along the pipe. If the slope is S0 = 0.005, determine the flowrate by using open-channel flow methods. Compare this result with that obtained by using pipe flow methods of Chapter 8. 10.43 A round concrete storm sewer pipe used to carry rainfall runoff from a parking lot is designed to be half full when the rainfall rate is a steady 1 in./hr. Will this pipe be able to handle the flow from a 2-in./hr rainfall without water backing up into the parking lot? Support your answer with appropriate calculations. 10.44 Find the discharge per unit width for a wide channel having a bottom slope of 0.00015. The normal depth is 0.003 m. Assume laminar flow and justify the assumption. The fluid is 20°C water. 10.45 Water flows down a wide rectangular channel having Manning’s n = 0.015 and bottom slope = 0.0015. Find the rate of discharge and normal depth for critical flow conditions. 10.48 The channel in Fig. P10.48 has two floodplains as shown. Find the discharge if the center channel is lined with brick and the two floodplains are lined with cobblestones. The slope S0 is 0.00025. n2 n2 n1 30 m 15 m 6m 6m 30 m ■ Figure P10.48 10.49 Determine the flowrate for the symmetrical channel shown in Fig. P10.66 if the bottom is smooth concrete and the sides are weedy. The bottom slope is S0 = 0.001. 10.50 (See The Wide World of Fluids article titled “Done without GPS or Lasers,” Section 10.4.3.) Determine the number of gallons of water delivered per day by a rubble masonry, 1.2-m-wide aqueduct laid on an average slope of 14.6 m per 50 km if the water depth is 1.8 m. Section 10.4.3 Uniform Flow—Determine Depth or Size 10.51 An old, rough-surfaced, 2-m-diameter concrete pipe with a Manning coefficient of 0.025 carries water at a rate of 5.0 m3/s when it is half full. It is to be replaced by a new pipe with a 600 Chapter 10 ■ Open-Channel Flow Manning coefficient of 0.012 that is also to flow half full at the same flowrate. Determine the diameter of the new pipe. the channel need fewer bricks if the channel were 2 ft wide, 6 ft wide, or 10 ft wide? Explain. 10.52 Four sewer pipes of 0.5-m diameter join to form one pipe of diameter D. If the Manning coefficient, n, and the slope are the same for all of the pipes, and if each pipe flows half full, determine D. 10.62 Two canals join to form a larger canal as shown in Video V10.6 and Fig. P10.62. Each of the three rectangular canals is lined with the same material and has the same bottom slope. The water depth in each is to be 2 m. Determine the width of the merged canal, b. Explain physically (i.e., without using any equations) why it is expected that the width of the merged canal is less than the combined widths of the two original canals (i.e., b < 4 m + 8 m = 12 m). 10.53 The spillway of a dam is 20.0 ft wide and has a flow rate of 5000 ft3/s. The spillway makes an angle of 30° with the horizontal. Find the vertical water depth down the spillway. See Problem 10.72. 10.54 The flowrate in the clay-lined channel (n = 0.025) shown in Fig. P10.54 is to be 300 ft3/s. To prevent erosion of the sides, the velocity must not exceed 5 ft/s. For this maximum velocity, determine the width of the bottom, b, and the slope, S0. Q2 Q1 2 ft Q3 8m b 4m 45° 30° ■ Figure P10.62 b ■ Figure P10.54 10.55 The rate of discharge through the canal in Fig. P10.55 is to be 10 m3/s. Find the width b and the bottom slope. The velocity will not exceed 5.0 m/s. 2.0 m 45° 10.63 Water flows uniformly at a depth of 1 m in a channel that is 5 m wide as shown in Fig. P10.63. Further downstream, the channel cross section changes to that of a square of width and height b. Determine the value of b if the two portions of this channel are made of the same material and are constructed with the same bottom slope. 1m b b ■ Figure P10.55 10.56 A rectangular, unfinished concrete channel of 28-ft width is laid on a slope of 8 ft/mi. Determine the flow depth and Froude number of the flow if the flowrate is 400 ft3/s. 10.57 An engineer is to design a channel lined with planed wood to carry water at a flowrate of 2 m3/s on a slope of 10 m/800 m. The channel cross section can be either a 90° triangle or a rectangle with a cross section twice as wide as its depth. Which would require less wood and by what percent? 10.58 Find the diameter required for reinforced concrete pipe laid at a slope of 0.001 and required to carry a uniform flow of 19.3 ft3/sec when the depth is 75% of the diameter. 10.59 A major river is divided into three parts or courses—the upper course, the middle course, and the lower course. The slope is 70 ft per mile in the upper course, 10 ft per mile in the middle course, and 1.0 ft per mile in the lower course. All three courses have rectangular cross sections. The upper course has a normal depth of 40.0 ft. The river width is 120 ft. Find the normal depths of the middle and lower courses. 10.60 An 8-ft-diameter concrete drainage pipe that flows half full is to be replaced by a concrete-lined V-shaped open channel having an interior angle of 90°. Determine the depth of fluid that will exist in the V-shaped channel if it is laid on the same slope and carries the same discharge as the drainage pipe. 10.61 A rectangular brick-lined channel has a bottom slope of 0.0025 and is designed to carry a uniform water flow rate of 300 ft3/s. Would Width = 5 m ■ Figure P10.63 Section 10.4.3 Uniform Flow—Determine Slope 10.64 Water flows 1 m deep in a 2-m-wide finished concrete channel. Determine the slope if the flowrate is 3 m3/s. 10.65 Uniform flow in a sluggish channel having a nearly rectangular cross section that is 498 ft wide and 16.5 ft deep carries a flow of 8,250 ft3/s. Approximately how much does the water surface elevation drop along a river mile? 10.66 To prevent weeds from growing in a clean earthen-lined canal, it is recommended that the velocity be no less than 2.5 ft/s. For the symmetrical canal shown in Fig. P10.66, determine the minimum slope needed. 12 ft 3 ft 4 ft ■ Figure P10.66 10.67 The symmetrical channel shown in Fig. P10.66 is dug in sandy loam soil with n = 0.020. For such surface material it is 601 Problems Section A–A recommended that to prevent scouring of the surface the average velocity be no more than 1.75 ft/s. Determine the maximum slope allowed. ϕ 10.68 Figure P10.68 shows a cross section of an aqueduct that carries water at 50 m3/s. The value of Manning’s n is 0.015. Find the bottom slope. A y ϕ b θ g θ A 4.0 m 45° y cos θ x 7.0 m ■ Figure P10.71 ■ Figure P10.68 10.69 The depth downstream of a sluice gate in a rectangular wooden channel of width 5 m is 0.60 m. If the flowrate is 18 m3/s, determine the channel slope needed to maintain this depth. Will the depth increase or decrease in the flow direction if the slope is (a) 0.02; (b) 0.01? 10.72 Consider the flow down a prismatic channel having a rectangular cross section of width b. The channel bottom makes an angle θ with the horizontal. Show that 2 tan θ − (n2Q2 )(A2R4/3 dy h κ ) = dx 1 − Q2(A2gy cos2θ)   where y is the vertical fluid depth, A = by cos θ, and Q is the vol10.70 A 50-ft-long aluminum gutter (Manning coefficient n = 0.011) ume flow rate. Discuss the form of the equation for small values of on a section of a roof is to handle a flowrate of 0.15 ft3/s during a θ (θ ∼ 1°). heavy rainstorm. The cross section of the gutter is shown in Fig. *10.73 Water flows at 150 ft3/s in a 3-ft-wide rectangular cleanP10.70. Determine the vertical distance that this gutter must be earth irrigation canal. The canal slope is 0.275°. At one point, the pitched (i.e., the difference in elevation between the two ends of the water depth is 3 ft. (a) Accurately compute the water depth at a gutter) so that the water does not overflow the gutter. Assume location 200 ft downstream. (b) Is the flow at the upstream location uniform depth channel flow. subcritical or supercritical? At the downstream location? (c) Sketch the canal and the water surface profile. Section 10.6.1 The Hydraulic Jump 10.74 Water flows upstream of a hydraulic jump with a depth of 0.5 m and a velocity of 6 m/s. Determine the depth of the water downstream of the jump. 10.75 A 5.0-m-wide channel has a slope of 0.004, a 8.0-m3/s. water flow rate, and a water depth 1.5 m after a hydraulic jump. Find the water depth before the jump. 5 in. 60° 3 in. ■ Figure P10.70 Section 10.5 Gradually Varied Flow 10.71 Consider the flow down a prismatic channel having a trapezoidal cross section of base width b and top width b + 2y cos θ cot ϕ. The channel bottom makes an angle θ with the horizontal, and y is the vertical fluid depth. (See Fig. P11.71.) Show that dy = dx 10.76 The water depths upstream and downstream of a hydraulic jump are 0.3 and 1.2 m, respectively. Determine the upstream velocity and the power dissipated if the channel is 50 m wide. 10.77 Under appropriate conditions, water flowing from a faucet, onto a flat plate, and over the edge of the plate can produce a circular hydraulic jump as shown in Fig. P10.77 and Video V10.12. Consider a situation where a jump forms 3.0 in. from the center of the plate with depths upstream and downstream of the jump of 0.05 in. and 0.20 in., respectively. Determine the flowrate from the faucet. tan θ − (n2Q2A2Rh 4/3κ2 ) Jump . b + 2y cos θ cot ϕ 1−[ 2 A gy cos2 θ ][ b + y cos θ cot ϕ ] Q2 3 in. Jump Note that the average fluid velocity is given by V= Q Q = A y cos θ(b + y cos θ cot ϕ) Discuss the form of the equation for small values of θ (θ ∼ 1°). 0.05 in. 0.20 in. ■ Figure P10.77 602 Chapter 10 ■ Open-Channel Flow 10.78 At a given location in a 12-ft-wide rectangular channel the flowrate is 900 ft3/s and the depth is 4 ft. Is this location upstream or downstream of the hydraulic jump that occurs in this channel? Explain. 10.79 A rectangular channel 3.0 m wide has a flow rate of 5.0 m3/s with a normal depth of 0.50 m. The flow then encounters a dam that rises 0.25 m above the channel bottom. Will a hydraulic jump occur? Justify your answer. 10.80 A 90° triangular flume has sides 2.0 m wide, a water flow rate of 1.0 m3/s, and a depth of 0.50 m. Find the depth after a hydraulic jump and the power loss in the jump. 10.81 Water flows in a rectangular channel with velocity V = 6 m/s. A gate at the end of the channel is suddenly closed so that a wave (a moving hydraulic jump) travels upstream with velocity Vw = 2 m/s. Determine the depths ahead of and behind the wave. Note that this is an unsteady problem for a stationary observer. However, for an observer moving to the left with velocity Vw, the flow appears as a steady hydraulic jump. 10.82 A hydraulic engineer wants to analyze steady flow in a rectangular channel featuring a hydraulic jump immediately downstream from a sluice gate that is open to a vertical clearance of 3 ft. The flow depth upstream from the sluice gate 8.7 ft, and the flow velocity beyond the sluice gate and prior to the hydraulic jump is 21.6 ft/s. Assume that the flow upstream from the sluice gate is subcritical. Find: (a) The discharge in the channel; (b) The flow depth before and after the hydraulic jump; (c) The flow velocities upstream from the sluice gate and beyond the hydraulic jump; (d) The energy loss rate in the hydraulic jump; (e) The force the sluice gate exerts on the fluid. 1.5 ft 2 ft 1.5 ft ■ Figure P10.86 10.87 Water flows over a broad-crested weir that has a width of 4 m and a height of Pw = 1.5 m. The free-surface well upstream of the weir is at a height of 0.5 m above the surface of the weir. Determine the flowrate in the channel and the minimum depth of the water above the weir block. 10.88 An engineering laboratory experiment uses a triangular weir in an open channel to measure flow rate. The nominal weir angle is 90°. In a certain test, the head of water above the weir was 4 in. The uncertainty in the weir angle is ±2° and the uncertainty in the depth measurement is ±0.05 in. Compute the flow rate and the uncertainty in the flow rate. Section 10.6.4 Underflow (Sluice) Gates 10.89 Water flows under a sluice gate in a 60-ft-wide finished concrete channel as is shown in Fig. P10.89. Determine the flowrate. If the slope of the channel is 2.5 ft/200 ft, will the water depth increase or decrease downstream of the gate? Assume Cc = y2  a = 0.65. Explain. Q 10 ft How does this compare with the force computed assuming a hydrostatic pressure distribution? 10.83 (See The Wide World of Fluids article titled “Grand Canyon Rapids Building.” Section 10.6.1.) During the flood of 1983, a large hydraulic jump formed at “Crystal Rapid” on the Colorado River. People rafting the river at that time report “entering the rapid at almost 30 mph, hitting a 20-ft-tall wall of water, and exiting at about 10 mph.” Is this information (i.e., upstream and downstream velocities and change in depth) consistent with the principles of a hydraulic jump? Show calculations to support your answer. Section 10.6.2, 3 Sharp-Crested and Broad-Crested Weirs 10.84 A rectangular sharp-crested weir is used to measure the flowrate in a channel of width 10 ft. It is desired to have the upstream channel flow depth be 6 ft when the flowrate is 50 cfs. Determine the height, Pw, of the weir plate. 10.85 Water flows over a sharp-crested triangular weir with θ = 90°. The head range covered is 0.20 ≤ H ≤ 1.0 ft and the accuracy in the measurement of the head, H, is δH = ±0.01 ft. Plot a graph of the percent error expected in Q as a function of Q. 10.86 Water flows over the sharp-crested weir shown in Fig. P10.86. The weir plate cross section consists of a semicircle and a rectangle. Plot a graph of the estimated flowrate, Q, as a function of head, H. List all assumptions and show all calculations. H y2 a = 2 ft ■ Figure P10.89 10.90 Water flows under a sluice gate in a channel of 10-ft width. If the upstream depth remains constant at 5 ft, plot a graph of flowrate as a function of the distance between the gate and the channel bottom as the gate is slowly opened. Assume free outflow. 10.91 A flow of 873 ft3/s passes under a sluice gate in a rectangular channel having a gradual contraction in width from 80 ft to 52 ft. The channel bed has scoured to a level that is 6 in. below the upstream channel bed, owing to the increased velocity caused by the sluice gate and the contraction. If the upstream flow depth is 4 ft and the sluice gate opening is 1.5 ft, compute the depth, velocity, and Froude number of the flow at the end of the contraction. ■ Lifelong Learning Problems 10.1 LL With the increased usage of low-lying coastal areas and the possible rise in ocean levels because of global warming, the potential for widespread damage from tsunamis is increasing. Obtain information about new and improved methods available to predict the occurrence of these damaging waves and how to better use coastal areas so that massive loss of life and property does not occur. Summarize your findings in a brief report. Problems 10.2 LL Recent photographs from NASA’s Mars Orbiter Camera on the Mars Global Surveyor provide new evidence that water may still flow on the surface of Mars. Obtain information about the possibility of current or past open-channel flows on Mars and other planets or their satellites. Summarize your findings in a brief report. 10.3 LL Hydraulic jumps are normally associated with water flowing in rivers, gullies, and other such relatively high-speed open 603 channels. However, recently, hydraulic jumps have been used in various manufacturing processes involving fluids other than water (such as liquid metal solder) in relatively small-scale flows. Obtain information about new manufacturing processes that involve hydraulic jumps as an integral part of the process. Summarize your findings in a brief report. 11 Compressible Flow Learning Objectives After completing this chapter, you should be able to: ■ explain speed of sound and Mach number and their practical significance. ■ distinguish between incompressible and compressible flows, and know when the approximations associated with assuming fluid incompressibility are acceptable. ■ describe important features of different categories of compressible flow. ■ discuss various responses of compressible flow to area change in a duct ■ discuss and recognize choking in compressible flow ■ solve problems involving isentropic and nonisentropic flows including flows with area change, flows with normal shocks, flows with friction, and flows with heating or cooling. ■ appreciate the similarities between compressible flow of gases and open-channel flow of liquids. ■ move on to advanced study of compressible flow. Most first courses in fluid mechanics concentrate on constant density (incompressible) flows. In earlier chapters of this book, we mainly considered incompressible flow behavior. The model of an incompressible fluid is convenient because when constant density and constant (including zero) viscosity are assumed, problem solutions are greatly simplified. Also, fluid incompressibility allows us to build on the Bernoulli equation as was done, for example, in Chapter 5. Examples in Chapters 5–10 should have convinced you that nearly incompressible flows are common in everyday experiences. Any study of fluid mechanics would, however, be incomplete without a brief introduction to compressible flow. Fluid compressibility is very important in numerous engineering applications of fluid mechanics. For example, the measurement of high-speed flow velocities requires compressible flow theory. The flows in gas turbine engine components are generally compressible. Many aircraft fly fast enough to encounter compressible flow. The variation of fluid density for compressible flows requires a relationship between density and other fluid properties, especially pressure and temperature. This relationship, the fluid equation of state, often unimportant for incompressible flows, is vital in the analysis of compressible flows. Also, temperature variations for compressible flows are usually significant, and thus the energy equation is necessary. Curious phenomena can occur with compressible flows. For example, we can have fluid acceleration because of friction, fluid deceleration in a converging duct, fluid temperature decrease with heating, and the formation of abrupt discontinuities in flows across which fluid properties change appreciably. 604 11.1 We consider mostly ideal gas flows. 11.1 VIDEO Ideal Gas Thermodynamics 605 For simplicity, in this introductory study of compressibility effects we mainly consider the steady, one-dimensional, constant (including zero) viscosity, compressible flow of an ideal gas. We limit our study to compressibility due to high-speed flow. In this chapter, one-dimensional flow refers to flow involving uniform distributions of fluid properties over any flow crosssectional area. Both frictionless (μ = 0) and frictional (μ ≠ 0) compressible flows are considered. If the change in density associated with a change of pressure is considered a measure of compressibility, our experience suggests that gases and vapors are much more compressible than liquids. We focus our attention on the compressible flow of a gas because such flows occur often. We limit our discussion to ideal gases, since the equation of state for an ideal gas is uncomplicated, yet representative of actual gases at pressures and temperatures of engineering interest, and because the flow trends associated with an ideal gas are generally applicable to other compressible fluids. Another advantage of emphasizing ideal gases is that the equations used for solving many problems are closed-form algebraic equations. A disadvantage is that there are many equations and they are almost all “messy” (nonlinear, and complicated in appearance). In practice, calculations are often made with the aid of tables of numerical values or by reading values from graphs. Because of the large number of equations and because compressible flow analysis has its own special “jargon,” it is easy for students to get lost. The most important objective is to understand the physics of compressible flow rather than how to manipulate the myriad equations, charts, and tables. Toward this end, we will explore the unique phenomena of each type of compressible flow qualitatively, using the minimum number of basic equations before developing and using the computational equations, charts, and tables. An excellent film about compressible flow is available (see Ref. 1). You really should watch this film, ideally both before and after studying this chapter. Ideal Gas Thermodynamics V11.1 Lighter flame Concepts from thermodynamics and concepts from fluid mechanics are equally important in compressible flow analysis. Before we begin the study of compressible flow, we should consider a few basic concepts from thermodynamics. If you have already studied thermodynamics, this section will serve as a review. In this section, we present all the concepts and equations needed to understand the compressible flow of an ideal gas but the developments and explanations are brief. Consult a thermodynamics textbook such as (Ref. 2) for more detail. Before we can develop compressible flow equations, we must learn how to evaluate ideal gas property changes. The equation of state for an ideal gas is ρ= p RT (11.1) We have already discussed fluid pressure, p, density, ρ, and temperature, T, in earlier chapters. The gas constant, R, is a constant for each distinct ideal gas or mixture of ideal gases, where R= λ Mgas (11.2) With this notation, λ is the universal gas constant and Mgas is the molecular weight of the ideal gas or gas mixture. Listed in Tables 1.7 and 1.8 are values of the gas constants of some common gases. Knowing any two of pressure, temperature, or density, we can calculate the third. For an ideal gas, internal energy, uˇ , is part of the stored energy of the gas as explained in Section 5.3 and is a function of temperature only (Ref. 2). Thus, the ideal gas specific heat at constant volume, cυ, can be expressed as cυ = ( ∂uˇ duˇ = ) ∂T υ dT (11.3) 606 Chapter 11 ■ Compressible Flow where the subscript υ on the partial derivative refers to differentiation at constant specific volume, υ = 1ρ. For a particular ideal gas, cυ is a function of temperature only. Equation 11.3 can be rearranged to yield duˇ = cυ dT Thus, uˇ 2 − uˇ 1 = ∫ T2 cυ dT (11.4) T1 Equation 11.4 allows us to evaluate the change in internal energy, uˇ 2 − uˇ 1, associated with ideal gas flow from section (1) to section (2). For simplicity, we can assume that cυ is constant for a particular ideal gas and obtain from Eq. 11.4 uˇ 2 − uˇ 1 = cυ (T2 − T1 ) (11.5) Actually, cυ for a particular gas varies with temperature (see Ref. 2). However, for moderate changes in temperature, the constant cυ assumption is reasonable. The fluid property enthalpy, hˇ , is defined as p hˇ = uˇ + ρ (11.6) It combines internal energy, uˇ , and pressure energy (flow work), pρ, and is useful when working with the energy equation (Eq. 5.69). For an ideal gas, we have already stated that uˇ = uˇ (T) From the equation of state (Eq. 11.1) p = RT ρ Thus, it follows that hˇ = hˇ (T) Since for an ideal gas, enthalpy is a function of temperature only, the ideal gas specific heat at constant pressure, cp, can be expressed as ∂hˇ dhˇ cp = ( ) = (11.7) ∂T p dT where the subscript p on the partial derivative refers to differentiation at constant pressure, and cp is a function of temperature only. The rearrangement of Eq. 11.7 leads to dhˇ = c dT p and hˇ 2 − hˇ 1 = ∫ T2 cp dT (11.8) T1 For moderate temperature changes, specific heat values can be considered constant. Equation 11.8 allows us to evaluate the change in enthalpy, hˇ 2 − hˇ 1, associated with ideal gas flow from section (1) to section (2). For simplicity, we can assume that cp is constant for a specific ideal gas and obtain from Eq. 11.8 hˇ 2 − hˇ 1 = cp (T2 − T1 ) (11.9) As is true for cυ, the value of cp for a given gas varies with temperature. Nevertheless, for moderate changes in temperature, the constant cp assumption is reasonable. 11.1 Ideal Gas Thermodynamics 607 From Eqs. 11.5 and 11.9 we see that changes in internal energy and enthalpy are related to changes in temperature. We turn our attention now to developing relationships for determining cυ and cp. Combining Eqs. 11.6 and 11.1 we get hˇ = uˇ + RT (11.10) Differentiating Eq. 11.10 with respect to temperature leads to dhˇ duˇ = +R dT dT (11.11) From Eqs. 11.3, 11.7, and 11.11 we conclude that cp − cυ = R (11.12) Equation 11.12 indicates that the difference between cp and cυ is constant for each ideal gas regardless of temperature. Also cp > cυ. The specific heat ratio, k, is defined as k= cp cυ (11.13) so combining Eqs. 11.12 and 11.13 leads to The four gas constants are inter-related. cp = kR k−1 (11.14) cυ = R k−1 (11.15) and Actually, cp, cυ, and k are all somewhat temperature dependent for any ideal gas. We will assume constant values for these variables in this book. Values of k and R for some commonly used gases at nominal temperatures are listed in Tables 1.7 and 1.8. These tabulated values can be used with Eqs. 11.13 and 11.14 to determine the values of cp and cυ. Example 11.1 demonstrates how internal energy and enthalpy changes can be calculated for a flowing ideal gas having constant cp and cυ. Density, pressure, temperature, internal energy, and enthalpy are all important properties that describe the state of an ideal gas but they do not form a complete set. One more property is necessary. This property is the entropy and is given the symbol s for entropy per unit mass. The units of s are J/kgK in SI, ft-lb/slug °R in BG and Btu/lbm °R in EE. We dealt briefly with entropy in Sec. 5.4 and used it to express the Second Law of Thermodynamics but we did not define it either mathematically or verbally. One useful verbal definition is as follows: Entropy is the property that does not change in any process which has no friction and no heat transfer. Such a process is said to be reversible (no friction) and adiabatic (no heat transfer). A shorter label for such a process is isentropic (iso means “same”). It follows that entropy does change if a process has friction or heat transfer, or both. It can be shown (Ref. 2) that a proper mathematical definition of entropy is s2 − s 1 = ∫ 1 Changes in entropy are important because they are related to loss of available energy. 2 δqreversible T (11.16) where “1” and “2” are the endpoints of a process, qreversible is the heat transfer per unit mass that would occur in a reversible (frictionless) process between “1” and “2”, and T is the absolute temperature during the process. Note that entropy is defined only in terms of its change in a process. It can also be shown (Ref. 2) that the Second Law of Thermodynamics can be expressed in mathematical form as Eq. 5.94 for a system and Eq. 5.97 for a control volume. Finally, it is shown in Sec. 5.4 that friction loss of available energy causes the entropy change in a flow process to exceed the actual heat transfer (δq) divided by the temperature ds ≥ δq . T (11.17) 608 Chapter 11 ■ Compressible Flow Because entropy is a property, it can be related to other fluid properties. For any pure substance, including ideal gases, the “first T ds equation” is (see Ref. 2) 1 T ds = duˇ + pd( ) ρ (11.18) where T is absolute temperature, s is entropy, uˇ is internal energy, p is absolute pressure, and ρ is density. Differentiating Eq. 11.6 leads to 1 1 dhˇ = duˇ + pd( ) + ( ) dp ρ ρ (11.19) By combining Eqs. 11.18 and 11.19, we obtain 1 T ds = dhˇ − ( ) dp ρ (11.20) Equation 11.20 is often referred to as the “second T ds equation.” For an ideal gas, Eqs. 11.1, 11.3, and 11.18 can be combined to yield ds = cυ dT R 1 + d( ) ρ T 1ρ (11.21) and Eqs. 11.1, 11.7, and 11.20 can be combined to yield Changes in entropy can be related to changes in temperature, pressure, and density. ds = cp dp dT −R p T (11.22) If cp and cυ are assumed to be constant for a given gas, Eqs. 11.19 and 11.20 can be integrated to get s2 − s1 = cυ ln ρ1 T2 + R ln ρ2 T1 (11.23) s2 − s1 = cp ln p2 T2 − R ln p T1 1 (11.24) and Equations 11.23 and 11.24 allow us to calculate the change of entropy of an ideal gas flowing from one section to another with constant specific heat values (cp and cυ). EXAMPLE 11.1 Internal Energy, Enthalpy, Density, and Entropy for an Ideal Gas GIVEN Air flows steadily between two sections in a long, straight portion of 4-in.-diameter pipe as is indicated in Fig. E11.1. The uniformly distributed temperature and pressure at each section are T1 = 540 °R, p1 = 100 psia, and T2 = 453 °R, p2 = 18.4 psia. Pipe Flow Control volume Section (1) Section (2) D D1 = D2 = 4 in. ■ Figure E11.1 SOLUTION FIND Calculate the (a) change in internal energy between sec- (a) Assuming air behaves as an ideal gas, we can use Eq. 11.5 to evaluate the change in internal energy between sections (1) and (2). Thus uˇ 2 − uˇ 1 = cυ (T2 − T1 ) (1) From Eq. 11.15 we have cυ = R k−1 tions (1) and (2), (b) change in enthalpy between sections (1) and (2), (c) change in density between sections (1) and (2), and (d) change in entropy between sections (1) and (2). (2) and from Table 1.7, R = 1716 (ft ∙ lb)(slug ∙ °R) and k = 1.4. Throughout this book, we use the nominal values of k for common gases listed in Tables 1.7 and 1.8 and consider these values as being representative. Since English Engineering System units 11.1 are used more often than British Gravitational System units in compressible flow discussions, we use 1 slug = 32 .174 lbm Ideal Gas Thermodynamics (d) We can calculate the entropy change between sections by using either Eq. 11.23 or Eq. 11.24. We use both to demonstrate that the same result is obtained either way. From Eq. 11.23, to get 1716 (ft ∙ lb)(slug ∙ °R) = 53.3 (ft ∙ lb)(lbm ∙ °R) R= 32.174 (lbm/slug) From Eq. 2 we obtain 53.3 (ft ∙ lb)(lbm ∙ °R) (1.4 − 1) = 133 (ft ∙ lb)(lbm ∙ °R) 609 s2 − s1 = cυ ln ρ1 T2 + R ln ρ2 T1 (7) To evaluate s2 − s1 from Eq. 7 we need the density ratio, ρ1ρ2, which can be obtained from the ideal gas equation of state (Eq. 11.1) as cυ = ρ1 p1 T2 = ρ2 ( T1 )( p2 ) (3) (8) and thus from Eqs. 7 and 8, Combining Eqs. 1 and 3 yields uˇ 2 − uˇ 1 = cυ (T2 − T1 ) = 133 (ft ∙ lb)(lbm ∙ °R) × (453 °R − 540 °R) = −11,600 ft ∙ lblbm (Ans) Or, if Btu are wanted as energy units, we note that s2 − s1 = cυ ln p1 T2 T2 + R ln [( )( )] p2 T1 T1 (9) By substituting values already identified in the problem statement into Eq. 9 with 100 psia p1 T2 453 °R = = 4.56 ( T1 )( p2 ) ( 540 °R )( 18.4 psia ) 778 ft ∙ lb = 1 Btu and so we get 11,600 (ft ∙ lb)lbm = −14.9 Btu/lbm uˇ 2 − uˇ 1 = − 778 (ft ∙ lb)Btu 453 °R 540 °R ) + [53.3 (ft ∙ lb)(lbm ∙ °R)] ln 4.56 s2 − s1 = [133 (ft ∙ lb)(lbm ∙ °R)] ln( (b) For enthalpy change we use Eq. 11.9. Thus hˇ 2 − hˇ 1 = cp (T2 − T1 ) (4) or s2 − s1 = 57.5 (ft ∙ lb)(lbm ∙ °R) where since k = cp cυ we obtain cp = kcυ = (1.4)[133 (ft ∙ lb)(lbm ∙ °R)] = 186 (ft ∙ lb)(lbm ∙ °R) From Eq. 11.24, (5) s2 − s1 = cp ln From Eqs. 4 and 5 we obtain hˇ 2 − hˇ 1 = cp (T2 − T1 ) = 186 (ft ∙ lb)(lbm ∙ °R) × (453 °R − 540 °R) = −16,200 ft ∙ lblbm (Ans) (c) For density change we use the ideal gas equation of state (Eq. 11.1) to get p2 p1 p1 1 p2 ρ2 − ρ 1 = − = − ) ( RT2 RT1 R T2 T1 (10) 453 °R 540 °R ) 18.4 psia − [53.3 (ft ∙ lb)(lbm ∙ °R)] ln ( 100 psia ) s2 − s1 = [186 (ft ∙ lb)(lbm ∙ °R)] ln( or s2 − s1 = 57.5 (ft ∙ lb)(lbm ∙ °R) (Ans) COMMENT There is a significant change in density when compared with the upstream density (100 psia) (144 in.2ft2 ) p1 = RT1 [53.3 (ft ∙ lb)(lbm ∙ °R)](540 °R) = 0.499 lbmft3 1 53.3 (ft ∙ lb)(lbm ∙ °R) (18.4 psia) (144 in.2ft2 ) ×[ 453 °R (100 psia) (144 in.2ft2 ) − ] 540 °R ρ1 = As anticipated, both Eqs. 11.23 and 11.24 yield the same result for the entropy change, s2 − s1. The entropy change is positive although the change of all other properties is negative. Note that, both the pressures and temperatures used must be absolute. or ρ2 − ρ1 = −0.389 lbmft3 p2 T2 − R ln p1 T1 By substituting known values into Eq. 10 we obtain (6) Using the pressures and temperatures given in the problem statement we calculate from Eq. 6 ρ2 − ρ 1 = (Ans) (Ans) 610 Chapter 11 ■ Compressible Flow If internal energy, enthalpy, and entropy changes for ideal gas flow with variable specific heats are desired, Eqs. 11.4, 11.8, and 11.21 or 11.22 must be used as explained in Ref. 2. Detailed tables (see, for example, Ref. 3) are available for variable specific heat calculations. For the isentropic flow of an ideal gas with constant cp and cυ, we get from Eqs. 11.23 and 11.24 cυ ln ρ1 p2 T2 T2 + R ln = cp ln − R ln = 0 ρ2 p1 T1 T1 (11.25) By combining Eq. 11.25 with Eqs. 11.14 and 11.15 we obtain ρ2 k p2 T2 k(k−1) = ( T1 ) ( ρ1 ) = ( p1 ) , (11.26) a useful relationship between temperature, density, and pressure for an ideal gas undergoing an isentropic process. From Eq. 11.26 we can conclude that p ρk = constant (11.27) for an ideal gas with constant cp and cυ flowing isentropically, a result already used without proof earlier in Chapters 1, 3, and 5. THE WIDE WORLD OF FLUIDS Hilsch tube (Ranque vortex tube) Years ago (around 1930) a portion of the swirling air flowed out the opposite end of the French physics student (George Ranque) discovered that appre- tube. Rudolph Hilsch, a German physicist, improved on this dis- ciably warmer and colder portions of rapidly swirling airflow could covery (ca. 1947). Hot air temperatures of 260 °F (127 °C) and cold be separated in a simple apparatus consisting of a tube open at air temperatures of −50 °F (−46 °C) have been claimed in an both ends into which was introduced, somewhere in between optimized version of this apparatus. Thus far the inefficiency of the two openings, swirling air at high pressure. Warmer air near the process has prevented it from being widely adopted for the outer portion of the swirling air flowed out one open end practical use. (See Problem 11.3LL.) of the tube through a simple valve and colder air near the inner 11.2 Stagnation Properties 1 V1 p1, T1, s1, etc. z1 The thermodynamic state of a fluid particle is defined by its properties (p, ρ, T, uˇ , hˇ , sˇ ); but we must also know the velocity of the particle and, possibly, its position in a gravitational field. The thermodynamic properties are called static properties; they are the values that would be measured by instruments that are static with respect to the fluid. The static properties represent the molecular structure of the fluid and obey all equations of state and other property-related laws and thermodynamic equations. The particle’s velocity and elevation are specified separately. A useful approach is to combine the (static) thermodynamic properties with the velocity and elevation to obtain equivalent thermodynamic properties that represent the total (thermodynamic and mechanical) state. This is done by using stagnation properties which are the properties that the fluid would obtain if it were brought to a condition of zero velocity and zero elevation in a frictionless process with no heat transfer and no work. We can obtain equations for stagnation properties with the aid of the figure in the margin which shows a (possibly imaginary) process that brings a fluid particle to rest and zero elevation. Applying the energy pO, TO, s O, etc. equation to the streamline between 1 and 0 gives z =0 O VO = 0 0 V 20 V 21 q + ws = hˇ 0 − hˇ 1 + − + gz0 − gz1. 2 2 11.2 where q= Stagnation Properties 611 ˙ shaft net in ˙ net in W Q and ws = m ˙ m ˙ By definition, q, ws, V0, and z0 are all zero, so we get V 21 hˇ 0 = hˇ 1 + + gz1, 2 where hˇ 0 is called the stagnation enthalpy. In compressible flow, potential energy and gravity force are almost always negligible, and the definition of stagnation enthalpy is usually simplified to V2 hˇ 0 ≡ hˇ + . 2 (11.28) The subscript 1 has been dropped to obtain a general definition of stagnation enthalpy. Note that we did not need to use the specification of frictionless flow to obtain the stagnation enthalpy. If the fluid is an ideal gas with constant specific heat, Eq. 11.28 becomes cpT0 = cpT + V2 , 2 and the stagnation temperature is given by T0 = T + V2 . 2cp (11.29) We can obtain equations for the stagnation pressure and stagnation density by noting that, by definition, the stagnation process is isentropic (frictionless with no heat transfer). For an ideal gas, T0 k(k−1) p0 =( p T) (11.30) T0 1/(k−1) ρ0 =( . ρ T) (11.31) T0 V2 =1+ . T 2cpT (11.32) and Dividing Eq. (11.29) by T, we obtain Substituting into Eqs.11.30 and 11.31, we have V 2 k(k−1) p0 = p(1 + 2cpT ) (11.33) V 2 1(k−1) ρ0 = ρ(1 + . 2cpT ) (11.34) and Equations (11.28)–(11.34) show that the stagnation properties are determined by the static properties and the fluid velocity. Because any fluid particle always has a pressure, a temperature, and a velocity 612 Chapter 11 ■ Compressible Flow (including V = 0), any fluid particle can be described by both its static properties and its stagnation properties. We may say that “the fluid has a pressure of 100 kPa and a stagnation pressure of 125 kPa.” The static properties actually are the properties of the fluid, and the stagnation properties are the properties that the fluid would have if it were brought to zero velocity in an isentropic process with no work. If (and only if) the fluid particle has zero velocity, its static and stagnation properties are equal. Stagnation properties represent the maximum possible values of the fluid properties unless energy is added to the fluid by heat transfer or work. The postulated stagnation process represents the ideal conversion of the fluid’s kinetic and potential energies into pressure and internal energy. Stagnation properties are sometimes called total properties. Because the process of stagnation (that is, bringing the fluid to rest) can be accomplished by process with friction as well as by a frictionless process, there is some possible ambiguity in using the term stagnation in cases where viscous flow processes are considered. This ambiguity can also be remedied by use of the term “isentropic stagnation properties.” 11.3 Mach Number and Speed of Sound The Mach number, Ma, was introduced in Chapters 1 and 7 as a dimensionless measure of compressibility in a fluid flow. In this and subsequent sections, we develop some useful relationships involving the Mach number. The Mach number is defined as the ratio of the local flow velocity, V, to the local speed of sound, c. Mach number is the ratio of local flow and sound speeds. The changes in fluid properties across a sound wave are extremely small compared to their local values. Ma = V c What we perceive as sound generally consists of weak pressure pulses that move through air with a Mach number of one. When our eardrums respond to a succession of these pulses, we hear sounds. To derive an equation for the speed of sound, we analyze the one-dimensional fluid mechanics of an infinitesimally thin, weak pressure pulse moving through a fluid at rest (see Fig. 11.1a). Ahead of the pressure pulse, the fluid velocity is zero, and the fluid pressure and density are p and ρ. Behind the pressure pulse, the fluid velocity has changed by an amount δV, and the pressure and density of the fluid have also changed by amounts δp and δρ. We select an infinitesimally thin control volume that moves with the pressure pulse as is sketched in Fig. 11.1a. The speed of the weak pressure pulse is considered constant and in one direction only; thus, our control volume is inertial. For an observer moving with this control volume (Fig. 11.1b), it appears as if fluid is entering the control volume through surface area A with speed c at pressure p and density ρ and leaving the control volume through surface area A with speed c − δV, pressure p + δp, and density ρ + δρ. When the continuity equation (Eq. 5.16) is applied to the flow through this control volume, the result is ρAc = (ρ + δρ)A(c − δV) (11.35) ρc = ρc − ρ δV + c δρ − (δρ)(δV) (11.36) or Weak pressure pulse Control volume p + δρ p ρ c ρ + δρ V=0 Weak pressure pulse Control volume p p + δρ ρ ρ + δρ c c – δV δV A A (a) A A (b) ■ Figure 11.1 (a) Weak pressure pulse moving through a fluid at rest. (b) The flow relative to a control volume containing a weak pressure pulse. 11.3 Mach Number and Speed of Sound 613 Since (δρ)(δV) is much smaller than the other terms in Eq. 11.36, we drop it from further consideration and keep ρ δV = c δρ (11.37) The linear momentum equation (Eq. 5.29) can also be applied to the flow through the control volume of Fig. 11.1b. The result is −cρcA + (c − δV)(ρ + δρ)(c − δV)A = pA − (p + δp)A (11.38) Frictional forces are negligibly small because the edges of the control volume have no area. We again neglect higher order terms such as (δV) 2 compared to c δV and combine Eqs. 11.35 and 11.38 to get −cρcA + (c − δV)ρAc = −δpA or ρδV = δp c (11.39) From Eqs. 11.37 (continuity) and 11.39 (linear momentum) we obtain c2 = δp δρ or c= δp B δρ (11.40) This expression for the speed of sound results from application of the conservation of mass and linear momentum principles to the flow through the control volume of Fig. 11.1b. These principles were similarly used in Section 10.2.1 to obtain an expression for the speed of waves traveling on the surface of a liquid in a channel. The conservation of energy principle can also be applied to the flow through the control volume of Fig. 11.1b. If the energy equation (Eq. 5.103) is used for the flow through this control volume, the result is δp V2 + δ( + g δz = δ(loss) ρ 2) (11.41) For gas flow we can consider g δz as being negligibly small in comparison to the other terms in the equation. Also, if we assume that the flow is frictionless, then δ(loss) = 0 and Eq. 11.41 becomes δp (c − δV) 2 c2 + − =0 ρ 2 2 or, neglecting (δV) 2 compared to c δV, we obtain ρ δV = c= 4000 0 Air Hydrogen Helium 1000 δp B δρ which is identical to Eq. 11.40. Thus, the linear momentum and conservation of energy principles lead to the same result. If we further assume that the frictionless flow through the control volume of Fig. 11.1b is adiabatic (no heat transfer), then the flow is isentropic. In the limit, as δp and δρ become vanishingly small (δp → δp → 0) 3000 2000 (11.42) By combining Eqs. 11.37 (continuity) and 11.42 (energy) we again find that 5000 R, J/kg⋅K δp c c= ∂p B ( ∂ρ )s (11.43) where the subscript s is used to designate that the partial differentiation occurs at constant entropy. 614 Chapter 11 ■ Compressible Flow Speed of sound is larger in fluids that are more difficult to compress. The bulk modulus of elasticity, Eυ, of any fluid including liquids is defined as (see Section 1.7.1) Eυ = dp ∂p = ρ( ) dρρ ∂ρ s (11.44) Thus, in general, from Eqs. 11.43 and 11.44, c= 2.0 Eυ Bρ (11.45) Equation 11.43 informs us that we can calculate the speed of sound by determining the partial derivative of pressure with respect to density at constant entropy. For the isentropic flow of an ideal gas (with constant cp and cυ), we learned earlier (Eq. 11.27) that Helium p = (constant) ρk Air, hydrogen k 1.0 and thus 0 6000 c, ft/s 4000 0 Air 0 (11.46) c = √kRT (11.47) Thus, for an ideal gas Water 2000 p ∂p p k−1 k−1 = (constant) kρ = kρ = k = kRT k ( ∂ρ )s ρ ρ 100 200 T, deg F From Eq. 11.47 and the charts in the margin we conclude that for a given temperature, the speed of sound, c, in hydrogen and in helium, is higher than in air. Values of the speed of sound are tabulated in Tables B.1 and B.2 for water and in Tables B.3 and B.4 for air. From experience we know that air is more easily compressed than water. Note from the values of c in Tables B.1 through B.4 and the graph in the margin that the speed of sound in air is much less than it is in water. From Eq. 11.44, we can conclude that if a fluid is truly incompressible, its bulk modulus would be infinitely large, as would be the speed of sound in that fluid. Thus, an incompressible flow is always an approximation of reality. THE WIDE WORLD OF FLUIDS Sonification The normal human ear is capable of detecting the representation of data in the form of sound. With this even very subtle sound patterns produced by sound waves. emerging technology, pathologists may soon learn to “hear” Most of us can distinguish the bark of a dog from the meow of a abnormalities in tissue samples, engineers may “hear” flaws in cat or the roar of a lion, or identify a person’s voice on the tele- gas turbine engine blades being inspected, and scientists may phone before they identify who is calling. The number of “hear” a desired attribute in a newly invented material. Perhaps “things” we can identify from subtle sound patterns is enor- the concept of hearing the trends in data sets may become as mous. Combine this ability with the power of computers to commonplace as seeing them. Analysts may listen to the stock transform the information from sensor transducers into varia- market and make decisions. Of course, none of this can happen tions in pitch, rhythm, and volume and you have sonification, in a vacuum. If the Mach number is large, the effects of fluid compressibility are significant, but if it is small, compressibility effects are negligible. If we square the Mach number and assume an ideal gas, we get Ma2 = 1.0 (11.48) The criteria for neglecting fluid compressibility suggested in Sections 1.7.2 and 3.8.1 are equivalent to specifying that the Mach number is small. The importance of Mach number as a parameter in compressible flow can be illustrated by considering the ideal gas equations for stagnation properties. The stagnation-to-static-temperature ratio, Eq. (11.29), gives T __ T0 0.5 0.0 0.1 ρV 2 V2 = . kRT kp 1.0 Ma 10 T0 V2 =1+ . T 2cpT 11.3 1.0 Multiplying and dividing the second term on the right by kR and using Eqs. (11.12), (11.47), and 11.48, we obtain p __ 0.5 p0 0.0 0.1 1.0 10.0 V2 k − 1 V2 k−1 = =( Ma2, 2) ( 2cpT 2 2 ) c (11.49) T0 k−1 =1+ Ma2. T 2 (11.50) so Ma 1.0 Using Eqs. (11.33) and (11.34), we get ρ __ ρ 0 0.5 0.0 0.1 615 Mach Number and Speed of Sound 1.0 Ma For isentropic flows the temperature, pressure, and density ratios are functions of the Mach number. 10.0 k(k−1) p0 k−1 = (1 + Ma2) p 2 (11.51) 1(k−1) ρ0 k−1 = (1 + Ma2) . ρ 2 (11.52) and The figures in the margin show the variation of these ratios with Mach number for a gas with k = 1.4 (the value for air). Note that the ratios are graphed in the form (static property)/(stagnation property). The graphs show that pressure is most strongly affected by compressibility and by temperature the least. Because of the exponential form of Eqs.11.51 and 11.52, calculations involving these equations are tedious. Traditionally, numerical values of the property ratios as functions of Mach number are evaluated using graphs or tables. Appendix D presents both graphs and brief tables for a gas with k = 1.4. p When using graphs or tables, an issue with notation arises. The expression might refer to p0 k − k−1 k−1 the ratio of two distinct quantities, p and p0, or to the function (1 + Ma2) or to a 2 single numerical value from a graph, table (Appendix D), or software. In this book we shall resolve the dilemma as follows: When solving problems, ■ p will refer to either the concept “static-to-stagnation pressure ratio” or to the ratio of two p0 p 6 distinct quantities (if p = 6 and p0 = 10, then = ); p0 10 k ■ − p p k−1 k−1 [Ma] will refer to the Mach number function ( [Ma] = (1 + Ma2) ); p0 p0 2 ■ p will refer to a single number found from a graph, table, or software for argument “z” p0 )z p = 0.528. p0 )Ma=1 T A and . Note that T/T0 and pp0 T0 A* (numbers less than 1.0) are tabulated and graphed with Ma as a independent parameter in Appendix D. There is no column or graph for density ratio because This notation will apply to other ratios such as ρ p T = . ρ0 ( p0 )/( T0 ) 616 Chapter 11 ■ Compressible Flow The remaining variable (A/A*) will be explained later. Even use of the tables can be cumbersome because a different set of tables must be prepared for each different value of k. Fortunately, any number of compressible flow function calculators can be found on the Internet or as a smartphone app. EXAM PLE 11.2 Mach Number, Stagnation Properties, and Mach Number Tables GIVEN An airliner cruises at 250 m/s at an altitude of 10km above sea level. FIND (a) Calculate the airliner’s Mach number and the stagna- V tion pressure and temperature relative to the airliner. (b) compute the same quantities for an airliner speed of 500 m/s. (c) compute the same quantities for a speed of 250 m/s but change the altitude to 1 km above sea level. ■ Figure E11.2 An airliner. SOLUTION (a) We assume that the atmospheric conditions are those of Standard Atmosphere. The atmospheric properties can be calculated from Eqs. (2.7)–(2.9) or read from Fig. 2.6 or Table C.2: p10 km = 26.50 kPa and T10 km = 223.3K; p1 km = 89.88 kPa and T1 km = 281.7K. 223.3K ; 0.878 T0 = 254.3K at 10 km and 250 m/s. T0 = V . c V 500 m/s ; = c 299.6 m/s Ma = 1.669 at 10 km and 500 m/s. Ma = From Eq. (11.47), c = √kRT = 20.05√T(K) m/s. c = 20.05√223.3 m/s = 299.6 m/s. p T = 0.212 and  ) = 0.642. p0 )Ma=1.669 T0 Ma=1.669 Thus when V = 250 m/s, Calculating p0 and T0, we have Ma = p T = 0.633 and  ) = 0.878. p0 )Ma=0.834 T0 Ma=0.834 26.50 kPa , 0.212 or p0 = 125.0 kPa at 10 km and 500 m/s; 223.3K , 0.642K or T0 = 347.8K at 10 km and 500 m/s. p10 km T10 km  and T0 = . p/p0 ) T/T0 ) (Ans) and T0 = The static properties are those of the atmosphere. The stagnation properties relative to the airliner are (Ans) (c) Finally, consider 250 m/s at 1 km. The speed of sound is now Substituting, we have 26.50 kPa ; 0.633 p0 = 41.86 kPa at 10 km and 250 m/s. p0 = (Ans) The airliner is cruising at subsonic speed. From Table D.I, the pressure and temperature ratios at this Mach number are p0 = (Ans) This is a supersonic cruise condition (Ma > 1). At this Mach number, the pressure and temperature ratios are At 10 km, 250 m/s ; 299.6 m/s Ma = 0.834 at 10 km and 250 m/s. (Ans) (b) Next, consider 500 m/s at 10 km. The speed of sound and the static pressure and temperature are still the values appropriate to the 10-km altitude. The Mach number is The Mach number is Ma = Then c = √kRT = 20.05√281.7K m/s = 336.5 m/s. p0 = (Ans) 11.4 617 Compressible Flow Regimes and The Mach number is V 250 m/s ; = c 336.5 m/s Ma = 0.743 at 1 km and 250 m/s. T0 = T 1 + ( Ma = (Ans) Again, this is a subsonic cruise condition. To calculate the stagnation pressure and temperature, we must use the atmospheric properties at 1 km. For some variety we use Eqs. (11.50) and (11.51). They give k−1 Ma2 = 281.7K[1 + 0.2(0.743) 2], ) 2 so T0 = 312.8K at 1 km and 250 m/s. (Ans) COMMENTS The Mach number changed with altitude, even for the same velocity. The reason is that the temperature and, hence, the speed of sound changed. The stagnation properties k−1 p0 = p 1 + Ma2) = 89.88 kPa [1 + 0.2(0.743) 2]3.5, changed with speed and altitude. A pocket calculator or computer ( 2 is often easier to use than to look up and interpolate values in a so table. Calculating the temperature ratio first is usually best, p0 = 129.7 kPa at 1 km and 250 m/s; (Ans) because this ratio is then raised to various powers to obtain pressure and density ratios. For rapid, though imprecise, calculations, use Graph D.I. k/(k−1) 11.4 Compressible Flow Regimes Experience has demonstrated that compressibility can have a large influence on important flow variables. For example, in Fig. 11.2 the variation of drag coefficient with Reynolds number and Mach number is shown for airflow over a sphere. It is clear from this figure that any effects of compressibility become more pronounced as the Mach number increases. The discussion in Sec. 3.8.1 indicates that compressibility effects increase as the fluid velocity (and hence Mach number) increases. Clearly compressibility effects can be of considerable importance. 0.8 Ma = 1.0 Ma = 0.9 0.7 0.6 0.5 Ma = 0.6 CD 0.4 0.3 Ma = 0.3 0.2 0.1 0 2 3 4 5 6 7 8 × 105 Re ■ Figure 11.2 The variation of the drag coefficient of a sphere with Reynolds number and Mach number. 618 Chapter 11 ■ Compressible Flow VIDEO To further illustrate some curious features of compressible flow, a simplified example is considered. Imagine the emission of weak pressure pulses from a point source. These pressure waves are spherical and expand radially outward from the point source at the speed of sound, c. If a pressure wave is emitted at different times, twave, we can determine where several waves will be at a common instant of time, t, by using the relationship V11.2 Waves on a water surface r = (t − twave )c where r is the radius of the sphere-shaped wave emitted at time = twave. For a stationary point source, the symmetrical wave pattern shown in Fig. 11.3a results. When the point source moves to the left with a constant velocity, V, the wave pattern is no longer symmetrical. In Figs. 11.3b, 11.3c, and 11.3d are illustrated the wave patterns at t = 3 s for different values of V. Also shown with a “+” are the positions of the moving point source at values of time, t, equal to 0 s, 1 s, 2 s, and 3 s. Knowing where the point source has been at different instants is important because it indicates to us where the different waves originated. From the pressure wave patterns of Fig. 11.3, we can draw some useful conclusions. Before doing this we should recognize that if instead of moving the point source to the left, we held the point source stationary and moved the fluid to the right with velocity V, the resulting pressure wave patterns would be identical to those indicated in Fig. 11.3. When the point source and the fluid are stationary, the pressure wave pattern is symmetrical (Fig. 11.3a) and an observer anywhere in the pressure field would hear the same sound frequency The wave pattern from a moving source is not symmetrical. 3c 2c c c 3c V 2c 2V 3V (b) (a) Zone of silence Zone of action Tangent plane (Mach wave) c 2c 3c 3c Zone of silence 2c c α V=c Zone of action 2V = 2c Mach cone 3V = 3c V 3V 2V (d) (c) Wave emitted at t = 0 s Wave emitted at t = 1 s Source at t = 0 s Source at t = 1, 2, or 3 s ■ Figure 11.3 (a) Pressure waves at t = 3 s, V = 0; (b) pressure waves at t = 3 s, V < c; (c) pressure waves at t = 3 s, V = c; (d) pressure waves at t = 3 s, V > c. Wave emitted at t = 2 s 11.4 V11.3 Jet noise VIDEO VIDEO V11.4 Speed boat from the point source. When the velocity of the point source (or the fluid) is very small in comparison with the speed of sound, the pressure wave pattern will still be nearly symmetrical. The speed of sound in an incompressible fluid is infinitely large. Thus, the stationary point source and stationary fluid situation is representative of incompressible flows. For truly incompressible flows, the communication of pressure information throughout the flow field is unrestricted and instantaneous (c = ∞). When the point source moves in fluid at rest (or when fluid moves past a stationary point source), the pressure wave patterns vary in asymmetry, with the extent of asymmetry depending on the ratio of the point source (or fluid) velocity and the speed of sound. When Vc < 1, the wave pattern is similar to the one shown in Fig. 11.3b. This flow is subsonic and compressible. A stationary observer will hear a higher frequency sound ahead of the source and a lower frequency behind it because the wave pattern is asymmetrical. We call this phenomenon the Doppler effect. Pressure information can still travel unrestricted throughout the flow field, but not symmetrically or instantaneously. When Vc = 1, pressure waves are not present ahead of the moving point source. The flow is sonic. If you were positioned to the left of the moving point source, you would not hear the point source until it was coincident with your location. For flow moving past a stationary point source at the speed of sound (Vc = 1), the pressure waves are all tangent to a plane that is perpendicular to the flow and that passes through the point source. The concentration of pressure waves in this tangent plane suggests the formation of an abrupt pressure variation across the plane. This plane is often called a Mach wave. Note that communication of pressure information is restricted to the region of flow downstream of the Mach wave. The region of flow upstream of the Mach wave is called the zone of silence, and the region of flow downstream of the tangent plane is called the zone of action. When V > c, the flow is supersonic, and the pressure wave pattern resembles the one depicted in Fig. 11.3d. A cone (Mach cone) that is tangent to the pressure waves can be constructed to represent the Mach wave that separates the zone of silence from the zone of action. The communication of pressure information is restricted to the zone of action. From the sketch of Fig. 11.3d, we can see that the angle of this cone, α, is given by sin α = 90 α α 60 30 0 1 1.5 2 Ma 2.5 3 619 Compressible Flow Regimes c 1 = V Ma (11.53) This relationship between Mach number, Ma, and Mach cone angle, α, shown by the figure in the margin, is valid for Vc > 1 only. The concentration of pressure waves at the surface of the Mach cone suggests an abrupt (but still infinitesimal) pressure and density variation across the cone surface. An abrupt density change can be visualized in a flow field by using special optical techniques. Examples of flow visualization methods include the schlieren, shadowgraph, and interferometer techniques (see Ref. 4). A schlieren photo of a flow for which V > c is shown in Fig. 11.4. The airflow through the row of compressor blade airfoils is as shown with the arrow. The flow enters supersonically (Ma1 = 1.14) and leaves subsonically (Ma2 = 0.86). The center two airfoils have pressure tap hoses connected to them. Regions of significant changes in fluid density appear in the supersonic portion of the flow. Also, the region of separated flow on each airfoil is visible. It should ■ Figure 11.4 The schlieren visualization of flow (supersonic to subsonic) through a row of compressor airfoils. (Photograph courtesy of TAGT-1.2 Cascade, R. Fuchs, H. A. Schreiber, W. Steinert, German Aerospace Center (DLR), Institute of Propulsion Technology, 1996.) 620 VIDEO Chapter 11 ■ Compressible Flow V11.5 Compressible flow visualization be noted that waves seen in the photograph are not exactly infinitesimal Mach waves but weak finite strength shock waves generated by the finite-size airfoils. This discussion about pressure wave patterns suggests the following categories of fluid flow: 1. Incompressible flow: Ma ≤ 0.3. Unrestricted, nearly symmetrical and instantaneous pressure communication. 2. Compressible subsonic flow: 0.3 < Ma < 1.0. Unrestricted but noticeably asymmetrical pressure communication. 3. Compressible supersonic flow: Ma ≥ 1.0. Formation of Mach wave; pressure communication restricted to zone of action. In addition to the above-mentioned categories of flows, two other regimes are commonly referred to: namely, transonic flows (0.9 ≤ Ma ≤ 1.2) and hypersonic flows (Ma > 5). Modern aircraft are mainly powered by gas turbine engines with transonic internal flows. When a spacecraft reenters the Earth’s atmosphere, the flow is hypersonic. Future aircraft may be expected to operate from subsonic to hypersonic speeds. THE WIDE WORLD OF FLUIDS Supersonic and compressible flows in gas turbines Modern shocks and seriously affect blade aerodynamic and structural gas turbine engines commonly involve compressor and turbine performance. It is possible to have supersonic flows past blades blades that are moving so fast that the fluid flows over the near the outer diameter of a rotor with subsonic flows near the blades are locally supersonic (See Fig. 11.4). Density varies con- inner diameter of the same rotor. These rotors are considered to siderably in these flows so they are also considered to be com- be transonic in their operation. Very large aero gas turbines can pressible. Shock waves can form when these supersonic flows involve thrust levels exceeding 100,000 lb. Two of these engines are sufficiently decelerated. Shocks formed at blade leading are sufficient to carry over 350 passengers halfway around the edges or on blade surfaces can interact with other blades and world at high subsonic speed. (See Problem 11.4LL) EXAM PLE 11.3 Mach Cone GIVEN An aircraft cruising at 1000-m elevation, z, above you moves past in a flyby. It is moving with a Mach number equal to 1.5 and the ambient temperature is 20 °C. FIND How many seconds after the plane passes overhead do you expect to wait before you hear the aircraft? SOLUTION Since the aircraft is moving supersonically (Ma > 1), we can imagine a Mach cone originating from the forward tip of the craft as is illustrated in Fig. E11.3a. A photograph of this phenomenon is shown in Fig. E11.3b. When the surface of the cone reaches the observer, the “sound” of the aircraft is perceived. The angle α in Fig. E11.4 is related to the elevation of the plane, z, and the ground distance, x, by α = tan −1 z 1000 = tan −1 x Vt (1) Also, assuming negligible change of Mach number with elevation, we can use Eq. 11.53 to relate Mach number to the angle α. Thus, Ma = 1 sin α Mach cone Aircraft moving with velocity V and Mach number Ma z α x = Vt (2) ■ Figure E11.3a 11.5 Shock Waves 621 or t = 2.17 s (Ans) COMMENT By repeating the calculations for various values of Mach number, Ma, the results shown in Fig. E11.3c are obtained. Note that for subsonic flight (Ma < 1) there is no delay since the sound travels faster than the aircraft. You can hear a subsonic aircraft approaching. 3 2.5 ■ Figure E11.3b NASA Schlieren photograph of shock waves from a T-38 aircraft at Mach 1.1, 13,000 ft. (Photo courtesy of NASA) t, s Combining Eqs. 1 and 2 we obtain 1 Ma = −1 sin [tan (1000Vt)] (3) (4) where c is the speed of sound. From Table B.4, c = 343.3 ms. Using Ma = 1.5, we get from Eqs. 3 and 4 1.5 = 1 1000 m sin{tan−1[ (1.5) (343.3 ms)t ]} 1.5 1 The speed of the aircraft can be related to the Mach number with V = (Ma)c (1.5, 2.17 s) 2 0.5 0 0 0.5 1 1.5 2 Ma 2.5 3 3.5 4 ■ Figure E11.3c The answer above is only an estimate. In reality, the wave emanating from the aircraft will be a finite strength shock wave and will have a shallower angle. You would hear the sound of the aircraft as a “sonic boom.” 11.5 Shock Waves Oblique shock in supersonic flow over a right circular cone. (Courtesy of Launch and Flight Division. Ballistic Research Laboratory/ ARRADCOM, Aberdeen Proving Ground, Md.) In the preceding section, we saw that introducing an infinitesimal disturbance into a supersonic flow generates a stationary (with respect to the disturbance) Mach wave. Placing a finite disturbance (say, a sharp wedge or cone) in a steady supersonic flow generates a finite-strength wave as shown in the photographs in the margin. Finite-strength disturbance waves are called shock waves. Shock waves are extremely thin, and fluid properties and velocity change by large amounts (tens or even hundreds of percent) across a shock wave. Shock waves are actually rather common occurrences; examples of shock waves include explosion waves from an extra-powerful Fourth of July firecracker, sonic “booms,” which are actually the shock waves generated by supersonic aircraft, and water hammer, which can occur when a water faucet or valve is opened or closed rapidly. Shock waves may be moving, like the wave that results from rapidly closing a valve in a pipe, or stationary, as in flow over a stationary wedge or blunt body. In either case, a shock wave adjusts the flow to some downstream condition. A normal shock is perpendicular to the upstream flow. An oblique shock is inclined at a constant angle to the upstream flow. A curved shock has a variable angle between the wave and the upstream flow. Flows that contain oblique or curved shocks are at least two-dimensional, because the shock changes the direction of the flow. One-dimensional flow can contain only normal shocks. For the sake of simplicity, our consideration is limited to normal shocks. For the same reason, we consider shocks only in an ideal gas. 622 Chapter 11 ■ Compressible Flow ρy py Ty ρx px Tx Fluid velocity Vx Shock velocity ■ Figure 11.5 Normal shock in a duct: (a) flow schematic; (b) flow view from control volume fixed to shock. Vy Thickness ≈ 0 Control volume (a) x (b) y 11.5.1 Normal Shock Figure 11.5(a) shows a normal shock in a duct. The shock may be moving or stationary with respect to the duct. We develop analysis that applies to either case by using a control volume fixed to the shock. Velocities and stagnation properties are defined relative to the shock. Figure 11.5(b) shows the flow into and out of the control volume. We assume that the velocity is uniform. We assume that the shock has zero thickness, so Ax = Ay = A Normal shock waves are assumed to be infinitesimally thin discontinuities. and the volume of the control volume is zero. The continuity equation is ρx AVx = ρy AVy. (11.54) From the ideal gas law, ρ= p . RT (11.55) We can write the velocity as V = Ma(c) = Ma√kRT. (11.56) Substituting into Eq. 11.54 and cancelling A, k, and R, we obtain pxMax √Tx Curved shock standing ahead of a blunt body in supersonic flow. (Courtesy of Launch and Flight Division, Ballistic Research Laboratory/ ARRADCOM, Aberdeen Proving Ground, Md.) = pyMay √Ty , or May Max = px Ty . p y B Tx (11.57) The linear momentum equation for the flow direction x is ∑ F = ρyV 2y A − ρxV 2x A. Because the sides of the control volume in contact with the duct wall have zero thickness, there is no force on them. The only forces are from pressure on the inlet and outlet planes, so the equation becomes pxA − py A = ρyV 2y A − ρxV 2x A. From Eq. (11.48), ρV 2 = kpMa2, and the momentum equation reduces to VIDEO px (1 + kMa 2x ) = py (1 + kMa 2y ), V11.8 Blast waves     or py px = 1 + kMa 2x . 1 + kMa 2y     (11.58) 11.5 Shock Waves 623 Next, we apply the general energy equation, ˙ +W ˙ s = m˙ hˇ y − hˇ x + Q ( V 2y    2 − V 2x . 2)    Because the control volume has zero thickness, there can be neither heat transfer nor work, so the energy equation reduces to hˇ y + V 2y V 2x = hˇ x + , 2 2 or, alternatively, in terms of stagnation enthalpy, hˇ 0y = hˇ 0x. (11.59) For an ideal gas, with constant cp the energy equation becomes T0y = T0x. (11.60) Substituting Eq. (11.50) into Eq. (11.60), we have Ty (1 + k−1 k−1 Ma2y ) = Tx(1 + Ma2x ), 2 2 or Ty Tx = 1 + [(k − 1)2]Ma2x . 1 + [(k − 1)2]Ma2y (11.61) Equations (11.57), (11.58), and (11.61) relate the ratio of temperature and pressure across a shock to the Mach numbers upstream and downstream from the shock. If we select a single parameter, say, upstream Mach number Max, we can express the other three parameters in terms of the selected parameter. To do this, we square Eq. (11.57): Ma2y Ma2x p x 2 Ty = ( ) ( ). py Tx Then we substitute Eqs. (11.58) and (11.61) into the right side. The result is a quadratic equation relating Ma2y to M2x ; k appears in the equation as a parameter. Because the equation is quadratic, there are two solutions for Ma2y : Ma2y = Ma2x (11.62) and Ma2y = Ratios of thermodynamic properties across a normal shock are functions of the Mach numbers. Ma2x + 2(k − 1) . [2k(k − 1)]Ma2x − 1 (11.63) The first solution seems trivial, but it shows that the Mach number need not change discontinuously at a plane. This solution is valid if there is no shock. If there is a shock, the second solution, Eq. (11.63), gives the downstream Mach number in terms of the upstream Mach number. For any realistic value of k (that is, k > 1), calculations with Eq.(11.63) show that Max and May always lie on opposite sides of Ma = 1.0; that is, if Max > 1, then May < 1 and vice versa. We can now substitute Eqs. (11.62) and (11.63) into the pressure ratio and temperature ratio equations, Eqs. (11.58) and (11.61). The constant Mach number solution gives pressure and temperature ratios equal to unity; that is, there can be no discontinuous change of pressure and temperature at a plane if there is no shock at the plane. If there is a shock, the pressure and temperature ratios are py px = 2k k−1 Ma2x − k+1 k+1 (11.64) 624 Chapter 11 ■ Compressible Flow The flow changes from supersonic to subsonic across a normal shock. ■ Table 11.1 Relative Changes of Properties as Predicted by the Normal Shock Equations May py px Ty Tx T0y T0x p0y p0x sy − sx Max > 1 (Supersonic Upstream) Max < 1 (Subsonic Upstream) < 1 (subsonic) > 1 (Not physically possible) (supersonic) >1 < 1 (Not physically possible) >1 < 1 (Not physically possible) =1 =1 <1 > 1 (Not physically possible) >0 < 0 (Not physically possible) and Ty Tx = (1 + 2(k − 1) k−1 2k Ma2x )( Ma2x − 1)[ . 2 k−1 (k + 1) 2Ma2x ] (11.65) A shock may change stagnation properties as well as static properties. According to Eq. (11.60), the stagnation temperature ratio across a shock is T0y T0 x = 1. (11.66) We obtain the stagnation pressure ratio across a shock by first expressing stagnation pressure in terms of static pressure and Mach number. Using Eq. (11.51), we have p0 y p0 x = py{1 + [(k − 1)2]Ma2y }k(k−1) px{1 + [(k − 1)2]Ma2x }k(k−1) . We reduce this expression to an equation involving only Ma2x and k by substituting Eq. (11.64) for the static pressure ratio and Eq. (11.63) for Ma2y to get p0 y −k k −1 k−1 (k−1) k+1 k−1 2k k − 1 (k−1) 2 = . Ma2x ) (1 + Ma2x ) ( k + 1 Max − k + 1 ) p0 x ( 2 2 Across a normal shock the values of some parameters increase, some remain constant, and some decrease. (11.67) Equations (11.63)–(11.67) enable us to calculate properties on the downstream side of a shock if we know the upstream properties and Mach number Max. Table 11.1 indicates typical results of calculations made with the equations. Note the second column, which tabulates results for a (hypothetical) shock that occurs in a subsonic flow. Our discussion of the different nature of subsonic and supersonic flow indicates that discontinuous changes of fluid properties do not occur in subsonic flow. To prove this, consider the second law of thermodynamics, in mathematical form: sy − sx ≥ ∫ x y δq . T With no heat transfer to or from the gas as it passes through a shock wave, sy − sx ≥ 0. Working with a dimensionless entropy change is convenient. Dividing Eq. 11.24 by R gives sy − sx R = Ty py k ln( ) − ln( ). px k−1 Tx 11.5 Shock Waves 625 Substituting Eqs. (11.64) and (11.65) for the pressure and temperature ratios yields an expression for entropy change in terms of Mach number. By selecting various values of Max, we can determine (sy − sx). The entropy change is negative for all subsonic inlet Mach numbers. Shock waves can exist only if the upstream flow (relative to the shock) is supersonic. Shock waves always increase the static pressure, density, and temperature and decrease the stagnation pressure (relative to the shock). If you need to calculate property changes across a shock wave, you can use Eqs. (11.63)– (11.67); however, these equations are quite tedious. This is especially true if you know, say, Ty Tx and are trying to calculate Mach number. Appendix D has tables and a graph of the normal shock functions for k = 1.4. A quick and accurate calculation can be made with a compressible flow function calculator from the Internet or as a smartphone app. E X AMPLE 11.4 Stagnation Pressure Drop across a Normal Shock GIVEN Designers involved with fluid mechanics work hard at minimizing loss of available energy in their designs. Shock waves produce significant losses when they occur. In addition to an increase in entropy, shocks result in a reduction in stagnation pressure. FIND For normal shocks, show that the stagnation pressure drop (and thus loss) is larger for higher Mach numbers. SOLUTION We assume that air (k = 1.4) behaves as a typical gas and use Table D.S to respond to the above-stated requirements. Since p0, x − p0, y p0, x =1− 1 p0, y 0.8 p0, x we can construct the following table with values of p0,yp0, x from Table D.S. p0,x – p0,y _________ 0.6 p0,x 0.4 p0, x − p0, y Max p0, yp0, x p0, x 1.0 1.2 1.5 2.0 2.5 3.0 3.5 4.0 5.0 1.0 0.99 0.93 0.72 0.50 0.33 0.21 0.14 0.06 0 0.01 0.07 0.28 0.50 0.67 0.79 0.86 0.94 COMMENT When the Mach number of the flow entering the shock is low, say Max = 1.2, the flow across the shock is nearly isentropic and the loss in stagnation pressure is small. However, as shown in Fig. E11.4, at larger Mach numbers, the entropy change across the normal shock rises dramatically and the stagnation pressure drop across the shock is appreciable. If a shock occurs at Max = 2.5, only about 50% of the upstream stagnation pressure is recovered. In devices where supersonic flows occur, for example, highperformance aircraft engine inlet ducts and high-speed wind tunnels, designers attempt to prevent shock formation, or if shocks must occur, they design the flow path so that shocks are positioned where they are weak (small Mach number). 0.2 0 0 1 2 3 4 5 6 Max ■ Figure E11.4 Of interest also is the static pressure rise that occurs across a normal shock. These static pressure ratios, pypx, obtained from Table D.S are shown in the table for a few Mach numbers. For a developing boundary layer, any pressure rise in the flow direction is considered as an adverse pressure gradient that can possibly cause flow separation (see Section 9.2.6). Thus, shock–boundary layer interactions are of great concern to designers of high-speed flow devices. Max pypx 1.0 1.2 1.5 2.0 3.0 4.0 5.0 1.0 1.5 2.5 4.5 10 18 29 626 Chapter 11 ■ Compressible Flow EXAMPLE 11.5 Supersonic Flow Pitot Tube GIVEN A total pressure probe is inserted into a supersonic air flow. A shock wave forms just upstream of the impact hole and head as illustrated in Fig. E11.5. The probe measures a total pressure of 60 psia. The stagnation temperature at the probe head is measured with a thermocouple and found to be 1000 °R. The static pressure upstream of the shock is measured with a wall tap to be 12 psia. Wall static pressure tap Supersonic flow Shock wave Stagnation pathline x y Total pressure probe FIND Determine the Mach number and velocity of the flow. ■ Figure E11.5 SOLUTION We assume that the flow along the stagnation pathline is isentropic except across the shock. Also, the shock is treated as a normal shock. Thus, in terms of the data we have p0,y px =( p0,y p0,x (1) p0,x )( px ) where p0,y is the stagnation pressure measured by the probe, and px is the static pressure measured by the wall tap. The stagnation pressure upstream of the shock, p0,x, is not measured. Combining Eqs. 1, 11.51, and 11.67 we obtain p0,y px = {[(k + 1)2]Ma2x }k(k−1) {[2k(k + 1)]Ma2x − [(k − 1)(k + 1)]} 1(k−1) px = we use Graph D.S (or Eq. 2) to ascertain that (Ans) To determine the flow velocity we need to know the static temperature upstream of the shock, since Vx = Max cx = Max √kRTx 11.6 Flow with no friction and no heat transfer is isentropic. Since the stagnation temperature remains constant across a normal shock, T0,x = T0,y = 1000 °R For the isentropic flow upstream of the shock, Eq. 11.50 or Graph D.I can be used. For Max = 1.9, Tx = 0.59 T0,x or Tx = (0.59) (1000 °R) = 590 °R With Eq. 3 we obtain 60 psia =5 12 psia Max = 1.9 T0,y = 1000 °R (2) which is called the Rayleigh Pitot-tube formula. Values of p0,ypx from Eq. 2 are considered important enough to be included in Graph D.S for k = 1.4. Thus, for k = 1.4 and p0,y The stagnation temperature downstream of the shock was measured and found to be (3) Vx = 1.87 √[53.3 (ft ˙ lb)(lbm ˙ °R)](590 °R) (1.4) = 392 (ft ˙ lblbm) 12[(32.2 lbm ˙ fts2 )lb]12 = 2220 fts (Ans) COMMENT Application of the incompressible flow Pitot tube results (see Section 3.5) would give highly inaccurate results because of the large pressure and density changes involved. Isentropic Flow In this section, we consider in further detail steady, one-dimensional, isentropic flow. As explained earlier, one-dimensionality implies velocity and fluid property change in the streamwise direction only. We consider flows through ducts with uniformly distributed velocities and fluid properties at each cross-section. Much of what we develop can also apply to the flow of a fluid particle along its pathline or flow through an infinitesimal streamtube. Isentropic flow has constant entropy and was discussed briefly in Section 11.1, where we learned that adiabatic and frictionless (reversible) flow is the most common form of isentropic flow. Some ideal gas relationships for isentropic flows were developed in Section 11.1. An isentropic flow is not achievable with actual fluids because of friction. Nonetheless, the study of isentropic flow is useful because it helps us to gain an understanding of actual compressible flow phenomena. In many situations, the isentropic flow model is sufficiently accurate to use for engineering analysis and design. 11.6 Isentropic Flow 627 11.6.1 Steady Isentropic Flow of an Ideal Gas For a one-dimensional flow of a compressible fluid we make the following assumptions: ■ ■ ■ ■ ■ The flow is steady. There are no body forces (gravity, electromagnetic, or others) There is no friction. There is no heat transfer. The flow does not pass through a shock wave. The last three assumptions taken together imply that the flow is adiabatic and reversible, therefore, the entropy of the fluid is constant: s = Constant; or, in terms of two locations in the flow s 1 = s2 . For steady flow, the energy equation is V 22 V 21 q + wshaft = (hˇ 2 + ) − (hˇ 1 + ). 2 2 According to our assumptions, q is zero, but wshaft is not necessarily zero, and the energy equation becomes V 22 V 21 wshaft = (hˇ 2 + ) − (hˇ 1 + ). 2 2 When we use the definition of stagnation enthalpy, wshaft = hˇ 02 − hˇ 01. If the fluid is an ideal gas, wshaft = cp (T02 − T01 ). (11.68) Doing work on the fluid increases its stagnation temperature, even if there is no heat transfer. Now we consider changes of stagnation pressure in isentropic flow. By definition of the stagnation state, s01 = s1 and s02 = s2. For isentropic flow, s02 − s01 = s2 − s1 = 0; thus p0 2 =( p0 1 T0 2 cpR T0 1 ) =( T0 2 k/(k−1) T0 1 ) . (11.69) Using Eq. (11.68), we can also write p0 2 p0 1 = (1 − wshaft k/(k−1) . cpT01 ) (11.70) A situation of considerable interest in compressible flow is the case with no shaft work. The following statement applies to this case: In isentropic flow with no shaft work, all stagnation properties are constant. This special case (wshaft = 0) is so important in compressible flow that most engineers and authors use the phrase “isentropic flow” to mean “isentropic flow with no work.” We will do the same, thus we add the assumption that there is no work to the assumptions of no heat transfer and no friction. 628 Chapter 11 ■ Compressible Flow Equations relating property and velocity changes in isentropic flow of an ideal gas can easily be developed from the stagnation property definitions and the constancy of stagnation properties. Consider, for example, the calculation of the change in temperature between two points in an isentropic flow. Because T0 is constant, we can write T0 1 = T1 + V 21 V 22 = T2 + = T02. 2cp 2cp (11.71) Using Mach number instead of velocity is sometimes convenient. From Eq. (11.50). T02{1 + [(k − 1)2]Ma21} T2 = , T1 T01{1 + [(k − 1)2]Ma22} but, as T0 is constant, T2 1 + [(k − 1)2]Ma21 = . T1 1 + [(k − 1)2]Ma22 (11.72) A third approach involves use of compressible flow functions (from tables, graphs, or software) relating static-stagnation property ratios and Mach number. We calculate the temperature ratio by T2 T T = ) / ) T1 T0 Ma2 T0 Ma1 (11.73) Recall that T T0 )Ma is a numerical value read from a table or graph or computed by software at Mach number Ma. In a fourth method, we make the calculation in two steps: T0 2 = T0 1 = T1 T T0 )Ma1 and T T2 = T02[ ) ]. T0 Ma2 (11.74) Any of Eqs. (11.71)–(11.74) could be used to relate temperature and velocities (or Mach numbers) at point 1 and 2. Which equation is most convenient depends on the exact information available. Do not fall into the trap of always using Mach number functions; direct calculations are sometimes easier. Similarly, pressure and velocity (or Mach number) in isentropic flow can be related by any of the following equations: p1(1 + V 21 k(k−1) V 22 k(k−1) = p 1 + ; 2 ( 2cpT1 ) 2cpT2 ) p2 {1 + [(k − 1)2]Ma21}k(k−1) = ; p1 {1 + [(k − 1)2]Ma22}k(k−1) p p2 p = ) / ) ; p1 p0 Ma2 p0 Ma1 p p2 = p02[ ) ]; p0 Ma2 where p0 2 = p0 1 = p1 . p p0 )Ma1 (11.75) (11.76) (11.77) (11.78) 11.6 E X AMPLE 11.6 629 Isentropic Flow Use of Isentropic Flow Functions, Graphs, and Tables GIVEN Air flows in a duct. At one point in the flow, the pressure, temperature, and velocity are 50 psia, 100 °F, and 500 ft/sec. At a downstream point, the Mach number is 1.80. The flow is steady and isentropic. Find Air pressure, temperature, and velocity at the downstream point. Do the calculation twice, once using isentropic flow equations and once using Table D.I or Graph D.I. Flow 1 2 P1 = 50 psia T1 = 100°F V1 = 500 ft/s Ma2 = 1.80 ■ Figure E11.6 Flow between two points in a duct. SOLUTION The first step is to convert all given data into absolute units. The only conversion necessary is temperature: Now we repeat the calculation using Table D.I and Graph D.I. The Mach number at 1 is T1 = 100 °F + 459.7 °F = 559.7 °R. Ma1 = We first do the calculations with isentropic flow equations. The energy equation for adiabatic flow with no work between 1 and 2 is cpT1 + Next, we use Table D.I to obtain the pressure and temperature ratio functions at 1 and at 2. At Ma = Ma1 = 0.43, V 21 V 22 = cpT2 + . 2 2 p T = 0.8806  and  ) = 0.9643. p0 )Ma=0.43 T0 Ma=0.43 The velocity at 2 is V 21 kRMa22 V2 = Ma2c2 = Ma2√kRT2, so cpT1 + = (cp + T, 2 2 ) 2 (Graph D.I can also be used to find values for p/p0 and T/ T0 but the values would be less precise, 0.88 and 0.96 respectively) At Ma = Ma2 = 1.80, p T = 0.1740 and  ) = 0.6068. p0 )Ma=1.80 T0 Ma=1.80 or T2 = cpT1 + V 212 cp + kRMa222 Because the flow is isentropic, . p2 = p1 Substituting numbers, we have (186.7 ft ∙ lb/lbm ∙ °R) (559.7 °R) + T2 = 186.7 ft ∙ lb/lbm ∙ °R + V1 V1 500 ft/s = = = 0.43. c1 √kRT1 (49.02√559.7) ft/s (500) 2 ft2sec2 2(32.2 ft ∙ lbmlb ∙ sec2 ) 1.4(53.35 ft ∙ lblbm ∙ R) (1.80) 2 T2 = 352.2 °R. 2 p2 = 9.88 psia. ; (Ans) The velocity at 2 is V2 = Ma2c2 = [(1.80) (49.02)]√352.2] ft/s; V2 = 1660 ft/s. (Ans) Now T2 = T1 T T 0.6068 = 559.7 °R( ; T0 )Ma2 / T0 )Ma1 0.9643 ) T2 = 352.2 °R. (Ans) The velocity at 2 is (Ans) We compute the pressure at 2 from the isentropic process equation, Eq. (11.26): T2 k(k−1) 352.2 3.5 p2 = p1 ( ) = 50 psia( ; T1 559.7 ) p2 = 9.88 psia. p p 0.1740 ; = 50 psia ( ) / p0 Ma2 p0 )Ma1 0.8806 ) V2 = Ma2√kRT2 = (1.80) (49.02)√352.2 ft/sec; V2 = 1660 fts (Ans) COMMENTS Of course the answers determined a by using (Ans) Table D.I are the same as those determined by using equations. We leave the decision to you as to which method is easier. Always convert pressures and temperatures to absolute units as the first step in solving a compressible flow problem. 630 Chapter 11 ■ Compressible Flow T p0 T0 p T ■ Figure 11.6 The (T–s) diagram relating stagnation and s static states. A very useful means of keeping track of the states of an isentropic flow of an ideal gas is a temperature–entropy (T–s) diagram, as is shown in Fig. 11.6. Experience has shown (see, for example, Refs. 2 and 3) that lines of constant pressure are generally as are sketched in Fig. 11.6. An isentropic flow is confined to a vertical line on a T–s diagram. The vertical line in Fig. 11.6 is representative of flow between the stagnation state and any state. Equation 11.72 shows that fluid temperature decreases with an increase in Mach number. Thus, the lower temperature levels on a T–s diagram correspond to higher Mach numbers. Fluid pressure also decreases with an increase in Mach number. Thus, lower fluid temperatures and pressures are associated with higher Mach numbers. 11.6.2 Incompressible Flow and Bernoulli’s Equation One of the most widely used and abused equations in fluid mechanics is Bernoulli’s equation: p+ 1 2 ρV + γz = Constant 2 The assumptions used in deriving Bernoulli’s equation are similar to those we just used in deriving the isentropic flow equations, except that whereas Bernoulli’s equation assumes incompressible flow, the isentropic equations assume adiabatic flow of an ideal gas with no shocks and no work. In this section, we show that, for a very small Mach number, the isentropic equation for pressure becomes identical to Bernoulli’s equation. In addition, we develop a criterion for deciding whether a gas flow can be treated as incompressible. Consider a steady flow with no shear stress, shaft work, or heat transfer. Under these conditions, the stagnation pressure is constant. We also assume that elevation changes are negligible. If the fluid is incompressible, we can calculate the pressure at any location from Bernoulli’s equation: 1 p = p0 − 2 ρV 2. (11.79) If the fluid is compressible and an ideal gas, the static pressure and stagnation pressure are related by p0 = p(1 + k(k−1) k−1 Ma2) . 2 (11.80) If we restrict consideration to Mach numbers less than 1, we can expand the Mach number term into an infinite series using the binomial theorem: (1 + x) n = 1 + nx + n(n − 1) 2 n(n − 1) (n − 2) 3 x + x + …. 2! 3! Putting x= k−1 k Ma2 and n = , 2 k−1 (11.81) 11.6 Compressibility effects are more important at higher Mach numbers. Isentropic Flow 631 we get (1 + k(k−1) k−1 k k Ma2) = 1 + Ma2 + Ma4 + H.O.T. , 2 2 8 where H.O.T. involves higher even powers of Ma. Substituting into Eq. (11.81) gives p0 = p(1 + kMa2 kMa4 + + H.O.T.). 2 8 Because we are interested in small values of Ma, the higher-order terms are negligible, and we can write the equation p0 ≈ p + kpMa2 Ma2 1 + . 2 ( 4 ) (11.82) From Eq. (11.48), we have kpMa2 = ρV2. Substituting and rearranging, we get p ≈ p0 − 1 Ma2 ρV 2(1 + . 2 4 ) (11.83) If the Mach number is small, then Ma2/4 is small compared to 1, and we can write 1 p ≈ p0 − 2 ρV 2; (11.84) therefore Bernoulli’s equation is an approximation to the isentropic flow pressure relation for small Mach number. The accuracy of this approximation depends on the “smallness” of the Mach number. Equation (11.83) shows that the error is proportional to Ma2 4 at low Mach numbers. If we want to limit the error in using Bernoulli’s equation for calculation of pressure to no more than, say, 2 percent, then Ma ≤ √4(0.02) = 0.283. There is nothing unique about 2 percent error. For rough estimates, 5 percent error may be acceptable, in which case the Mach number must be less than about 0.45. The most widely used criterion for the boundary between compressible and incompressible flow is: A flow with Ma < 0.3 everywhere can be assumed to be incompressible. Recall that we arrived at essentially this same criterion in Sec. 3.8.1. 11.6.3 The Critical State We have shown that the nature of a compressible flow changes drastically when the fluid velocity exceeds the speed of sound. The flow state that corresponds to exactly sonic flow is called the critical state. The fluid properties at the critical state are called the critical properties. Consider acceleration of a gas from zero velocity to the local sonic speed in an isentropic process. As the fluid’s velocity increases, its temperature and speed of sound decrease. The stagnation temperature is constant for this process, so we can write T0 = T(1 + k−1 Ma2). 2 At the critical state, the Mach number is unity. Using the superscript * to denote conditions at the critical state, we have T0 = T *(1 + k−1 (1) 2), 2 or T* 2 = . T0 k+1 (11.85) 632 Chapter 11 ■ Compressible Flow T p0 T0 ( 2 p* = ______ k+1 ( ) k ( ______ k–1 ) p 0 ) 2 T* = ______ T0 k+1 s ■ Figure 11.7 The relationship between the stagnation and critical states. For the specified isentropic process, the pressure and density at the critical state are given by k(k−1) p* T * k(k−1) 2 =( ) =( p0 T0 k + 1) (11.86) 1(k−1) ρ* T * 1(k−1) 2 =( ) =( . ρ0 T0 k + 1) (11.87) and The stagnation and critical states are at the same entropy level. 11.7 For an isentropic flow, properties at the critical state are uniquely determined by the stagnation properties and k. The critical properties T *, p*, and ρ* are constant if the stagnation properties are constant. The stagnation and critical pressures and temperatures are shown on the T–s diagram of Fig. 11.7. One-Dimensional Flow in a Variable Area Duct A compressible flow in a duct may respond to four different types of inputs: cross-sectional area variation, viscous friction between the fluid and the duct wall, heat addition or removal, and shaft work. Work exchange is the subject of Chapter 12 so we will remove it from further consideration here. For simplicity and clarity, we will consider each of the remaining inputs in the absence of the other two. Here we will examine the effects of variation of cross-sectional area in the absence of friction or heat; the flow will be frictionless and adiabatic (meaning no heat addition or removal). With neither friction nor heat transfer, the flow will be isentropic except for any shocks that are present. When fluid flows steadily through a duct that has a flow cross-sectional area that varies with axial distance, the conservation of mass (continuity) equation . m = ρAV = constant (11.85) can be used to relate the velocity and density to the area. For incompressible flow, the fluid density remains constant and the flow velocity from section to section varies inversely with cross-sectional area. However, when the flow is compressible, density, cross-sectional area, and flow velocity can all vary from section to section. We proceed to determine how fluid properties and flow velocity change with axial location in a variable area duct when the flow through the duct is steady and isentropic. Our preliminary analysis will be done without assuming that the fluid is an ideal gas. 11.7.1 General Considerations In Chapter 3, Newton’s second law was applied to the inviscid (frictionless) and steady flow of a fluid particle. For the streamwise direction, the result (Eq. 3.5) for either compressible or incompressible flows is 1 dp + 2 ρ d(V 2 ) + γ dz = 0 (11.86) The frictionless flow from section to section through a control volume is also described by Eq. 11.86, if the flow is one-dimensional, because every particle of fluid will have the same experience. For gas flow, the potential energy term, γ dz, can be dropped because of its small size in comparison to the 11.7 One-Dimensional Flow in a Variable Area Duct 633 other terms. Thus, an appropriate equation of motion in the streamwise direction for the steady, onedimensional, and isentropic (adiabatic and frictionless) flow of gas is obtained from Eq. 11.86 as dp dV =− V ρV 2 (11.87) If we form the logarithm of both sides of the continuity equation (Eq. 11.85), the result is ln ρ + ln A + ln V = constant. Density, crosssectional area, and velocity may all vary for compressible flow in a duct. (11.88) Differentiating Eq. 11.88 we get dρ dA dV + + =0 ρ A V or − dρ dA dV + = ρ V A (11.89) Now we combine Eqs. 11.87 and 11.89 together with dρ = dp/(dp/dρ) to obtain dp V2 dA 1− = . 2( dpdρ ) A ρV (11.90) Since the flow being considered is isentropic, the speed of sound is related to variations of pressure with density by c= ∂p . ( B ∂ρ )s This equation, combined with the definition of Mach number Ma = V c and Eq. 11.90 yields dp dA (1 − Ma2 ) = . 2 A ρV (11.91) Combining Eqs. 11.87 and 11.91 gives dV dA 1 =− . V A (1 − Ma2 ) A converging duct will decelerate a supersonic flow and accelerate a subsonic flow. A converging– diverging duct is required to accelerate a flow from subsonic to supersonic flow conditions. (11.92) We can use Eq. 11.92 to conclude that when the flow is subsonic (Ma < 1), velocity and section area changes are in opposite directions. In other words, the area increase associated with subsonic flow through a diverging duct like the one shown in Fig. 11.8a is accompanied by a velocity decrease. Subsonic flow through a converging duct (see Fig. 11.8b) involves an increase of velocity. These trends are consistent with incompressible flow behavior, which we described earlier in this book, for instance, in Chapters 3 and 8. Equation 11.92 also serves to show us that when the flow is supersonic (Ma > 1), velocity and area changes are in the same direction. A diverging duct (Fig. 11.8a) will accelerate a supersonic flow. A converging duct (Fig. 11.8b) will decelerate a supersonic flow. These trends are the opposite of what happens for incompressible and subsonic compressible flows. These observations are summarized in Table 11.2. In isentropic flow, an increase in velocity always corresponds to a Mach number increase and vice versa; therefore a converging passage always drives the Mach number toward one and a diverging passage always drives the Mach number away from one. These conditions imply that the Mach number cannot pass through one (either accelerating or decelerating) in a passage that is only convergent or only divergent. If we consider the process of accelerating a gas from rest or a low subsonic Mach number in a simply converging passage, the maximum possible Mach number that can be achieved is one. This maximum value can occur only at the end of the converging passage. 634 Chapter 11 ■ Compressible Flow Subsonic flow (Ma < 1) Supersonic flow (Ma > 1) dA > 0 dV < 0 dA > 0 dV > 0 dA < 0 dV > 0 dA < 0 dV < 0 Flow (a) Flow ■ Figure 11.8 (a) A diverging duct. (b) (b) A converging duct. Now suppose that the Mach number is one at some point in a flow. Since infinite acceleration (dV → ± ∞) is not possible unless there is a shock wave, which is specifically excluded by the assumptions, Eq. (11.92) implies that dA = 0  when Ma = 1. In isentropic flow, sonic flow (Ma = 1) can occur only at an area minimum (Fig. 11.9). The occurrence of sonic flow at an area maximum is excluded by the considerations identified in the preceding paragraph. To accelerate a fluid from rest (or a subsonic Mach number) to a supersonic Mach number, a converging-diverging nozzle is required. A rocket engine such as the one shown in the figure in the margin uses a converging-diverging nozzle to accelerate hot gases to a very large Mach number with corresponding high velocity in order to generate sizeable thrust. Finally, suppose that a compressible fluid is flowing in a passage that actually has an area minimum (called a throat). Is it always possible to conclude that the Mach number is one at the throat? To investigate this question, we rearrange Eq. (11.92) as follows: dA dV = (Ma2 − 1) . A V If dA is zero, then two options are possible: Ma = 1 or dV = 0. Simply knowing that dA = 0 does not allow us to determine which condition occurs. Figure 11.10 illustrates possible velocity (or Mach number) distributions in a passage with a minimum area. If the throat Mach number is not one, the velocity must have a local maximum or minimum at the throat ■ Table 11.2 Effects of area change in isentropic compressible flow. Converging passage dA < 0 Diverging passage dA > 0 M<1 M>1 Velocity (V) Temperature (T) Pressure (p) Density (ρ) Passage acts as a Increases Decreases Decreases Decreases Nozzle Decreases Increases Increases Increases Diffuser Velocity (V) Temperature (T) Pressure (p) Density (ρ) Passage acts as a Decreases Increases Increases Increases Diffuser Increases Decreases Decreases Decreases Nozzle 11.7 One-Dimensional Flow in a Variable Area Duct Ma = 1 cannot occur here. Ma = 1 can occur only here. Flow 635 Flow Area minimum dA = 0 Area maximum dA = 0 (a) (b) ■ Figure 11.9 (a) A converging–diverging duct. (b) A diverging–converging duct. (dVth = 0). This behavior is observed in a Venturi tube in incompressible flow as discussed in Chapters 3 and 8. If the throat Mach number is one, dVth, can be either positive or negative, and the velocity can either increase or decrease downstream from the throat, depending on downstream conditions. The following points summarize our findings about isentropic flow: ■ ■ ■ The response of the flow to a specific type of area change is exactly opposite for subsonic and supersonic flow. Details are given in Table 11.2. Sonic flow (Ma = 1) can occur only at a minimum area. Minimum areas occur at the inlet of a simply diverging passage, the outlet of a simply converging passage, and the throat of a converging–diverging passage. It is possible, but not necessary, that the Mach number at the throat of a convergingdiverging passage be one. If the Mach number at a throat is not one, the velocity must pass through a maximum or minimum. If the throat Mach number is one, the fluid may either accelerate or decelerate downstream from the throat. We used a completely general property relation (the second Tds equation), so these conclusions are valid for one-dimensional isentropic flow of any compressible fluid, ideal gas or not. y V or Ma dVth = 0 Math > 1 Math = 1, dVth > 0 V = c* Mae Math = 1 Math = 1, dVth < 0 Ma = 1 Flow x Math < 1 dVth = 0 ■ Figure 11.10 Possible velocity (or Mach number) distributions in passage with an area minimum. x Throat (a) (b) 11.7.2 Isentropic Flow of an Ideal Gas With Area Change A1 m• A2 V1 1 V2 2 Isentropic flow between planes 1 and 2 in a variable-area duct. By assuming that the fluid is an ideal gas, we can obtain working equations for flow in variable-area ducts. The figure in the margin illustrates the general flow geometry. The typical problem involves relating changes of fluid properties and velocity to the passage area. Calculation of mass flow rate or force on the duct walls is also important. For isentropic flow, we can use Eqs. (11.71)–(11.78) and Table D.I or Graph D.I to relate pressure, temperature, velocity, and Mach number for flow between any two planes in the duct. All we need for a complete analysis are equations relating Mach number to passage area and mass flow rate to Mach number, flow properties, and passage area. 636 Chapter 11 ■ Compressible Flow Area–Mach Number Relation. To obtain an equation relating area and Mach number, we use the continuity equation for a control volume with inflow at plane 1 and outflow at plane 2 (see the figure in the margin): ρ1V1 A1 = ρ2V2 A2. For an ideal gas, ρ= p  and V = Ma(c) = Ma√kRT. RT Substituting these equations into the continuity equation gives k p1A1Ma1 k p2 A2Ma2 = . B R ( √T1 ) B R ( √T2 ) Canceling k and R and solving for the area ratio, we have A2 Ma1 p2 −1 T2 12 = . A1 Ma2 ( p1 ) ( T1 ) For isentropic flow, we can substitute Eqs. (11.72) and (11.76) for the temperature and pressure ratios, giving A2 Ma1 1 + [(k − 1)2]Ma22 (k+1)2(k−1) = . A1 Ma2 { 1 + [(k − 1)2]Ma21 } If any three of the quantities A1, Ma1, A2, and Ma2 are specified, the fourth can be calculated. Usually A1 and Ma1 are known and either Ma2 or A2 is to be found. If Ma2 is given and A2 is to be calculated, the calculation is simple; however, if A2 is given and Ma2 must be calculated, that is not so easy. We can simplify the calculations by making a table or graph. To tabulate area ratio, we select a reference area with known Mach number. The obvious choice is the area corresponding to a Mach number of one. We set Ma1 = 1, A1 = A* and drop the subscript 2 to get (k+1)2(k−1) A 1 2 k−1 2 = 1 + Ma . )] A* Ma [ k + 1 ( 2 The ratio of flow area to the critical area is a useful concept for isentropic duct flow. In Eq. (11.93), A is the area at the location where the Mach number is Ma, and A* (the critical area) is the area that would correspond to a Mach number of one. Note that A* is a reference area; the actual area need not be A* anywhere in the flow. Values of A/A* for k = 1.4 are included in the isentropic flow table and the isentropic flow function graphs in Appendix D. The relationship between A/A* and Ma for any value of k can usually be obtained from “calculators” on the Internet or a smartphone. The area ratio function can be used to relate areas and Mach numbers at two planes in a duct by A2 A A = A1 A* )Ma2/ A* )Ma1. 2.0 A ___ A* 1.0 0 (11.93) 1.0 Ma The variation of area ratio with Mach number for isentropic flow of an ideal gas (k = 1.4, linear coordinate scales). (11.94) A plot of the area ratio function versus Mach number (the figure in the margin) reveals some interesting information and confirms some of the observations made in the preceding section. First, we note that A/A* is always greater than one except at Ma = 1, where A/A* is 1. This confirms that in isentropic flow, the Mach number can be one only at the minimum area. Next, we note that if Mach number is given, there is a single corresponding area ratio, so the problem of calculating the area required to obtain a given Mach number seems easy. We must be careful, however. If the initial Mach number (Ma1) is less than one and the desired final Mach number (Ma2) is greater than one, a minimum area 11.7 One-Dimensional Flow in a Variable Area Duct 637 exactly equal to A* must occur between planes 1 and 2; otherwise the flow cannot pass through Ma = 1. Finally, we note that if A/A* is known, there are two corresponding Mach numbers. Choice of the proper Mach number for a known area ratio requires more information than just the area ratio itself. Generally, we must know the pressure or some other item of information in addition to the area ratio to obtain the Mach number. We can sometimes logically eliminate one of the possibilities. If we know that the flow is subsonic at plane 1 and that a minimum area exactly equal to A* does not occur between planes 1 and 2, the flow at 2 cannot be supersonic. Before concluding this section on area ratio, we offer a few comments and warnings. The critical area A* is the area necessary to accelerate or decelerate a flow to sonic conditions. We do not necessarily identify A* with any real area in a flow passage; it is a reference condition, just like the stagnation state is a reference condition. If we know that the flow is sonic somewhere in a passage, the passage area equals the critical area at the sonic location. We also know that it must be the minimum area if the flow is isentropic. However, the converse is not always true. If there is a minimum area in a flow passage, we cannot assume that the minimum area is equal to A* without further evidence that the flow actually is sonic at the minimum area. Stated simply, in a passage with a throat, the throat area (Ath) does not necessarily equal the critical area (A*) unless other evidence indicates it is. E X AMPLE 11.7 Area Ratio Function and Flow in a ConvergingDiverging Passage GIVEN The small laboratory wind tunnel (similar to one used in Ref. 1) shown in Fig. E11.7 draws air from the atmosphere. The pressure and temperature of the laboratory air are 100 kPa and 20 °C (293.3 K). Schlieren photographs reveal a shock wave in the nozzle, as shown. Find The air temperature and pressure at the throat and at planes 1, 2, and 3. Assume isentropic flow except for the shock. Laboratory air Shock th ■ Figure E11.7 Small laboratory wind tunnel. To vacuum Test section t 0 1 2 Control valve 3 Ath = 10–3 m2, A1 = 2 × 10–3 m2, A3 = 3 × 10–3 m2 SOLUTION The passage is converging–diverging, so we must be alert to the possibilities of sonic flow at the throat and supersonic flow in the diverging section. There is a shock in the diverging section of the nozzle so the flow is supersonic at 1. Because the flow is subsonic (Ma ≈ 0) in the laboratory, the air must pass through Ma = 1 somewhere upstream from plane 1. The only place this can occur is the throat, so pressure and temperature (since Ma ≈ 0 in the laboratory air). The critical area for the flow upstream from the shock is A*0→1 = Ath. We calculate the throat pressure and temperature with the aid of Table D.I. p pth = p0( ) = 100 kPa (0.5283), p0 Ma=1 Math = 1.0. The process is isentropic ahead of the shock, so the stagnation properties are constant up to the shock and equal to the laboratory or pth = 52.8 kPa; (Ans) 638 Chapter 11 ■ Compressible Flow p0 and A* than the flow upstream from the shock. In particular, A*3, which is equal to A*2, is not equal to the throat area. We calculate the area ratio function at plane 3 from and T = 293.2K(0.8333), Tth = T0( ) T0 Ma=1 A3 A A = ( )( ) . ) A* 3 A2 A* Ma2 or Tth = 244.3K. (Ans) From Table D.I A = 1.259, A* )Ma=0.547 The flow at plane 1, just ahead of the shock, has −3 A*1 = Ath = 1.0 × 10 2 m. so The area ratio for plane 1 is A 3 × 10−3 m2 = (1.259) = 1.888. A* )3 ( 2 × 10−3 m2 ) A1 2 × 10−3 m2 = 2.000. = A* 1 × 10 −3 m2 From Table D.I, we find the supersonic Mach number corresponding to this area ratio: This value of A/A* corresponds to a supersonic Mach number and a subsonic Mach number. As there is no throat between planes 2 and 3, the flow at plane 3 must be subsonic. From Table D.I, Ma1 = 2.20. Ma2 = 0.547,   The pressure and temperature ratios (from Table D.I) are Ma3 = 0.326,   p T = 0.0935 and  ) = 0.5081. p0 )Ma=2.20 T0 Ma=2.20 We now calculate p1 = 0.0935p01 = 0.0935(100 kPa), p = 0.9290, and p0 )Ma=0.326 T = 0.9792. T0 )Ma=0.326 We calculate p3: or p1 = 9.35 kPa; p3 = p 2 (Ans) p p 0.9290 = 51.2 kPa( , p0 )3 / p0 )2 0.8159 ) or and p3 = 58.3 kPa. T1 = 0.5081T01 = 0.5081(293.2K), (Ans) The expression for T3 is or T1 = 149.0K. p = 0.8159, po )A/A*=1.888 T T3 = T0( ) , T0 3 (Ans) Plane 2 lies downstream from the shock. We must use the shock functions in Table D.S to relate properties at plane 2 to those at plane 1. For Ma1 = 2.20, but as the shock does not affect T0, Ty = 5.480,   ) = 1.857, and  px )Max =2.20 Tx Max =2.20 May = 0.547. or Then COMMENTS Note the logic used to decide that A* upstream py p2 = T3 = (293.2 K) (0.9792), T3 = 287.1 K. of the shock is equal to Ath, the geometric area of the throat. The value of A* on the down-stream side of the shock can be calculated as py p = (5.480) (9.35 kPa), px ) 1 or p2 = 51.2 kPa; A*2 = (Ans) Ty T = (1.857) (149.0 K), Tx ) 1 T3 = T2 or T2 = 276.7K. A2 = 1.59 × 10−3 m2. (A/A*) Ma2 Note that A*2 > A*1. There are alternative ways to manipulate the functions and tables to obtain the answers. For example, we could use and T2 = (Ans) (Ans) To calculate the properties at plane 3, we note that the flow from plane 2 to plane 3 is isentropic, although with different values of p0y p T T   and  p3 = p01 ) ( ) ) / ) p0x Ma1 p0 3 T0 3 T0 2 instead of the equations we used. This example demonstrates how to use the shock functions and tables to patch together different isentropic flows at a shock. 11.7 One-Dimensional Flow in a Variable Area Duct 639 Mass Flow Relations and Choking. Several alternative equations can be used for calculation of mass flow rate in a duct. The “primitive variable” equation for mass flow rate is m ˙ = ρVA. (11.95) If the fluid is an ideal gas, ρ= p  and V = Ma(c) = Ma√kRT, RT and an alternative equation for mass flow rate is m ˙= k pAMa . B R ( √T ) (11.96) Both Eqs. 11.95 and 11.96 evaluate the mass flow rate in terms of static properties. By substituting p = p0[1 + ( −k(k−1) k−1 2 Ma ] 2 ) and T = T0[1 + ( −1 k−1 2 Ma ] , 2 ) we obtain the mass flow rate in terms of stagnation properties: m ˙ = √k( − (k+1)2(k−1) k−1 2 Ma 1 + Ma . [ ( 2 ) ] √RT0 ) p0 A (11.97) This equation has certain advantages for isentropic flow because p0 and T0 are constants. Any of Eqs. (11.95)–(11.97) can be used to calculate mass flow rate. The choice between them depends on what data are available. By setting Ma = 1 and A = A* in Eq. (11.97), we get m ˙= (k+1)2(k−1) 2 √k . (k + 1) √RT0 p0 A* (11.98) Equation 11.98 shows that A* is constant in steady isentropic flow with constant values of m ˙ , p0, T0, R, and k. We can obtain a dimensionless mass flow function from Eq. (11.97): − (k+1)2(k−1) ˙ √RT0 m k−1 = √kMa (1 + Ma2) . p0 A 2 (11.99) A plot of this function is shown in Fig. 11.11. The key fact about Fig. 11.11 for this discussion is the occurrence of the maximum at Ma = 1: ˙ √RT0 m = f (k only). p0A )max We can draw several conclusions from this fact. ■ ■ Choked flow occurs when the Mach number is 1.0 at the minimum crosssectional area. ■ If p0 and T0 are fixed (isentropic flow), a specified mass flow rate can be forced through a limiting minimum area. No smaller area can pass that flow. If p0 and T0 are fixed (isentropic flow), any particular fixed-geometry duct, with fixed minimum area, can pass only a certain maximum mass flow rate. This maximum mass flow rate occurs when the Mach number at the minimum area is one. Once the Mach number at the minimum area becomes one, it is not possible to increase the mass flow rate. It is possible to increase the limiting mass flow rate in a fixed geometry duct or to force a specified mass flow through a smaller area by increasing p0 √T0 . Doing work on a gas increases both p0 and T0; however, the ratio p0 √T0 is usually increased by compression. The preceding statements describe the phenomenon called choking. 640 Chapter 11 ■ Compressible Flow 0.8 0.7267 0.6847 0.7 Mass flow function 0.6 0.6673 0.5 0.4 k = 1.3 k = 1.4 k = 1.67 0.3 0.2 0.1 0 0 0.5 1 1.5 2 2.5 3 Mach number ■ Figure 11.11 Dimensionless mass flow function [(m ˙ √RT0 )(p0 A)]. The choking phenomenon occurs in compressible duct flow when the local Mach number reaches 1 at the minimum area in the duct. When this occurs, the mass flow rate through the duct cannot be increased unless the ratio of stagnation pressure to square root of stagnation temperature is increased. EXAMPLE 11.8 Design of a Compressible Flow Passage pe = 101.33 kPa GIVEN The duct between the turbine and exhaust nozzle of a jet engine is 0.60 m in diameter. The gas flowing in the duct has stagnation pressure and temperature of 275 kPa and 700 K, respectively. The gas flow rate is 82 kg/s. Find The relevant areas of a nozzle for expanding the gas to p0 = 275 kPa T0 = 700 K 0.6 m standard atmospheric pressure. Calculate the force required to hold the nozzle to the duct. The gas has k = 1.4 and R = 265 N ∙ m/kg ∙ K. Flow m• = 82 kg/s Inlet i Unknown nozzle shape Exit e ■ Figure 11.8a Schematic diagram of nozzle SOLUTION design problem. First establish what the nozzle geometry should be. The desired static-stagnation pressure ratio at the exit plane is The nozzle exit area is that required to produce a Mach number of 1.285 is: A Ae = Ath . A* )Ma p pe 101.33 kPa = = = 0.3684. p0 )e p0e 275.0 kPa From Table D.I, the desired exit Mach number is e Mae = 1.285. From Table D.I, We assume that the inlet Mach number will be less than 1, so we need a minimum area (throat) between the entrance and the exit. We calculate the throat area from Eq. (11.98), with k = 1.4: m ˙ = 0.6847 For Ath = A*, Ath = ˙ √RT0 m 0.6847 p0 = p0 A* √RT0 . 82 kgs√(265 N ∙ mkg ∙ K) (700K) 0.6847(275,000 Nm2 ) Ath = 0.1876 m2. ; To calculate the force, we apply the linear moment equation to a control volume enclosing the nozzle, as shown in Fig. E11.8b. We assume that the anchoring force, F, acts in the flow direction. The momentum equation gives F = ( pe + ρeV 2e )Ae − ( pi + ρiV 2i )Ai + patm (Ai − Ae ). (Ans) From Eq. 11.48, The inlet area of the nozzle is π π Ai = D2i = (0.6 m) 2; 4 4 Ai = 0.2874 m2. A = 1.0601 so Ae = 0.1876 m2 (1.0601); A* )Ma=1.285 Ae = 0.1989 m2. (Ans) ρV 2 = kpMa2, so (Ans) F = pe Ae (1 + kMa2e ) − pi Ai (1 + kMa2i ) + patm (Ai − Ae ). 11.7 One-Dimensional Flow in a Variable Area Duct 641 The force F is Force Vi F Ve Control surface ■ Figure 11.8b Control volume for nozzle force analysis. F = (101,303 Nm2 ) (0.1989 m2 )[1 + 1.4(1.285) 2] − (243,700 Nm2 ) (0.2874 m2 )[1 + 1.4(0.419) 2] + 101,330 N/m2 (0.2874 m2 − 0.1989 m2 ) = −11,500 N. F = 11,500 N Opposite the flow direction. (Ans) COMMENTS A ‘‘force function,’’ pA(1 + k Ma2 ) F ≡ = F* p*A*(1 + k) 1 + kMa2   We find Mai and pi from Ai 0.2874 m2 A = = = 1.5319. A* )Mai Ath 0.1876 m2 From Table D.I (using subsonic Mach number), p Mai = 0.419 and  ) = 0.8862. p0 i Then pi = (0.8862) (275 kPa) = 243.7 kPa . Ma B 2(k + 1) (1 + k−1 2 Ma ) 2 is listed in some compressible flow tables [Ref. 4] We treated patm and pe as separate entities in the momentum equation even though they are equal for this problem. In some cases, the pressure at the exit of a supersonic nozzle is not equal to the pressure of the surrounding atmosphere (Section 11.7.4). The nozzle flow was calculated as isentropic because the exit Mach number was to be supersonic; therefore no shocks should occur in the nozzle. The flow in the nozzle is choked. The given flow rate (82 kg/s) with the given stagnation properties (275 kPa, 700K) cannot pass through any area smaller than 0.1876 m2. THE WIDE WORLD OF FLUIDS Liquid knife A supersonic stream of liquid nitrogen is capable of cracks and crevices of the material being cut, then expands rap- cutting through engineering materials like steel and concrete. idly and breaks up the solid material it has penetrated. After Originally developed at the Idaho National Engineering Laboratory doing its work, the nitrogen gas simply becomes part of the for cutting open barrels of waste products, this technology is atmosphere, which is mostly nitrogen already. This technology is now more widely available. The fast-moving nitrogen enters the also useful for stripping coatings even from delicate surfaces. 11.7.3 Operation of a Converging Nozzle We can show how the information about compressible flow in variable-area ducts fits together by examining the flow in a converging nozzle. Figure 11.12 shows a converging nozzle that exhausts gas from a large reservoir to a variable-pressure receiver. We assume that the pressure and temperature in the reservoir are constant. We can vary the pressure in the exhaust region (called the back pressure) by means of the control valve that connects the receiver to a downstream vacuum pump. By opening or closing the control valve, we decrease or increase the back pressure and therefore expect to increase or decrease the nozzle flow rate. We assume that there is no wall friction and no heat transfer or work. We can make several preliminary observations about the flow in the nozzle. ■ ■ ■ ■ The nozzle is only convergent, so the flow cannot pass through Ma = 1. The flow at the nozzle inlet is subsonic (Ma ≈ 0), so the flow in the entire nozzle is subsonic, with the possible exception of the nozzle exit. The flow cannot be supersonic in the nozzle, so there can be no shocks and the flow is isentropic everywhere in the nozzle. The stagnation properties are constant and equal to the gas properties in the reservoir. The maximum possible Mach number in the nozzle is one. This value can occur only at the nozzle exit (minimum area). 642 Chapter 11 ■ Compressible Flow Exhaust receiver Control valve Gas Pressure p0 Temperature T0 Supply reservoir ■ To Vacuum pump ■ Figure 11.12 Apparatus for studying flow Converging nozzle in a converging nozzle. There is a maximum possible mass flow rate that can occur. This maximum is determined by the gas constants (k and R), the values of the reservoir (stagnation) properties, and the exit area. The maximum occurs only when the Mach number is one at the nozzle exit. The flow through the nozzle is described in terms of the pressure and Mach number distribution in the nozzle and the (dimensionless) mass flow rate (m ˙ √RT0 /p0 Ae ), all as functions of the ratio of back pressure to supply pressure and k. We also plot the ratio of pressure in the nozzle exit plane (pe) to the back pressure (pe pB ). The various curves are shown in Fig. 11.13. The curves and points labeled a correspond to a closed control valve. There is no flow. The pressure equals the reservoir pressure everywhere, and the Mach number is zero everywhere. Case b corresponds to a slight opening of the control valve. The back pressure is less than the supply pressure and there is flow. The gas accelerates from the reservoir to the nozzle exit, and the pressure drops. Minimum pressure and maximum Mach number are at the nozzle exit. The pressure at the nozzle exit is equal to the back pressure. If we know the ratio pBp0 , the equations pe pB =  and  Mae = Ma) pep0 p0 p0 enable us to determine the Mach number at the nozzle exit. We can then calculate the mass flow at the exit (and hence in the entire nozzle) from Eq. 11.98. p0 T0 a 1 pB b x Ae Mae f pe /p0 pe c e d p*/p0 a f Mae = 1 e d 1 c m• √RT0 p0 Ae b c d p*/p0 b e f p*/p0 a x (a) pB /p0 (b) p/p0 (c) ■ Figure 11.13 Performance of a converging nozzle at various pressure ratios. pB /p0 1 11.7 One-Dimensional Flow in a Variable Area Duct 643 Case c is similar to case b, except that a larger control valve opening permits a lower back pressure with correspondingly higher mass flow and exit Mach number and a lower exit pressure. Note that the maximum Mach number still occurs at the nozzle exit and is less than 1. In case d, the control valve has been opened a sufficient amount to bring the exit Mach number to a value of 1.0. The exit pressure now equals the critical pressure (p*) and also equals the back pressure. Knowing Mae = 1.0 allows calculation of the mass flow rate. Case e corresponds to opening the control valve farther than in case d. When we try this, we find that no changes occur in the nozzle. In case d, we reached the limit of the nozzle’s capability. The exit Mach number cannot exceed 1, the exit pressure cannot drop below the critical pressure, which is determined solely by the supply pressure and the specific heat ratio (Eq. 11.96), and the mass flow cannot exceed the choked value. The only difference between case d and case e is that the back pressure and exit pressure are no longer equal. The flow must adjust to the lower back pressure after leaving the nozzle. The downstream flow is multidimensional, so the pressure curve is shown as a wavy line downstream from the nozzle exit. Opening the control valve with further lowering of the back pressure does not change the nozzle flow. At first, these results may seem strange; however, they have a simple physical explanation. Once the exit flow becomes sonic, pressure waves from the exhaust region cannot propagate upstream into the nozzle and supply reservoir to cause any flow adjustment. There are two flow regimes in a simple converging nozzle: unchoked flow and choked flow. Which flow regime occurs in a given case depends on the relative values of the back pressure and critical pressure. ■ ■ If pBp0 > p*p0 , the flow is unchoked. If pBp0 ≤ p*p0 , the flow is choked. The critical pressure ratio (p*p0 ) depends only on specific heat ratio. For a conversing nozzle, the occurrence of choked flow does not depend on nozzle geometry. You might be curious about the generality of these results. Although it appears that we have been discussing a very specific system, we did so only for the purpose of illustration. The large reservoir represents a supply of gas with specific stagnation conditions. The exhaust receiver/ control valve/vacuum pump represents the resistance to flow downstream from the nozzle. In principle, considering a compressor and control valve upstream from the nozzle and a constant-pressure exhaust region would be equally valid. In this case, all of the dimensionless plots of Fig. 11.13 would still be valid; however, if we permit p0 to increase, we can increase the dimensional mass flow rate (m ˙ ), even at choked conditions. Finally, we have shown a long, smoothly contoured nozzle. The principles apply to converging nozzles of any length, even orifices; however, the departure from one-dimensional flow conditions becomes significant in short nozzles. 11.7.4 Operation of a Converging–Diverging Nozzle We complete our consideration of compressible flow in variable-area ducts by discussing the flow in a converging–diverging nozzle. Figure 11.14 shows a converging–diverging nozzle that exhausts gas from a large reservoir at constant pressure and temperature to a variable-pressure receiver. The back pressure can be varied by opening or closing a control valve (similar to Fig. 11.12). Generalizations of this arrangement are similar to those discussed for the converging nozzle. We assume that there is no wall friction, heat transfer, or work. Exhaust receiver Gas Pressure p0 1 Temperature T0 Supply reservoir Converging–diverging nozzle Control valve To vacuum pump ■ Figure 11.14 Apparatus for studying flow in a converging–diverging nozzle. 644 Chapter 11 ■ Compressible Flow Preliminary observations about the flow in the converging–diverging nozzle are as follows: ■ ■ ■ ■ ■ ■ The nozzle is convergent–divergent, so the flow can pass through Ma = 1. The flow also can be subsonic everywhere. If Ma = 1 anywhere, it must be at the throat. There may be supersonic flow in the divergent portion of the nozzle, so there may be shocks in the flow. If there are shocks, the flow is not completely isentropic. If there are no shocks, the flow is isentropic. If there are shocks, the flow from the reservoir up to the first shock is isentropic. Flow downstream from a shock also is isentropic but with different values of entropy, stagnation pressure, and critical area. The maximum possible Mach number that can occur anywhere in the passage corresponds to acceleration of the fluid in an isentropic process from the reservoir to the nozzle exit. The maximum possible Mach number can occur only at the exit and is determined by the ratio of exit area to throat area. The maximum possible mass flow rate in the nozzle is determined by the gas constants, the reservoir properties, and the minimum area, which occurs at the throat. The flow in the converging–diverging nozzle can be described by plots of pressure and Mach number distributions in the nozzle, the dimensionless mass flow rate (m ˙ √RT0p01Ath ), and the ratio of nozzle exit pressure to supply pressure (pe p01 ), all as functions of the ratio of back pressure to supply pressure (pBp01 ). The nozzle performance curves are shown in Fig. 11.15. The case labeled a corresponds to complete closure of the control valve. There is no flow, the pressure is uniform, and the Mach number is zero everywhere. Curves and points labeled b represent a slightly open control valve. The gas accelerates in the convergent portion of the nozzle, and the pressure drops. The throat Mach number is considerably less than 1 (subsonic flow). The flow decelerates in the divergent portion of the nozzle, and pressure rises. The pressure and Mach number distributions are roughly symmetrical about the throat. There are no shocks, because the flow is subsonic everywhere. The flow at the nozzle exit is subsonic, and the exit and back pressures are equal. The exhaust Mach number can be found from p pB = , p0 )Mae p01 and the mass flow can be calculated from Eq. (11.97). A value of A* can be calculated from Eq. (11.93) or Eq. (11.98) applied at the exhaust plane. A* will be less than Ath. Curve c represents a slightly larger control valve opening. The situation is qualitatively similar to case b, but the mass flow is larger. The pressure and Mach number distributions are still roughly symmetrical about the throat. The maximum Mach number occurs at the throat. p1 Ma1 Shock p2 (case e) pe pe /p0 1 Mae Ma2 (b) Ath p*/p0 1 Ma < 1 Ma = 1 Ma > 1 (a) f ab c d e f g h i j Length rpc 1 m• √RT0 p0 Ae e d c b g h rpc rpc 3 1 1 g j i Ae p/p0 a 1 pB variable 2 rpc 1 1 j i h gf e d c pB /p0 b 1 rpc 2 rpc a 3 (c) rpc 1 ■ Figure 11.15 Performance of converging–diverging nozzle at various pressure ratios. 1 pB /p0 1 11.7 One-Dimensional Flow in a Variable Area Duct 645 In case d, the control valve has been opened just enough to bring the throat Mach number exactly to 1.0. The flow is still subsonic everywhere except exactly at the throat, where Ma = 1.0. The flow is isentropic, but now pth = p*  and  Ath = A*. The mass flow has reached its maximum possible value and the flow has just become choked. The pressure rises downstream from the throat, so the back pressure at which the converging–diverging nozzle chokes is greater than p*. The pressure ratio that causes choking to first occur in a converging– diverging nozzle is called the first critical pressure ratio (rpc1 ) and can be calculated by rpc1 = A variety of flow situations can occur for flow in a converging– diverging duct. VIDEO V11.6 Rocket engine start-up p p0 )AA*=Ae /Ath, Ma < 1 (11.100) that is, by finding the pressure ratio corresponding to the subsonic value of A/A*. What happens when the valve is opened beyond the case d point? Because the throat is choked, the mass flow cannot be increased, and conditions upstream of the throat cannot be affected. There is a response in the flow downstream from the throat, because, at case d, the downstream flow is subsonic. When the valve is opened and the back pressure is lowered, the fluid begins to accelerate as it enters the divergent portion of the nozzle; that is, the flow becomes supersonic. If the control valve is opened only slightly past the case d point, the downstream resistance is too large for complete acceleration in the divergent portion of the nozzle, so a shock occurs in the divergent portion. Curves and points labeled e represent the flow for a slightly more open control valve. The flow accelerates in the convergent portion of the nozzle, reaches sonic speed at the throat, and accelerates to supersonic speed downstream from the throat. The supersonic acceleration terminates in a shock wave. Downstream from the shock, the flow experiences subsonic deceleration and exits the passage with Mae < 1. The exit pressure and back pressure are equal. The exact position of the shock depends on the back pressure, and for a given back pressure, is fixed. The mass flow rate for case e, as well as for all lower values of back pressure, is the same as for case d. Opening the control valve and lowering the back pressure causes the shock to move downstream (case f ). Note that once shock passes a plane, the flow up to that plane is no longer affected by lowering the back pressure. As the valve is opened farther, the shock is eventually drawn to the nozzle exit. The flow accelerates isentropically all the way from the reservoir to the exit shock. Exactly at the exit, the pressure jumps to the back pressure as the fluid exits through the shock. This situation is shown as case g. The exit pressure is double-valued at case g. Lowering the back pressure further causes the shock to move out of the nozzle and become multidimensional (case h). The flow accelerates isentropically from the reservoir to the exit. The gas exits the nozzle supersonically with Mae corresponding to the nozzle’s exit-to-throat area ratio. The exit pressure is determined only by the stagnation pressure (p01 ) and the exit Mach number and is not equal to the back pressure. The gas adjusts to the back pressure externally. If we continue to lower the back pressure, it eventually reaches equality with the exit pressure (case i). At this condition, there is no pressure adjustment in the exiting gas and the flow is said to be perfectly expanded. Lowering the pressure further (case j) requires external expansion pressure adjustments but does not affect the nozzle flow. We can summarize this information about converging–diverging nozzle flow as follows. There are four regimes of flow in a converging–diverging nozzle. ■ ■ ■ ■ Venturi regime (cases a–d). The flow is subsonic and isentropic everywhere. The flow accelerates in the convergent portion and decelerates in the divergent portion. Maximum Mach number and minimum pressure occur at the throat. Shock regime (cases d–g). The flow is subsonic in the convergent portion, sonic at the throat, and partly supersonic in the divergent portion. The acceleration terminates in a shock that stands in the divergent portion at a location determined by the exact value of the back pressure. The flow experiences subsonic deceleration from the shock to the exit. The flow is choked. Overexpanded regime (cases g–i). The flow accelerates throughout the nozzle. The throat flow is sonic, and the exit flow is supersonic. The pressure of the gas increases to the back pressure downstream from the nozzle exit. Underexpanded regime (cases i–j). This case is similar to the overexpanded regime, except that the external pressure adjustments are expansive rather than compressive. 646 VIDEO Chapter 11 ■ Compressible Flow V11.7 Supersonic nozzle flow The boundaries between the four flow regimes are indicated by curves d, g, and i. The ratios of back pressure to supply pressure that correspond to these cases are called the first critical pressure ratio (rpc1 ), the second critical pressure ratio (rpc2 ), and the third critical pressure ratio (rpc3 ). These ratios are calculated in the following manner. ■ ■ First critical pressure ratio: See Eq. (11.100). Second critical pressure ratio: Assume that a shock stands at the nozzle exit and calculate the ratio of downstream static to upstream stagnation pressure: rpc2 = where py px )Mae py px )Mae × rpc3 (11.101) is the shock static pressure ratio function and Mae is given by Mae = Ma) A/A*=Ae /Ath, Ma > 1 ■ Third critical pressure ratio: rpc3 = p . p0 )Mae (11.102) Calculation of these ratios is easier than it looks. Note that the three critical pressure ratios for a converging–diverging nozzle depend only on the nozzle geometry and the specific heat ratio of the flowing gas. The first step in calculating flow through a converging–diverging nozzle is to classify the flow. You do so by calculating the three critical pressure ratios and comparing them to the actual ratio of back pressure to supply pressure. After you have identified the proper flow regime, the rest of the problem falls into place easily. THE WIDE WORLD OF FLUIDS Rocket nozzles To develop the massive thrust needed for exiting into the atmosphere supersonically. At launch, the Space Shuttle liftoff, the gas leaving the rocket nozzles had to static pressure of the gas flowing from the nozzle exit is less be moving supersonically. For this to happen, the nozzle flow than atmospheric and so the flow is overexpanded. At higher path must first converge, then diverge as shown in the figure elevations where the atmospheric pressure is much less than in the margin. Entering the nozzle at very high pressure and at launch level, the static pressure of the gas flowing from the temperature, the gas accelerates in the converging portion of nozzle exit is greater than atmospheric and so now the flow is the nozzle until the flow chokes at the nozzle throat. underexpanded. The result is expansion or divergence of the Downstream of the throat, the gas further accelerates in the exhaust gas plume as it exits into the atmosphere. (See diverging portion of the nozzle (area ratio of 77.5 to 1), finally Problem 11.49.) EXAMPLE 11.9 Converging and Converging—Diverging Nozzles GIVEN A thick-walled pressure vessel has a hole in its side. p = 14.6 psia Figure E11.9 shows two possible configurations for the hole. The pressure and temperature of the air inside the vessel are 20 psia and 560°R, respectively. The pressure outside the vessel is 14.6 psia. Ainside = 0.02 in2 FIND The leak rate of air for each hole configuration. SOLUTION First recognize that hole a acts like a simply converging nozzle and hole b acts like a converging–diverging nozzle. The “converging portion’’ of hole b is formed by streamlines within the tank. Both holes have the same minimum area and the supply temperature and pressure are the same, so the maximum possible flow rate is the same for both holes. If both holes are choked, they have the same flow rate. We first check both holes for choking. Aoutside = 0.01 in2 P = 20 psia T = 100.3°F Hole a or Aoutside = 0.02 in2 Ainside = 0.01 in 2 Hole b Figure E11.9 Thick-walled pressure vessel with two possible hole configurations. 11.8 Hole a Hole a will be choked if Constant-Area Duct Flow With Friction To calculate the leak rate in hole a, we need the Mach number at the exit. For pB p* < = 0.5283. p01 p0 pe pb = = 0.730, p0 p0 For hole a, Table D.I gives Mae = 0.686. 14.6 psia pB = = 0.730. p01 20 psia Using Eq. (11.99), we have Hole a will not be choked. ˙a = m Hole b Hole b will be choked if where p Ae where A/A* = = 0.020.01 = 2.0, Ma < 1. p0 )AA* Ath rpc3 = rpc1 = 0.9371. The back pressure ratio is still 0.730, so hole b is choked. Next, calculate the mass flow through hole b using Eq. (11.98), with k = 1.4 and Ma = 1: m ˙ b = 0.6847( √RT ) 0 11.8 (20 lb/in2 ) (0.01 in2 )√32.2 ft ∙ lbm/lb ∙ s2 (Ans) p Ae where AA* = = 2, Ma > 1 = rpc3 = 0.0935 ) p0 A/A* Ath (note: Mae = 2.20) and rpc2 = p0 Amin √(53.35 ft ∙ lb/lbm ∙ °R) (560 °R) ˙ b = 0.00450 lbm/s. m √(53.35 ft ∙ lb/lbm ∙ °R) (560°R) COMMENTS These answers are only approximations, because the flow through real holes in the wall of a pressure vessel would probably not be one-dimensional. The third and second critical pressure ratios for hole b are From Table D.I = 0.6847[ (20 lb/in2 ) (0.01 in2 )√32.2 ft ∙ lbm/lb ∙ s2 × √1.4(0.686)[1 + (0.2) (0.686) 2] −3; m ˙ a = 0.00407 lbm/s. pB < rpc1, p01 rpc1 = 647 py px )Ma=2.20 × rpc3. Using Table D.S, we have ]; (Ans) rpc2 = (5.480) (0.0935) = 0.512. Hole b is operating in the shock regime. Constant-Area Duct Flow With Friction We now examine the effects of wall friction in one-dimensional compressible flow in a duct. To simplify our discussion, we limit consideration to flow in ducts of constant cross-sectional area. We assume that the flow is one-dimensional, with the velocity varying only with duct length but not with distance across the duct. This situation is somewhat artificial, because a transverse velocity gradient is always associated with shear stress. The situation that we have assumed is most closely approximated by turbulent flow. The single value of fluid velocity associated with any duct cross section should be interpreted as an average velocity. 11.8.1 Preliminary Consideration: Comparison with Incompressible Duct Flow Flow of an incompressible fluid in a constant-area duct is one of the most important practical problems in fluid mechanics and was considered extensively in Chapter 8. Incompressible flow in long ducts eventually becomes fully developed, and the velocity profile does not change shape. Continuity requires that the average velocity remain constant, so neither velocity profile shape nor magnitude vary in fully developed incompressible flow. Therefore the fluid’s kinetic energy and momentum are constant. The retarding effect of fluid friction is balanced by a pressure drop in the flow direction. This pressure drop is directly proportional to the irreversible loss of mechanical energy caused by fluid friction. What changes does fluid compressibility introduce? If the density changes, the average velocity must also change to satisfy the continuity equation. Changing the average velocity changes the magnitude of the entire velocity profile. More important, if the velocity changes, the fluid’s kinetic 648 Chapter 11 ■ Compressible Flow Insulated wall Adiabatic flow Section (1) Section (2) Control volume Fanno flow involves wall friction with no heat transfer and constant crosssectional area. ■ Figure 11.16 Adiabatic constant area flow. energy and momentum also change. Even if the flow is fully developed, the retarding effect of fluid friction is balanced by both pressure and momentum changes. Although fluid friction causes an irreversible loss of mechanical energy, this loss cannot be interpreted in terms of a (static) pressure drop. Pressure changes in compressible flow, cannot be calculated from the Darcy–Weisbach equation. In compressible flow, friction appears explicitly only in the momentum equation. The effects of friction on flow can be determined only by simultaneous solution of the equations of continuity, momentum, energy, and state. We must use the general energy equation, so we make an explicit assumption about work and heat transfer. In keeping with the approach of considering inputs one at a time, in addition to constant area, we assume that there is no shaft work and that there is no heat transfer. Adiabatic flow with friction in a constant area duct is called Fanno flow after Gino Girolamo Fanno, an Italian mechanical engineer who developed the analysis. If the duct is reasonably short, the adiabatic flow assumption is reasonable. Because compressibility effects usually occur only if the fluid velocity is large and because, on dimensional grounds, mechanical energy loss increases as the square of the velocity, ducts in which friction and compressibility are both significant are usually kept short and heat transfer can be neglected. 11.8.2 The Fanno Line As we did with variable-area flow, we will first investigate what we can learn about Fanno flow without specifying any particular fluid. Consider the steady, one-dimensional, and adiabatic flow of a compressible fluid through the constant area duct shown in Fig. 11.16. This is Fanno flow. For the control volume indicated, the energy equation (Eq. 5.69) leads to 0(negligibly small for gas flow) 0(flow is adiabatic) 0(no compressors or turbines) V 2 − V 21 m ˙ [hˇ 2 − hˇ 1 + 2 + g(z2 − z1 ) ] = Q̇ net + Ẇ shaft 2 in. net in or V2 hˇ + = hˇ 0 = constant 2 (11.103) where hˇ 0 is the stagnation enthalpy, which is constant for flow with no work and no heat transfer. The continuity equation for the control volume shown in Fig. 11.16 is ρ1V1A1 = ρ2V2A2 . Since the area is constant, this can be reduced to ρV = G = constant (11.104) where G, the density-velocity product, is called the mass flow density. Combining Eqs. 11.103 and 11.104 we get (ρV) 2 = hˇ 0 hˇ + 2ρ2 or G2 hˇ + 2 = hˇ 0 . 2ρ (11.105) 11.8 T Constant-Area Duct Flow With Friction 649 Constant entropy line 1 pa 2 a T0 Ta Fanno line ■ Figure 11.17 The T–s diagram for s hˇ ρ Enthalpy-Density Curve for Fanno Flow Entropy increases in Fanno flows because of wall friction. Fanno flow. With G and hˇ 0 constant, Eq. 11.105 is a relationship between hˇ and ρ (or hˇ and v, since v = 1/ ρ), which, for specific values of G and hˇ 0, can be graphed as a line, as shown on the figure in the margin. This line is called a Fanno line. It represents all possible states of the fluid for specific values of G and hˇ 0. For any simple compressible fluid, if two properties are known, the other properties can be determined. If we pick several points on the hˇ – ρ curve, we can determine corresponding values of temperature and entropy and graph the Fanno line on a T –s diagram, as shown in Fig. 11.17. We can observe many important features of Fanno flow by examining Fig. 11.17. Any flow state, such as that at Section 1 in Fig. 11.16, can be shown as a single point on the corresponding Fanno line. The action of friction between the duct wall and the fluid between Section 1 and Section 2 causes the fluid properties to change and the state point on the Fanno line to move from point 1 to point 2. The Second Law of Thermodynamics tells us that entropy cannot decrease in an adiabatic process so the state point must always move from left to right, toward maximum entropy, point a. The fluid kinetic energy (actually, V 22cp) is represented by the vertical distance between the “T0” line and the Fanno line. We can draw two conclusions from this. First, the “upper branch” of the Fanno line represents slower speeds and the “lower branch” represents faster speeds. Second, along the lower branch, friction causes the fluid to slow down, while along the upper branch, friction actually causes the fluid to speed up! Both of these actions are different from incompressible flow where the fluid speed remains constant in fully developed flow. Point a, the point of maximum entropy, is apparently quite significant. We proceed to find the fluid speed at that point. Fanno flow is governed by the continuity equation, Eq. 11.104, and the energy equation, Eq. 11.105. Differentiating these equations, ρdV + Vdρ = 0 (11.106) dhˇ + VdV = 0. (11.107) and From Eq. 11.106 dρ ρ (11.108) dρ dhˇ = V 2 ρ (11.109) dp dhˇ = Tds − ρ (11.110) dV = −V Substituting Eq. 11.108 into Eq. 11.107 gives From the second Tds equation, Eq. 11.20 The equations above, Eqs. 11.106–11.110, apply anywhere along the Fanno line. At the maximum entropy point, point a, ds = 0 and Eq. 11.110 becomes dp dhˇ = ρ (11.111) 650 Chapter 11 ■ Compressible Flow T T T * Nor mal shoc 2 k 2 1 * 2 1 * 1 s s (a) (b) s (c) ■ Figure 11.18 (a) Subsonic Fanno flow. (b) Supersonic Fanno flow. (c) Normal shock occurrence in Fanno flow. Substituting Eq. 11.111 into 11.109 and rearranging V2 = dp dρ )a But since point a is the point of maximum entropy, s is locally constant there, so we find V 2a = Friction accelerates a subsonic Fanno flow. ∂p = c2 at point a ∂ρ )s That is, the fluid speed equals the sonic speed and Ma = 1 at the point of maximum entropy. It follows that the flow described by the upper (lower speed) branch is subsonic and the flow described by the lower (higher speed) branch is supersonic. We can replace the “a” designation of the maximum entropy point with the more familiar designation of sonic flow, *, thus the temperature is T *, the pressure is p*, and so on. Because the entropy can only increase in an adiabatic process, Fanno flow is always driven toward Ma = 1 by friction. It also follows that the flow cannot pass through Ma = 1. This is another manifestation of choking; when the flow reaches sonic speed in a duct, the flow cannot respond to further frictional forces. This in turn means that sonic speed can only occur at the end of a frictional duct. One might ask “What happens when a Fanno flow is choked and more frictional duct is added?” The answer is that the mass flowrate m ˙ and the mass flow density G will be reduced and the flow will be described by another Fanno line, one shifted to the right. From the Fanno line itself, the effect of friction on the fluid temperature and the fluid entropy can be noted. The effect of friction on static pressure is not surprising; when velocity increases, pressure decreases and vice versa. Lastly, although the flow cannot pass through the point of maximum entropy, it is possible to “jump” from the supersonic branch to the subsonic branch by passing through a normal shock wave. Some examples of Fanno flow behavior are shown in Fig. 11.18. A case involving subsonic Fanno flow that is accelerated by friction to a higher Mach number without choking is illustrated in Fig. 11.18a. A supersonic flow that is decelerated by friction to a lower Mach number without choking is illustrated in Fig. 11.18b. In Fig. 11.18c, an abrupt change from supersonic to subsonic flow in the Fanno duct is represented. This sudden deceleration occurs across a standing normal shock. The qualitative aspects of Fanno flow that we have already discussed are summarized in Table 11.3 and Fig. 11.19. Movement of the state point along the Fanno line is caused by friction and governed by the linear momentum equation. Applying this equation (Eq. 5.22) to the flow through the control volume sketched in Fig. 11.20a gives p1A1 − p2A2 − Rx = m ˙ (V2 − V1 ) 11.8 p0, I ■ Table 11.3 Summary of Fanno Flow Behavior p0, a T0 Flow Parameter Subsonic Flow Supersonic Flow Stagnation temperature Ma Constant Increases (maximum is 1) Accelerates flow Decreases Decreases Constant Decreases (minimum is 1) Decelerates flow Increases Increases Friction Pressure Temperature 651 Constant-Area Duct Flow With Friction p1 T1 a T* T1' p1' ■ Figure 11.19 Fanno line. where Rx is the frictional force exerted by the inner pipe wall on the fluid. Since A1 = A2 = A and m ˙ = ρAV = constant (i.e. ρV = ρ1V1 = ρ2V2 ), we obtain Rx = ρV(V2 − V1 ) A p1 − p2 − (11.112) The differential form of Eq. 11.112, which is valid for Fanno flow through the semi-infinitesimal control volume shown in Fig. 11.20b, is −dp − τw π D dx = ρV dV A (11.113) The wall shear stress, τw, is related to the wall friction factor, f, by Eq. 8.20 as Friction forces in Fanno flow are given in terms of the friction factor. f= 8τw ρV 2 (11.114) By substituting Eq. 11.114 and A = πD2 4 into Eq. 11.113, we obtain −dp − fρ V 2 dx = ρV dV. 2 D (11.115) Dividing by p gives the differential momentum equation in dimensionless form ρV 2 f dx dp ρ d(V 2 ) + + =0 p p 2 2 p D (11.116) which relates fluid properties (p, ρ) and velocity to the duct friction parameter, f dx/D. For a noncircular duct, the diameter D is replaced by the hydraulic diameter, Dh = 4A/P, where P is the “wetted” perimeter. Section (1) Flow Section (2) Control volume p1A1 p2A2 Rx (a) Semi-infinitesimal control volume Flow (p + δ p)A pA D τw πD δ x δx (b) ■ Figure 11.20 (a) Finite control volume. (b) Semi-infinitesimal control volume. 652 Chapter 11 ■ Compressible Flow 11.8.3 Adiabatic Frictional Flow (Fanno Flow) of an Ideal Gas To obtain quantitative results, it is necessary to specify a particular working fluid. As usual, we will consider an ideal gas. Combining the ideal gas equation of state (Eq. 11.1), the ideal gas speed-of-sound equation (Eq. 11.46), and the Mach number definition with Eq. 11.116 leads to dp f dx Ma2 d(V 2 ) + Ma2 k + k =0 p 2 D 2 V2 (11.117) Since V = Ma c = Ma √kRT, then V 2 = Ma2 kRT or d(V 2 ) d(Ma2 ) dT = + (11.118) 2 T V Ma2 2 The application of the energy equation (Eq. 5.69) to Fanno flow gives T + V 2cp. = T0 = constant. Differentiating and dividing by T gives d(V 2 ) dT + =0 T 2cpT (11.119) Substituting Eqs. 11.14, 11.47, into Eq. 11.119 yields d(V 2 ) dT k−1 + Ma2 =0 T 2 V2 (11.120) which can be combined with Eq. 11.118 to form d(Ma2 )Ma2 d(V 2 ) = 2 V 1 + [(k − 1)2]Ma2 (11.121) We can combine the continuity and state equations with Eq. 11.118 to get dp d(Ma2 ) 1 d(V 2 ) = − p 2 V2 Ma2 (11.122) Consolidating Eqs. 11.122 and 11.117 leads to fk d(V 2 ) d(Ma2 ) 1 dx (1 + kMa2 ) − + Ma2 =0 2 2 2 D V Ma2 (11.123) Finally, incorporating Eq.11.121 into Eq.11.123 yields (1 − Ma2 ) d(Ma2 ) dx =f 2 4 D {1 + [(k − 1)2]Ma }kMa (11.124) Equation 11.124 can be integrated from one section to another in a Fanno flow duct. We elect to use the critical (*) state as a reference and to integrate Eq. 11.124 from an upstream state to the critical state. Thus ∫ Ma*=1 Ma For Fanno flow, the Mach number is a function of the distance to the critical state. (1 − Ma2 ) d(Ma2 ) = {1 + [(k − 1)2] Ma2}kMa4 ∫ ℓ ℓ* f dx D (11.125) where ℓ is length measured from an arbitrary but fixed upstream reference location to a section in the Fanno flow. For an approximate solution, we can assume that the friction factor is constant at an average value over the integration length, ℓ* − ℓ. We also consider a constant value of k. Thus, we obtain from Eq. 11.125 f(ℓ* − ℓ) [(k + 1)2]Ma2 1 (1 − Ma2 ) k+1 + ln{ = 2 k 2k D Ma 1 + [(k − 1)2]Ma2 } (11.126) 11.8 Reference section Frictionless and adiabatic duct Constant-Area Duct Flow With Friction Imagined choked flow section Section (2) Section (1) 653 D = constant Flow ℓ1 Imagined duct friction factor = f Actual duct with friction factor = f ℓ2 ℓ* (a) Reference section Frictionless and adiabatic duct Section (1) Section (2) Actual choked flow section D = constant Flow ℓ1 ℓ2 Actual duct with friction factor = f ℓ* (b) ■ Figure 11.21 (a) Unchoked Fanno flow. (b) Choked Fanno flow. For a given value of k, values of f (ℓ* − ℓ)D can be tabulated as a function of Mach number for Fanno flow. For example, values of f (ℓ* − ℓ)D for air (k = 1.4) Fanno flow are graphed as a function of Mach number in Graph D.F in Appendix D and in the figure in the margin and tabulated in Table D.F. Note that the critical state does not have to exist in the actual Fanno flow being considered, since for any two sections in a given Fanno flow D f (ℓ* – ℓ) ________ 5.0 0.0 0.1 1.0 10 f(ℓ* − ℓ2 ) f(ℓ* − ℓ1 ) f − = (ℓ1 − ℓ2 ) D D D Ma For Fanno flow, the length of duct needed to produce a given change in Mach number can be determined. (11.127) The sketch in Fig. 11.21 illustrates the physical meaning of Eq. 11.127. Recall that for a non-circular duct, D is replaced by Dh. For a given Fanno flow (constant specific heat ratio, duct diameter, and friction factor) the length of duct required to change the Mach number from Ma1 to Ma2 can be determined from Eqs. 11.126 and 11.127 or a graph such as Graph D.F or a table such as Table D.F or even software. To get the values of other fluid properties in Fanno flow we need to develop more equations. By consolidating Eqs. 11.118 and 11.120 we obtain (k − 1) dT d(Ma2 ) =− T 2{1 + [(k − 1)2]Ma2} (11.128) Integrating Eq. 11.128 from any state upstream in a Fanno flow to the critical (*) state leads to (k + 1)2 T = T* 1 + [(k − 1)2]Ma2 (11.129) Ma √kRT V T = = Ma V* T * A √kRT * (11.130) The velocity ratio is 654 Chapter 11 ■ Compressible Flow 5.0 Substituting Eq. 11.129 into 11.130 yields 12 [(k + 1)2]Ma2 V ={ V* 1 + [(k − 1)2]Ma2 } T __ T* (11.131) 1.0 0.0 0.1 1.0 10 Ma Equations 11.129 and 11.131 are graphed in the margin for air. From the continuity equation we get for Fanno flow ρ V* = ρ* V (11.132) ρ 1 + [(k − 1)2]Ma2 12 ={ ρ* [(k + 1)2]Ma2 } (11.133) 5.0 Combining 11.129 and 11.131 results in V ___ V* 1.0 0.0 0.1 1.0 10 The ideal gas equation of state (Eq. 11.1) leads to Ma p ρ T = p* ρ* T* 5.0 and merging Eqs. 11.134, 11.133, and 11.129 gives 12 p (k + 1)2 1 = p* Ma { 1 + [(k − 1)2]Ma2 } p/p* 1.0 0.0 0.1 (11.134) 1.0 10 Ma This relationship is graphed in the margin for air. Finally, the stagnation pressure ratio can be written as p0 p0 p p* = ( )( )( ) p p*0 p* p*0 For Fanno flow, thermodynamic and flow properties can be calculated as a function of Mach number. (11.135) (11.136) which by use of Eqs. 11.51 and 11.135 yields [(k+1)2(k−1)] p0 1 2 k−1 2 = 1 + Ma )] p*0 Ma [( k + 1 )( 2 (11.137) Values of f (ℓ* − ℓ)D, TT *, VV *, pp*, and p0 p*0 for Fanno flow of air (k = 1.4) are graphed as a function of Mach number in Graph D.F of Appendix D. The usefulness of Graph D.F and Table D.F is illustrated in Examples 11.10 and 11.11. See Ref. 7 for additional information about compressible internal flow. EXAMPLE 11.10 The Fanno Line ft ∙ lb (k = 1.4, R = 53.3 lbm ∙ °R ) enters an insulated constant cross-section duct with the following properties: T0 = 518.67 °R, T1 = 514.55 °R, p1 = 14.3 psia GIVEN Air FIND Plot a Fanno line on a T–s diagram SOLUTION We will plot the Fanno line from a table of values of T and s – s1 where, from Eq. 11.22, the entropy change is given by p T kR s − s1 = cp ln ( T1 ) − R ln ( p1 ) with cp = k − 1 = (1.4) (53.3 ft ∙ lblbm ∙ °R) = 187 ft ∙ lb/lbm ∙ °R. Perhaps the 1.4 − 1 simplest way to generate a table is to use the Mach number as a parameter. For any value of the Mach number, p T )Ma ) p p p* T * T T Ma = and = where and are ) p1 p T1 T* Ma p* )Ma T T * )Ma1 p* )Ma1 Fanno flow functions given by Table D.F or Graph D.F or Eqs. 11.129 and 11.135.     11.8 T01 T0 = = T1 T )Ma1 1 T01 2 2 518.67 = 1.0080 so that Ma1 = [ − 1)] = 514.55 k − 1 ( T1 The initial value Ma1 is found from Constant-Area Duct Flow With Friction 655 The Fanno line is plotted in Figure E11.10. 600 1 500 400 T, °R 2 2 (1.0080 − 1) ] = 0.20. From Table D.F [ 1.4 − 1 p T = 5.4555 and = 1.1905. Now choosing Ma = 0.4, ) p* Ma1 T* )Ma1 p T = 2.6958 and ) = 1.1628 so from Table D.F ) T * Ma=0.4 p* Ma=0.4 1.1628 o that T = 514.55 oR ( 1.1905 ) = 502.58 R and s − s1 = 300 200   ft ∙ lb 1.1628 ft ∙ lb 2.6958 (187 lbm°R ) ln ( 1.1905 ) − (53.3 lbm°R ) ln ( 5.4555 ) = ft ∙ lb 33.17 . lbm°R Repeating these calculations for other Mach numbers (using a spreadsheet) gives the table below: Ma 0.1 0.4 0.6 0.8 1 1.2 1.4 1.6 2 2.5 3.5 TT* pp* TT1 pp1 T [oR] 1.1976 10.9435 1.0060 2.0060 517.63 1.1628 2.6958 0.9767 0.4942 502.58 1.1194 1.7634 0.9403 0.3232 483.83 1.0638 1.2893 0.8936 0.2363 459.81 1.0000 1.0000 0.8400 0.1833 432.22 0.9317 0.8044 0.7826 0.1474 402.69 0.8621 0.6632 0.7241 0.1216 372.61 0.7937 0.5568 0.6667 0.1021 343.03 0.6667 0.4082 0.5600 0.0748 288.15 0.5333 0.2921 0.4480 0.0535 230.52 0.3478 0.1685 0.2922 0.0309 150.34 E X AMPLE 11.11 s – s1 [ft*lblbmR] –35.99 33.17 48.69 55.85 57.83 56.20 51.96 45.82 29.75 5.87 –44.74 100 –50.00 0.00 50.00 100.00 s – s1 ft*lb/lbm°R ■ Figure E.11.10 COMMENTS As can be seen from the figure and the table, the Mach number at the “nose” of the Fanno line is 1. The upper branch represents subsonic flow and the lower branch represents supersonic flow. According to the Second Law of Thermodynamics, since Fanno flow is an adiabatic process, the entropy can only increase; therefore, the flow state must always move toward the “nose.” Movement along the Fanno line is caused by friction in the flow, so friction always drives the Mach number toward one. Choked and Non-choked Fanno Flow Atmospheric air (T0 = 288 K, p0 = 101 kPa) is drawn through a converging nozzle into a 2-m-long, 0.1-m-diameter circular duct as shown in Fig. E11.11. The friction factor for the duct is 0.02. Assume that the flow is adiabatic (without heat transfer) everywhere and that friction is negligible in the nozzle flow. GIVEN p0 = 101 kPa (abs) Frictionless and adiabatic nozzle Section (1) Adiabatic duct with friction factor f = 0.02 Control volume D = 0.1 m ℓ=2m FIND (a) What is the maximum mass flowrate in this system? For this flowrate, determine the static and stagnation temperatures and pressures at the duct inlet and at the exit. (b) What will be the mass flowrate if the duct is shortened to 1 m and the same duct exit pressure is maintained? (c) What back pressure would be required to obtain the flowrate from Part (a) through the shorter duct of Part (b)? Section (2) Standard atmospheric air T0 = 288 K ■ Figure E11.11 Adiabatic flow in a long duct 656 Chapter 11 ■ Compressible Flow SOLUTION Part (a) The flow in the nozzle will be isentropic flow with area change and the flow in the duct will be Fanno flow. When the pressure at the duct exit (back pressure) is lowered, flow will begin in the system. As the pressure decreases, the flowrate will increase. For sufficiently high back pressures, the flow will be subsonic throughout. For any flowrate, the velocity will increase in the nozzle because of the area reduction and it will increase further in the duct because of friction. The highest velocity will always occur at the end of the duct. Eventually, as the back pressure is lowered, the velocity at the end of the duct will reach the speed of sound. The flow will then become choked and further decreases in back pressure will have no effect on the nozzle-duct system. The Mach number at the end of the duct will be 1; the Mach number at any other location in the duct (including the duct entrance/nozzle exit) will be subsonic. The key to solving this problem is to determine the Mach number at station 1, the nozzle exit/duct inlet. At section 1 f ℓ* − ℓ1 ℓ 2m = f = 0.02 = 0.4 D )1 D 0.1m (1) because the flow is choked (Ma = 1) at section 2. We can now determine Ma1 three different ways: we could solve Eq. 11.126 (which would be quite unpleasant); we could use Graph D.F (which would be quick but with limited precision); or we could use Table D.F (which would be precise but somewhat cumbersome). Here will use either the graph or the table. Either one gives Ma1 = 0.63. We now calculate p1, p01, T1, and T01 by considering the nozzle and using isentropic flow functions. The stagnation properties are those in the atmosphere upstream from the nozzle From Graph D.I or Table D.I p T = 0.76 and ) = 0.93 p0 )1 T0 1 p1 = (0.76) (101 kPa) = 77 kPa(abs) and T1 = (0.93) (288 K) = 268 K (Ans) The mass flowrate can be calculated using any one of Eqs. 11.95–11.97, with 11.96 being easiest for the data available k pAMa B R √T 1.4 B 286.9 Nm/kgK (77,000 N/m2 ) π (0.1 m) 2 (0.63) 4 √268 K p1 T1 63 kPa = = 45 kPa and T2 = T* = p 1.7 T p* )1 T * )1 268 K = = 240 K (Ans) 1.1   and p01 101 kPa = = 87.1 kPa and T02 = T *0 = p0 1.16 p*0 )1 (Ans) T01 = 288 K (T0 is constant in Fanno flow) p02 = p*0 =   Part b Now, ℓ = 1 m and f ℓ = 0.2 Also pBack = 45 kPa, D p01 = 101 kPa, and T0 = 288 K The solution to this problem hinges on whether the flow is still choked at the duct exit. If it is, we can proceed just like we did for part a. If not, we would probably have to iterate to try to match the exit pressure to the back pressure. We begin by assuming the flow is choked. We can check this assumption by comparing the duct exit pressure to the back pressure. p ℓ* − ℓ1 Taking f = 0.2 gives Ma1 = 0.70, ) = 1.5 and D p* 1 p = 0.72. Then we can calculate p0 )1 p2 = p* = p0 T2 = T0 Then = p2 = p* = p p 0.72 = (101 kPa) ( = 48.5 kPa(abs) p0 )1/ p* )1 1.5 ) We see that the back pressure (45 kPa) is less than the exit plane pressure (48.5 kPa), so the flow will be choked at the exit plane and the flow will adjust to the backpressure in a multidimensional process downstream from the duct exit. To calculate the flowrate, we will use Eq. 11.96 again but this time at the exit. The exit temperature is p01 = 101 kPa and T01 = 288 K. m ˙= Then T2 T* 2 = T0 = T0 = (288 K) (0.833) = 240 K (k + 1) T02 T0   and the mass flowrate is ˙= m k pAMa B R √T 1.4 = B 286.9 Nm/kgK kg ∙ m2 ( N ∙ m ∙ s2 ) = 1.65 kg/s 1 2 (Ans) To calculate the air properties at station 2, we use Fanno flow functions and note that Ma2 = 1 so all properties at 2 are critical properties. From Graph D.F or Table D.F at Ma1 = 0.63 p0 p T = 1.7  ) = 1.1  ) = 1.16 p* )1 T* 1 p*0 1 = 1.73 kg/s. (48,500 N/m2 ) π (0.1 m) 2 (1.0) 4 √240 K kg ∙ m2 12 ( N ∙ m ∙ s2 ) (Ans) The flowrate goes up when the duct length (and hence flow resistance) goes down. This should be no surprise. kg ℓ Part c Now we know m ˙ = 1.65 and f = 0.2 and we wish s D to find the exit plane pressure. Because the desired flowrate (1.65 kg/s) is less than the maximum (choking) flowrate (1.73 kg/s), we know that the flow is not choked and the exit plane pressure and the back pressure are equal. Because the inlet stagnation 11.8 properties and the flowrate are the same as in Part a, we know that conditions at the nozzle exit/duct entrance are Ma1 = 0.63 f p ℓ* − ℓ1 = 0.4 p1 = 77 kPa  ) = 1.7 D p* 1 Then f ℓ* − ℓ2 ℓ* − ℓ1 ℓ =f − f = 0.4 − 0.2 = 0.2 D D D From Graph D.F or Table D.F Ma2 = 0.70  p = 1.5 p* )2 so p2 = pBack = p1 p p 1.5 = 77 kPa = 68 kPa(abs) (Ans) p* )2/ p* )1 1.7 657 Constant-Area Duct Flow With Friction COMMENTS For a given duct cross-section and inlet stagna- ℓ D) determine the flowrate. Both provide resistance to the flow. The back pressure controls the flow only if it is above the choking value. This example verifies that decreasing the duct friction leads to a larger flowrate even if the duct is choked and that raising the backpressure (above the choking value) decreases the flowrate. The reverse is also true. In Part 3, we calculated the back pressure that would be necessary to produce a specified flowrate in a particular duct. How that pressure is achieved was of no concern. The problem of determining the flowrate for a specified back pressure (greater than the choking value) is more complicated, usually requiring iteration. tion pressure, the back pressure and duct friction (in the form f An important question for many practical applications is how to find f. From dimensional analysis, we expect that f = f (Re, Ma, ε . D) Based on a rather small amount of experimental evidence, dependence of f on Ma appears to be weak, at least for subsonic flow; accordingly, values of f for Ma = 0 (incompressible flow) are usually sufficient. For fully developed flow, we obtain values for f from the Moody chart (Fig. 8.20), the Colebrook equation, Eq. (8.53a), or the laminar flow friction factor equation, Eq. (8.19). Because f depends on Reynolds number, our assumption of a constant value along the duct seems questionable. The Reynolds number can be written Re = ρVD m ˙ D m˙ D = = . μ μ A A μ (11.138) The mass flow rate, diameter, and cross-sectional area are all constant, so the variation of Reynolds number along the duct results only from viscosity. For a gas, viscosity varies rather slowly with temperature (μ ∼ T 0.7 ). In addition, the dependence of f on Re is weak in turbulent flow, which is the usual condition at velocities high enough for compressibility to be significant so the assumption of constant f is surprisingly good. The following example illustrates determination of the friction factor during a Fanno flow calculation. E X AMPLE 11.12 Duct Sizing in Fanno Flow GIVEN An astronaut breathes oxygen supplied to a spacesuit FIND The required inside diameter of the umbilical to supply through an umbilical cord (Fig. E11.12a). The umbilical is 7 m long. The oxygen is supplied from a storage tank at 100 kPa and 283 K. The suit pressure is 20 kPa. 0.05 kg/s of oxygen. The inside of the umbilical is lined with a material with roughness of 0.01 mm. 658 Chapter 11 ■ Compressible Flow We can speed up the calculations considerably if we recognize that the quantity in brackets is Oxygen supply tank k + 1 (k+1)2(k−1) A/A*) Ma1 = 1.728 A/A*) Ma1, ( 2 ) for k = 1.4. Thus D2 = 4 (0.05 kg/s)√(260 N ∙ m/kg ∙ K) (283 K) ](1.728) A/A*) Ma π[ (200,000 N/m2 )√1.4 and D = √D2 = 0.0112(√A/A*) Ma1 ). (2) The ratio of back pressure to supply pressure is pBack 20 kPa = = 0.20. p01 100 kPa Umbilical To determine the diameter, we must use Eqs. 1 and 2, together with the isentropic flow table (for A/A*), the Fanno flow table (to relate duct friction and Mach number), and the Moody Chart or Colebrook formula (for f ). Clearly this will require iteration. It is most convenient to use Ma1 as the iteration parameter. We begin with ■ Figure E11.12a Astronaut being supplied with oxygen through an umbilical. Ma1(1) = 0.5. SOLUTION Figure E11.12b is a schematic diagram of the duct flow problem. The flow in the umbilical obviously is subsonic (no converging–diverging nozzle at the inlet). The properties of oxygen are R = 260 N ∙ m/kg ∙ K, k = 1.4, and μ = 1.97 × 10 −5 N ∙ s/m2. For a circular duct, from Eq. (11.138), Re = m ˙ D ˙ 4m . = μ πD24 πμD The superscript (1) indicates that this is our first guess. From Table D.I A/A*) Ma1 = 1.340. From Eq. (2), D = 0.0112√1.340 = 0.013 m. From Eq. (1), Re = Substituting numbers where possible, we have Re = (4) (0.05 kgs) (π) (1.97 × 10 −5 N ∙ sm )D 2 3230 , D = (1) where D is in meters. To calculate the flow rate at plane 1 we use Eq. 11.97. As we know m ˙ but do not know the area A, we rearrange the equation and substitute A = πD2/4 to get the following: D2 = ˙ √RT0 4 m 1 k−1 1+ Ma21) π ( p0 √k )[ Ma1 ( 2 (k+1)2(k−1) ]. Also, 1 × 10 −5 m ε = = 0.00077. D 1.3 × 10 −2 m From Fig. 8.20 (the Moody chart), f = 0.020. The duct length parameter is 1 200 kPa 283 K O2 m• 1 D=? pB = 20 kPa 2 Space suit ■ Figure E11.12b Schematic diagram of space suit umbilical. ℓ 7m = (0.020) ( = 10.77. D 0.013 m ) Next, we calculate the choking pressure ratio for this duct length parameter. We let Ma*1 be the inlet Mach number corresponding to choking. From Table D.F for f (ℓ* − ℓ1 )D = 10.77, Umbilical Tank f 7m 0.05 kg/s 3230 = 2.48 × 105. 0.013 Ma*1 = 0.227 and  p = 4.801. p* )Ma*1 For Table D.I p = 0.9648. p0 )Ma*1 11.8 The choking pressure ratio is Constant-Area Duct Flow With Friction 659 The results of this calculation are p1 p01 pB* 0.9648 = = = 0.201. ( p0 ) p 4.801 1 1 p* Because pBack p*B = 0.10 < = 0.201, p01 p01 the duct seems to be choked. Setting f(ℓ* − ℓ1 )D = f(ℓ)D = 10.77, we find that the “new” Mach number is Ma1, new = Ma*1 . As Ma*1 = 0.227 ≠ 0.5 = Ma1(1) , we now try Ma1(2) = 0.227 From Table D.I, ε = 5.92 × 10 −4, D f = 0.019, f(ℓ)D = 7.994, Ma*1 = 0.256, and D = 0.0169 m,  Re = 1.93 × 105,  p*B = 0.225 (Choked). p01 For the next iteration Ma1(4) = 0.256. The results for this trial are ε = 5.81 × 10 −4, D f = 0.019, f(ℓ)D = 7.855, Ma*1 = 0.258, and pB* = 0.226 (Choked). p01 D = 0.0172 m,  Re = 1.89 × 105,   AA*) Ma1 = 2.629 It is clear that the Mach number, to two decimal places, converges to From Eqs. (1) and (2), D = 0.0182 m and Re = 1.79 × 105. Also, Ma1 = 0.26. ε 10 −5 m = = 5.49 × 10 −4 D 1.82 × 10 −2 m and, from the Moody chart f = 0.0193. This gives f(ℓ)D = 0.0193( 7m = 7.38. 0.0182 m ) For this Mach number, the calculated parameters are D = 0.017 m, R = 1.90 × 105, ε = 5.88 × 10 −4, f = 0.019, f (ℓ)D = 7.82, D p*B and  = 0.226 (The flow is choked at the duct exit) p01 We specify that the duct size is D = 0.017 m = 1.7 cm (Ans) From Tables D.F and D.I, Ma*1 = 0.264, pp*) Ma*1 = 4.117, and pp0 ) Ma*1 = 0.9526 so pB* 0.9526 = = 0.231 (> 0.2) p01 4.117 The duct seems to be choked. We now repeat the procedure with Ma1(3) = 0.264. COMMENTS Just as with rigid pipes, the actual duct would probably not be exactly 1.7 cm in diameter. Considering the conservative design approach used in a space program, the actual duct may be twice the size, say 3.5-4.0 cm or even larger. The system would be fitted with some sort of control valve or pressure regulator. Don’t be fooled into thinking this flow could be modeled as incompressible because the inlet Mach number is low (0.26). The Mach number at the spacesuit is 1 and the flow is choked. Compressibility is very significant. A Note about Supersonic Flow. Our discussion has concentrated on subsonic frictional flow. A discussion of supersonic frictional flow would be interesting but of limited practical value compared to the effort expended for three reasons. First, Eq. (11.137) shows that loss of stagnation pressure is very large if the Mach number is greater than 1, so supersonic flow is usually avoided if friction is significant. Second, interaction between shocks and boundary layers is a complex multidimensional process. Third, the maximum duct length in supersonic flow is so short ( f (ℓ* − ℓ1 )D < 0.822) that the flow is never fully developed. The one-dimensional Fanno flow model is not sufficiently accurate to be of value. If you are interested in discussions of supersonic Fanno flow, see Reference 4. 660 Chapter 11 ■ Compressible Flow 11.9 Rayleigh flow involves heat transfer with no wall friction and constant crosssectional area. Frictionless Flow in a Constant-Area Duct with Heating or Cooling Compressibility is often significant in flows with heating or cooling. Examples include flow in steam generators, gas turbine combustion chambers, and moisture evaporation or condensation in a high-speed air stream. In order to focus on the effects of heating or cooling alone, we assume that the flow is steady, without friction, and the duct area is constant. As we have done implicitly in previous sections, we also assume that the fluid has fixed composition and constant specific heat. This is called Rayleigh Flow after the British physicist J. W. Strutt, Lord Rayleigh. Because of the nature of real flows with heating or cooling, the Rayleigh flow model is further from reality than isentropic flow or Fanno flow. Heat exchange with a surface in contact with a fluid is intimately involved with fluid friction. Combustion of fuel in air results in a change of fluid composition, as does evaporation or condensation. Combustion typically results in high temperatures, which lead to changes in cp , cv, and k. Nevertheless, the Rayleigh flow model is reasonably accurate when the temperature difference between a surface and the fluid is large so that significant heat transfer occurs over a short length of duct, when the fuel/air ratio in combustion is small, and when the ratio of liquid mass evaporated or condensed to fluid mass is small. Even if these conditions are not met, Rayleigh flow yields useful insight into the effect of heating or cooling on a compressible flow. 11.9.1 The Rayleigh Line Consider steady, one-dimensional, frictionless flow through the finite control volume shown in Fig. 11.22. Writing the linear momentum equation gives 0 p1 A − p2 A + F f = m ˙ (V2 − V1 ) = ρVA(V2 − V1 ) Dividing by the area and regrouping p1 + (ρV)V1 = p2 + (ρV)V2 (11.139) The continuity equation for the control volume is ρ1V1 A = ρ2V2 A. Dividing by the area ρV = G = constant (11.140) where G, the density-velocity product, is called the mass flow density. Combining Eqs. 11.139 and 11.140 we get p+ G2 = constant = I ρ   (11.141a) where I is called the Impulse function. Using the specific volume v = 1 we can write ρ p + G2 v = I Frictionless and adiabatic duct (11.141b) Frictionless duct Section (1) D = constant ■ Figure 11.22 Rayleigh flow. Flow Finite control volume Section (2) 11.9 p v The Rayleigh Line in p-v Coordinates Frictionless Flow in a Constant-Area Duct with Heating or Cooling 661 With G and I constant, Eq. 11.141b is a relationship between p and v, which, for specific values of G and I, can be graphed as a straight line, as shown on the figure in the margin. This line is called a Rayleigh line. It represents all possible states of the fluid for specific values of G and I. For any simple compressible fluid, if two properties are known, the other properties can be determined. If we pick several points on the p-v curve, we can determine corresponding values of temperature and entropy and graph the Rayleigh line on a T–s diagram, as shown in Fig. 11.23. We can observe many important features of Rayleigh flow by examining Fig. 11.23. Any flow state, such as that at Section 1 in Fig. 11.22, can be shown as a single point on the corresponding Rayleigh line. The addition or removal of heat between Section 1 and Section 2 causes the fluid properties to change and the state point on the Rayleigh line to move along the line. Since there is no friction, the Second Law of Thermodynamics, Eq. 5.101, in the form ds = δq T tells us that the state point will move to the right (ds > 0) if the fluid is heated (δq > 0) and the state point will move to the left if the fluid is cooled (δq < 0). Thus heating always drives the fluid toward maximum entropy (point a) and cooling drives the fluid away from the maximum entropy point. What is the fluid speed at point a, the point of maximum entropy? Rayleigh flow is governed by the continuity equation, Eq. 11.140, and the linear momentum equation, Eq. 11.141a. Differentiating the momentum equation dp − G2 dρ = 0. ρ2 Using Eq. 11.140 V2 = dp dρ (11.141) Eq. 11.141 gives the flow speed at any point on the Rayleigh line. At point a, ds = 0 and the entropy is locally constant, therefore V 2a = The maximum entropy state on the Rayleigh line corresponds to sonic conditions. ∂p = c2 at point a ∂ρ )s That is, the fluid speed equals the sonic speed and Ma = 1 at the point of maximum entropy. It follows that the flow described by the upper branch of the Rayleigh line is subsonic and the flow described by the lower branch of the Rayleigh line is supersonic. We can replace the “a” designation of the maximum entropy point with the more familiar designation of sonic flow, *, thus the temperature is T *, the pressure is p*, and so on. (It must be strongly stated that the * properties for Rayleigh flow are different from the * properties for isentropic flow and Fanno flow.) T (Ma = √1_k ) b b a (Maa = 1) Ma < 1 Ma > 1 s ■ Figure 11.23 Rayleigh line. 662 Chapter 11 ■ Compressible Flow Fluid temperature reduction can accompany heating in a subsonic Rayleigh flow. Rayleigh flow is always driven toward Ma = 1 by heating and away from Ma = 1 by cooling. The flow can pass through Ma = 1 but only by heating to precisely Ma = 1 and then switching to cooling; however, this would be nearly impossible in practice. Furthermore, if we restrict the process to heat addition, it is not possible to pass beyond point a; from any particular starting location, there is a maximum amount of heating possible. This is yet another manifestation of choking, in this case thermal choking; when the flow is heated to sonic speed in a duct, the flow cannot respond to further heating. There is no limit to cooling, other than absolute zero. One might ask “What happens when a Rayleigh flow is choked and more heat is added?” The answer is that the mass flowrate m ˙ and the mass flow density G will be reduced and the flow will be described by a different Rayleigh line, one shifted to the right. There are two final points to note about the Rayleigh line. First, the subsonic branch has a 1 point of maximum temperature (point b in Fig. 11.23). In an ideal gas, Mab = (See Problem Bk 11.73). Between this point and the maximum-entropy point, the temperature actually drops when heat is added! Lastly, although it is impossible for the flow to pass through the point of maximum entropy, it is possible to “jump” from the supersonic branch to the subsonic branch by passing through a normal shock wave. A summary of the qualitative aspects of Rayleigh flow is given in Table 11.4 and Fig. 11.24. 11.9.2 Frictionless Flow of an Ideal Gas with Heating or Cooling (Rayleigh Flow) To obtain quantitative results, it is necessary to specify a particular working fluid. As usual, we will consider an ideal gas. We elect to use the state of the fluid at point a of Fig. 11.23 as the reference state. As shown earlier, the Mach number at point a is 1 so, as done for isentropic flow and Fanno flow, we will designate this state by the superscript (*). Even though the Rayleigh flow being considered may not choke and state a is not achieved by the flow, this reference state is useful. If we apply the linear momentum equation to Rayleigh flow between any upstream section and the section, actual or imagined, where state * is attained, we get p + ρV 2 = p* + ρ*V 2* or p ρV 2 ρ* 2 + =1+ V * p* p* p* (11.142) ■ Table 11.4 Summary of Rayleigh Flow Characteristics Heating V Ma T T0 p p0 Cooling Subsonic Supersonic Subsonic Supersonic Increase Increase Increase for 0 ≤ Ma ≤ √1k Decrease for √1k ≤ Ma ≤1 Increase Decrease Decrease Decrease Decrease Increase Decrease Decrease Decrease for 0 ≤ Ma ≤ √1k Increase for √1k ≤ Ma ≤1 Decrease Increase Increase Increase Increase Decrease Increase Increase Decrease Decrease Decrease Increase 11.9 663 Frictionless Flow in a Constant-Area Duct with Heating or Cooling T T T b b ng ati He a ng oli g Co n ati He g in at ng He s g in at oli Co s (a) a ng oli Co He b Norm shoc al k a s (b) (c) ■ Figure 11.24 (a) Subsonic Rayleigh flow. (b) Supersonic Rayleigh flow. (c) Normal shock in a Rayleigh flow. By substituting the ideal gas equation of state (Eq. 11.1) into Eq. 11.142 and making use of the ideal gas speed-of-sound and the definition of Mach number, we obtain p 1+k = p* 1 + kMa2 (11.143) This relationship is graphed in the margin for k = 1.4. From the ideal gas equation of state (Eq. 11.1) we conclude that 2.0 p __ p* p ρ* T = T* p* ρ 1.0 (11.144) Conservation of mass (Eq. 11.140) with constant A gives 0.0 0.1 1.0 ρ* V = ρ V* 10 Ma (11.145) which when combined with Eqs. 11.47 (ideal gas speed of sound) and the Mach number definition gives ρ* T = Ma ρ BT* (11.146) Combining Eqs. 11.144 and 11.146 leads to 2 p T = ( Ma) T* p* which when combined with Eq. 11.143 gives (also see margin figure) 2.0 T __ T* 1.0 0.0 0.1 (1 + k)Ma 2 T =[ T* 1 + kMa2 ] 1.0 10 ρ* (1 + k)Ma V = = Ma [ ρ V* 1 + kMa2 ] V* ρ ρ (11.149) This relationship is graphed in the margin for k = 1.4. The energy equation (Eq. 5.69) tells us that because of the heat transfer involved in Rayleigh flows, the stagnation temperature varies. The heat and the stagnation temperature are related by 2.0 * _V ___ , __ (11.148) From Eqs. 11.145, 11.146, and 11.148 we see that Ma 1.0 0.0 0.1 (11.147) 1.0 Ma 10 q1−2 = cp (T02 − T01 ). (11.150) T0 T0 T T* = T0,* ( T )( T * )( T0,* ) (11.151) We note that     664 Chapter 11 ■ Compressible Flow We can use Eq. 11.50 (developed earlier for steady, isentropic, ideal gas flow) to evaluate T0 T and T *T 0,* because these two temperature ratios, by definition of the stagnation state, involve isentropic processes. Equation 11.148 can be used for TT *. Thus, consolidating Eqs. 11.151, 11.50, and 11.148 we obtain Unlike Fanno flow, the stagnation temperature in Rayleigh flow varies. T0 = T0,* 2.0 T0 ____ T0,* (1 + (11.152) This relationship is graphed in the margin for k = 1.4. Finally, we observe that 1.0 0.0 0.1 1.0 10 Ma p0 p0 p p* = ( )( )( p p0,* p* p0,* ) (11.153) We can use Eq. 11.51 developed earlier for steady, isentropic, ideal gas flow to evaluate p0 p and p*p0,* because these two pressure ratios, by definition, involve isentropic processes. Equation 11.143 can be used for pp*. Together, Eqs. 11.51, 11.143, and 11.153 give 2.0 k(k−1) p0 (1 + k) k−1 2 2 = 1 + Ma 2 [( )] p0,* k + 1 )( 2 (1 + kMa ) p0 ____ p0,* k−1 Ma2) 2 (1 + kMa2 ) 2 2(k + 1)Ma2 1.0 0.0 0.1 1.0 10 Ma (11.154) This relationship is graphed in the margin for k = 1.4. Values of pp*, TT*, ρρ* or VV*, T0 T 0 ,* and p0 p0,* are graphed in Graph D.R of Appendix D and tabulated in Table D.R as a function of Mach number for Rayleigh flow with k = 1.4. See Ref. 7 for a more advanced treatment of internal flows with heat transfer. EXAMPLE 11.13 The Rayleigh Line ft ∙ lb (k = 1.4, R = 53.3 lbm ∙ °R ) enters a constant cross-section duct with the following properties: T0 = GIVEN Air 518.67 °R, T1 = 514.55 °R, p1 = 14.3 psia (These are the same conditions as Example 11.10) FIND Plot a Rayleigh line on a T–s diagram SOLUTION We will plot the Rayleigh line from a table of values of T and s – s1 where, from Eq. 11.24, the entropy change is given by s – s1 = p T kR cp ln = ( T1 ) − R ln ( p1 ) with cp = k − 1 (1.4) (53.3 ft ∙ lb/lbm ∙ °R) = 187 ft ∙ lb/lbm ∙ °R. 1.4 − 1 3000 2000 T, °R 1 1 Rayleigh flow functions given by Eqs. 11.148 and 11.143 or Table D.R or Graph D.R. 2500 The initial value Ma1 is found from T01 T0 518.67 = = = 1.0080 so that Ma1 = T )Ma1 T1 514.55 1 1 T01 2 2 2 2 (1.0080 − 1) − 1 = ] = 0.20. )] [ 1.4 − 1 [ k − 1 ( T1 ng ati He 1500 ng oli Co 1000 g in at He 500 0 Perhaps the simplest way to generate a table is to use the Mach number as a parameter. For any value of the Mach number, p T )Ma ) p p p* T * T T Ma = and = where and are ) p1 p T1 T * Ma p* )Ma T T * )Ma p* )Ma p T = 2.2727 and = 0.2066. Now p* )Ma1 T * )Ma1 p choosing Ma = 0.4, from Table D.R = 1.9608 and p* )Ma=0.4 From Table D.R 0 100 200 (s – s1) ft*lb/lbm*°R ■ Figure E11.13 300 400 11.9 Frictionless Flow in a Constant-Area Duct with Heating or Cooling T 0.6152 = 0.6152 so that T = 514.55 °R = 1532.2 °R 0.2066 T* )Ma=0.4 and s − s1 = 665 The Rayleigh line is plotted in Figure E11.13. COMMENTS As can be seen from the figure and the table, the Mach number at the “nose” of the Rayleigh line is 1. The upper branch represents subsonic flow and the lower branch repft ∙ lb 0.6152 ft ∙ lb 2.2727 ft ∙ lb 187 ln − 53.3 ln = 196.2 . resents supersonic flow. Movement along the Rayleigh line is lbm °R 0.2066 lbm °R 1.9608 lbm °R caused by heating or cooling. Heating always drives the Mach Repeating these calculations for other Mach numbers (using a number toward one. spreadsheet) gives the table below: Ma T2 T* T1 T* p2 p* p1 p* T2 T1 p2 p1 T [°R] 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.5 3 3.5 0.207 0.615 0.917 1.025 1.000 0.912 0.805 0.702 0.609 0.529 0.379 0.280 0.214 0.207 0.207 0.207 0.207 0.207 0.207 0.207 0.207 0.207 0.207 0.207 0.207 0.207 2.273 1.961 1.596 1.266 1.000 0.796 0.641 0.524 0.434 0.364 0.246 0.176 0.132 2.273 2.273 2.273 2.273 2.273 2.273 2.273 2.273 2.273 2.273 2.273 2.273 2.273 1.000 2.977 4.437 4.963 4.840 4.413 3.898 3.396 2.947 2.560 1.833 1.357 1.037 1.000 0.863 0.702 0.557 0.440 0.350 0.282 0.230 0.191 0.160 0.108 0.078 0.058 515 1532 2283 2554 2490 2271 2006 1748 1517 1317 943 698 533 EX AMPLE 11.14 s – s1 [ft*lb / lbmR] 0 212 297 331 339 334 322 307 290 273 232 193 158 Rayleigh Flow in a Combustion Chamber GIVEN Air enters a constant-area combustion chamber at 400 °R and 12 psia. The air inlet velocity is 400 ft/s. Fuel is burned in the chamber, providing 50 Btu/lbm of heat input. FIND (a) Determine the pressure, temperature, Mach number, and fluid velocity at the exit. (b) What is the maximum possible heat that could be added? SOLUTION Part a A schematic of the combustion chamber is shown in Figure E11.14. At section 1 c = √kRT = √1.4(53.3 ft ∙ lblbm ∙ °R) (400 °R) (32.2 ft ∙ lbmlbf°R) = 980 ft/s 400 Ma1 = = 0.41 980 T1. 400 T01 = = = 413°R 0.9675 T T0 )Ma 1 T0 p T = 1.943  ) = 0.6345  ) = 0.5465 p* )Ma1 T * Ma1 T *0 Ma1 V = 400 ft/s q = 50 Btu/lbm 1 p = 10 psia T = 400 R 2 ■ Figure E11.14 A Combustion Chamber The energy equation can be used to find the downstream stagnation temperature T02 = T01 + q 50 Btu/lbm = 413 °R + = 621 °R cp 0.24 Btu/lbm °R Now T02 T *0 = T02 T0 621 = (0.5465) = 0.8217 T01 T *0 )1 413 666 Chapter 11 ■ Compressible Flow Part b The maximum possible heat addition would cause thermal choking, with the Mach number at section 2 equal to 1.0. This would correspond to T02 = T *0. From Part a From Table D.R at T0 p T = 0.8217;  = 1.5904,  = 0.9196 and T *0 )2 p* )2 T * )2 Ma2 = 0.602 (Ans) T01 = 413 °R and  Therefore Then p T )2 p* )2 T* 1.5902 = 10 psia = 8.2 psia T2 = T1 p2 = p1 p 1.9428 T p* )1 T* )1 0.9196 = 400 °R = 580 °R (Ans) 0.6345 T02, max = T *0 = T01 1.4 (53.3 ft ∙ lb ft ∙ lbm ft (580 °R) = 711 32.2 s lbm°R ) lb ∙ s2 1 413 °R = = 756 °R T0 0.5465 T *0 )1 Finally qmax = cp (T *0 − T01 ) = 0.24 Btu/lbm °R(756 °R − 413 °R) = 83.2 Btu/lbm (Ans) V2 = Ma2√kRT2 = 0.602 B T0 = 0.5465 T *0 )1 (Ans) COMMENTS Note that in Part a, heat addition leads to a substantial velocity increase. In Rayleigh flow, there is no way to determine the length of the duct. 11.9.3 Rayleigh Lines, Fanno Lines, and Normal Shocks From the analyses of the Rayleigh line in Section 11.9.1, the Fanno line in Section 11.8.1, and the normal shock in Section 11.5, it is apparent that the steady flow of an ideal gas across a normal shock is governed by some of the same equations used for describing both Fanno and Rayleigh flows (energy equation for Fanno flows, momentum equation for Rayleigh flow, and continuity equation for both). Thus, for a given density–velocity product (ρV), gas (R, k), and conditions at the inlet of the normal shock (Tx, px, and sx ), the conditions downstream of the shock (state y) will be on both a Fanno line and a Rayleigh line that pass through the inlet state (state x), as is illustrated in Fig. 11.25. In other words, the upstream and downstream states are represented by the intersection of a Fanno line and a Rayleigh line. The second law of thermodynamics requires that entropy must increase across a normal shock wave. This law and sketches of the Fanno line and Rayleigh line intersection, like that of Fig. 11.25, persuade us to conclude that flow across a normal shock can only proceed from the lower branch intersection to the upper branch intersection; that is, from supersonic to subsonic flow. Also, we observe that a shock can occur in either Fanno flow or Rayleigh flow, but only from supersonic flow to subsonic flow. T lin e ne gh yle i no li Ra Fan shock y Normal The energy equation for Fanno flow and the momentum equation for Rayleigh flow are valid for flow across normal shocks. x s ■ Figure 11.25 The relationship between a normal shock and Fanno and Rayleigh lines. 11.10 11.10 Compressible gas flows and openchannel liquid flows are strikingly similar in several ways. Analogy between Compressible and Open-Channel Flows 667 Analogy between Compressible and Open-Channel Flows During a first course in fluid mechanics, students rarely study both open-channel flows (Chapter 10) and compressible flows. This is unfortunate because these two kinds of flows are strikingly similar in several ways. Furthermore, the analogy between open-channel and compressible flows is useful because important two-dimensional compressible flow phenomena can be simply and inexpensively demonstrated with a shallow, open-channel flow field in a ripple tank or water table. The propagation of weak pressure pulses (sound waves) in a compressible flow can be considered to be comparable to the movement of small-amplitude gravity waves on the surface of an open-channel flow. In each case—one-dimensional compressible flow and open-channel flow— the influence of flow velocity on wave pattern is similar. When the flow velocity is less than the wave speed, wave fronts can move upstream of the wave source and the flow is subsonic (compressible flow) or subcritical (open-channel flow). When the flow velocity is equal to the wave speed, wave fronts cannot move upstream of the wave source and the flow is sonic (compressible flow) or critical (open-channel flow). When the flow velocity is greater than the wave speed, the flow is supersonic (compressible flow) or supercritical (open-channel flow). Normal shocks can occur in supersonic compressible flows. Hydraulic jumps can occur in supercritical open-channel flows. Comparison of the characteristics of normal shocks (Section 11.5) and hydraulic jumps (Section 10.6.1) suggests a strong resemblance and thus analogy between the two phenomena. For compressible flows a meaningful dimensionless variable is the Mach number, where Ma = V c In open-channel flows, an important dimensionless variable is the Froude number, where Fr = Voc √gy (11.155) The velocity of the channel flow is Voc, the acceleration of gravity is g, and the depth of the flow is y. Since the speed of a small-amplitude wave on the surface of an open-channel flow, coc, is (see Section 10.2.1) coc = √gy (11.156) we conclude that Fr = Voc coc (11.157) From these equations we see the similarity between Mach number (compressible flow) and Froude number (open-channel flow). For compressible flow, the continuity equation is ρAV = constant (11.158) where V is the flow velocity, ρ is the fluid density, and A is the flow cross-sectional area. For an open-channel flow, conservation of mass leads to ybVoc = constant (11.159) where Voc is the flow velocity, and y and b are the depth and width of the open-channel flow. Comparing Eqs.11.158 and 11.159 we note that if flow velocities are considered similar and flow area, A, and channel width, b, are considered similar, then compressible flow density, ρ, is analogous to open-channel flow depth, y. This analogy between depth and density can be used to interpret water table results in terms of compressible flow results. It should be pointed out that the similarity between Mach number and Froude number is generally not exact. If compressible flow and open-channel flow velocities are considered to be similar, then it follows that for Mach number and Froude number similarity the wave speeds c and coc must also be similar. 668 Chapter 11 ■ Compressible Flow From the development of the equation for the speed of sound in an ideal gas (see Eqs. 11.43 and 11.46) we have for the compressible flow c = √(constant) kρk−1 (11.160) From Eqs. 11.160 and 11.156, we see that if y is to be completely similar to ρ as suggested by comparing Eqs. 11.158 and 11.159, then k should be equal to 2. Typically k = 1.3 or 1.4 or 1.67, not 2. This limitation to exactness is, however, usually not serious enough to compromise the benefits of the analogy between compressible and open-channel flows. 11.11 Two-Dimensional Supersonic Flow Supersonic flows accelerate across expansion Mach waves. A very brief introduction to two-dimensional supersonic flow is included here so that you are aware of the basic principles that are associated with this type of flow. We begin with a consideration of supersonic flow over a wall with a small change of direction as sketched in Fig. 11.26. We apply the component of the linear momentum equation (Eq. 5.22) parallel to the Mach wave to the flow across the Mach wave. (See Eq. 11.53 for the definition of a Mach wave.) The result is that the component of velocity parallel to the Mach wave is constant across the Mach wave. That is, Vt1 = Vt2. Thus, from the simple velocity triangle construction indicated in Fig. 11.26, we conclude that the flow accelerates because of the change in direction of the flow. If several changes in wall direction are involved as shown in Fig. 11.27, then the supersonic flow accelerates (expands) because of the changes in flow direction across the Mach waves (also called expansion waves). Each Mach wave makes an appropriately smaller angle α with the upstream wall because of the increase in Mach number that occurs with each direction change (see Section 11.4). A rounded expansion corner may be considered as a series of infinitesimal changes in direction. Conversely, even sharp corners are actually rounded when viewed on a small enough scale. Thus, expansion fans as illustrated in Fig. 11.28 are commonly used for supersonic flow around a “sharp” corner. If the flow across the Mach waves is considered to be isentropic, then the increase in flow speed is accompanied by a decrease in static pressure. Therefore, for the flow of an ideal gas, the stagnation properties remain unchanged as the flow passes through an expansion fan. Expansion Mach wave Vn1 Vt1 V1 p1 VIDEO V11.9 Wedge oblique shocks and expansions Expansion Mach waves p2 < p1 α1 V2 > V1 Vn2 α2 α3 α4 Vt2 = Vt1 ■ Figure 11.26 Flow acceleration across a ■ Figure 11.27 Flow acceleration across Mach wave. Mach waves. Expansion fan Compression Mach wave V1 Vn1 Vt1 V1 V2 > V1 p1 p2 > p1 V2 < V1 Vn2 Vt2 = Vt1 ■ Figure 11.29 Flow deceleration across a ■ Figure 11.28 Corner expansion fan. Mach wave. 11.11 Two-Dimensional Supersonic Flow 669 Oblique shock wave Compression Mach waves V2 < V1 Compression Mach waves V1 VIDEO ■ Figure 11.30 Flow deceleration across ■ Figure 11.31 Mach waves coalesce to form an Mach waves. oblique shock wave. V11.10 Bullet oblique shocks Detached curved shock Attached oblique shock (a) (b) ■ Figure 11.32 Supersonic flow over a wedge: (a) Smaller wedge angle results in attached oblique shock. (b) Large wedge angle results in detached curved shock. VIDEO V11.11 Space Shuttle oblique shocks When the change in supersonic flow direction involves the change in wall orientation that is sketched in Fig. 11.29, compression rather than expansion occurs. The flow decelerates and the static pressure increases across the Mach wave. For several changes in wall direction, as indicated in Fig. 11.30, several Mach waves occur, each at an appropriately larger angle α with the upstream wall. A rounded compression corner may be considered as a series of infinitesimal changes in direction and even apparently sharp corners are actually rounded. Mach waves or compression waves can coalesce to form an oblique shock wave as shown in Fig. 11.31. Unlike the expansion fan in which each Mach wave is of infinitesimal strength, the coalescence of Mach waves produces an oblique shock across which there is a change of static and stagnation properties (and hence entropy) and Mach number. The above discussion of compression waves can be usefully extended to supersonic flow impinging on an object. For example, for supersonic flow incident on a sharp wedge-shaped leading edge (see Fig. 11.32), an attached oblique shock can form as suggested in Fig. 11.32a. For the same incident Mach number but with a larger wedge angle, a detached curved shock as sketched in Fig. 11.32b can result. In Example 11.5, we considered flow along a stagnation pathline across a detached curved shock to be identical to flow across a normal shock wave. From this brief look at two-dimensional supersonic flow, one can easily conclude that the extension of these concepts to flows over immersed objects and within ducts can be exciting, especially if two- and three-dimensional effects are considered. References 4, 5, and 6 provide much more on this subject than could be included here. 670 Chapter 11 ■ Compressible Flow 11.12 Chapter Summary and Study Guide compressible flow ideal gas internal energy enthalpy specific heat ratio entropy adiabatic isentropic Mach number speed of sound stagnation pressure subsonic sonic Mach wave supersonic Mach cone transonic flows hypersonic flows converging– diverging duct throat temperature– entropy (T–s) diagram choked flow critical state critical pressure ratio normal shock wave oblique shock wave expansion wave overexpanded underexpanded nonisentropic flow Fanno flow Rayleigh flow This chapter considers the flow of fluid with substantial changes in density associated with high speeds. While the flow of liquids may most often be considered incompressible over a wide range of speeds, the flow of gases and vapors may involve substantial fluid density changes at higher speeds. At lower speeds, with Mach number less than about 0.3, gas and vapor flows may be treated as incompressible. Since fluid density and other fluid property changes are significant in compressible flows, property relationships are important. An ideal gas, with well-defined fluid property relationships, is often used as an approximation of an actual gas. This simplification still allows useful conclusions to be made about compressible flows. The Mach number is a key variable in compressible flow. Most easily understood as the ratio of the local speed of flow and the speed of sound in the flowing fluid, it is a measure of the extent to which the flow is compressible. It is used to define regimes of compressible flows, which range from subsonic (Mach number less than 1) to supersonic (Mach number greater than 1). Liquids are nearly incompressible so the speed of sound is large and Mach number is very small in liquid flows. There are profound differences between subsonic and supersonic flows. In subsonic flow, disturbances at a point are propagated throughout the entire flow field, but in supersonic flow, disturbances affect only a limited region downstream from the point itself. In supersonic flow, the effect of a specific point is concentrated in a Mach wave. Finite disturbances in supersonic flow often result in a shock wave—a discontinuous jump in fluid properties and velocity. The notion of an isentropic (constant entropy) flow is introduced. The most important isentropic flow is one that is adiabatic (no heat transfer) and frictionless. Isentropic stagnation properties are defined in order to account for the effects of the fluid’s velocity on the thermodynamic properties. The Temperature–Entropy (T–s) diagram is introduced as an aid in visualizing property changes in compressible flow. The important practical problem of compressible flow in a duct is considered next. Three different inputs are examined: area variation, friction, and heat addition or removal. In all cases, the phenomenon of choking is observed. Choking occurs when the Mach number reaches 1 at some location in a duct. Choking limits the amount of area decrease, duct friction, or heat addition that can occur. Three major nonisentropic compressible flows considered in this chapter are Fanno flows, Rayleigh flows, and flows across normal shock waves. Unusual outcomes include the conclusions that friction can accelerate a subsonic Fanno flow, heating can result in fluid temperature reduction in a subsonic Rayleigh flow, and a flow can decelerate from supersonic flow to subsonic flow across a shock wave. The value of temperature–entropy (T–s) diagramming of flows to better understand them is demonstrated. Numerous formulas describing a variety of ideal gas compressible flows are derived. These formulas can be easily solved with computers. However, to provide the learner with a better grasp of the details of a compressible flow process, a graphical or tabular approach is often used. The striking analogy between compressible and open-channel flows leads to a brief discussion of the usefulness of a ripple tank or water table to simulate compressible flows. Expansion and compression Mach waves associated with two-dimensional compressible flows are introduced, as is the formation of oblique shock waves from compression Mach waves. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises is completed, you should be able to ■ ■ ■ ■ write out the meanings of the terms listed here in the margin and understand each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. estimate the change in ideal gas properties in a compressible flow. calculate Mach number value for a specific compressible flow. estimate when a flow may be considered incompressible and when it must be considered compressible to preserve accuracy. 11.12 ■ ■ ■ Chapter Summary and Study Guide 671 estimate details of isentropic flows of an ideal gas though converging, diverging, and converging–diverging passages. estimate details of Fanno and Rayleigh flows and flows across normal shock waves. explain the analogy between compressible and open-channel flows. Some of the important equations in this chapter are: Ideal gas equation of state ρ= p RT (11.1) uˇ 2 − uˇ 1 = cυ (T2 − T1 ) (11.5) p hˇ = uˇ + ρ (11.6) Enthalpy change hˇ 2 − hˇ 1 = cp (T2 − T1 ) (11.9) Specific heat difference cp − cυ = R Internal energy change Enthalpy Specific heat ratio k= (11.12) cp (11.13) cυ Specific heat at constant pressure cp = kR k−1 (11.14) Specific heat at constant volume cυ = R k−1 (11.15) First Tds equation 1 T ds = duˇ + pd( ) ρ (11.18) Second Tds equation 1 T ds = dhˇ − ( ) dp ρ (11.20) Entropy change in ideal gas s2 − s1 = cυ ln ρ1 T2 + R ln ρ T1 2 (11.23) Entropy change in ideal gas s2 − s1 = cp ln p2 T2 − R ln p1 T1 (11.24) Isentropic process in ideal gas p ρk = constant ∂p ( B ∂ρ )s Speed of sound c= Speed of sound in gas c = √kRT Speed of sound in liquid c= Mach cone angle Mach number sin α = Ma = (11.27) (11.40) (11.47) Eυ Bρ (11.45) c 1 = V Ma (11.53) V c (7.xx) Isentropic flow T 1 = T0 1 + [(k − 1)2]Ma2 (11.50) Isentropic flow k(k−1) p 1 ={ p0 1 + [(k − 1)2]Ma2 } (11.51) 672 Chapter 11 ■ Compressible Flow Isentropic flow 1(k−1) ρ 1 ={ ρ0 1 + [(k − 1)2]Ma2 } (11.52) Isentropic flow-critical pressure ratio k(k−1) p* 2 =( p0 k + 1) (11.87) Isentropic flow-critical temperature ratio T* 2 = T0 k+1 (11.85) Ma2y = Normal shock py Normal shock Normal shock Normal shock px Ty Tx p0,y p0,x = = = Ma2x + [2(k − 1)] [2k(k − 1)]Ma2x − 1 (11.62) 2k k−1 Ma2x − k+1 k+1 (11.64) {1 + [(k − 1)2]Ma2x } {[2k(k − 1)]Ma2x − 1} { (k + 1) 2[2(k − 1)]}Ma2x k(k−1) k(1−k) k+1 k−1 2 2 Ma 1 + Ma x) x) ( 2 ( 2 2k k − 1 1(k−1) 2 Ma − x (k + 1 k + 1) (11.65) (11.67) Isentropic flow with area change dV dA 1 =− V A (1 − Ma2 ) (11.92) Isentropic flow with area change 1 + [(k − 1)2]Ma2 (k+1)[2(k−1)] A 1 = A* Ma { 1 + [(k − 1)2] } (11.93) Fanno flow f(ℓ* − ℓ) [(k + 1)2]Ma2 1 (1 − Ma2 ) k+1 + ln = (11.126) 2 2} { k 2k D Ma 1 + [(k − 1)2]Ma Fanno flow (k + 1)2 T = T* 1 + [(k − 1)2]Ma2 (11.129) Fanno flow 12 [(k + 1)2]Ma2 V ={ V* 1 + [(k − 1)2]Ma2 } (11.131) Fanno flow 12 p (k + 1)2 1 = p* Ma { 1 + [(k − 1)2]Ma2 } (11.135) Fanno flow [(k+1)2(k−1)] p0 1 2 k−1 2 = 1 + Ma )] p*0 Ma [( k + 1 )( 2 (11.137) Rayleigh flow p 1+k = p* 1 + kMa2 Rayleigh flow (1 + k)Ma 2 T =[ T* 1 + kMa2 ] (11.148) Rayleigh flow ρ* (1 + k)Ma V = = Ma [ ρ V* 1 + kMa2 ] (11.149) (111.143) k−1 Ma2) 2 (1 + kMa2 ) 2 2(k + 1)Ma2 (1 + Rayleigh flow T0 = T 0,* Rayleigh flow k(k−1) p0 (1 + k) 2 k−1 2 = 1 + Ma )] p0,* 2 (1 + kMa2 ) [( k + 1 )( (11.152) (11.154) Problems 673 References 1. Coles, D., “Channel Flow of a Compressible Fluid,” Available on the Internet at http://web.mit.edu/hml/ ncfmf.html and also on You Tube® Summary description of film in Illustrated Experiments in Fluid Mechanics, The NCFMF Book of Film Notes, MIT Press, Cambridge, Mass., 1972. 2. Moran, M. J., and Shapiro, H. N., Fundamentals of Engineering Thermodynamics, 6th Ed., Wiley, New York, 2008. 3. Keenan, J. H., Chao, J., and Kaye, J., Gas Tables, 2nd Ed., Wiley, New York, 1980. 4. Shapiro, A. H., The Dynamics and Thermodynamics of Compressible Fluid Flow, Vol. 1, Wiley, New York, 1953. 5. Liepmann, H. W., and Roshko, A., Elements of Gasdynamics, Dover Publications, 2002. 6. Anderson, J. D., Jr., Modern Compressible Flow with Historical Perspective, 3rd Ed., McGraw-Hill, New York, 2003. 7. Greitzer, E. M., Tan, C. S., and Graf, M. B., Internal Flow Concepts and Applications, Cambridge University Press, UK, 2004. Problem is related to a chapter video available in WileyPLUS. * Problem to be solved with aid of programmable calculator or computer. Note: A programmable calculator, computer, or smartphone can be used to evaluate compressible flow functions for any problem † Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are no unique answers to these problems. Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. Unknown flow Ideal Gas k = 1.4 Adiabatic, no work Section 11.1 Ideal Gas Thermodynamics 11.1 Distinguish between flow of an ideal gas and inviscid flow of a fluid. 11.2 Air flows steadily between two sections in a duct. At section (1), the temperature and pressure are T1 = 80 °C, p1 = 301 kPa(abs), and at section (2), the temperature and pressure are T2 = 180 °C, p2 = 181 kPa(abs). Calculate the (a) change in internal energy between sections (1) and (2), (b) change in enthalpy between sections (1) and (2), (c) change in density between sections (1) and (2), (d) change in entropy between sections (1) and (2). How would you estimate the loss of available energy between the two sections of this flow? 11.3 Consider the flow process in Fig. P11.3. Does the fluid flow from left to right or from right to left? Justify your answer. Assume as ideal gas with constant specific heats. †11.4 As demonstrated in Video V11.1 fluid density differences in a flow may be seen with the help of a schlieren optical system. 2 1 p1 = 4.0 psia V1 = 2000 ft/s T1 = 194°R p2 = 40.0 psia V2 = 545 ft/s T2 = 496°R ■ Figure P11.3 Discuss what variables affect fluid density and the different ways in which a variable density flow can be achieved. Describe briefly how a schlieren optical visualization 11.5 system (Videos V11.1 and V11.5, also Fig. 11.4) works. How else might density changes in a fluid flow be made visible to the eye? 11.6 Methane is compressed adiabatically from 100 kPa(abs) and 25 °C to 200 kPa(abs). What is the minimum compressor exit temperature possible? Explain. 674 Chapter 11 ■ Compressible Flow 11.7 Air expands adiabatically through a turbine from a pressure and temperature of 180 psia, 1600 °R to a pressure of 14.7 psia. If the actual temperature change is 85% of the ideal temperature change, determine the actual temperature of the expanded air and the actual enthalpy and entropy differences across the turbine. A = 1 ft2 11.8 An expression for the value of cp for carbon dioxide as a function of temperature is cp = 286 − 2.49 × 106 1.15 × 105 + T T2 where cp is in (ft ∙ lb)/(lbm ∙ °R) and T is in °R. Compare the change in enthalpy of carbon dioxide using the constant value of cp (see Table 1.7) with the change in enthalpy of carbon dioxide using the expression above, for T2 − T1 equal to (a) 10 °R, (b) 1000 °R, (c) 3000 °R. Set T1 = 540 °R. 11.9 Air flows in a 15-cm-diameter horizontal pipe. At section 1, p = 600 kPa, T = 70 °C, and V = 35 m/s. At section 2: T = 42 °C and V = 115 m/s. Determine (a) The pressure at section 2 (b) The total friction force on the pipe wall (c) The heat transfer 11.10 An air heater in a large coal-fired steam generator heats fresh air entering the steam generator by cooling flue gas leaving the steam generator. One million lbm/hr of air at 100 °F and 1.1 million lbm/hr of flue gas at 720 °F enters the air heater. The flue gas leaves at 310 °F. Flue gas has cp = 0.26 Btu/lbm °F and k = 1.39. Pressure changes are small and may be neglected. Calculate the temperature of the air leaving the air heater and the total entropy change for the process. Section 11.2 Stagnation Properties 11.11 Determine the static pressure to stagnation pressure ratio associated with the following motion in standard air: (a) a runner moving at the rate of 10 mph, (b) a cyclist moving at the rate of 40 mph, (c) a car moving at the rate of 65 mph, (d) an airplane moving at the rate of 500 mph. 11.12 The air velocity in the duct in Fig. P11.12 is 750 ft/s. The air static temperature is 100 °F. Use the mercury manometer measurement to calculate the static and stagnation pressures of the flowing air. 40°F ps = 3 psig p0 = 15 psig ■ Figure P11.13 11.14 Steam (H2O vapor) flows in a pipeline in a power station. The steam pressure is 150 psia, its temperature is 500 °F, and it flows with velocity 750 ft/s. Calculate the stagnation pressure and stagnation temperature. If you are familiar with Steam Tables or steam property software, use these tools to make an “exact” calculation. If you are not familiar with these tools, model the steam as an ideal gas with molecular weight of 18 and k = 1.3. 11.15 An ideal gas flows with velocity V, pressure p, temperature T, and density ρ. Determine a set of equations for stagnation properties, including entropy, if the stagnation process is defined to be isothermal (T = constant) rather than isentropic (s = constant). Section 11.3 Mach Number and Speed of Sound 11.16 If the observed speed of sound in steel is 5300 m/s, determine the bulk modulus of elasticity of steel in N/m3. The density of steel is nominally 7790 kg/m3. How does your value of Ev for steel compare with Eυ for water at 15.6 °C? Compare the speeds of sound in steel, water, and air at standard atmospheric pressure and 15 °C and comment on what you observe. 11.17 Using information provided in Table C.1, develop a table of speed of sound in ft/s as a function of elevation for U.S. standard atmosphere. 11.18 An airplane is flying at a flight (or local) Mach number of 0.70 at 10,000 m in the Standard Atmosphere. Find the ground speed (a) if the air is not moving relative to the ground and (b) if the air is moving at 30 km/hr in the opposite direction from the airplane. 11.19 The stagnation pressure in a Mach 2 wind tunnel operating with air is 900 kPa. A 1.0-cm-diameter sphere positioned in the wind tunnel has a drag coefficient of 0.95. Calculate the drag force on the sphare. V = 750 ft/s 11.20 Calculate the speed of sound in air, helium, and hydrogen. The temperature is 70 °F. T = 100°F h = 8 in. ■ Figure P11.12 11.13 Air is flowing in a duct as shown in Fig. P11.13. A Pitotstatic tube and a thermocouple are inserted into the flow stream as shown. Calculate the air velocity and the air-mass flow rate. 11.21 Determine the Mach number of a car moving in standard air at a speed of (a) 25 mph, (b) 55 mph, and (c) 100 mph. 11.22 A jet-propelled airliner flies at an altitude of 15,000 m and a velocity of 265 m/s. What is the stagnation pressure and stagnation temperature of the air entering the engine? If the airliner is stationary on the ground, the engines are running with the same Mach number flow into the engines, and the air flow to each engine is 40 kg/s, what is the airflow to each engine at the above flight conditions? Problems Section 11.4 Subsonic and Supersonic Flow A schlieren photo of a bullet moving through air (see 11.23 Video V11.10) at 14.7 psia and 68 °F indicates a Mach cone angle of 28°. How fast was the bullet moving in: (a) m/s, (b) ft/s, (c) mph? 11.24 At a given instant of time, two pressure waves, each moving at the speed of sound, emitted by a point source moving with constant velocity in a fluid at rest, are shown in Fig. P11.24. Determine the Mach number involved and indicate with a sketch the instantaneous location of the point source. 10 in. 2 in. 5 in. 675 11.31 An aircraft cruises at a Mach number of 2.0 at an altitude of 15 km. Inlet air is decelerated to a Mach number of 0.4 at the engine compressor inlet. A normal shock occurs in the inlet diffuser upstream of the compressor inlet at a section where the Mach number is 1.2. For isentropic diffusion, except across the shock, and for standard atmosphere, determine the stagnation temperature and pressure of the air entering the engine compressor. 11.32 At some point for air flow in a duct, p = 20 psia, T = 500 °R, and V = 500 ft/s. Can a normal shock occur at this point? 11.33 A normal shock propagates at 2000 ft/s into the still air in a tube. The temperature and pressure of the air are 80 °F and 14.7 psia before “hit” by the shock. Calculate the air temperature, pressure, and velocity after the shock, the stagnation temperature and pressure relative to the shock ahead of and behind the shock, and the stagnation temperature and pressure relative to the tube ahead of and behind the shock. 11.34 Air at V1 = 800 m/s, p1 = 100 kPa, and T1 = 300 K passes through a normal shock. Calculate the velocity V2, temperature T2, and pressure p2 after the shock. What would be the values of T2 and p2 if the same velocity change were accomplished isentropically? 11.35 A normal shock occurs in a perfect gas. Sketch a Temperature – Entropy (T–s) diagram of the process and show the following: Static and stagnation pressure and temperature before and after the shock and gas velocity before and after the shock. ■ Figure P11.24 11.25 Sound waves are very small-amplitude pressure pulses that travel at the “speed of sound.” Do very large-amplitude waves such as a blast wave caused by an explosion (see Video V11.8) travel less than, equal to, or greater than the speed of sound? Explain. 11.26 If a new Boeing 787 Dreamliner cruises at a Mach number of 0.87 at an altitude of 30,000 ft, how fast is this in (a) mph, (b) ft/s, (c) m/s? 11.27 Explain how you could vary the Mach number but not the Reynolds number in airflow past a sphere. For a constant Reynolds number of 300,000, estimate how much the drag coefficient will increase as the Mach number is increased from 0.3 to 1.0. Section 11.6 Isentropic Flow 11.36 The stagnation pressure and temperature of air flowing past a probe are 120 kPa(abs) and 100 °C, respectively. The air pressure is 80 kPa(abs). Determine the airspeed and the Mach number considering the flow to be (a) incompressible, (b) compressible. 11.37 Air flows isentropically through a duct as shown in Fig. P11.37. For the conditions shown, find the Mach number at both stations 1 and 2 and the flow rate. The diameter is not necessarily constant. Ma1 Ma2 11.28 Air flows in a constant-area, insulated duct. The air enters the duct at 520 °R, 50 psia, and Ma = 0.45. At a downstream location, the Mach number is one. Find: (a) The pressure and temperature at the downstream location (b) The change in specific entropy (c) The frictional force if the duct is circular and 1 ft in diameter. Section 11.5 Shock Waves 11.29 A normal shock occurs in a stream of oxygen. The oxygen flows at Ma = 1.8 and the upstream pressure and temperature are 15 psia and 85 °F. (a) Calculate the following on the downstream side of the shock: static pressure, stagnation pressure, static temperature, stagnation temperature, static density, and velocity. (b) If the Mach number is doubled to 3.6, what will be the resulting values of the parameters listed in part (b)? 11.30 The Pitot tube on a supersonic aircraft (see Video V3.8) cruising at an altitude of 30,000 ft senses a stagnation pressure of 12 psia. If the atmosphere is considered standard, determine the airspeed and Mach number of the aircraft. A shock wave is present just upstream of the probe impact hole. 1 D1 = 0.15 m V1 = 40 m/s p1 = 4 atm abs. 2 T2 = 225 K p2 = 2 atm abs. ■ Figure P11.37 11.38 A simplified schematic diagram of a carburetor of a gasoline engine is shown in Fig. P11.38. The throat area is 0.5 in2. The engine draws air downward through the carburetor Venturi and maintains a throat pressure of 14.3 psia. This low throat pressure draws fuel from the float chamber and into the air stream. The energy losses in the 0.06-in. I.D. fuel line and valve are given by hL = KV 2 , 2g with K = 6.0. The fuel specific gravity is 0.75. Assume an atmospheric pressure of 14.7 psia. Find the air–fuel ratio (that is, the ratio of the air mass flow rate to the fuel mass flow rate) based on an (a) ideal (constant density and inviscid) air flow and (b) an isentropic air flow. 676 Chapter 11 ■ Compressible Flow Air the mass flowrate through the duct if the receiver pressure is (a) 10 psia, (b) 5 psia. Sketch temperature–entropy diagrams for situations (a) and (b). Verify results obtained with values from the appropriate graph in Appendix D with calculations involving ideal gas equations. Is condensation of water vapor a concern? Explain. Choke Air vent 11.44 Helium at 68 °F and 14.7 psia in a large tank flows steadily and isentropically through a converging nozzle to a receiver pipe. The cross-sectional area of the throat of the converging passage is 0.05 ft2. Determine the mass flowrate through the duct if the receiver pressure is (a) 10 psia, (b) 5 psia. Sketch temperature– entropy diagrams for situations (a) and (b). Fuel 0.5 in2 Float chamber Throttle ■ Figure P11.38 11.39 An engineering student wants to satisfy her curiosity about the compressibility of air in motion. She has set up a converging nozzle in which air discharges into the atmosphere. Figure P11.39 shows the nozzle with the necessary information. For these conditions, find the velocity at the exit by using the incompressible Bernoulli’s equation, a compressible isothermal process, and a compressible isentropic process. Comment on your results. 11.46 The flow blockage associated with the use of an intrusive probe can be important. Determine the percentage increase in section velocity corresponding to a 0.5% reduction in flow area due to probe blockage for airflow if the section area is 1.0 m2, T0 = 20 °C, and the unblocked flow Mach numbers are (a) Ma = 0.2, (b) Ma = 0.8, (c) Ma = 1.5, (d) Ma = 30. 11.47 At a certain point in a pipe, air flows steadily with a velocity of 150 m/s and has a static pressure of 70 kPa and a static temperature of 4 °C. The flow is adiabatic and frictionless. p0 = 175 kPa T0 = 300 K Tb = 300 K V0 ≈ 0 ρ 0 = 2.04 kg/m3 11.45 An ideal gas is to flow isentropically from a large tank where the air is maintained at a temperature and pressure of 59 °F and 80 psia to standard atmospheric discharge conditions. Describe in general terms the kind of duct involved and determine the duct exit Mach number and velocity in ft/s if the gas is air. V1 pb = 100 kPa ρb = 1.16 kg/m3 (a) Calculate the maximum possible reduction in area and the following quantities for that minimum area: stagnation pressure, stagnation temperature, static pressure, static temperature, velocity, and Mach number. (b) Calculate the quantities listed in (a) at a point where the area is 15% smaller than the initial area. ■ Figure P11.39 11.48 A tank of oxygen has a hole of area 0.5 cm2 in its wall. The temperature of the oxygen in the tank is 25 °C. Calculate the rate (kg/s) at which oxygen leaks to the atmosphere for tank pressures of 135 kPa and 375 kPa. Assume frictionless flow. 11.40 A nozzle for a supersonic wind tunnel is designed to achieve a Mach number of 3.0, with a velocity of 2000 m/s, and a density of 1.0 kg/m3 in the test section. Find the temperature and pressure in the test section and the upstream stagnation conditions. The fluid is helium. 11.49 The gas entering a rocket nozzle has a stagnation pressure of 1500 kPa and a stagnation temperature of 3000 °C. The rocket is traveling in the still Standard Atmosphere at 30,000 m. Find the throat and exit area for a flow rate of 10 kg/s. Assume k = 1.35, R = 287.0 N ∙ m/kg ∙ K. The gas is perfectly, expanded to the ambient pressure. 11.41 Air flows isentropically through a duct to a section where p1 = 25 kPa, T1 = 300K, and V1 = 900 m/s. For these conditions: (a) Determine the stagnation conditions for the flow. (b) What is the Mach number at station 1? Show a T–s diagram displaying stagnation and static conditions. (c) Is the flow choked? Is the throat behind or ahead of section 1? Label this state on the T–s diagram. 11.42 The static pressure to stagnation pressure ratio at a point in a gas flow field is measured with a Pitot-static probe as being equal to 0.6. The stagnation temperature of the gas is 20 °C. Determine the flow speed in m/s and the Mach number if the gas is air. What error would be associated with assuming that the flow is incompressible? Section 11.7 One Dimensional Flow in a Variable Area Duct 11.43 Air flows steadily and isentropically from standard atmospheric conditions to a receiver pipe through a converging duct. The cross-sectional area of the throat of the converging duct is 0.05 ft2. Determine 11.50 Air flows in the channel in Fig. P11.50. Determine the Mach number, static pressure, and stagnation pressures at station 3. Assume isentropic flow except for the normal shock wave. 1 3 2 Shock wave A2 = 75 in2 A1 = 100 in2 Ma1 = 2.1 p1 = 65 psia ■ Figure P11.50 A3 = 120 in2 Problems 677 11.51 A jet engine is to be designed for an altitude of 12,000 m, where the atmospheric pressure is 19.3 kPa. The jet nozzle has a supersonic exit Mach number and is perfectly expanded. The stagnation pressure and temperature of the gas are 100 kpa and 600 °C. The flow rate of gas is 45 kg/s. Calculate the throat area, exit area, and exit velocity. Use k = 1.4 and R = 260 J/kg ∙ K for the gas. 11.57 Air flows adiabatically between two sections in a constant area pipe. At upstream section (1), p0,1 = 100 psia, T0,1 = 600 °R, and Ma1 = 0.5. At downstream section (2), the flow is choked. Estimate the magnitude of the force per unit cross-sectional area exerted by the inside wall of the pipe on the fluid between sections (1) and (2). 11.52 Air is flowing in the converging–diverging nozzle shown in Fig. P11.52. Determine the three critical pressure ratios and the Mach numbers immediately upstream and immediately downstream from the shock. 11.58 Supersonic airflow enters an adiabatic, constant area (inside diameter = 1 ft) 30-ft-long pipe with Ma1 = 3.0. The pipe friction factor is estimated to be 0.02. What ratio of pipe exit pressure to pipe inlet stagnation pressure would result in a normal shock wave standing at (a) x = 5 ft, or (b) x = 10 ft, where x is the distance downstream from the pipe entrance? Determine also the duct exit Mach number and sketch the temperature–entropy diagram for each situation. A = 15 cm2 Ae = 25 cm2 Shock At = 10 cm2 11.59 Supersonic airflow enters an adiabatic, constant area pipe (inside diameter = 0.1 m) with Ma1 = 2.0. The pipe friction factor is 0.02. If a standing normal shock is located right at the pipe exit, and the Mach number just upstream of the shock is 1.2, determine the length of the pipe. 11.60 Consider the flow of air through the piping system shown in Fig. P11.60. If the system is choked, determine the location where the Mach number is 1. Calculate the pressure ratio pB p01 that causes choking in this system. ■ Figure P11.52 11.53 A convergent–divergent nozzle has an exit throat area ratio of 3.0. It is to be supplied with air. Find: (a) The first, second, and third critical pressure ratios; (b) The exit plane Mach number in each case; (c) The throat Mach number in each case. 11.54 An ideal gas flows isentropically through a converging– diverging nozzle. At a section in the converging portion of the nozzle, A1 = 0.1 m2, p1 = 600 kPa(abs), T1 = 20 °C, and Ma1 = 0.6. For section (2) in the diverging part of the nozzle, determine A2, p2, and T2 if Ma2 = 3.0 and the gas is air. 11.55 Air is supplied to a convergent–divergent nozzle from a reservoir where the pressure is 100 kPa. The air is then discharged through a short pipe into another reservoir where the pressure can be varied. The cross-sectional area of the pipe is twice the area of the throat of the nozzle. Friction and heat transfer may be neglected throughout the flow. If the discharge pipe has constant cross-sectional area, determine the range of static pressure in the pipe for which a normal shock will stand in the divergent section of the nozzle. If the discharge pipe tapers so that its cross-sectional area is reduced by 25%, show that a normal shock cannot be drawn to the end of the divergent section of the nozzle. Find the maximum strength of shock (as expressed by the upstream Mach number) that can be formed. p01 T0 100 in 2 1 1.5 in diameter f= pB 100 in 3 4 1 in diameter Short nozzle 0.0030 in 1.5 in pipe 0.0035 in 1 in pipe ■ Figure P11.60 11.61 Estimate the maximum mass flow rate of air that can be passed by the duct shown in Fig. P11.61 for μ = 4 × 10−7 lb ∙ s /ft2. Insulated steel pipe p0 = 100 psia T0 = 140°F 2 in diameter 20 in ■ Figure P11.61 Section 11.8 Constant Area Duct Flow with Friction 11.56 Standard atmospheric air (T0 = 59 °F, p0 = 14.7 psia) is drawn steadily through a frictionless and adiabatic converging nozzle into an adiabatic, constant cross-sectional area duct. The duct is 10 ft long and has an inside diameter of 0.5 ft. The average friction factor for the duct may be estimated as being equal to 0.03. What is the maximum mass flowrate in slugs/s through the duct? For this maximum flowrate, determine the values of static temperature, static pressure, stagnation temperature, stagnation pressure, and velocity at the inlet [section (1)] and exit [section (2)] of the constant area duct. Sketch a temperature–entropy diagram for this flow. 11.62 Waste gas (CO2) is vented to outer space from a spacecraft through a circular pipe 0.2 m long. The pressure and temperature in the spacecraft are 35 kPa and 25 °C. The gas must be vented at the rate of 0.01 kg/s. The friction factor for the flow in the pipe is given by f = 64/Re, Re < 5000, f = 0.013, Re > 5000, Re = 4m ˙ . πμD The viscosity (μ) is 4 × 10−4 N ∙ s/m2. Determine the required pipe diameter. 678 Chapter 11 ■ Compressible Flow 11.63 Air enters a 4-cm-square galvanized steel duct with p0 = 150 kPa, T0 = 400K, and V1 = 120 m/s. (Note: μ = 2.2 × 10−5 N ∙ s/m2). (a) Compute the maximum possible duct length for these conditions. (b) If the actual duct length is 0.75 times the maximum value, calculate the mass flow rate, the exit pressure, and the stagnation pressure. (c) If the actual duct length is 1.3 times the maximum value for the stated conditions, compute the new mass flow rate and inlet velocity. Assume a low back pressure and use the same value of f as used in the maximum possible duct length case. 11.64 Figure P11.64 shows an insulated pipe attached to a tank of air. Estimate the maximum mass flow rate that the pipe could exhaust from the tank. 1.5. Downstream of the normal shock, the subsonic flow is accelerated with heating until it chokes at the duct exit. Determine the static temperature and pressure, the stagnation temperature and pressure, and the fluid velocity at the duct entrance, just upstream and downstream of the normal shock, and at the duct exit. Sketch the temperature–entropy diagram for this flow. 11.69 Air enters a frictionless, constant area duct with Ma = 2.5, T0 = 20 °C, and p0 = 101 kPa(abs). The gas is decelerated by heating until a normal shock occurs where the local Mach number is 1.3. Downstream of the shock, the subsonic flow is accelerated with heating until it exits with a Mach number of 0.9. Determine the static temperature and pressure, the stagnation temperature and pressure, and the fluid velocity at the duct entrance, just upstream and downstream of the normal shock, and at the duct exit. Sketch the temperature–entropy diagram for this flow. 11.70 Air enters a 15-cm pipe with velocity 120 m/s, 1 atmosphere pressure, and T = 100 °C. How much heat must be added to bring the air to the maximum static temperature? What are the values of temperature and pressure at this point? p = 3 MPa T = 298 K 11.71 Show that for Rayleigh flow, the maximum amount of heat that may be added to the gas is given by: D = 10 cm f = 0.02 L = 25 m ■ Figure P11.64 Section 11.9 Frictionless Flow in a Constant Area Duct with Heating or Cooling 11.65 Standard atmospheric air [T0 = 288 K, p0 = 101 kPa(abs)] is drawn steadily through an isentropic converging nozzle into a frictionless diabatic (q = 500 kJ/kg) constant area duct. For maximum flow, determine the values of static temperature, static pressure, stagnation temperature, stagnation pressure, and flow velocity at the inlet [section (1)] and exit [section (2)] of the constant area duct. Sketch a temperature–entropy diagram for this flow. 11.66 Air enters a 0.5-ft inside diameter duct with p1 = 20 psia, T1 = 80 °F, and V1 = 200 ft/s. What frictionless heat addition rate in Btu/s is necessary for an exit gas temperature T2 = 1500 °F? Determine p2, V2, and Ma2 also. 11.67 Air enters a length of constant area pipe with p1 = 200 kPa(abs), T1 = 500 K, and V1 = 400 m/s. If 500 kJ/kg of energy is removed from the air by frictionless heat transfer between sections (1) and (2), determine p2, T2, and V2. Sketch a temperature–entropy diagram for the flow between sections (1) and (2). 11.68 Air enters a frictionless, constant area duct with Ma1 = 2.0, T0,1 = 59 °F, and p0,1 = 14.7 psia. The air is decelerated by heating until a normal shock wave occurs where the local Mach number is qmax (Ma21 − 1) 2 = cpT1 2(k + 1)Ma21 11.72 Air is stored in a tank where the pressure is 40 psia and the temperature is 500 °R. A converging-diverging nozzle with an exitto-throat area ratio of 2.5 attaches the tank to a duct where heat is exchanged with the air. The exit pressure is 15 psia and a normal shock stands at the exit of the nozzle. Determine the magnitude and direction of the heat exchange. 11.73 Prove that, in Rayleigh flow, the Mach number at the point of maximum temperature is 1/√k. ■ Lifelong Learning Problems 11.1 LL Is there a limit to how fast an object can move through the atmosphere? Explain. 11.2 LL Discuss the similarities between hydraulic jumps in open-channel flow and shock waves in compressible flow. Explain how this knowledge can be useful. 11.3 LL [See The Wide World of Fluids article titled “Hilsch Tube (Ranque Vortex Tube),” Section 11.1.] Explain why a Hilsch tube works and cite some high and low gas temperatures actually achieved. What is the most important limitation of a Hilsch tube, and how can it be overcome? 11.4 LL [See The Wide World of Fluids article titled “Supersonic and Compressible Flows in Gas Turbines,” Section 11.3.] Using typical physical dimensions and rotation speeds of manufactured gas turbine rotors, consider the possibility that supersonic fluid velocities relative to blade surfaces are possible. How do designers use this knowledge? 12 Turbomachines Learning Objectives After completing this chapter, you should be able to: ■ explain how a turbomachine works. ■ describe the basic differences between a turbine and a pump. ■ discuss the importance of minimizing loss in a turbomachine. ■ select an appropriate turbomachine for a particular application. ■ determine the operating point for a turbomachine installed in a system. ■ sketch typical turbomachine blades and velocity diagrams. ■ apply similarity laws to predict the effects of size and/or speed on a family of turbomachines. ■ perform engineering work or undertake advanced study involving the fluid mechanics of turbomachinery (e.g., design, development, research). Turbomachines are dynamic fluid machines that add (for pumps) or extract (for turbines) flow energy In previous chapters we often used “black boxes” to represent fluid machines such as pumps or turbines. The purpose of this chapter is to present the fluid mechanics of these devices when they are turbomachines. Pumps and turbines (often turbomachines) are used in a wide variety of configurations. In general, pumps add energy to the fluid—they do work on the fluid to move and/or increase the pressure of the fluid; turbines extract energy from the fluid—the fluid does work on them. The term “pump” will be used to refer to all pumping machines, including pumps, fans, blowers, and compressors. The major components of a typical turbomachine are a rotor attached to an input/output shaft and blades (sometimes called “buckets” or “vanes”) attached to that rotor to form flow passages. (The rotor is often called an impeller [in a pump], a runner [in a hydraulic turbine], or a wheel [in a fan].) A fluid that is moving can force rotation and produce shaft power. In this case we have a turbine. On the other hand, we can exert a shaft torque, typically with a motor, and by using blades on a rotor, force the fluid to move. In this case we have a pump. Figure 12.1 shows the turbine and compressor (pump) rotors of an automobile turbocharger. Examples of turbomachine-type pumps include simple window fans, propellers on ships or airplanes, squirrel-cage fans on home furnaces, axial-flow water pumps used in deep wells, and compressors in automobile turbochargers. Examples of turbines include the turbine portion of gas turbine engines on aircraft, steam turbines used to drive generators at electrical generation stations, and the small, high-speed air turbines that power dentist drills. 679 680 Chapter 12 ■ Turbomachines Compressed airflow Exhaust gas flow Exhaust gas flow Airflow Airflow Shaft rotation Exhaust gas flow Turbine rotor Exhaust gas flow Compressor rotor Compressed airflow ■ Figure 12.1 Automotive turbocharger turbine and compressor rotors. (Photograph courtesy of Concepts NREC.) Turbomachines serve in an enormous array of applications in our daily lives and thus play an important role in modern society. These machines can have a high-power density (power per unit volume), relatively few moving parts, and reasonable efficiency. The following sections provide an introduction to the fluid mechanics of these important machines. References 1–6 are a few examples of the many books that offer much more information about turbomachines. 12.1 Turbomachines exchange fluid energy and mechanical energy using blades mounted on a rotor. Turbomachines are classified as axial-flow, mixedflow, or radial-flow machines. Introduction Turbomachines are mechanical devices consisting of several blades attached to a rotor that either extract energy from a fluid (turbine) or add energy to a fluid (pump) as a result of dynamic interactions between the device and the fluid. Although the actual design and construction of these devices often require considerable insight and effort, their basic operating principles are quite simple. The fluid involved can be either a gas (as with a window fan or a gas turbine engine), a vapor (steam turbine), or a liquid (as with the water pump on a car or a turbine at a hydroelectric power plant). While the basic operating principles are the same whether the fluid is a liquid or a gas, important differences in the fluid dynamics involved can occur. For example, cavitation may be an important design consideration when liquids are involved if the pressure at any point within the flow is reduced to the vapor pressure. Compressibility effects may be important when gases are involved if the Mach number becomes large enough. Many turbomachines contain some type of housing or casing that surrounds the rotor, thus forming an internal flow passageway through which the fluid flows (see Fig. 12.2). Others, such as a windmill or a window fan, are unducted. Some turbomachines include stationary blades or vanes in addition to rotor blades. These stationary vanes can be arranged to accelerate the flow with a drop in pressure and thus serve as nozzles. Or these vanes can be set to slow the flow with a pressure rise and thus act as diffusers. Turbomachines are classified as axial-flow, mixed-flow, or radial-flow machines depending on the predominant direction of the fluid motion relative to the rotor axis as the fluid passes the blades (see Fig. 12.2). For an axial-flow machine the fluid maintains a significant axial-flow component from the inlet to outlet of the rotor. For a radial-flow machine the flow across the blades involves a substantial radial-flow component at the rotor inlet, exit, or both. In other machines, designated as mixed-flow machines, there may be significant radial- and axial-flow velocity components for the flow through the rotor row. Each type of machine has advantages and disadvantages for different applications and in terms of fluid-dynamic performance. 12.2 Basic Energy Considerations 681 Outlet Rotor Inlet ω Housing or casing (a) Radial-flow fan Rotor Inlet Stator or strut Outlet Housing or casing (b) Axial-flow fan ■ Figure 12.2 (a) A radial-flow turbomachine, (b) an axial-flow turbomachine. (Photographs courtesy of Comair Rotron, Inc.) 12.2 Basic Energy Considerations An understanding of the work transfer in turbomachines can be obtained by considering the basic operation of a household fan (pump) and a windmill (turbine). Although the actual flows in such devices are very complex (i.e., three-dimensional and unsteady), the essential phenomena can be illustrated by use of simplified flow considerations and velocity diagrams. Consider a fan blade driven at constant angular velocity, ω, by a motor as is shown in Fig. 12.3a. We denote the blade speed as U = ωr, where r is the radial distance from the axis of the fan. The absolute fluid velocity (that seen by a person sitting stationary at the table on which the fan rests) is denoted V, and the relative velocity (that seen by a person riding on the fan blade) is denoted W. As shown by the figure in the margin, the actual (absolute) fluid velocity is the vector sum of the relative velocity and the blade velocity W U V=W+U (12.1) V A simplified sketch of the fluid velocity as it “enters” and “exits” the fan at radius r is shown in Fig. 12.3b. The shaded surface labeled a–b–c–d is a portion of the cylindrical surface (including a “slice” through the blade) shown in Fig. 12.3a. In this section, and throughout this chapter, we assume for simplicity that the flow moves smoothly along the blade so that relative to the moving blade the velocity is parallel to the leading and trailing edges (points 1 and 2) of the blade. In other words, we assume that the fluid is perfectly guided by the blade. Obviously, this assumption becomes more accurate as the ratio of blade length to blade-to-blade spacing increases. For now we assume that the fluid enters and leaves the fan at the same distance from the axis of rotation; thus, U 1 = U 2 = ωr. In actual turbomachines, the entering and leaving flows are not necessarily tangent to the blades, and the fluid pathlines can change radius. These considerations are important at 682 Chapter 12 ■ Turbomachines Q b a r ω c d V1 (a) U1 W1 a b (1) Blade motion Exit surface Tangential Blade cross section Inlet surface Axial V2 U2 (2) c W2 When blades move because of the fluid force, we have a turbine; when blades move fluid, we have a pump. d (b) ■ Figure 12.3 Idealized flow through a fan: (a) fan blade geometry; (b) absolute velocity, V; relative velocity, W; and blade velocity, U at the inlet and exit of the fan blade section. design and off-design operating conditions. Interested readers are referred to Refs. 1, 2, and 3 for more information about these aspects of turbomachine flows. With this information we can construct the velocity diagrams shown in Fig. 12.3b. Note that this view is from the top of the fan, looking radially down toward the axis of rotation. The motion of the blade is toward the top of the page; the motion of the incoming air is assumed to be directed along the axis of rotation. The important concept to grasp from this sketch is that the fan blade (because of its shape and motion) “pushes” the fluid, causing it to change direction. The absolute velocity vector, V, is turned during its flow across the blade from section (1) to section (2). Initially the fluid had no component of absolute velocity in the direction of the motion of the blade, the θ (or tangential) direction. When the fluid leaves the blade, this tangential component of absolute velocity is nonzero. For this to occur, the blade must push on the fluid in the tangential direction. That is, the blade exerts a tangential force on the fluid in the direction of the motion of the blade. This tangential force and the blade motion are in the same direction—the blade does work on the fluid. This device is a pump. On the other hand, consider the windmill shown in Fig. 12.4a (also see Video V12.1). Rather than the rotor being driven by a motor, the blades move in the direction of the force exerted on each blade by the wind blowing through the rotor. We again note that because of the blade shape and motion, the absolute velocity vectors at sections (1) and (2), V1 and V2, have different directions. For this to happen, the blades must have pushed up on the fluid—opposite to the direction of blade motion. Alternatively, because of equal and opposite forces (action/reaction) the fluid must have 12.2 VIDEO Basic Energy Considerations 683 V12.1 Windmills (a) Blade motion W2 U2 V2 W1 (1) (2) U1 V1 ω (1) (b) (2) (c) ■ Figure 12.4 Idealized flow through a windmill: (a) windmill; (b) windmill blade geometry; (c) absolute velocity, V; relative velocity, W; and blade velocity, U at the inlet and exit of the windmill blade section. pushed on the blades in the direction of their motion—the fluid does work on the blades. This extraction of energy from the fluid is the purpose of a turbine. These examples involve work transfer to or from a flowing fluid in two axial-flow turbomachines. Similar concepts hold for other turbomachines including mixed-flow and radial-flow configurations. THE WIDE WORLD OF FLUIDS Current from currents The use of large, efficient wind tur- occur at many locations in the 70% of the Earth’s surface that bines to generate electrical power is becoming more common- is water. Clearly, a 4-knot (2.5 m/s) tidal current is not as fast place throughout the world. “Wind farms” containing numerous as a 40-mph (70 km/hr) wind driving a wind turbine. However, turbines located at sites that have proper wind conditions can since turbine power output is proportional to the fluid density, produce a significant amount of electrical power. Recently, and since seawater is more than 800 times as dense as air, researchers in the United States, the United Kingdom, and significant power can be extracted from slow, but massive, Canada have been investigating the possibility of harvesting ocean currents. One promising configuration involves blades the power of ocean currents and tides by using current tur- twisted in a helical pattern. This technology may provide elec- bines that function much like wind turbines. Rather than being trical power that is both ecologically friendly and economically driven by wind, they derive energy from ocean currents that viable. 684 Chapter 12 ■ Turbomachines EXAMPLE 12.1 Basic Difference between a Pump and a Turbine GIVEN The rotor shown in Fig. E12.1a rotates at a constant angular velocity ω = 100 rad/s. Although the fluid initially approaches the rotor in an axial direction, the flow across the blades is primarily outward (see Fig. 12.2a). Measurements indicate that the absolute velocity at the inlet and outlet are V1 = 12 m/s and V2 = 15 m/s, respectively. FIND Is this device a pump or a turbine? SOLUTION To answer this question, we need to know if the tangential component of the force of the blade on the fluid is in the direction of the blade motion (a pump) or opposite to it (a turbine). We assume that the blades are tangent to the incoming relative velocity and that the relative flow leaving the rotor is tangent to the blades as shown in Fig. E12.1b. We calculate the inlet and outlet blade speeds as U1 = ωr1 = (100 rad/s) (0.1 m) = 10 m/s and U2 = ωr2 = (100 rad/s) (0.2 m) = 20 m/s With the known, absolute fluid velocity and blade velocity at the inlet, we can draw the velocity diagram (the graphical representation of Eq. 12.1) at that location as shown in Fig. E12.1c. β2 = 60° Note that we have assumed that the absolute flow at the blade row inlet is radial (i.e., the direction of V1 is radial). At the outlet we know the blade velocity, U2; the outlet speed, V2; and the relative velocity direction, β2 (because of the blade geometry). Therefore, we can graphically (or trigonometrically) construct the outlet velocity diagram as shown in the figure. By comparing the velocity diagrams at the inlet and outlet, it can be seen that as the fluid flows across the blade row, the absolute velocity vector turns in the direction of the blade motion. At the inlet there is no component of absolute velocity in the direction of rotation; at the outlet this component is not zero. That is, the blade pushes and turns the fluid in the direction of the blade motion, thereby doing work on the fluid, adding energy to it. (Ans) This device is a pump. (2) (2) U2 = 20 m /s r2 = 0.2m (1) Blade motion r1 = 0.1m (1) W2 60° (2) ω = 100 rad/s U1 = 10 m /s (1) Blade W1 + (b) (a) U2 = 20 m /s 30° W2 β 2 = 60° V2 = 15 m /s Outlet U1 W1 Radial V2 Circumferential U2 U1 = 10 m /s V1 = 12 m /s (c) ■ Figure E12.1 V1 W1 W2 ω Inlet (d) 12.3 COMMENT On the other hand, by reversing the direction of flow from larger to smaller radii, this device can become a radial-flow turbine. In this case (Fig. E12.1d) the flow direction is reversed (compared to that in Figs. E12.1a, b, and c) and the velocity diagrams are as indicated. Stationary vanes around the perimeter of the rotor would be needed to achieve V1 as shown. Note that the component of the absolute velocity, V, in the 12.3 685 direction of the blade motion is smaller at the outlet than at the inlet. The blade must push against the fluid in the direction opposite the motion of the blade to cause this. Hence (by equal and opposite forces), the fluid pushes against the blade in the direction of blade motion, thereby doing work on the blade. There is a transfer of work from the fluid to the blade—a turbine operation. Angular Momentum Considerations When shaft torque and rotation are in the same direction, we have a pump; otherwise we have a turbine. VIDEO Angular Momentum Considerations V12.2 Selfpropelled lawn sprinkler In the previous section we indicated how work transfer to or from a fluid flowing through a pump or a turbine occurs by interaction between moving rotor blades and the fluid. Since all of these turbomachines involve the rotation of a rotor about a central axis, it is appropriate to discuss their performance in terms of torque and angular momentum. Recall that work can be written as force times distance or as torque times angular displacement. Hence, if the shaft torque (the torque that the shaft applies to the rotor) and the rotation of the rotor are in the same direction, energy is transferred from the shaft to the rotor and from the rotor to the fluid—the machine is a pump. Conversely, if the torque exerted by the shaft on the rotor is opposite to the direction of rotation, the energy transfer is from the fluid to the rotor—a turbine. The amount of shaft torque (and, hence, shaft work) can be obtained from the moment-of-momentum equation derived formally in Section 5.2.3. Consider a fluid particle traveling outward through the rotor in the radial-flow machine shown in Figs. E12.1a, b, and c. For now, assume that the particle enters the rotor with a radial velocity only (i.e., no “swirl”). After being acted upon by the rotor blades during its passage from the inlet [section (1)] to the outlet [section (2)], this particle exits with radial (r) and circumferential (θ) components of velocity. Thus, the particle enters with no angular momentum about the axis of rotation but leaves with nonzero angular momentum about that axis. (Recall that the axial component of angular momentum for a particle is its mass times the distance from the axis times the θ component of absolute velocity.) In a turbomachine a series of particles (a continuum) passes through the rotor. Thus, the moment-of-momentum equation applied to a control volume as derived in Section 5.2.3 is valid. For steady flow (or for turbomachine rotors with steady-in-the-mean or steady-on-average cyclical flow), Eq. 5.42 gives ∑ (r × F) = ∫ (r × V) ρV ∙ nˆ dA cs Recall that the left side of this equation represents the sum of the external torques (moments) acting on the contents of the control volume, and the right side is the net rate of flow of moment-ofmomentum (angular momentum) through the control surface. The axial component of this equation applied to the one-dimensional simplification of flow through a turbomachine rotor with section (1) as the inlet and section (2) as the outlet results in Tshaft = −m ˙ 1 (r1Vθ1 ) + m˙ 2 (r2Vθ2 ) (12.2) where Tshaft is the shaft torque applied to the contents of the control volume. The “−” is associated with mass flowrate into the control volume, and the “+” is used with the outflow. The sign of the Vθ component depends on the direction of Vθ and the blade motion, U. If Vθ and U are in the same direction, then Vθ is positive. The sign of the torque exerted by the shaft on the rotor, Tshaft, is positive if Tshaft is in the same direction as rotation, and negative otherwise. As seen from Eq. 12.2, the shaft torque is directly proportional to the mass flowrate, m˙ = ρQ. (It takes considerably more torque and power to pump water than to pump air with the same volume flowrate.) The torque also depends on the tangential component of the absolute velocity, Vθ. Equation 12.2 is often called the Euler turbomachine equation, a name also frequently assigned to Eqs. 12.4, 12.5, and 12.8. 686 Chapter 12 ■ Turbomachines The Euler turbomachine equation is derived from the axial component of the momentof-momentum equation. Also recall that the shaft power, Ẇ shaft, is related to the shaft torque and angular velocity by Ẇ shaft = Tshaft ω (12.3) By combining Eqs. 12.2 and 12.3 and using the fact that U = ωr, we obtain ˙ 1 (U1Vθ1 ) + m˙ 2 (U2Vθ2 ) Ẇ shaft = −m (12.4) Again, the value of Vθ is positive when Vθ and U are in the same direction and negative otherwise. Also, Ẇ shaft is positive when the shaft torque and ω are in the same direction and negative otherwise. Thus, Ẇ shaft is positive when power is supplied to the contents of the control volume (pumps) and negative otherwise (turbines). This outcome is consistent with the sign convention involving the work term in the energy equation considered in Chapter 5 (see Eq. 5.67). Finally, in terms of work per unit mass, wshaft = Ẇ shaft m ˙ , we obtain wshaft = −U1Vθ1 + U2Vθ2 (12.5) where we have used the fact that by conservation of mass, m ˙ 1 = m˙ 2. Equations 12.3, 12.4, and 12.5 are the basic equations for pumps or turbines whether the machines are radial-, mixed-, or axialflow devices and for compressible and incompressible flows. Note that neither the axial nor the radial component of velocity enter into the specific work (work per unit mass) equation. Another useful but more laborious form of Eq. 12.5 can be obtained by writing the right-hand side in a slightly different form based on the velocity diagrams at the entrance or exit as shown generically in Fig. 12.5. The velocity component Vx is the generic through-flow component of velocity and it can be axial, radial, or in-between depending on the rotor configuration. From the large right triangle we note that V 2 = V 2θ + V 2x or V x2 = V 2 − V 2θ (12.6) From the small right triangle we note that V x2 + (Vθ − U) 2 = W 2 (12.7) By combining Eqs. 12.6 and 12.7 we obtain V2 + U2 − W 2 2 which when written for the inlet and exit and combined with Eq. 12.5 gives VθU = wshaft = Turbomachine work is related to changes in absolute, relative, and blade velocities. V 22 − V12 + U 22 − U12 − (W 22 − W12 ) 2 (12.8) Thus, the power and the shaft work per unit mass can be calculated from the speed of the blade, U, the absolute fluid speed, V, and the fluid speed relative to the blade, W. This is an alternative to using fewer components of the velocity as suggested by Eq. 12.5. Equation 12.8 contains more terms than Eq. 12.5; however, it is an important concept equation because it shows that work is x θ V W U Vθ Vx ■ Figure 12.5 Velocity triangle: V = absolute velocity, W = relative velocity, U = blade velocity. 12.4 The Centrifugal Pump 687 associated with a change in kinetic energy (V2/2), work done by “centrifugal force” (U2/2), and reaction to acceleration or diffusion of the flow in the blade passage (W2/2). Thus the most effective pump will have a kinetic energy increase, flow radially outward, and flow decelerated within blade passages. The most effective turbine would have kinetic energy decrease, flow radially inward, and flow accelerated within blade passages. Because of the general nature of the velocity diagram in Fig. 12.5, Eq. 12.8 is applicable for axial-, radial-, and mixed-flow rotors. THE WIDE WORLD OF FLUIDS 1948 Buick Dynaflow started it Prior to 1948 almost all cars converter is similar to blowing air from a fan onto another fan had manual transmissions, which required the use of a clutch that is unplugged. One can hold the blade of the unplugged pedal to shift gears. The 1948 Buick Dynaflow was the first fan and keep it from turning, but as soon as it is let go, it will automatic transmission to use the hydraulic torque converter begin to speed up until it comes close to the speed of the and was the model for present-day automatic transmissions. powered fan. The torque converter uses transmission fluid Currently, in the United States over 84% of the cars have auto- (not air) and consists of a pump (the powered fan) driven by matic transmissions. The torque converter replaces the clutch the engine driveshaft, a turbine (the unplugged fan) connected found on manual shift vehicles and allows the engine to con- to the input shaft of the transmission, and a stator (absent in tinue running when the vehicle comes to a stop. In principle, the fan model) to efficiently direct the flow between the pump but certainly not in detail or complexity, operation of a torque and turbine. 12.4 VIDEO The Centrifugal Pump V12.3 Windshield washer pump One of the most common radial-flow turbomachines is the centrifugal pump. This type of pump has two main components: an impeller attached to a rotating shaft, and a stationary casing, housing, or volute enclosing the impeller. The impeller consists of a number of blades (usually curved), also sometimes called vanes, arranged in a regular pattern around the shaft. Figure 12.6 shows the essential features of a centrifugal pump. As the impeller rotates, fluid is sucked in through the eye of the casing and flows radially outward. Energy is added to the fluid by the rotating blades, and both pressure and absolute velocity are increased as the fluid flows from the eye to the periphery of the blades. For the simplest type of centrifugal pump, the fluid discharges directly into a volute-shaped casing. The casing shape reduces the velocity as the fluid leaves the impeller, and this decrease in kinetic energy is converted into an increase in pressure. The volute-shaped casing, with its increasing area in the direction of flow, is used to produce an essentially uniform velocity distribution as the fluid moves around the casing into the discharge opening. For large centrifugal pumps, a different design is often used in which diffuser guide vanes surround the impeller. The diffuser vanes decelerate the flow as the fluid is directed into the pump casing. This type of centrifugal pump is referred to as a diffuser pump. Discharge Impeller Hub plate Eye Inflow Blade Casing, housing, or volute (a) (b) ■ Figure 12.6 Schematic diagram of basic elements of a centrifugal pump. 688 Chapter 12 ■ Turbomachines (a) (b) ■ Figure 12.7 (a) Open impeller, (b) enclosed or shrouded impeller, (c) complete pump assembly. (Courtesy of Flowserve Corporation.) Centrifugal pumps involve radially outward flows. (c) Impellers are generally of two types. For one configuration the blades are arranged on a hub or backing plate and are open on the other (casing or shroud) side. A typical open impeller is shown in Fig. 12.7a. For the second type of impeller, called an enclosed or shrouded impeller, the blades are covered on both hub and shroud ends as shown in Fig. 12.7b. Pump impellers can also be single or double suction. For the single-suction impeller the fluid enters through the eye on only one side of the impeller, whereas for the double-suction impeller the fluid enters the impeller along its axis from both sides. The double-suction arrangement reduces end thrust on the shaft, and also, since the net inlet flow area is larger, inlet velocities are reduced. Figure 12.7c shows a double suction pump assembly. Pumps can be single or multistage. For a single-stage pump, only one impeller is mounted on the shaft, whereas for multistage pumps, several impellers are mounted on the same shaft. The stages operate in series; that is, the discharge from the first stage flows into the eye of the second stage, the discharge from the second stage flows into the eye of the third stage, and so on. The flowrate is the same through all stages, but each stage develops an additional pressure rise. Thus, a very large discharge pressure, or head, can be developed by a multistage pump. Centrifugal pumps come in a variety of arrangements (open or shrouded impellers, volute or diffuser casings, single- or double-suction, single- or multistage), but the basic operating principle remains the same. Work is done on the fluid by the rotating blades (centrifugal action and tangential blade force acting on the fluid over a distance), creating a large increase in kinetic energy as well as an increase in pressure of the fluid flowing through the impeller. The kinetic energy is converted into a further increase in pressure as the fluid flows from the impeller into the casing enclosing the impeller. A simplified theory describing the behavior of the centrifugal pump is developed in the following section. 12.4.1 Theoretical Considerations Although flow through a pump is very complex (unsteady and three-dimensional), the basic theory of operation of a centrifugal pump can be developed by considering the average one-dimensional flow of the fluid as it passes between the inlet and the outlet sections of the rotating impeller. As shown in Fig. 12.8, for a typical blade passage, the absolute velocity, V1, of the fluid entering the passage is the vector sum of the velocity of the blade, U1, rotating in a circular path with angular velocity ω, and the relative velocity, W1, within the blade passage so that V1 = W1 + U1. 12.4 The Centrifugal Pump 689 Similarly, at the exit V2 = W2 + U2. Note that U1 = r1ω and U2 = r2ω. Fluid velocities are taken to be average velocities over the inlet and exit sections of the blade passage. The relationship between the various velocities is shown graphically in Fig. 12.8. (In Fig. 12.8, the velocity vectors are shown in their “true” position, with each vector originating at a single point [fluid particle] at the impeller inlet and outlet. That these are equivalent to the previous triangular diagrams can be seen from the parallelogram rule for vector addition. Alternatively, one may imagine the W vector moved to the opposite side of the diagram). As discussed in Section 12.3, the moment-of-momentum equation indicates that the shaft torque, Tshaft, required to rotate the pump impeller is given by equation Eq. 12.2 applied to a pump with m ˙ 1 = m˙ 2 = m˙ . That is, Tshaft = m ˙ (r2Vθ2 − r1Vθ1 ) (12.9) Tshaft = ρQ(r2Vθ2 − r1Vθ1 ) (12.10) or where Vθ1 and Vθ2 are the tangential components of the absolute velocities, V1 and V2 (see Figs. 12.8b,c). For a rotating shaft, the power transferred, Ẇ shaft, is given by Ẇ shaft = Tshaftω Centrifugal pump impellers have an increase in blade velocity along the flow path (U2 > U1). and, therefore, from Eq. 12.10 Ẇ shaft = ρQω(r2Vθ2 − r1Vθ1 ) Since U1 = r1ω and U2 = r2ω we obtain Ẇ shaft = ρQ(U2Vθ2 − U1Vθ1 ) (12.11) W2 β2 V2 Vθ 2 α2 Vr 2 (c) U2 W2 β2 V2 α2 U2 r2 W1 W1 β1 β1 r1 α1 + ω (a) U1 V1 Vθ 1 α1 Vr 1 U1 (b) ■ Figure 12.8 Velocity diagrams at the inlet and exit of a centrifugal pump impeller. V1 690 Chapter 12 ■ Turbomachines Equation 12.11 shows how the power supplied to the shaft of the pump is transferred to the flowing fluid. It also follows that the shaft work per unit mass of flowing fluid is wshaft = Ẇ shaft = U2Vθ2 − U1Vθ1 ρQ (12.12) For incompressible flow in a pump, we get from Eq. 5.82 2 pout pin Vout V 2in + wshaft = ( + gzout) − ( + + gzin) + loss ρ ρ 2 2 Combining Eq. 12.12 with this, we get pout pin V 2out V 2in + U2Vθ2 − U1Vθ1 = ( + gzout) − ( + + gzin) + loss ρ ρ 2 2 Dividing both sides of this equation by the acceleration of gravity, g, we obtain U2Vθ 2 − U1Vθ1 = Hout − Hin + hL g where H is total head defined by H= p V2 + +z ρg 2g and hL = loss/g is head loss. From this equation we see that (U2Vθ2 − U1Vθ1 )g is the shaft work head added to the fluid by the pump. Head loss, hL, reduces the actual head rise, Hout − Hin , achieved by the fluid. Thus, the ideal head rise possible, hi, is hi = The pump actual head rise is less than the pump ideal head rise by an amount equal to the head loss in the pump. U2Vθ2 − U1Vθ1 g (12.13) The actual head rise, Hout − Hin = ha , is always less than the ideal head rise, hi, by an amount equal to the head loss, hL, in the pump. Some additional insight into the meaning of Eq. 12.13 can be obtained by using the following alternate version (see Eq. 12.8 and the associated discussion): hi = 1 [(V 22 − V 21 ) + (U 22 − U 21 ) + (W 21 − W 22 )] 2g (12.14) A useful relationship between the flowrate and the pump ideal head rise can be obtained as follows. Often the fluid has no tangential component of velocity Vθ1, or swirl, as it enters the impeller; that is, the angle between the absolute velocity and the tangential direction is 90° (α1 = 90° in Fig. 12.8(b)). In this case, Eq. 12.13 reduces to hi = U2Vθ2 g (12.15) From Fig. 12.8c cot β2 = U2 − Vθ2 Vr2 so that Eq. 12.15 can be expressed as hi = U2Vr2 cot β2 U 22 − g g (12.16) The flowrate, Q, is related to the radial component of the absolute velocity through the equation Q = 2πr2b2Vr2 (12.17) 12.4 The Centrifugal Pump 691 where b2 is the impeller blade height at the radius r2. Thus, combining Eqs. 12.16 and 12.17 yields hi = hi U2 cot β2 U 22 Q − g 2πr2b2g (12.18) This equation is graphed in the margin and shows that the ideal head rise for a centrifugal pump varies linearly with Q for a given blade geometry, diameter, and angular velocity. For actual pumps, the blade angle β2 falls in the range of 15°−35°, with a normal range of 20° < β2 < 25°, and with 15° < β1 < 50° (Ref. 11). Blades with β2 < 90° are called backward curved, whereas blades with β2 > 90° are called forward curved and blades with β2 = 90° are called radial. Pumps are not usually designed with forward-curved vanes because such pumps tend to experience unstable flow. Only small and inexpensive pumps are designed with radial vanes. Q E X AMPLE 12.2 Centrifugal Pump Performance Based on Inlet/Outlet Velocities GIVEN Water is pumped at the rate of 1400 gpm through a centrifugal pump operating at a speed of 1750 rpm. The impeller has a uniform blade height, b, of 2 in. with r1 = 1.9 in. and r2 = 7.0 in., and the exit blade angle β2 is 23° (see Fig. 12.8). Assume ideal flow conditions and that the tangential velocity component, Vθ1, of the water entering the blade is zero (α1 = 90°). FIND Determine (a) the tangential velocity component, Vθ2, at the exit, (b) the ideal head rise, hi, and (c) the power, Ẇ shaft, transferred to the fluid. Discuss the difference between ideal and ˙ shaft , ideal or actual? Explain. actual head rise. Is the power, W SOLUTION (a) At the exit the velocity diagram is as shown in Fig. 12.8c, where V2 is the absolute velocity of the fluid, W2 is the relative velocity, and U2 is the tip velocity of the impeller with (1750 rpm) U2 = r2ω = (7/12 ft) (2π rad/rev) (60 s/min) = 107 ft/s = (107 ft/s) 2 32.2 ft/s2 = 316 ft − (107 ft/s) (5.11 ft/s) cot 23° 32.2 ft/s2 (Ans) (c) From Eq. 12.11, with Vθ1 = 0, the power transferred to the fluid is given by the equation Ẇ shaft = ρQU2Vθ2 Since the flowrate is given, it follows that = Q = 2πr2b2Vr2 (1.94 slugs/ft3 )(1400 gpm)(107 ft/s)(95.0 ft/s) [1(slug ∙ ft/s2 )/lb](7.48 gal/ft3 )(60 s/min) = (61,500 ft ∙ lb/s)(1 hp/550 ft ∙ lb/s) = 112 hp or Q Vr2 = 2πr2b2 Note that the ideal head rise and the power transferred to the fluid are related through the relationship 1400 gpm = Ẇ shaft = ρgQhi 3 (7.48 gal/ft ) (60 s/min) (2π) (7/12 ft) (2/12 ft) = 5.11 ft/s COMMENT It should be emphasized that results given in the From Fig. 12.8c we see that cot β2 = U2 − Vθ2 Vr2 so that Vθ2 = U2 − Vr2 cot β2 = (107 − 5.11 cot 23°) ft/s = 95.0 ft/s (Ans) (b) From Eq. 12.15 the ideal head rise is given by U2Vθ2 (107 ft/s) (95.0 ft/s) = g 32.2 ft/s2 = 316 ft hi = Alternatively, from Eq. 12.16, the ideal head rise is hi = (Ans) U2Vr2 cot β2 U22 − g g (Ans) previous equation involve the ideal head rise. The actual head-rise performance characteristics of a pump are usually determined by experimental measurements obtained in a testing laboratory. The actual head rise is always less than the ideal head rise for a specific flowrate because of the loss of available energy associated with actual flows. Also, it is important to note that even if actual values of U2 and Vr2 are used in Eq. 12.16, the ideal head rise is calculated. The only idealization used in this example problem is that the exit flow angle is identical to the blade angle at the exit. If the actual exit flow angle was made available in this example, it could have been used in Eq. 12.16 to calculate the ideal head rise. The pump power, Ẇ shaft, is the actual power required to achieve a blade speed of 107 ft/s, a flowrate of 1400 gpm, and the tangential velocity, Vθ2, associated with this example. If pump losses could somehow be reduced to zero (every pump designer’s dream), the actual and ideal head rise would have been identical at 316 ft. As is, the ideal head rise is 316 ft and the actual head rise something less, probably about 275 ft. 692 Chapter 12 ■ Turbomachines Head Theoretical or ideal head, hi Other losses Friction losses Actual head, ha ■ Figure 12.9 Effect of losses on the pump head–flowrate curve. Flowrate Ideal and actual head-rise levels differ by the head loss in the pump. Figure 12.9 shows the ideal head versus flowrate curve (Eq. 12.18) for a centrifugal pump with backward-curved vanes (β2 < 90°). Since there are simplifying assumptions (i.e., zero losses) associated with the equation for hi, we would expect that the actual rise in head of fluid, ha, would be less than the ideal head rise, and this is indeed the case. As shown in Fig. 12.9, the ha versus Q curve lies below the ideal head-rise curve and shows a nonlinear variation with Q. The differences between the two curves (as represented by the shaded areas between the curves) arise from several sources. These differences include losses due to fluid skin friction in the blade passages, which vary as Q2, and other losses due to such factors as flow separation, impeller bladecasing clearance flows, and other three-dimensional flow effects. Near the design flowrate, some of these other losses are minimized. Centrifugal pump design is a highly developed field, with much known about pump theory and design procedures (see, for example, Refs. 4–6). However, due to the general complexity of flow through a centrifugal pump, the actual performance of the pump cannot be accurately predicted on a completely theoretical basis as indicated by the data of Fig. 12.9. Actual pump performance is determined experimentally through tests on the pump. From these tests, pump characteristics are determined and presented as pump performance curves. It is this information that is most helpful to the engineer responsible for incorporating pumps into a given flow system. 12.4.2 Pump Performance Characteristics The actual head rise, ha, and the pump flowrate, Q, are determined experimentally using an arrangement like that shown in Fig. 12.10. The flowrate is measured by a flowmeter of the type discussed in Sec. 8.6. The pump head rise is ha = p2 − p1 V 22 − V 21 + z2 − z1 + γ 2g (12.19) with sections (1) and (2) at the pump inlet and exit, respectively. Typically, the differences in elevation and velocity heads are small so that ha ≈ p2 − p1 γ (12.20) Valve for varying system resistance Flowrate meter Q (2) z2 – z1 (1) ■ Figure 12.10 Typical Behind pump (out of sight): Drive shaft and motor with speed and power measurement experimental arrangement for determining the performance of a pump. 12.4 The Centrifugal Pump 693 The power, 𝒫f, transferred to the fluid is given by the equation 𝒫f = γQha (12.21) and this quantity, expressed in terms of horsepower, is traditionally called the water horsepower. Thus, 𝒫f = water horsepower = γQha 550 (12.22) with γ expressed in lb/ft3, Q in ft3/s, and ha in ft. Note that the γ appearing in Eq. 12.22 must be the specific weight of the fluid moving through the pump. In addition to the head or power transferred to the fluid, the overall efficiency, η, is of interest, where η= 𝒫f power gained by the fluid = shaft power driving the pump Ẇ shaft (12.23) The denominator of this relationship represents the total power applied to the shaft of the pump and is often referred to as brake horsepower (bhp). Thus, η= ha Falling head curve Rising head curve Q The overall pump efficiency is diminished by the hydraulic losses in the pump, as previously discussed, and further reduced by the mechanical losses in the bearings and seals. There may also be some power loss due to leakage of the fluid between the back surface of the impeller hub plate and the casing, or through other pump components. This leakage contribution to the overall efficiency is called the volumetric loss. Thus, the overall efficiency arises from three sources, the hydraulic efficiency, ηh, the mechanical efficiency, ηm, and the volumetric efficiency, ηυ, so that η = ηhηmηυ. Performance characteristics for a given pump geometry and operating speed are usually given in the form of plots of ha, η, and bhp versus Q (commonly referred to as capacity) as illustrated in Fig. 12.11. Actually, only two curves are needed since ha, η, and bhp are related through Eq. 12.23. For convenience, all three curves are usually provided. Note that for the pump characterized by the data of Fig. 12.11, the head curve continuously rises as the flowrate decreases, and in this case the pump is said to have a rising head curve. As shown by the figure in the margin, pumps may also have ha − Q curves that initially rise as Q is decreased from the design value and then fall with a continued decrease in Q. These pumps have a falling (or drooping) head curve. The head developed by the pump at zero discharge is called the shutoff head, and it represents the rise in pressure head across the pump with the discharge valve closed. Since there is no flow with the valve closed, the related efficiency is zero, and the power supplied by the pump (bhp at Q = 0) is simply dissipated as heat. Although centrifugal pumps can be operated for short periods of time with the discharge valve closed, damage will occur due to overheating and large mechanical stress with any extended operation with the valve closed. As can be seen from Fig. 12.11, as the discharge is increased from zero the brake horsepower increases, with a subsequent fall as the maximum discharge is approached. As previously noted, Shutoff head Head Head, ha Brake horsepower, bhp Efficiency, η Overall pump efficiency is the ratio of power actually gained by the fluid to the shaft power supplied. γQha 550 bhp Efficiency Brake horsepower Normal or design flowrate 0 0 Flowrate, Q ■ Figure 12.11 Typical performance characteristics for a centrifugal pump of a given size operating at a constant impeller speed. 65 63 65 7 300 63 60 6 55 40 b 50 hp 200 30 25 15 100 20 NPSH R 0 15 10 5 0 40 80 120 160 200 240 Capacity, gal/min 280 320 0 NPSHR, ft Head, ft 400 60 8 in. dia 500 55 Chapter 12 ■ Turbomachines 50% 694 ■ Figure 12.12 Performance curves for a two-stage centrifugal pump operating at 3500 rpm. Data given for three different impeller diameters. with ha and bhp known, the efficiency can be calculated. As shown in Fig. 12.11, the efficiency is a function of the flowrate and reaches a maximum value at some particular value of the flowrate, commonly referred to as the normal or design flowrate or capacity for the pump. The points on the various curves corresponding to the maximum efficiency are denoted as the best efficiency points (BEP). When selecting a pump for a particular application, it is usually desirable to have the pump operate near its maximum efficiency. Thus, performance curves of the type shown in Fig. 12.11 are very important to the engineer responsible for the selection of pumps for a particular flow system. Matching the pump to a particular flow system is discussed in Section 12.4.4. Pump performance characteristics from pump manufacturers are also presented in charts of the type shown in Fig. 12.12. Impellers with different diameters may be used in a given casing, so performance characteristics for several impeller diameters can be provided with corresponding lines of constant efficiency and brake horsepower as illustrated in Fig. 12.12. Thus, the same information can be obtained from this type of graph as from the curves shown in Fig. 12.11. It is to be noted that an additional curve is given in Fig. 12.12, labeled NPSHR, which stands for net positive suction head required. As discussed in the following section, the significance of this curve is related to conditions on the suction side of the pump, which must also be carefully considered when selecting and positioning a pump. 12.4.3 Net Positive Suction Head (NPSH) Cavitation, which may occur in liquid pumps and hydraulic turbines, should be avoided. On the suction side of a pump, low pressures are commonly encountered, with the concomitant possibility of cavitation occurring within the pump. As discussed in Section 1.8, cavitation occurs when the liquid pressure at a given location is reduced to the vapor pressure of the liquid. When this occurs, vapor bubbles form (the liquid starts to “boil”); this phenomenon can cause a loss in efficiency as well as structural damage to the pump. To characterize the potential for cavitation, the difference between the total head on the suction side, near the pump impeller inlet, psγ + V 2s 2g, and the liquid vapor pressure head, pυγ, is used. This difference is called the net positive suction head (NPSH) NPSH = ps V 2s pυ + zs − + γ γ 2g (12.24) The position reference for the elevation head usually passes through the centerline of the pump impeller inlet so that zs = 0. There are actually two values of NPSH of interest. The first is the required NPSH, denoted NPSHR, that must be maintained, or exceeded, so that cavitation will not occur. The lowest pressure in the pump inlet occurs in the impeller eye, not at the suction flange, so it is necessary to determine the NPSHR by experiment. This is the curve shown in Fig. 12.12. Pumps are tested to determine the value for NPSHR, as defined by Eq. 12.24, by either directly detecting cavitation or observing a change in the head–flowrate curve or the efficiency (Ref. 7). The second value for NPSH of concern is the available NPSH, denoted NPSHA, which represents the suction head that actually occurs for 12.4 The Centrifugal Pump 695 (2) Reference plane z1 p1 = patm (1) ■ Figure 12.13 Schematic of a pump installation in which the pump must lift fluid from one level to another. the particular flow system. This value can be determined experimentally, or calculated if the system parameters are known. A typical flow system is shown in Fig. 12.13. The energy equation applied between the free liquid surface, where the pressure is atmospheric, patm, and a point on the suction side of the pump near the impeller inlet yields ps V 2s patm + ∑ hL + z1 = + γ γ 2g where ∑hL represents head losses between the free surface and the pump inlet. The elevation, z1, is measured from the reference plane, usually passing through the pump shaft. For the situation shown in Fig. 12.13, z1 is a negative number. Thus, the head available at the pump inlet is ps V 2s patm = + + z1 − ∑ hL γ γ 2g so that NPSHA = For proper pump operation, the available net positive suction head must be greater than the required net positive suction head. (12.25) For this calculation, absolute pressures are normally used since the vapor pressure is usually specified as an absolute pressure. For proper pump operation it is necessary that NPSHA ≥ NPSHR It is noted from Eq. 12.25 that as the height of the pump above the fluid surface is increased (i.e., z1 becomes more negative), the NPSHA is decreased. Therefore, there is some maximum height that the liquid can be lifted above which the pump cannot operate without cavitation. The specific value depends on the head losses and the value of the vapor pressure. It is further noted that if the supply tank or reservoir is above the pump, z1 will be positive in Eq. 12.25, and the NPSHA will increase as this height is increased. E X AMPLE 12.3 GIVEN patm pυ + z1 − ∑ hL − γ γ Net Positive Suction Head A centrifugal pump is to be placed above a large, open water tank, as shown in Fig. 12.13, and is to pump water at a rate of 0.5 ft3/s. At this flowrate the required net positive suction head, NPSHR, is 15 ft, as specified by the pump manufacturer. The water temperature is 80 °F and atmospheric pressure is 14.7 psi. Assume that the major head loss between the tank and the pump inlet is due to a filter at the pipe inlet having a minor loss coefficient KL = 20. Other losses can be neglected. The pipe on the suction side of the pump has a diameter of 4 in. FIND Determine the maximum height that the pump can be located above the water surface without cavitation. If you were required to place a valve in the flow path, would you place it upstream or downstream of the pump? Why? 696 Chapter 12 ■ Turbomachines SOLUTION From Eq. 12.25 the available net positive suction head, NPSHA, is given by the equation NPSHA = From Table B.1 the water vapor pressure at 80 °F is 0.5069 psia and γ = 62.22 lb/ft3. Equation (1) can now be written as patm pυ + z1 − ∑ hL − γ γ (−z1 ) max = − 10.2 ft 62.22 lb/ft3 (0.5069 lb/in.2 ) (144 in.2/ft2 ) − − 15 ft 62.22 lb/ft3 = 7.65 ft (Ans) and the limiting value for z1 will occur when NPSHA = NPSHR. Thus, (−z1 ) max = patm pυ − ∑ hL − − NPSHR γ γ Since the only head loss to be considered is the loss ∑ hL = KL V2 2g (1) Thus, to prevent cavitation, with its accompanying poor pump performance and long-term damage, the pump should not be located higher than 7.65 ft above the water surface. COMMENT If the valve is placed upstream of the pump, it with V= (14.7 lb/in.2 ) (144 in.2/ft2 ) 3 Q 0.5 ft /s = = 5.73 ft/s A (π/4) (4/12 ft) 2 would now operate with a lower inlet pressure because of the additional upstream loss and could now suffer cavitation with its usually negative consequences. So, placing the valve on the downstream side of the pump is normally the better choice. it follows that ∑ hL = (20) (5.73 ft/s) 2 2(32.2 ft/s2 ) = 10.2 ft 12.4.4 System Characteristics, Pump-System Matching, and Pump Selection The system equation relates the head required by the system to the flowrate through the system. A typical flow system in which a pump is used is shown in Fig. 12.14. The mechanical energy equation applied between points (1) and (2) indicates that ha = (p2 − p1 )γz2 − z1 + ∑ hL where ha is the actual head supplied by the pump, and ∑hL represents all friction losses in the pipe and minor losses for pipe fittings and valves. From our study of pipe flow, we know that hL varies approximately as the flowrate squared; that is, hL ∝ Q2 (see Section 8.4). For the system shown, the net pressure change is zero, but this is not always true. Thus, Eq. 12.26 can be written in the form ha = z2 − z1 + KQ2 ha z2 – z1 ha = z2 – z1 + KQ2 Q (12.26) (12.27) where K depends on the pipe sizes and lengths, friction factors, and minor loss coefficients. Equation 12.27 is the system equation and shows how the actual head gained by the fluid from the pump is related to the system parameters. A plot of the system equation like the figure in the margin is called a system curve. In this case the system parameters include the change in elevation head, z2 − z1, and the losses due to friction as expressed by KQ2. Each flow system has its own specific system equation. If the flow is laminar, the pipe frictional losses will be proportional to Q rather than Q2 (see Section 8.2). (2) (1) z2 z1 Pump ■ Figure 12.14 Typical flow system. 12.4 Best efficiency System point (BEP) curve (B) (A) Actual pump head, ha Pump performance curve Efficiency curve Operating point Elevation (static) head = z 2 – z1 ■ Figure 12.15 Utilization of the system curve and the pump performance curve to obtain the operating point for the system. Flowrate, Q The intersection of the pump performance curve and the system curve is the operating point. 697 Efficiency Change in system equation The Centrifugal Pump There is also a unique relationship between the actual pump head gained by the fluid and the flowrate, which is governed by the pump design (as indicated by the pump performance curve). To determine the operating point when a specific pump is installed in a specific system, it is necessary to utilize both the system curve, as determined by the system equation, and the pump performance curve. If both curves are plotted on the same graph, as illustrated in Fig. 12.15, their intersection (point A) represents the operating point for the system. That is, this point gives the head and flowrate that satisfy both the system equation and the pump equation. On the same graph the pump efficiency is shown. Ideally, we want the operating point to be near the best efficiency point (BEP) for the pump. For a given pump, it is clear that as the system equation changes, the operating point will shift. For example, if the pipe friction increases due to pipe wall fouling, the system curve changes, resulting in the operating point A shifting to point B in Fig. 12.15 with a reduction in flowrate and efficiency. The following example shows how the system and pump characteristics can be used to determine an operating point and decide if a particular pump is suitable for a given application. THE WIDE WORLD OF FLUIDS Space Shuttle fuel pumps The fuel pump of your car engine is Pump speeds were as high as 35,360 rpm. The liquid oxygen vital to its operation. Similarly, the fuels (liquid hydrogen and was pumped from 100 to 7420 psia, the liquid hydrogen from 30 oxygen) of each Space Shuttle main engine (there are three per to 6515 psia. Liquid hydrogen and oxygen flowrates of about shuttle) relied on multistage turbopumps to get from storage 17,200 gpm and 6100 gpm, respectively, were achieved. These tanks to main combustors. High pressures were utilized through- pumps could empty your home swimming pool in seconds. The out the pumps to avoid cavitation. The pumps, some centrifugal hydrogen went from −423 °F in storage to +6000 °F in the com- and some axial, were driven by axial-flow, multistage turbines. bustion chamber! E X AMPLE 12.4 Use of Pump Performance Curves GIVEN Water is to be pumped from one large, open tank to a second large, open tank as shown in Fig. E12.4a. The pipe diameter throughout is 6 in., and the total length of the pipe between the pipe entrance and exit is 200 ft. Minor loss coefficients for the entrance, exit, and the elbow are shown, and the friction factor for the pipe can be assumed constant and equal to 0.02. A certain centrifugal pump having the performance characteristics shown in Fig. E12.4b is suggested as a good pump for this flow system. FIND With this pump, what would be the flowrate between the tanks? Do you think this pump would be a good choice? 698 Chapter 12 ■ Turbomachines (2) Water KL = 1.0 KL = 1.5 Δz = 10 ft Diameter of pipe = 6 in. Total pipe length = 200 ft Pump (1) KL = 0.5 (a) 100 ha, ft Efficiency, % 80 Head 60 40 Efficiency 20 0 0 400 800 1200 1600 2000 2400 Q, gal/min ■ Figure E12.4a, b (b) SOLUTION Application of the energy equation between the two free surfaces, points (1) and (2) as indicated, gives p2 p1 V 21 V 22 + z1 + ha = + z2 + + γ γ 2g 2g +f ℓ V2 V2 + ∑ KL D 2g 2g (1) Thus, with p1 = p2 = 0, V1 = V2 = 0, Δz = z2 − z1 = 10 ft, f = 0.02, D = 6/12 ft, and ℓ = 200 ft, Eq. 1 becomes ha = 10 + 0.02 [ (200 ft) (6/12 ft) + (0.5 + 1.5 + 1.0) ] V2 2(32.2 ft /s2 ) (2) where the given minor loss coefficients have been used. Since V= Q(ft3/s) Q = A (π/4) (6/12 ft) 2 Eq. 2 can be expressed as ha = 10 + 4.43 Q2 (3) where Q is in ft3/s, or with Q in gallons per minute ha = 10 + 2.20 × 10−5Q2 (4) Equation 3 or 4 represents the system equation for this particular flow system and reveals how much actual head the fluid will need from the pump to maintain a certain flowrate. Performance data shown in Fig. E12.4b indicate the actual head provided by this particular pump when it operates at a certain flowrate. Thus, when Eq. 4 is plotted on the same graph with performance data, the intersection of the two curves represents the operating point for the pump and the system. This combination is shown in Fig. E12.4c with the intersection observed to occur at Q = 1600 gal/min (Ans) with the corresponding actual head gained equal to 66.5 ft. Another concern is whether the pump is operating efficiently at the operating point. As can be seen from Fig. E12.4c, although this is not peak efficiency, which is about 86%, it is close (about 84%). Thus, this pump would be a satisfactory choice, assuming the 1600 gal/min flowrate is at or near the desired flowrate. 12.4 100 120 Efficiency Δz = 100 ft Δz = 10 ft 84% Operating point 66.5 84% 80 Head, ha, ft Efficiency, % Head, ha, ft Efficiency, % Δz = 80 ft 100 Head 699 The Centrifugal Pump System curve (Eq. 4) 500 gal/min 1600 gal/min 60 40 36% 0 0 1600 20 2400 Efficiency Q, gal/min ■ Figure E12.4c (Continued) 0 System 0 400 800 1200 1600 2000 2400 Q, gal/min The power needed to drive the pump is Ẇ shaft = = ■ Figure E12.4d γQha η (62.4 lb/ft3 )[(1600 gal/min)(7.48 gal/ft3 ) (60 s/min)](66.5 ft) 0.84 = 17,600 ft∙lb/s = 32.0 hp By repeating the calculations for Δz = z2 − z1 = 80 ft and 100 ft (rather than the given 10 ft), the results shown in Fig. E12.4d are obtained. Although the given pump could be used with Δz = 80 ft (provided that the 500 gal/min flowrate produced is acceptable), it would not be an acceptable pump for this application since its efficiency would be only 36%. Energy could be saved by using a different pump with a performance curve that more nearly matches the new system requirements (i.e., higher efficiency at the operating condition). On the other hand, the given pump would not work at all for Δz = 100 ft COMMENT Pumps can be arranged in series or in parallel to provide for additional head or flow capacity. When two pumps are placed in series, the resulting pump performance curve is obtained by adding heads at the same flowrate. As illustrated in Fig. 12.16a, for two identical pumps in series, both the actual head gained by the fluid and the flowrate are increased, but neither will be doubled if the system curve remains the same. The operating point is at (A) for one pump and moves to (B) for two pumps in series. For two identical pumps in parallel, the combined performance curve is obtained by adding flowrates at the same head, as shown in Fig. 12.16b. As illustrated, the flowrate for the system will not be doubled with the addition of two pumps in parallel (if the same system (B) (A) One pump P P Head, ha System curve Two pumps Head, ha For two pumps in series, add heads; for two in parallel, add flowrates. since its maximum head (ha = 88 ft when Q = 0) is not enough to lift the water 100 ft, let alone overcome head losses. This is shown in Fig. E12.4d by the fact that for Δz = 100 ft the system curve and the pump performance curve do not intersect. Note that head loss within the pump itself was accounted for with the pump efficiency, η. Thus, hs = ha η , where hs is the pump shaft work head and ha is the actual head rise experienced by the flowing fluid. If an engineer prefers an analytical procedure instead of a graphical procedure, all that needs to be done is to fit a least squares curve to the pump curve (a second order curve will usually suffice). Upon equating the expression for the pump curve to the expression for the system curve, the engineer has only to solve a polynomial equation. Of course a personal computer, scientific calculator, or even a smart phone makes these mathematical tasks almost trivial. System curve Two pumps (B) (A) P One pump P Flowrate, Q Flowrate, Q (a) (b) ■ Figure 12.16 Effect of operating pumps in (a) series and (b) in parallel. 700 Chapter 12 ■ Turbomachines curve applies). However, for a relatively flat system curve, as shown in Fig. 12.16b, a significant increase in flowrate can be obtained as the operating point moves from point (A) to point (B). 12.5 Dimensionless Parameters and Similarity Laws As discussed in Chapter 7, dimensional analysis is particularly useful in the planning and execution of experiments and in presenting the data. Since the characteristics of pumps are usually determined experimentally, it is expected that dimensional analysis and similitude considerations will prove to be useful in the study and documentation of these characteristics. From the previous section we know that the principal dependent pump variables are the actual head rise, ha, shaft power, Ẇ shaft, and efficiency, η. We expect that these variables will depend on the geometrical configuration, which can be represented by some characteristic diameter, D, other pertinent lengths, ℓi, and surface roughness, ε. In addition, the other important variables are flowrate, Q, the pump shaft rotational speed, ω, fluid viscosity, μ, and fluid density, ρ. We will only consider incompressible fluids presently, so compressibility effects need not concern us yet. Thus, any one of the dependent variables ha, Ẇ shaft, and η can be expressed as Dependent variable = f (D, ℓi, ε, Q, ω, μ, ρ) and a straightforward application of dimensional analysis leads to ℓi ε Q ρωD2 Dependent pi term = ϕ( , , , D D ωD3 μ ) Dimensionless pi terms and similarity laws are important pump considerations. (12.28) The dependent pi term involving the head is usually expressed as CH = gha ω2D2, where gha is the actual head rise in terms of energy per unit mass, rather than simply ha, which is energy per unit weight. This dimensionless parameter is called the head rise coefficient or, more simply, the head coefficient. The dependent pi term involving the shaft power is expressed as C𝒫 = Ẇ shaft ρω3D5, and this standard dimensionless parameter is termed the power coefficient. The power appearing in this dimensionless parameter is commonly based on the shaft (brake) horsepower, bhp, so that in BG and EE units, Ẇ shaft = 550 × (bhp). The rotational speed, ω, which appears in these dimensionless groups is expressed in rad/s. The final dependent pi term is the efficiency, η, which is already dimensionless. Thus, in terms of dimensionless parameters the performance characteristics are expressed as gha ℓi ε Q ρωD2 = ϕ 1 ( D , D , ωD3 , μ ) ω2D2 Ẇ shaft ℓi ε Q ρωD2 C𝒫 = = ϕ , , 2( , D D ωD3 μ ) ρω3D5 ρgQha ℓi ε Q ρωD2 η= = ϕ3( , , , D D ωD3 μ ) Ẇ CH = shaft The last pi term in each of the above equations is a form of Reynolds number that represents the relative influence of viscous effects. When the pump flow involves high Reynolds numbers, as is usually the case, experience has shown that the effect of the Reynolds number can be neglected. (This high Reynolds number effect can be observed in the Moody diagram, Fig. 8.20.) For simplicity, the relative roughness, ε/D, can also be neglected in pumps since the highly irregular shape of the pump chamber is usually the dominant geometric factor rather than the surface roughness. Thus, with these simplifications and for geometrically similar pumps (all pertinent dimensions, ℓi, scaled by a common length scale), the dependent pi terms are functions of only QωD3, so that gha Q = ϕ1( 3 ) 2 2 ωD ωD Ẇ shaft Q = ϕ2( 3 ) 3 5 ωD ρω D Q η = ϕ3( 3 ) ωD (12.29) (12.30) (12.31) 12.5 0.25 100% 40 40 80 30 60 40 20 10 0 0 1000 2000 3000 Q, gal/min 4000 0 5000 η 40 20 CH Brake horsepower Head, ha, ft 100 Horsepower 60 CH 0.15 0 50 20 80 0.20 20 Head 60 Efficiency 60 80 70 100% η 80 Efficiency 701 Dimensionless Parameters and Similarity Laws 0 0.10 0.016 CP (pump) 0.012 0.008 CP 0.05 0.004 0 0 (a) 0.025 0.050 CQ 0.075 0 0.100 (b) ■ Figure 12.17 Typical performance data for a centrifugal pump: (a) characteristic curves for a 12-in. centrifugal pump operating at 1000 rpm, (b) dimensionless characteristic curves. (Data from Ref. 8, IIHRHydroscience & Engineering, University of Iowa College of Engineering, used by permission.) The dimensionless parameter CQ = QωD3 is called the flow coefficient. These three equations provide the desired similarity relationships among a family of geometrically similar pumps. If two pumps from the family are operated at the same value of flow coefficient Q Q = 3 ( ωD )1 ( ωD3 )2 (12.32) gha gha ( ω2D2 )1 = ( ω2D2 )2 (12.33) Ẇ shaft Ẇ shaft = ( ρω3D5 )1 ( ρω3D5 )2 (12.34) η1 = η2 (12.35) it then follows that Pump scaling laws relate geometrically similar pumps. where the subscripts 1 and 2 refer to any two pumps from the family of geometrically similar pumps. With these so-called pump scaling laws it is possible to experimentally determine the performance characteristics of one pump in the laboratory and then use these data to predict the corresponding characteristics for other pumps within the family under different operating conditions. Figure 12.17a shows some typical curves obtained for a centrifugal pump. Figure 12.17b shows the results plotted in terms of the dimensionless coefficients, CQ, CH, C𝒫, and η. From these curves the performance of different-sized, geometrically similar pumps can be predicted, as can the effect of changing speeds on the performance of the pump from which the curves were obtained. It is to be noted that the efficiency, η, is related to the other coefficients through the relationship η = CQCH C −1 𝒫 . This follows directly from the definition of η. EX AMPLE 12.5 GIVEN Use of Pump Scaling Laws An 8-in.-diameter centrifugal pump operating at 1200 rpm is geometrically similar to the 12-in.-diameter pump having the performance characteristics of Figs. 12.17a and 12.17b while operating at 1000 rpm. The working fluid is water at 60 °F. FIND For peak efficiency, predict the discharge, actual head rise, and shaft horsepower for this smaller pump. 702 Chapter 12 ■ Turbomachines SOLUTION As is indicated by Eq. 12.31, for a given efficiency the flow coefficient has the same value for a given family of pumps. From Fig. 12.17b we see that at peak efficiency CQ = 0.0625. Thus, for the 8-in. pump and Ẇ shaft = C𝒫 ρω3D5 = (0.014) (1.94 slugs/ft3 ) (126 rad/s) 3 (8/12 ft) 5 = 7150 ft ∙ lb/s 7150 ft ∙ lb/s Ẇ shaft = = 13.0 hp (Ans) 550 ft ∙ lb/s/hp Q = CQωD3 = (0.0625) (1200/60 rev/s) (2π rad/rev) (8/12 ft) 3 Q = 2.33 ft3/s (Ans) or in terms of gpm Q = (2.33 ft3/s) (7.48 gal/ft3 ) (60 s/min) = 1046 gpm (Ans) The actual head rise and the shaft horsepower can be determined in a similar manner since at peak efficiency CH = 0.19 and C𝒫 = 0.014, so that with ω = 1200 rev/min (1 min/60 s) (2π rad/rev) = 126 rad/s 2 2 2 𝒫f = γQha = (62.4 lb/ft3 ) (2.33 ft3/s) (41.6 ft) = 6050 ft ∙ lb/s Thus, the efficiency, η, is η= 2 CH ω D (0.19) (126 rad/s) (8/12 ft) = 41.6 ft (Ans) = g 32.2 ft/s2 ha = COMMENT This last result gives the shaft horsepower, which is the power supplied to the pump shaft. The power actually gained by the fluid is equal to γQha, which in this example is 𝒫f 6050 = = 85% ̇ 7150 Wshaft which checks with the efficiency curve of Fig. 12.17b. 12.5.1 Special Pump Scaling Laws Two special cases related to pump similitude commonly arise. In the first case we are interested in how a change in the operating speed, ω, for a given pump, affects pump characteristics. It follows from Eq. 12.32 that for the same flow coefficient (and therefore the same efficiency) with D1 = D2 (the same pump) Effects of changes in pump operating speed and impeller diameter are often of interest. 8 Q1 ω1 = ω2 Q2 The subscripts 1 and 2 now refer to the same pump operating at two different speeds at the same flow coefficient. Also, from Eqs. 12.33 and 12.34 it follows that ⋅ ⋅ Wshaft1/Wshaft2 6 ha1/ha2 4 ha1 ω21 = 2 ha2 ω2 (12.37) Ẇ shaft1 ω31 = 3 ω2 Ẇ shaft2 (12.38) and 2 0 (12.36) Q1/Q2 0 0.5 1 ω 1/ω 2 1.5 2 Thus, for a given pump operating at a given flow coefficient, the flow varies directly with speed, the head varies as the speed squared, and the power varies as the speed cubed. These effects of angular velocity variation are illustrated in the sketch in the margin. These scaling laws are useful in estimating the effect of changing pump speed when some data are available from a pump test obtained by operating the pump at a particular speed. A question that often arises in connection with application of Eqs. 12.37 and 12.38 is “does the pump operating point at speed “1” transform into the pump operating point at speed “2”?”. Since the operating point depends on the system characteristics as well as the pump characteristics, the answer is generally “no.” There is one very important exception. If the system is unchanged and the system equation is of the form hs = KQ2 (i.e., no elevation or pressure head) then the operating point does transform directly. In the second special case we are interested in how a change in the impeller diameter, D, of a geometrically similar family of pumps, operating at a given speed, affects pump characteristics. As before, it follows from Eq. 12.32 that for the same flow coefficient with ω1 = ω2 Q1 D31 = 3 Q2 D2 (12.39) 12.5 Dimensionless Parameters and Similarity Laws 703 Similarly, from Eqs. 12.33 and 12.34 8 Q1/Q2 4 2 0 ha1/ha2 0 0.5 1 (12.40) Ẇ shaft1 D51 = 5 D2 Ẇ shaft2 (12.41) and ⋅ ⋅ Wshaft1/Wshaft2 6 ha1 D21 = 2 ha2 D2 1.5 D1/D2 Pump affinity laws relate the same pump at different speeds or geometrically similar pumps at the same speed. 2 Thus, for a family of geometrically similar pumps operating at a given speed and the same flow coefficient, the flow varies as the diameter cubed, the head varies as the diameter squared, and the power varies as the diameter raised to the fifth power. These strong effects of diameter variation are illustrated in the sketch in the margin. These scaling relationships are based on the condition that, as the impeller diameter is changed, all other important geometric variables are properly scaled to maintain geometric similarity. This type of geometric scaling is not always possible due to practical difficulties associated with manufacturing the pumps. It is common practice for manufacturers to put impellers of different diameters in the same size pump casing. In this case, complete geometric similarity is not maintained, and the scaling relationships expressed in Eqs. 12.39, 12.40, and 12.41 will not, in general, be valid. However, experience has shown that if the impeller diameter change is not too large, less than about 20%, these scaling relationships can still be used to estimate the effect of a change in the impeller diameter. The pump similarity laws expressed by Eqs. 12.36 through 12.41 are sometimes referred to as the pump affinity laws. The effects of viscosity and surface roughness have been neglected in the foregoing similarity relationships. However, it has been found that as the pump size decreases, these effects more significantly influence efficiency because of smaller clearances and blade size. An approximate, empirical relationship to estimate the influence of diminishing size on efficiency is 1 − η2 D1 n ≈( ) 1 − η1 D2 (12.42) The value of n is usually given as 1/4 or 1/5; however, values as small as 1/10 and as large as 1/2 have been proposed (Ref. 9). In general, it is to be expected that the similarity laws will not be very accurate if tests on a model pump with water are used to predict the performance of a prototype pump with a highly viscous fluid, such as oil, because at the much smaller Reynolds number associated with the oil flow, the fluid physics involved is different from the higher Reynolds number flow associated with water. Some empirical methods of correcting for Reynolds number effects are discussed in Ref. 10. 12.5.2 Specific Speed A useful dimensionless parameter can be obtained by eliminating diameter D between the flow coefficient and the head rise coefficient. This is accomplished by raising the flow coefficient to an exponent (1/2) and dividing this result by the head coefficient raised to another exponent (3/4) so that ω√Q (QωD3 ) 1/2 = = Ns (gha ω2D2 ) 3/4 (gha ) 3/4 (12.43) The dimensionless parameter Ns is called the specific speed. Specific speed varies with flow coefficient just as the other coefficients and efficiency discussed earlier do. However, it is customary to specify a value of specific speed at the flow coefficient corresponding to peak efficiency only, i.e., at the BEP. For pumps with low Q and high ha, the specific speed is low compared to a pump with high Q and low ha. Centrifugal pumps typically are low-capacity, high-head pumps, and therefore have low specific speeds. Specific speed as defined by Eq. 12.43 is dimensionless, and therefore independent of the system of units used in its evaluation as long as a consistent unit system is used. However, in the United States a modified, dimensional form of specific speed, Nsd, is commonly used, where Nsd = ω(rpm) √Q(gpm) [ha (ft)]3/4 (12.44) 704 Chapter 12 ■ Turbomachines 15,000 500 0.6 1000 0.7 10,000 0.8 4000 5000 ηBEP 0.9 In this case Nsd is said to be expressed in U.S. customary units. Typical values of Nsd are in the range 500 < Nsd < 4000 for centrifugal pumps. Both Ns and Nsd have the same physical meaning, but their magnitudes will differ by a constant conversion factor (Nsd = 2733 Ns ) when ω in Mixed flow Centrifugal Eq. 12.43 is expressed in rad /s. Axial pump flow Each family or class of pumps has a particular range of values of specific speed associated with it. Thus, pumps that have low-capacity, high-head characteristics will have specific speeds that are smaller than pumps that have high-capacity, low-head characteristics. The specific speed is very useful to engineers and designers, since if the required head, Ns flowrate, and speed are specified, it is possible to select an appropriate (most efficient) type of pump for a particular application. For example, as shown by the figure in the margin, as the specific speed, Nsd, increases beyond about 2000 the peak efficiency, η, of the purely radial-flow centrifugal pump starts to fall off, and other types of more efficient pump design are preferred. In addition to the centrifugal pump, the axial-flow pump is widely used. As discussed in Section 12.6, in an axial-flow pump the direction of flow is primarily parallel to the rotating shaft rather than radial as in the centrifugal pump. Axial-flow pumps are essentially high-capacity, lowhead pumps, and therefore have large specific speeds (Nsd > 9000). Mixed-flow pumps combine features of both radial-flow and axial-flow pumps and have intermediate values of specific speed. Figure 12.18 illustrates how the specific speed changes as the configuration of the pump changes from centrifugal or radial to axial. 2000 1.0 12.5.3 Suction Specific Speed With an analysis similar to that used to obtain the specific speed, the suction specific speed, Ss, can be expressed as Ss = (12.45) [g(NPSHR )]3/4 where ha in Eq. 12.43 has been replaced by the required net positive suction head (NPSHR ). This dimensionless parameter is useful in determining the required operating conditions on the suction side of the pump. As was true for the specific speed, Ns, the value for Ss commonly used is for peak efficiency. For a family of geometrically similar pumps, Ss should have a fixed value. If this value is known, then the NPSHR can be estimated for other pumps within the same family operating at different values of ω and Q. As noted for Ns, the suction specific speed as defined by Eq. 12.45 is also dimensionless, and the value for Ss is independent of the system of units used. However, as was the case for specific Impeller shrouds 20000 15000 9000 10000 8000 Vanes 8.0 7.0 Axis of rotation Axial flow 6.0 0.9 1.0 0.8 0.7 0.6 7000 6000 Vanes Mixed flow 0.5 0.4 0.3 Radial flow Impeller hub 5.0 Hub Vanes 4.0 Vanes Hub 3.0 Vanes Hub Impeller shrouds 2.0 Hub 5000 4000 3000 2000 1500 Impeller shrouds 900 1000 800 700 600 500 Specific speed, Nsd 0.2 Specific speed may be used to approximate what general pump geometry (axial, mixed, or radial) to use for maximum efficiency. ω√Q Specific speed, Ns ■ Figure 12.18 Variation in specific speed at maximum efficiency with type of pump. (Adapted from Ref. 11, used with permission. Courtesy of Hydraulic Institute, Parsippany, NJ 07054 www.Pumps.org.) 12.6 Axial-Flow and Mixed-Flow Pumps 705 speed, in the United States a modified dimensional form for the suction specific speed, designated as Ssd, is commonly used, where Ssd = ω(rpm) √Q(gpm) [NPSHR (ft)]3/4 (12.46) For double-suction pumps the discharge, Q, in Eq. 12.46 is one-half the total discharge. Typical values for Ssd fall in the range of 7000 to 12,000 with a “nominal” value of approximately 7900 (Ref. 12). If Ssd is specified, Eq. 12.46 can be used to estimate the NPSHR for a given set of operating conditions. However, this calculation would generally only provide an approximate value for the NPSHR, and the actual determination of the NPSHR for a particular pump should be made through a direct measurement whenever possible. Note that Ssd = 2733 Ss, with ω expressed in rad/s in Eq. 12.45. 12.6 Axial-flow pumps often have alternating rows of stator blades and rotor blades. Axial-Flow and Mixed-Flow Pumps As noted previously, centrifugal pumps are radial-flow machines that operate most efficiently for applications requiring high heads at relatively low flowrates. This head–flowrate combination typically yields specific speeds (Nsd ) that are less than approximately 4000. For many applications, such as those associated with drainage and irrigation, high flowrates at low heads are required and centrifugal pumps are not suitable. In this case, axial-flow pumps are commonly used. This type of pump consists essentially of a propeller confined within a cylindrical casing. Axial-flow pumps are often called propeller pumps. For this type of pump the flow is primarily in the axial direction (parallel to the axis of rotation of the shaft), as opposed to the radial flow found in the centrifugal pump. Whereas the head developed by a centrifugal pump includes a contribution due to centrifugal action, the head developed by an axial-flow pump is due primarily to the tangential force exerted by the rotor blades on the fluid. A schematic of an axial-flow pump arranged for vertical operation is shown in Fig. 12.19. The rotor is connected to a motor through a shaft, and as it rotates (usually at a relatively high speed) the fluid is sucked in through the inlet. Typically the fluid discharges through a row of fixed stator (guide) vanes used to straighten the flow leaving the rotor. Some axial-flow pumps also have inlet guide vanes upstream of the rotor row, and some are multistage in which pairs (stages) of rotating blades (rotor blades) and fixed vanes (stator blades) are arranged in series. Axial-flow pumps usually have specific speeds (Nsd ) in excess of 9000. Very high specific speed pumps do not have stator vanes. The definitions and broad concepts that were developed for centrifugal pumps are also applicable to axial-flow pumps. The actual flow characteristics, however, are quite different. In Fig. 12.20 typical head, power, and efficiency characteristics are compared for a centrifugal pump and an axial-flow pump. It is noted that at design capacity (maximum efficiency) the head and brake Shaft to motor Discharge Drive shaft Fixed stator blades Rotor blades Inlet ■ Figure 12.19 Schematic diagram of an axial-flow pump arranged for vertical operation. Chapter 12 ■ Turbomachines 50 40 Brake Horsepower 500 60 Head, ft Centrifugal pump Axial-flow pump 400 300 Brake horsepower 200 Head 30 100 20 10 Percent Efficiency 706 80 60 Efficiency 40 20 0 0 10 20 30 40 50 60 Q, 1000 gal/min ■ Figure 12.20 Comparison of performance characteristics for a centrifugal pump and an axialflow pump, each rated 42,000 gal/min at a 17-ft head. (Data from Ref. 13, used with permission.) horsepower are the same for the two pumps selected. But as the flowrate decreases, the power input to the centrifugal pump falls to 180 hp at shutoff, whereas for the axial-flow pump the power input increases to 520 hp at shutoff. This characteristic of the axial-flow pump can cause overloading of the drive motor if the flowrate is reduced significantly from the design capacity. It is also noted that the head curve for the axial-flow pump is much steeper than that for the centrifugal pump. Thus, with axial-flow pumps there will be a large change in head with a small change in the flowrate, whereas for the centrifugal pump, with its relatively flat head curve, there will be only a small change in head with large changes in the flowrate. It is further observed from Fig. 12.20 that, except at design capacity, the efficiency of the axial-flow pump is lower than that of the centrifugal pump. To improve operating characteristics, some axial-flow pumps are constructed with adjustable blades. The characteristics of centrifugal and axial-flow pumps have important ramifications for a strategy for starting the pumps. A centrifugal pump should be started with the system closed because the pump requires less power with no flow. An axial-flow pump, on the other hand, should be started with the system wide open to avoid the power spike at zero flow. For applications requiring specific speeds intermediate to those for centrifugal and axial-flow pumps, mixed-flow pumps have been developed that operate efficiently in the specific speed range 4000 < Nsd < 9000. As the name implies, the flow in a mixed-flow pump has both a radial and an axial component. Figure 12.21 shows some typical data for centrifugal, mixed-flow, and Runner vanes Runner vanes Diameter D Eye A B C D Type Centrifugal Centrifugal Mixed flow Axial flow Nsd 1,250 2,200 6,200 13,500 Gal/min 2,400 2,400 2,400 2,400 Head, ft 70 48 33 20 Rpm 870 1,160 1,750 2,600 D, in. 19 12 10 7 Deye /D 0.5 0.7 0.9 1.0 ■ Figure 12.21 Comparison of different types of impellers. Specific speed for centrifugal pumps based on single suction and identical flowrate. (Adapted from Ref. 13, used with permission.) 12.7 Fans 707 axial-flow pumps, each operating with the same flowrate. These data indicate that as we proceed from the centrifugal pump to the mixed-flow pump to the axial-flow pump, the specific speed increases, the head decreases, the speed increases, the impeller diameter decreases, and the eye diameter increases. The dimensionless parameters and scaling relationships developed in the previous sections apply to all three types of pumps—centrifugal, mixed-flow, and axial-flow—since the dimensional analysis used is not restricted to a particular type of pump. Additional information about pumps can be found in Refs. 4, 7, 9, 13, and 14. THE WIDE WORLD OF FLUIDS Mechanical heart assist devices As with any pump, the function of the heart’s left ventricle. (The left ventricle supplies human heart can suffer various malfunctions and problems dur- oxygenated blood to the entire body and performs about 80% of ing its useful life. Recent developments in artificial heart technol- the heart’s work.) Some LVADs are centrifugal or axial-flow ogy may be able to provide help to those whose pumps have pumps that provide a continuous flow of blood. The continuous- broken down beyond repair. One of the more promising tech- flow devices may take some adjustment on the part of patients niques is use of a left-ventricular assist device (LVAD), which who do not hear a pulse or a heartbeat. Despite advances in supplements a diseased heart. Rather than replacing a diseased artificial heart technology, it is probably still several years before heart, an LVAD pump is implanted alongside the heart and works fully implantable, quiet, and reliable devices will be considered in parallel with the cardiovascular system to assist the pumping for widespread use. 12.7 Fans are used to pump air and other gases and vapors. Fans When the fluid to be moved is air, or some other gas or vapor, fans are commonly used. Types of fans vary from the small fan used for cooling desktop computers to large fans used in many industrial applications such as ventilating of large buildings and providing mechanical draft for fossil fuel-fired steam generators in power stations. Fans typically operate at relatively low rotation speeds (a few hundred to a few thousand rpm) and are capable of moving large volumes of gas (more than a million cfm). Although the fluid of interest is a gas, the change in gas density through the fan does not usually exceed 7%, which for air represents a change in pressure of only about 1 psi (Ref. 15). Thus, in dealing with fans, the gas density is treated as a constant, and the flow analysis is based on incompressible flow concepts. Because of the low-pressure rise involved, fans are often constructed of lightweight sheet metal. Fans are also called blowers, boosters, and exhausters, depending on the location within the system; that is, blowers are located at the system entrance, exhausters are at the system exit, and boosters are at some intermediate position within the system. Turbomachines used to produce larger changes in gas density and pressure than are possible with fans are called compressors (see Section 12.9.1). As is the case for pumps, fan designs include centrifugal (radial-flow) fans, as well as mixedflow and axial-flow (propeller) fans, and the analysis of fan performance closely follows that previously described for pumps. The shapes of typical performance curves for centrifugal and axial-flow fans are quite similar to those shown in Fig. 12.20 for centrifugal and axial-flow pumps. However, fan performance data are often given in terms of pressure rise, either static or total, rather than the more conventional head rise commonly used for pumps. Fan static pressure rise, the most commonly used measure of fan performance, is a hybrid quantity, being the difference between outlet static pressure and inlet total pressure. Scaling relationships for fans are the same as those developed for pumps, that is, Eqs. 12.32 through 12.35 apply to fans as well as pumps. As noted above, for fans it is common to replace the head, ha, in Eq. 12.33 with pressure head, pa ρg, so that Eq. 12.33 becomes pa pa ( ρω2D2 )1 = ( ρω2D2 )2 (12.47) 708 Chapter 12 ■ Turbomachines where pa is either fan total pressure rise or fan static pressure rise and, as before, the subscripts 1 and 2 refer to any two fans from the family of geometrically similar fans. Specific speed for fans is defined by Eq. 12.43, with gha replaced by paρ. As for pumps, U.S. industry uses a dimensional specific speed Nsdf where Nsdf = ω(rpm)√Q(cfm) [pa (in.H2O)]3/4 12.47A and pa is relative to standard air density. Centrifugal fans have specific speed in the range 2000 < Nsdf < 20,000 and axial fans lie in the range 50,000 < Nsdf < 200,000. Additional information about fans can be found in Refs. 15–18. Equations 12.47, 12.32, and 12.34 are called the fan laws and can be used to scale performance characteristics between members of a family of geometrically similar fans. THE WIDE WORLD OF FLUIDS Hi-tech ceiling fans Energy savings of up to 25% can be real- By making the fan blade more propeller-like, it is possible to ized if thermostats in air-conditioned homes are raised by a have a more uniform, efficient distribution. However, since ceil- few degrees. This can be accomplished by using ceiling fans ing fans are restricted by law to operate at less than 200 rpm, and taking advantage of the increased sensible cooling ordinary airplane propeller design is not appropriate. After brought on by air moving over skin. If the energy used to run considerable design effort, a highly efficient ceiling fan capa- the fans can be reduced, additional energy savings can be ble of delivering the same airflow as the conventional design realized. Most ceiling fans use flat, fixed pitch, nonaerody- with only half the power has been successfully developed and namic blades with uniform chord length. Because the tip of a marketed. The fan blades are based on the slowly turning prop paddle moves through air faster than its root does, airflow over used in the Gossamer Albatross, the human-powered aircraft such fan blades is lowest near the hub and highest at the tip. that flew across the English Channel in 1979. 12.8 The two basic types of hydraulic turbines are impulse and reaction. Turbines As discussed in Section 12.2, turbines are devices that extract energy from a flowing fluid. The geometry of turbines is such that the fluid exerts a torque on the rotor in the direction of its rotation. The shaft power generated is available to drive generators or other devices. In the following sections we discuss mainly the operation of hydraulic turbines (those for which the working fluid is water) and to a lesser extent gas and steam turbines (those for which the density of the working fluid may be much different at the inlet than at the outlet). No matter what the working fluid, turbines can be broadly classified into two types: impulse turbines and reaction turbines. These classifications are based on the mechanism of the fluid interaction with the turbine blades. In an impulse turbine, the force on the blades is produced solely by turning the fluid, without appreciable pressure drop in the blade passage, with all of the pressure drop occurring in a fixed nozzle. The force on the vane in Examples 5.10 and 5.17 is solely from impulse. In a reaction turbine, some of the fluid-vane force is from fluid turning and some of the force is a reaction to acceleration of the fluid relative to the vane. In reaction blading, a pressure drop occurs in both a fixed nozzle and the moving vane. Turbine blading is characterized by the degree of reaction (R), which is the ratio of the drop in static pressure (or enthalpy) across the moving blade to the overall drop in static pressure (or enthalpy) across the fixed nozzle plus the moving blade. Impulse turbines have R = 0 while reaction turbines typically have 0.1 < R < 0.7. The Pelton wheel shown in Fig. 12.22 is a classical example of an impulse turbine. In these machines the total head of the incoming fluid (the sum of the pressure head, velocity head, and elevation head) is converted into a large velocity head at the exit of the supply nozzle (or nozzles if a multiple nozzle configuration is used). Both the pressure drop across the bucket (blade) and the change in relative speed of the fluid across the bucket are negligible. The space surrounding the rotor is not completely filled with fluid. It is the impulse of the individual jets of fluid striking the buckets that generates the torque. 12.8 Turbines 709 Rotor Nozzle Bucket (a) (b) ■ Figure 12.22 (a) Schematic diagram of a Pelton wheel turbine, (b) photograph of a Pelton wheel turbine. (Courtesy of Voith Hydro, York, PA.) For reaction turbines, on the other hand, the rotor is surrounded by a casing (or volute), which is completely filled with the working fluid. There is both a pressure drop and a fluid relative speed change across the rotor. As shown for the radial-inflow turbine in Fig 12.23, guide vanes act as nozzles to accelerate the flow and turn it in the appropriate direction as the fluid enters the rotor. Thus, part of the pressure drop occurs across the guide vanes and part occurs across the rotor. In many respects the operation of a reaction turbine is similar to that of a pump “flowing backward,” although such oversimplification can be quite misleading. Both impulse and reaction turbines can be analyzed using the moment-of-momentum principles discussed in Section 12.3. In general, impulse turbines are high-head, low-flowrate devices, while reaction turbines are low-head, high-flowrate devices. 12.8.1 Impulse Turbines Although there are various types of impulse turbine designs, perhaps the easiest to understand is the Pelton wheel (see Fig. 12.24). Lester Pelton (1829–1908), an American mining engineer during the California gold mining days, is responsible for many of the still-used features of this type of turbine. The Pelton wheel is essentially an improved water wheel. In one way, the Pelton wheel is not typical of other turbomachines; it is neither axial flow nor radial flow; it is “tangential flow.” A Pelton wheel is most efficient when operated with a large head (for example, a water source from a lake located significantly above the turbine nozzle), which is converted into a relatively large velocity at the exit of the nozzle. Among the many design considerations for such a turbine are the head loss 710 Chapter 12 ■ Turbomachines Adjustable guide vanes Rotor vanes Casing Draft tube Tail race (a) (b) ■ Figure 12.23 (a) Schematic diagram of a reaction turbine, (b) photograph of a reaction turbine. (Courtesy of Voith Hydro, York, PA.) Pelton wheel turbines operate most efficiently with a larger head and lower flowrates. that occurs in the pipe (the penstock) transporting the water to the turbine, the design of the nozzle, and the design of the buckets on the rotor. As shown in Fig. 12.24, a high-speed jet of water strikes the Pelton wheel buckets and is deflected. The water enters and leaves the control volume surrounding the wheel as free jets (at atmospheric pressure). In addition, a person riding on the bucket would note that the speed of the b rm a b V1 ■ Figure 12.24 Details of Pelton wheel a turbine. 12.8 711 Turbines ω Tangential rm Radial V1 a U b V2 VIDEO V12.4 Pelton wheel lawn sprinkler ■ Figure 12.25 Ideal fluid velocities for a Pelton wheel turbine. water does not change as it slides across the buckets (assuming viscous effects are negligible). That is, the magnitude of the relative velocity does not change, but its direction does. The change in direction of the velocity of the fluid jet causes a torque on the rotor, resulting in a power output from the turbine. Design of the optimum shape of the buckets to obtain maximum power is a very difficult matter. Ideally, the fluid enters and leaves the control volume shown in Fig. 12.25 with no radial component of velocity. The buckets would ideally turn the relative velocity vector through 180°, but physical constraints dictate that β, the angle of the exit edge of the blade, is less than 180°. Thus, the fluid leaves with an axial component of velocity as shown in Fig. 12.26. The inlet and exit velocity diagrams at the arithmetic mean radius, rm, are assumed to be as shown in Fig. 12.27. To calculate the torque and power, we must know the tangential components of the absolute velocities at the inlet and exit. From Fig. 12.27 we see that Vθ1 = V1 = W1 + U a (12.48) b Blade cross section W1 = V1 – U Tangential β a Axial W2 = W1 = V1 – U b a ■ Figure 12.26 Flow as viewed by an observer riding on the Pelton wheel—relative velocities. b U W1 W1 = W2 V2 V1 U β a b ■ Figure 12.27 Inlet and exit velocity diagrams for a Pelton wheel turbine. 712 Chapter 12 ■ Turbomachines ⋅ V (1 – cos β ) = mr m 1 |Tshaft| max ⋅ 2 |Wshaft | = 0.25m⋅ V 1 (1– cos β ) max Actual power ⋅ Wshaft Tshaft ⋅ –Wshaft –Tshaft Actual torque Umax power = 0.5V1 0 0.2 V1 0.4 V1 0.6 V1 0.8 V1 1.0 V1 U = ω rm ■ Figure 12.28 Typical theoretical and experimental power and torque for a Pelton wheel turbine as a function of bucket speed. and Vθ2 = W2 cos β + U In Pelton wheel analyses, we assume the relative speed of the fluid is constant (no friction). (12.49) Thus, with the assumption that W1 = W2 , we can combine Eqs. 12.48 and 12.49 to obtain Vθ2 − Vθ1 = (U − V1 )(1 − cos β) (12.50) This change in tangential component of velocity combined with Eqs. 12.2 and 12.4 gives Tshaft = m ˙ rm (U − V1 )(1 − cos β) . where m = ρQ is the mass flowrate through the turbine. Since U = ωRm, it follows that Ẇ shaft = Tshaft ω = m ˙ U(U − V1 )(1 − cos β) (12.51) These results are plotted in Fig. 12.28 along with typical experimental results. Note that V1 > U (i.e., the jet impacts the bucket), and Ẇ shaft < 0 (i.e., the turbine extracts power from the fluid). Several interesting points can be noted from these results. First, the power is a function of β. However, a typical value of β = 165° (rather than the optimum 180°) results in a relatively small (less than 2%) reduction in power since 1 − cos 165° = 1.966, compared to 1 − cos 180° = 2. Second, although the torque is maximum when the wheel is stopped (U = 0), there is no power under this condition—to extract power one needs both force and motion. Third, the power output is a maximum when U ∣ max power = V1 2 (12.52) This can be shown by using Eq. 12.51 and solving for U that gives dẆ shaft dU = 0. A bucket speed one-half the speed of the fluid coming from the nozzle gives maximum power. Fourth, the maximum speed occurs when Tshaft = 0 (i.e., the load is completely removed from the turbine, as would happen if the shaft connecting the turbine to the generator were to break and frictional torques were negligible). For this case U = ωR = V1, the turbine is “freewheeling,” and the water simply passes across the rotor without any force on the buckets. 12.8 E X AMPLE 12.6 Turbines 713 Pelton Wheel Turbine Characteristics GIVEN Water to drive a Pelton wheel is supplied through a (0) z0 = 200 ft pipe from a lake as indicated in Fig. E12.6a. The head loss due to friction in the pipe is important, but minor losses can be neglected. ℓ = 1000 ft, f = 0.02 β = 150 deg D = 8 in FIND (a) Determine the nozzle diameter, D1, that will give the maximum power output. 2R = 3 ft (1) (b) Determine the maximum power and the angular velocity of the rotor at the conditions found in part (a). z1 = 0 D1 ■ Figure E12.6a SOLUTION (a) As indicated by Eq. 12.51, the power output depends on the flowrate, Q = m ˙ /ρ, and the jet speed at the nozzle exit, V1, both of which depend on the diameter of the nozzle, D1, and the head loss associated with the supply pipe. That is Ẇ shaft = ρQU(U − V1 ) (1 − cos β) (1) The nozzle exit speed, V1, can be obtained by applying the energy equation (Eq. 5.85) between a point on the lake surface (where V0 = p0 = 0) and the nozzle outlet (where z1 = p1 = 0) to give z0 = V 21 + hL 2g Ẇ shaft = √1 + 152 ℓ V2 D 2g A1V1 D1 2 = ( ) V1 A D dẆ shaft 2 D1 = −1.04 × 106[ dD1 (1 + 152 D41 ) 3/2 4(152) D51 3 −( ) =0 2 (1 + 152 D41 ) 5/2 ] or 304 D41 = 1 35,000 (3) 25,000 −Wshaft, ft·lb/s 1/2 2(32.2 ft /s2 ) (200 ft) D1 4 1000 ft [ 1 + 0.02( 8/12 ft )( 8/12 ) ] 113.5 √1 + 152 where D1 is in feet. D 41 D1 = 0.239 ft 30,000 1/2 2gz0 V1 = 4 ℓ D1 [1 + f D ( D ) ] = √1 + 152 D41 ] (5) (6) The maximum power possible occurs when dẆ shaft dD1 = 0, which according to Eq. 6 can be found as or = 113.5 1.04 × 106 D21 Ẇ shaft = − (1 + 152 D41 ) 3/2 We have neglected minor losses associated with the pipe entrance and the nozzle. With the given data, Eq. 2 becomes ℓ D1 4 V 12 z0 = 1 + f [ D ( D ) ]2g U− D41 [ where U is in feet per second and Ẇ shaft is in ft ∙ lb/s. These results are plotted as a function of U for various values of D1 in Fig. E12.6b. As shown by Eq. 12.52, the maximum power (in terms of its variation with U) occurs when U=V12, which, when used with Eqs. 4 and 5, gives The speed, V, of the fluid in the pipe of diameter D is obtained from the continuity equation V= 323 UD21 (2) where the head loss is given in terms of the friction factor, f, as (see Eq. 8.34) hL = f By combining Eqs. 1 and 4 and using Q = πD21V14 we obtain the power as a function of D1 and U as 20,000 D1 = 0.200 ft 15,000 10,000 D1 = 0.300 ft 5,000 0 0 20 (4) 40 60 U, ft/s (b) ■ Figure E12.6b 80 100 714 Chapter 12 ■ Turbomachines COMMENT The reason that an optimum diameter nozzle Thus, the nozzle diameter for maximum power output is D1 = 0.239 ft (Ans) (b) The corresponding maximum power can be determined from Eq. 6 as 1.04 × 10 (0.239) = −3.25 × 104 ft ∙ lb/s Ẇ shaft = − [1 + 152(0.239) 4]3/2 6 2 or Ẇ shaft = −3.25 × 104 ft ∙ lb/s × = −59.0 hp 1 hp 550 ft ∙ lb/s (Ans) The rotor speed at the maximum power condition can be obtained from V1 U = ωR = 2 where V1 is given by Eq. 4. Thus, exists can be explained as follows. A larger-diameter nozzle will allow a larger flowrate but will produce a smaller jet velocity because of the head loss within the supply side. A smallerdiameter nozzle will reduce the flowrate but will produce a larger jet velocity. Since the power depends on a product combination of flowrate and jet velocity (see Eq. 1), there is an optimumdiameter nozzle that gives the maximum power. These results can be generalized (i.e., without regard to the specific parameter values of this problem) by considering Eqs. 1 and 3 and the condition that U = V12 to obtain π Ẇ shaft ∣U=V1/2 = − ρ(1 − cos β) 16 3/2 ℓ × (2gz0 ) 3/2 D21 1 + f 5 D41 /( ) D By setting dẆ shaft dD1 = 0, it can be shown that the maximum power occurs when D1 = D/(2f 113.5 ft/s √1 + 152(0.239) 4 V1 = ω= 2R 3 2( ft) 2 = 30.9 rad/s × 1 rev/2π rad × 60 s/min = 295 rpm Dentist drill turbines are usually of the impulse class. (Ans) ℓ 1/4 D) which gives the same results obtained earlier for the specific parameters of the example problem. Note that the optimum condition depends only on the friction factor and the length-todiameter ratio of the supply pipe. What happens if the supply pipe is frictionless or of essentially zero length? A second type of impulse turbine that is widely used (most often with air as the working fluid) is indicated in Fig. 12.29. A circumferential series of fluid jets strikes the rotating blades which, as with the Pelton wheel, alter both the direction and magnitude of the absolute velocity. As with the Pelton wheel, the inlet and exit pressures (i.e., on either side of the rotor) are equal, and the magnitude of the relative velocity is unchanged as the fluid slides across the blades (if frictional effects are negligible). Typical inlet and exit velocity diagrams are shown in Fig. 12.30. As discussed in Section 12.2, in order for the absolute velocity of the fluid to be changed as indicated during its passage across Control volume Nozzles Tshaft ω Fluid jets (a) (b) ■ Figure 12.29 (a) A multinozzle, non-Pelton wheel impulse turbine commonly used with air as the working fluid, (b) dental drill. 12.8 Turbines 715 U W2 U2 V2 W1 Section (2) Section (1) V1 U1 Vθ 1 ■ Figure 12.30 Inlet and exit velocity triangles for the impulse turbine shown in Fig. 12.29. VIDEO V12.5 Dental drill the blade, the blade must push on the fluid in the direction opposite of the blade motion. Hence, the fluid pushes on the blade in the direction of the blade’s motion—the fluid does work on the blade (a turbine). This turbine is an axial-flow device. E X AMPLE 12.7 Axial-Flow Impulse Turbine (Dental Drill) GIVEN An air turbine used to drive the high-speed drill used by your dentist is shown in Fig. 12.29 (also see Video V12.5). Air exiting from the upstream nozzle holes forces the turbine blades to move in the direction shown. The turbine rotor speed is 300,000 rpm, the tangential component of velocity out of the nozzle is twice the blade speed, and the tangential component of the absolute velocity out of the rotor is zero. The radius to the blade tip is 0.168 in., and the radius to the blade base is 0.133 in. FIND Estimate the shaft energy per unit mass of air flowing through the turbine. SOLUTION We use the fixed, nondeforming control volume that includes the turbine rotor and the fluid in the rotor blade passages at an instant of time (see Fig. E12.29a). The only torque acting on this control volume is the shaft torque. For simplicity we analyze this problem using an arithmetic mean radius, rm, where = (−310,000 ft2/s2 )(32.174(ft ∙ lbm)(lb ∙ s2 ) ) = −9640 ft ∙ lb/lbm (Ans) COMMENT For each lbm of air passing through the turbine A sketch of the velocity diagrams at the rotor entrance and exit is shown in Fig. 12.30. Application of Eq. 12.5 gives there is 9640 ft ∙ lb of energy available at the shaft to drive the drill. However, because of fluid friction, the actual amount of energy given up by each lbm of air will be greater than the amount available at the shaft. How much greater depends on the efficiency of the fluid-mechanical energy transfer between the fluid and the turbine blades. Recall that the shaft power, Ẇ shaft , is given by wshaft = −U1Vθ1 + U2Vθ2 ˙ wshaft Ẇ shaft = m 1 rm = (r0 + ri ) 2 (1) where wshaft is shaft energy per unit of mass flowing through the turbine. From the problem statement, Vθ1 = 2U and Vθ2 = 0, where U = ωrm = (300,000 rev/min) (1 min/60 s) (2π rad/rev) × (0.168 in. + 0.133 in.)2(12 in./ft) = 394 ft/s is the mean-radius blade velocity. Thus, Eq. (1) becomes wshaft = −U1Vθ1 = −2U 2 = −2(394 ft/s) 2 = −310,000 ft2/s2 (2) Hence, to determine the power, we need to know the mass flowrate, m ˙ , which depends on the size and number of the nozzles. Although the energy per unit mass is large (i.e., 9640 ft ∙ lb/lbm), the flowrate is small, so the power is not “large.” Obviously, compressibility effects are likely at the speeds involved in this problem. Turbine wheels of similar design are often used as the first stage of power-generating steam turbines. Of course, both the size of the wheel and the magnitude of the power extracted are very much larger than those in this dental drill. 716 Chapter 12 ■ Turbomachines ω ω Rotor blades Rotor ω Adjustable guide vane Adjustable guide vanes Plan view of guide vanes ω Draft tube (a) (b) ■ Figure 12.31 (a) Typical radial-flow Francis turbine, (b) typical axial-flow Kaplan turbine. 12.8.2 Reaction Turbines Reaction turbines are best suited for higher flowrate and lower head situations. As indicated in the previous section, impulse turbines are best suited (i.e., most efficient) for lower-flowrate and higher-head operations. Reaction turbines, on the other hand, are best suited for higher-flowrate and lower-head situations. In a reaction turbine the working fluid completely fills the passageways through which it flows. The angular momentum, pressure, and velocity of the fluid decrease as it flows through the turbine rotor—the turbine rotor extracts energy from the fluid. As with pumps, turbines are manufactured in a variety of configurations—radial-flow, mixedflow, and axial-flow types. Typical radial- and mixed-flow hydraulic turbines are called Francis turbines, named after James B. Francis, an American engineer. At very low heads the most efficient type of turbine is the axial-flow or propeller turbine. The Kaplan turbine, named after Victor Kaplan, a German professor, is an efficient axial-flow hydraulic turbine with adjustable blades. Cross sections of these different turbine types are shown in Fig. 12.31. As shown in Fig. 12.31a, flow across the rotor blades of a radial-inflow turbine has a major component in the radial direction. Adjustable inlet guide vanes direct the water into the rotor with a tangential component of velocity. The absolute velocity of the water leaving the rotor is essentially without a tangential component. Hence, the rotor decreases the angular momentum of the fluid, the fluid exerts a torque on the rotor in the direction of rotation, and the rotor extracts energy from the fluid. The Euler turbomachine equation (Eq. 12.2) and the corresponding power equation (Eq. 12.4) are equally valid for this turbine as they are for the centrifugal pump discussed in Section 12.4. As shown in Fig. 12.31b, for an axial-flow Kaplan turbine, the fluid flows through the inlet guide vanes, called wicket gates, and achieves a tangential velocity in a vortex (swirl) motion before it reaches the rotor. Flow across the rotor contains a major axial component. Both the inlet guide vanes and the turbine blades can be adjusted by changing their setting angles to produce the best match (optimum output) for the specific operating conditions. For example, the operating head available may change from season to season and/or the flowrate through the rotor may vary. 12.8 Actual head available for a turbine, ha, is always greater than shaft work head, hs, because of head loss, hL, in the turbine. Turbines 717 Pumps and turbines are often thought of as the “inverse” of each other. Pumps add energy to the fluid; turbines remove energy. The propeller on an outboard motor (a pump) and the propeller on a Kaplan turbine are in some ways geometrically similar, but they perform opposite tasks. Similar comparisons can be made for centrifugal pumps and Francis turbines. In fact, some large pump/turbines at “pumped storage” hydroelectric power plants are designed to be run as turbines during high-power demand periods (i.e., during the day) and as pumps to resupply the upstream reservoir from the downstream reservoir during low-demand times (i.e., at night). Thus, a pump type often has its corresponding turbine type. However, is it possible to have the “inverse” of a Pelton wheel turbine—an impulse pump? As with pumps, incompressible flow turbine performance is often specified in terms of appropriate dimensionless parameters. The flow coefficient, CQ = QωD3, the head coefficient, CH = gha ω2D2, and the power coefficient, C𝒫 = Ẇ shaft ρω3D5, are defined in the same way for pumps and turbines. On the other hand, turbine efficiency, η, is the inverse of pump efficiency. That is, the efficiency is the ratio of the shaft power output to the power available in the flowing fluid, or Ẇ shaft η= ρgQha For geometrically similar turbines and for negligible Reynolds number and surface roughness effects, the relationships between the dimensionless parameters are given functionally by those shown in Eqs. 12.29, 12.30, and 12.31. That is, CH = ϕ1 (CQ ), C𝒫 = ϕ2 (CQ ), and η = ϕ3 (CQ ) where the functions ϕ1, ϕ2, and ϕ3 are dependent on the type of turbine involved. Also, for turbines the efficiency, η, is related to the other coefficients according to η = C𝒫CHCQ. As indicated above, the design engineer has a variety of turbine types available for any given application. It is necessary to determine which type of turbine would best fit the job (i.e., be most efficient) before detailed design work is attempted. As with pumps, the use of a specific speed parameter can help provide this information. For hydraulic turbines, the rotor diameter D is eliminated between the flow coefficient and the power coefficient to obtain the power specific speed ω√Ẇ shaft ρ N′s = . (gha ) 5/4 We use the more common, but not dimensionless, definition of power specific speed N′sd = Specific speed may be used to approximate what kind of turbine geometry (axial to radial) would operate most efficiently. ω(rpm)√Ẇ shaft (bhp) [ha (ft)]5/4 (12.53) That is, N′sd is calculated with angular velocity, ω, in rpm; shaft power, Ẇ shaft, in brake horsepower, and actual head available, ha, in feet. Note that the flowrate, Q, does not appear in N′sd; it is a dependent variable for turbines. Optimum turbine efficiency (for large turbines) as a function of specific speed is indicated in Fig. 12.32. Also shown are representative rotor and casing cross sections. Note that impulse turbines are best at low specific speeds; that is, when operating with large heads and small flowrate. The other extreme is axial-flow turbines, which are the most efficient type if the head is low and if the flowrate is large. For intermediate values of specific speeds, radial- and mixed-flow turbines offer the best performance. The data shown in Fig. 12.32 are meant only to provide a guide for turbine-type selection. The actual turbine efficiency for a given turbine depends very strongly on the detailed design of the turbine. Considerable analysis, testing, and experience are needed to produce an efficient turbine. However, the data of Fig. 12.32 are representative. Much additional information can be found in the literature. 718 Chapter 12 ■ Turbomachines Impulse turbines Reaction turbines Radial-flow 10 Mixed-flow 20 Axial-flow 40 60 80 100 ′ Nsd 100 Kaplan Francis Impulse η% 90 80 70 10 20 40 60 ■ Figure 12.32 Typical turbine cross sections and maximum efficiencies as a function of specific speed. 80 100 ′ Nsd THE WIDE WORLD OF FLUIDS Cavitation damage in hydraulic turbines The occurrence or cavitation. Further along the flow path, the fluid can decel- of cavitation in hydraulic pumps seems to be an obvious pos- erate rapidly enough with accompanying increase in local sibility since low-suction pressures are expected. Cavitation pressure to make cavitation bubbles collapse with enough damage can also occur in hydraulic turbines, even though they intensity to cause blade surface damage or damage to the do not seem obviously prone to this kind of problem. Local draft tube in the form of material erosion. Over time, this ero- acceleration of liquid over blade surfaces can be sufficient to sion can be severe enough to require repair or replacement, result in local pressures low enough to cause fluid vaporization which is very expensive. EXAMPLE 12.8 Use of Specific Speed to Select Turbine Type GIVEN A hydraulic turbine is to operate at an angular velocity FIND What type of turbine should be selected? Explain. of 6 rev/s, a flowrate of 10 ft3/s, and a head of 20 ft. SOLUTION The most efficient type of turbine to use can be obtained by calculating the specific speed, N′sd , and using the information of Fig. 12.32. To use the dimensional form of the specific speed indicated in Fig. 12.32, we must convert the given data into the appropriate units. For the rotor speed we get ω = 6 rev/s × 60 s/min = 360 rpm To estimate the shaft power, we assume all of the available head is converted into power and multiply this amount by an assumed efficiency (94%). Ẇ shaft = γQzη = (62.4 lb/ft3 ) (10 ft3/s) [ Ẇ shaft = 21.3 hp 20 ft(0.94) 550 ft ˙ lb/s ˙ hp ] Thus for this turbine, N′sd = ω√Ẇ shaft (ha ) 5/4 = (360 rpm) √21.3 hp (20 ft) 5/4 = 39.3 According to the information of Fig. 12.32, A mixed-flow Francis turbine would probably give the highest efficiency and an assumed efficiency of 0.94 is appropriate. (Ans) COMMENT What would happen if we wished to use a Pelton wheel for this application? Note that with only a 20-ft head, the maximum jet velocity, V1, obtainable (neglecting viscous effects) would be V1 = √2 gz = √2 × 32.2 ft/s2 × 20 ft = 35.9 ft/s 12.9 As shown by Eq. 12.52, for maximum efficiency of a Pelton wheel the jet velocity is ideally two times the blade velocity. Thus, V1 = 2ωR, or the wheel diameter, D = 2R, is D= V1 35.9 ft/s = 0.952 ft = ω (6 rev/s × 2π rad/rev) To obtain a flowrate of Q = 10 ft3/s at a velocity of V1 = 35.9 ft/s, the jet diameter, d1, must be given by Q= π 2 d 1 V1 4 or d1 = [ 4Q 1/2 4(10 ft3/s) 1/2 = [ π(35.9 ft/s) ] = 0.596 ft πV1 ] Compressible Flow Turbomachines 719 tical design. Typically d1 << D (see Fig. 12.22). By using multiple jets it would be possible to reduce the jet diameter. However, even with 8 jets, the jet diameter would be 0.211 ft, which is still too large (relative to the wheel diameter) to be practical. Hence, the above calculations reinforce the results presented in Fig. 12.32—a Pelton wheel would not be practical for this application. If the flowrate were considerably smaller, the specific speed could be reduced to the range where a Pelton wheel would be the type to use (rather than a mixed-flow reaction turbine). It might seem that turbine speed is selected arbitrarily but it is not. Since most large hydraulic turbines drive electric generators, the rotational speed is determined by the number of magnetic poles in the generator and the frequency desired (60 Hz in North America and 50 Hz in Europe). A Pelton wheel with a diameter of D = 0.952 ft supplied with water through a nozzle of diameter d1 = 0.596 ft is not a prac- 12.9 Compressible Flow Turbomachines Compressible flow turbomachines are in many ways similar to the incompressible flow pumps and turbines described in previous portions of this chapter. The main difference is that the density of the fluid (a gas or vapor) changes significantly from the inlet to the outlet of compressible flow machines. This added feature sometimes has interesting consequences, benefits, and complications. Compressors are like pumps, adding energy to the fluid, causing a significant pressure rise and a corresponding significant increase in density and temperature. Compressible flow turbines, on the other hand, remove energy from the fluid, causing a lower pressure and a smaller density and temperature at the outlet than at the inlet. The information provided earlier about basic energy considerations (Section 12.2) and basic angular momentum considerations (Section 12.3) is directly applicable to these turbomachines. As discussed in Chapter 11, compressible flow study requires an understanding of the principles of thermodynamics. Similarly, an in-depth analysis of compressible flow turbomachines requires use of various thermodynamic concepts. For example, the fluid temperature change is of little consequence in liquid pumps and turbines, but it is of critical importance in compressible flow machines. In this section we provide only a brief discussion of some of the general properties of compressors and compressible flow turbines. The interested reader is encouraged to read some of the excellent references available for further information (e.g., Refs. 1–3, 19–21). 12.9.1 Compressors Turbocompressors operate with the continuous compression of gas flowing through the device. Since there is a significant pressure and density increase, there is also a considerable temperature increase. Radial-flow (or centrifugal) compressors are essentially centrifugal pumps (see Section 12.4) that use a gas (rather than a liquid) as the working fluid. They are typically high pressure rise, low flowrate, and are axially compact turbomachines. A photograph of a centrifugal compressor rotor is shown in Fig. 12.33. The amount of compression is typically given in terms of the total pressure ratio, PR = p02 p01, where the pressures are absolute. Thus, a radial-flow compressor with PR = 3.0 can compress standard atmospheric air from 14.7 psia to 3.0 × 14.7 = 44.1 psia, with a corresponding temperature rise of about 220 °F. Higher pressure ratios can be obtained by using multiple-stage devices in which flow from the outlet of the preceding stage proceeds to the inlet of the following stage. If each stage has the same pressure ratio, PR, the overall pressure ratio after n stages is PRn. Thus, as shown by the figure 720 Chapter 12 ■ Turbomachines ■ Figure 12.33 Centrifugal compressor rotor. (Photograph courtesy of Concepts NREC.) Multistaging is common in highpressure ratio compressors. P0 out, psia 500 PR = 2 400 P0 in = 14.7 psia 300 200 100 0 0 1 2 3 4 5 Stage, n in the margin, a four-stage compressor with individual stage PR = 2.0 can compress standard air from p0 in = 14.7 psia to p0 out = 24 × 14.7 = 235 psia. Adiabatic (i.e., no heat transfer) compression of a gas causes an increase in temperature and requires more work than isothermal (constant temperature) compression of a gas. An interstage cooler (i.e., an intercooler heat exchanger) as shown in Fig. 12.34 can be used to reduce the compressed gas temperature entering the second stage and thus the work required. Relative to centrifugal water pumps, radial compressors of comparable size rotate at much higher speeds, typically on the order of tens of thousands of revolutions per minute. It is not uncommon for the rotor blade tip speed and the speed of the absolute flow leaving the impeller to be greater than the speed of sound. Such high speeds generate very large values of head, typically a few miles (of air) (see Eq. 12.14). The axial-flow compressor is the other widely used configuration. This type of turbomachine has a lower pressure rise per stage and a higher flowrate, and is more radially compact than a centrifugal compressor. As shown in Fig. 12.35, axial-flow compressors usually consist of several stages, with each stage containing a rotor/stator row pair. For an 11-stage compressor, a pressure ratio of PR = 1.2 per stage gives an overall pressure ratio of p02 p01 = 1.211 = 7.4. As Intercooler Cooling coils Stagnation enthalpy or temperature Inlet to stage 1 Velocity Pressure Outlet of stage 2 Stator Rotor Blade Rotor Shaft Stage 1 Stage 2 Shaft ■ Figure 12.34 Two-stage centrifugal ■ Figure 12.35 Enthalpy, velocity, and pressure distribution compressor with an intercooler. in an axial-flow compressor. ω 12.9 Compressible Flow Turbomachines 721 Compressor stages Turbine stages ■ Figure 12.36 Rolls-Royce Trent 900 three-shaft propulsion system. (Courtesy of Rolls-Royce plc.) Axial-flow compressor multistaging requires less frontal area but more length than centrifugal compressors. VIDEO V12.6 Flow in a compressor stage Compressor blades can stall, and unstable flow conditions can subsequently occur. the gas is compressed and its density increases, a smaller annulus cross-sectional area is required and the flow channel size and blade height decrease from the inlet to the outlet of the compressor. The typical jet aircraft engine uses an axial-flow compressor as one of its main components (see Fig. 12.36 and Ref. 22). An axial-flow compressor can include a set of inlet guide vanes upstream of the first rotor row. These guide vanes optimize the relative velocity into the first rotor row by directing the flow away from the axial direction. Rotor blades push on the gas in the direction of blade motion and to the rear, adding energy (like in an axial pump) and moving the gas through the compressor. The stator blade rows act as diffusers, turning the fluid back toward the axial direction and increasing the static pressure. The stator blades cannot add energy to the fluid because they are stationary. Typical pressure, velocity, and enthalpy distributions along the axial direction are shown in Fig. 12.35. [If you are not familiar with the thermodynamic concept of enthalpy (see Section 11.1), you may replace “enthalpy” by temperature as a qualitative approximation.] The degree of reaction of the compressor stage is equal to the ratio of the rise in static enthalpy or temperature achieved across the rotor to the enthalpy or temperature rise across the entire stage. Most modern compressors involve 50% or higher reaction. The blades in an axial-flow compressor are airfoils carefully designed to produce appropriate lift and drag forces on the flowing gas. As occurs with airplane wings, compressor blades can stall (see Section 9.4). When the flowrate is decreased from the design amount, the velocity diagram at the entrance of the rotor row indicates that the relative flow meets the blade leading edge at larger angles of incidence than the design value. When the angle of incidence becomes too large, blade stall can occur and the result is compressor surge or stall—unstable flow conditions that can cause excessive vibration, noise, poor performance, and possible damage to the machine. The lower flowrate bound of compressor operation is related to the beginning of these instabilities (see Fig. 12.37). Other important compressible flow phenomena such as, shock waves (see Section 11.5.3) and choked flow (see Section 11.4.2) occur commonly in compressible flow turbomachines. These phenomena are very sensitive to even very small changes or variations of geometry. Shock strength is kept low to minimize shock loss, and choked flows limit the upper flowrate boundary of machine operation. See Fig. 12.37, where choking is indicated by the vertical portion of the pressure ratio and efficiency lines. Chapter 12 ■ Turbomachines 1.00 9 .80 8 .60 .40 4 8 12 m⋅ 16 std A 20 24 28 , (lbm/s)/ft 2 Overall total-pressure ratio, p02 / p01 Isentropic efficiency 722 7 Design point 6 110 Stall-limit line 5 100 90 4 3 70 2 1 Corrected 80 speed as percent of design value 60 50 4 8 12 m⋅ 16 std A 20 24 28 , (lbm/ s) / ft 2 ■ Figure 12.37 Performance characteristics of an axial-flow compressor (Ref. 20). The experimental performance data for compressors are systematically summarized with parameters prompted by dimensional analysis. As mentioned earlier, total pressure ratio, p02 p01, is used instead of the head-rise coefficient associated with pumps, blowers, and fans. Two different types of efficiency, isentropic efficiency and polytropic efficiency, are used to characterize compressor performance. Each of these compressor efficiencies involves a ratio of ideal work to actual work required to accomplish the compression. The isentropic efficiency involves a ratio of the ideal work required with an adiabatic and frictionless (no loss) compression process to the actual work required to achieve the same total pressure rise. The polytropic efficiency involves a ratio of the ideal work required to achieve the actual end state of the compression with a polytropic and frictionless process between the actual beginning and end stagnation states across the compressor and the actual work involved between these same states. A more detailed explanation of these efficiencies is beyond the scope of this text. Those interested in learning more about these parameters should study any of several available books on turbomachines (for example, Refs. 2 and 3). The flow parameter commonly used for compressors is based on the following dimensionless grouping from dimensional analysis ˙ √kRT01 m D2 p01 where R is the gas constant, m ˙ the mass flowrate, k the specific heat ratio, T01 the stagnation temperature at the compressor inlet, D a characteristic length, typically rotor diameter, and p01 the stagnation pressure at the compressor inlet. To account for variations in test conditions, the following strategy is employed. We set ˙ √kRT01 ˙ √kRT01 m m ( D2 p01 )test = ( D2 p01 )std where the subscript “test” refers to a specific test condition and “std” refers to the standard atmosphere ( p0 = 14.7 psia, T0 = 518.7 °R) condition. When we consider a given compressor operating on a given working fluid (so that R, k, and D are constant), the above equation reduces to m ˙ std = ˙ test √T01 testT0 std m p01 test p0 std (12.54) 12.9 Compressible Flow Turbomachines 723 In essence, m ˙ std is the compressor-test mass flowrate “corrected” to the standard atmosphere inlet condition. The corrected compressor mass flowrate, m ˙ std, is used instead of flow coefficient. Often, m ˙ std is divided by A, the frontal area of the compressor flow path or by D2. While for pumps, blowers, and fans, rotor speed was accounted for in the flow coefficient, it does not appear in the corrected mass flowrate. Thus, for compressors, rotor speed needs to be accounted for with an additional group. This dimensionless group is ND √kRT01 which is called the machine Mach number. For the same compressor operating on the same gas, we eliminate D, k, and R and, as with corrected mass flowrate, obtain a corrected speed, Nstd, where Nstd = N √T01 Tstd (12.55) Often, the percentage of the corrected speed design value is used. An example of how compressor performance data are typically summarized is shown in Fig. 12.37. 12.9.2 Compressible Flow Turbines A gas turbine engine generally consists of a compressor, a combustor, and a turbine. Turbines that use a gas or vapor as the working fluid are in many respects similar to hydraulic turbines (see Section 12.8). Compressible flow turbines may be impulse or reaction turbines, and mixed-, radial-, or axial-flow turbines. The fact that the gas may expand (compressible flow) in coursing through the turbine can introduce some important phenomena that do not occur in hydraulic turbines. (Note: It is tempting to label turbines that use a gas as the working fluid as gas turbines. However, the terminology “gas turbine” is commonly used to denote a gas turbine engine, as employed, for example, for aircraft propulsion or stationary power generation. As shown in Fig. 12.36, these engines typically contain a compressor, combustion chamber, and turbine.) Although for compressible flow turbines the axial-flow type is common, the radial-inflow type is also used for various purposes. As shown in Fig. 12.33, the turbine that drives the typical automobile turbocharger compressor is a radial-inflow type. The main advantages of the radialinflow turbine are: (1) It is robust and durable, (2) it is axially compact, and (3) it can be relatively inexpensive. A radial-flow turbine usually has a lower efficiency than an axial-flow turbine, but lower initial costs may be the compelling incentive in choosing a radial-flow turbine over an axial-flow one. Axial-flow turbines are widely used compressible flow turbines. Steam turbines used in electrical generating plants and marine propulsion and the turbines used in gas turbine engines are usually of the axial-flow type. Often they are multistage turbomachines, although singlestage compressible flow turbines are also produced. They may be either an impulse type or a reaction type. With compressible flow turbines, the ratio of static enthalpy (or temperature) drop across the rotor to this drop across the stage, rather than the ratio of static pressure differences, is used to determine reaction. Strict impulse (zero pressure drop) turbines have slightly negative reaction; the static enthalpy or temperature actually increases across the rotor. Zero-reaction turbines involve no change of static enthalpy or temperature across the rotor but do involve a slight pressure drop. The early stages of steam and gas turbines are usually axial-flow impulse type, similar to Fig. 12.30 but considerably larger. A two-stage, axial-flow impulse turbine is shown in Fig. 12.38a. The gas accelerates through the supply nozzles, has some of its energy removed by the first-stage rotor blades, accelerates again through the second-stage nozzle row, and has additional energy removed by the second-stage rotor blades. As shown in Fig. 12.38b, the static pressure remains constant across the rotor rows. Across the second-stage nozzle row, the static pressure decreases, absolute velocity increases, and the stagnation enthalpy (temperature) is constant. Flow across the second rotor is similar to flow across the first rotor. Since the working 724 Chapter 12 ■ Turbomachines Rotor blades (Rotor) Moving blades Moving blades (Rotor) Nozzles (Stator) (a) Fixed blades (Stator) Nozzles (a) (b) Stagnation enthalpy or temperature Static enthalpy or temperature Static pressure Static pressure Absolute velocity Absolute velocity (b) Turbine performance maps are used to display complex turbine characteristics. (c) ■ Figure 12.38 Enthalpy, velocity, and pressure ■ Figure 12.39 Enthalpy, pressure, and distribution in a two-stage impulse turbine. velocity distribution in a three-stage reaction turbine. fluid is a gas, the significant decrease in static pressure across the turbine results in a significant decrease in density—the flow is compressible. Hence, more detailed analysis of this flow must incorporate various compressible flow concepts developed in Chapter 11. Interesting phenomena such as shock waves and choking due to sonic conditions at the “throat” of the flow passage between blades can occur because of compressibility effects. The interested reader is encouraged to consult the various references available (e.g., Refs. 2, 3, 21) for fascinating applications of compressible flow principles in turbines. The rotor and nozzle blades in a three-stage, axial-flow reaction turbine, such as would appear in the later stages of steam and gas turbines, are shown in Fig. 12.39a. The axial variations of pressure and velocity are shown in Fig. 12.39c. Both the stationary and rotor blade (passages) act as flow-accelerating nozzles. That is, the static pressure and enthalpy (temperature) decrease in the direction of flow for both the fixed and the rotating blade rows. This distinguishes the reaction turbine from the impulse turbine (see Fig. 12.38b). Energy is removed from the fluid by the rotors only (the stagnation enthalpy or temperature is constant across the adiabatic flow stators). Because of the reduction of static pressure in the downstream direction, the gas expands, and the flow passage area must increase from the inlet to the outlet of this turbine. This is seen in Fig. 12.39b. Performance data for compressible flow turbines are summarized with the help of parameters derived from dimensional analysis. Isentropic and polytropic efficiencies (see Refs. 2, 3, and 21) are commonly used, as are inlet-to-outlet total pressure ratios (p01 p02 ), corrected rotor speed (see Eq. 12.55), and corrected mass flowrate (see Eq. 12.54). Figure 12.40 shows a compressible flow turbine performance “map.” Note the occurrence of choking (the curves become flat). 12.10 Chapter Summary and Study Guide 725 23.0 50 22.5 49 22.0 Corrected mass flow rate, kg/s 48 Corrected mass flow rate, lbm/s 47 46 45 44 43 42 41 Corrected speed as percent of design speed 40 50 60 70 80 90 100 110 20.5 20.0 19.5 19.0 18.0 39 turbomachine axial-, mixed-, and radial-flow velocity diagram angular momentum shaft torque Euler turbomachine equation shaft power centrifugal pump pump performance curve overall efficiency system equation head-rise coefficient power coefficient flow coefficient pump scaling laws specific speed impulse turbine reaction turbine Pelton wheel fan compressor choking 21.0 18.5 40 12.10 21.5 17.5 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 Ratio of inlet –total to exit –total pressure 2.2 2.3 ■ Figure 12.40 Typical compressible flow turbine performance “map.” (Ref. 21) Chapter Summary and Study Guide Various aspects of turbomachine flow are considered in this chapter. The connection between fluid angular momentum change and shaft torque is key to understanding how turbo-pumps and turbines operate. The shaft torque associated with change in the axial component of angular momentum of a fluid as it flows through a pump or turbine is described in terms of the inlet and outlet velocity diagrams. Such diagrams indicate the relationship among absolute, relative, and blade velocities. Performance characteristics for centrifugal pumps are discussed. Standard dimensionless pump parameters, similarity laws, and the concept of specific speed are presented for use in pump analysis. How to use pump performance curves and the system curve for proper pump operation and selection is presented. A brief discussion of axial-flow and mixed-flow pumps is given. An analysis of impulse turbines is provided, with emphasis on the Pelton wheel turbine. For impulse turbines there is negligible pressure difference across the blade; the torque is a result of the change in direction of the fluid jet striking the blade. Radial-flow and axial-flow reaction turbines are also briefly discussed. Brief discussions of fans and compressible flow machines are presented. The following checklist provides a study guide for this chapter. When your study of the entire chapter and end-of-chapter exercises has been completed, you should be able to ■ ■ ■ ■ ■ ■ write out meanings of the terms listed here in the margin and explain each of the related concepts. These terms are particularly important and are set in italic, bold, and color type in the text. draw appropriate velocity diagrams for flows entering and leaving given pump or turbine configurations. estimate the actual shaft torque, actual shaft power, and ideal pump head rise for a given centrifugal pump configuration. use pump performance curves and the system curve to predict pump performance in a given system. predict the performance characteristics for one pump or fan based on the performance of another pump or fan of the same family using scaling laws. use specific speed to determine whether a radial-flow, mixed-flow, or axial-flow machine would be most appropriate for a given situation. 726 Chapter 12 ■ Turbomachines ■ ■ ■ ■ estimate the actual shaft torque and actual shaft power for flow through an impulse turbine. estimate the actual shaft torque and actual shaft power for a given reaction turbine. use specific speed to determine whether an impulse or a reaction turbine would be most appropriate for a given situation. read a performance map for a compressor or compressible flow turbine and point out operating limits. Some of the important equations in this chapter are: Vector addition of velocities V=W+U (12.1) Shaft power Tshaft = −m ˙ 1 (r1Vθ1 ) + m˙ 2 (r2Vθ2 ) Ẇ shaft = Tshaft ω Ẇ shaft = −m ˙ 1 (U1Vθ1 ) + m˙ 2 (U2Vθ2 ) Shaft work wshaft = U2Vθ2 − U1Vθ1 (12.12) Shaft torque Shaft power (12.2) (12.3) (12.4) Pump ideal head rise hi = U2Vθ2 − U1Vθ1 g (12.13) Pump actual head rise ha = p2 − p1 V 22 − V 21 + z2 − z1 + γ 2g (12.19) Pump similarity relationship Pump similarity relationship Pump similarity relationship Pump scaling law Pump scaling law Pump scaling law Pump scaling law gha Q = ϕ1( 3 ) 2 2 ωD ωD ̇ Wshaft Q = ϕ2( 3 ) 3 5 ωD ρω D η = ϕ3( Q ωD3 ) Q Q ( ωD3 )1 = ( ωD3 )2 gha gha ( ω2D2 )1 = ( ω2D2 )2 Ẇ shaft Ẇ shaft = ( ρω3D5 )1 ( ρω3D5 )2 η1 = η2 Specific speed (pumps) Nsd = Specific speed (turbines) N′sd = (12.29) (12.30) (12.31) (12.32) (12.33) (12.34) (12.35) ω(rpm) √Q(gpm) [ha (ft)]34 ω(rpm)√Ẇ shaft (bhp) [ha (ft)]54 (12.44) (12.53) References 1. Cumpsty, N. A., Jet Propulsion, 2nd Ed., Cambridge University Press, Cambridge, UK, 2003. 2. Saravanamuttoo, H. I. H., Rogers, G. F. C., and Cohen, H., Gas Turbine Theory, 5th Ed., PrenticeHall, Saddle River, N.J., 2001. 3. Wilson, D. G., and Korakianitis, T., The Design of High-Efficiency Turbomachinery and Gas Turbines, 2nd Ed., Prentice-Hall, Saddle River, N.J., 1998. 4. Stepanoff, H. J., Centrifugal and Axial Flow Pumps, 2nd Ed., Wiley, New York, 1957. 5. Wislicenus, G. F., Preliminary Design of Turbopumps and Related Machinery, NASA Reference Publication 1170, 1986. Problems 727 6. Neumann, B., The Interaction Between Geometry and Performance of a Centrifugal Pump, Mechanical Engineering Publications Limited, London, 1991. 7. Garay, P. N., Pump Application Desk Book, Fairmont Press, Lilburn, Ga., 1990. 8. Rouse, H., Elementary Mechanics of Fluids, Wiley, New York, 1946. 9. Kittredge, C. P., “Estimating the Efficiency of Prototype Pumps From Model Tests,” ASME Paper 67-WA-FE6, 1967. 10. Wright, T., and Gerhart, P. M., Fluid Machinery: Application, Selection, and Design, CRC-Taylor and Francis, Boca Raton FL, 2010. 11. Hydraulic Institute, Hydraulic Institute Standards, 14th Ed., Hydraulic Institute, Cleveland, Ohio, 1983. 12. Heald, C. C., ed., Cameron Hydraulic Data, 17th Ed., Ingersoll-Rand, Woodcliff Lake, N.J., 1988. 13. Kristal, F. A., and Annett, F. A., Pumps: Types, Selection, Installation, Operation, and Maintenance, McGraw-Hill, New York, 1953. 14. Karassick, I. J., et al., Pump Handbook, 3rd Ed., McGraw-Hill, New York, 2000. 15. Stepanoff, A. J., Turboblowers, Wiley, New York, 1955. 16. Jorgensen, R., ed., Fan Engineering, 9th Ed., Howden Buffalo, Buffalo, New York, 1999 (also available as a CD-ROM). 17. Wallis, R. A., Axial Flow Fans and Ducts, Wiley, New York, 1983. 18. Reason, J., “Fans,” Power, Vol. 127, No. 9, 103–128, 1983. 19. Cumpsty, N. A., Compressor Aerodynamics, Longman Scientific & Technical, Essex, UK, and Wiley, New York, 1989. 20. Johnson, I. A., and Bullock, R. D., eds., Aerodynamic Design of Axial-Flow Compressors, NASA SP-36, National Aeronautics and Space Administration, Washington, D.C., 1965. 21. Glassman, A. J., ed., Turbine Design and Application, Vol. 3, NASA SP-290, National Aeronautics and Space Administration, Washington, D.C., 1975. 22. Saeed Farokhi, Aircraft Propulsion, Wiley, New York, 2009. Problem is related to a chapter video available in WileyPLUS. * † Problem to be solved with aid of programmable calculator or computer. Open-ended problem that requires critical thinking. These problems require various assumptions to provide the necessary input data. There are not unique answers to these problems. Problems Note: Unless specific values of required fluid properties are given in the problem statement, use the values found in the tables on the inside of the front cover. Section 12.1 Introduction and Section 12.2 Basic Energy Considerations ω V1 r1 12.1 The rotor shown in Fig. P12.1 rotates clockwise. Assume that the fluid enters in the radial direction and the relative velocity is tangent to the blades and remains constant across the entire rotor. Is the device a pump or a turbine? Explain. r2 ■ Figure P12.1 728 Chapter 12 ■ Turbomachines 12.2 The measured shaft torque on the turbomachine shown in Fig. P12.2 is −60 N∙m when the absolute velocities are as indicated. Determine the mass flowrate. What is the angular velocity if the magnitude of the shaft power is 1800 N∙m/s? Is this machine a pump or a turbine? Explain. radius for flow through an axial-flow turbomachine rotor. The axial component of velocity is 50 ft/s at sections (1) and (2). (a) Label each velocity vector appropriately. Use V for absolute velocity, W for relative velocity, and U for blade velocity. (b) Are you dealing with a turbine or a fan? (c) Calculate the work per unit mass involved. (d) Sketch a reasonable blade section. Do you think that the actual blade exit angle will need to be less or greater than 15°? Why? V1 = 6 m/s r2 = 0.1 m ω + 50° 30° 15° 15° (1) (2) 30° Axial direction V2 = 5 m/s r1 = 0.2 m ■ Figure P12.6 Section 12.4 The Centrifugal Pump and Section 12.4.1 Theoretical Considerations ■ Figure P12.2 12.3 Uniform horizontal sheets of water of 3-mm thickness issue from the slits on the rotating manifold shown in Fig. P12.3. The velocity relative to the arm is a constant 3 m/s along each slit. Determine the torque needed to hold the manifold stationary. What would the angular velocity of the manifold be if the resisting torque is negligible? 3 m/s 12.7 The radial component of velocity of water leaving the centrifugal pump sketched in Fig. P12.7 is 45 ft/s. The magnitude of the absolute velocity at the pump exit is 90 ft/s. The fluid enters the pump rotor radially. Calculate the shaft work required per unit mass flowing through the pump. Vr2 = 45 ft/s 0.5 ft V2 = 90 ft/s V1 ω + 0.2 ft 3000 rpm 0.1 m 0.3 m ■ Figure P12.3 12.4 At a given radial location, a 15-mph wind against a windmill (see Video V12.1) results in the upstream (1) and downstream (2) velocity triangles shown in Fig. P12.4. Sketch an appropriate blade section at that radial location and determine the energy transferred per unit mass of fluid. |W2| = |W1| W2 W1 U1 = 20 mph U2 = 20 mph V2 V1 = 15 mph ■ Figure P12.7 12.8 Water enters a centrifugal pump with an absolute velocity V1 = 10 m/s in the radial direction and leaves with an absolute velocity V2, which makes an angle of θ2 = 60° with the radial direction, as shown in Fig. P12.8. The impeller width (perpendicular to the paper) is b = 0.125 m, R1 = 0.125 m, and R2 = 0.35 m. Find the input torque 𝒯 required to drive the pump if there are no friction losses. V2 θ2 60° 𝒯,ω ■ Figure P12.4 †12.5 Sketch how you would arrange four 3-in.-wide by 12-in.long thin but rigid strips of sheet metal on a hub to create a windmill like the one shown in Video V12.1. Discuss, with the help of velocity triangles, how you would arrange each blade on the hub and how you would orient your windmill in the wind. 12.6 Sketched in Fig. P12.6 are the upstream [section (1)] and downstream [section (2)] velocity triangles at the arithmetic mean V1 R1 R2 Impeller ■ Figure P12.8 Problems 12.9 A centrifugal pump impeller is rotating at 1200 rpm in the direction shown in Fig. P12.9. The flow enters parallel to the axis of rotation and leaves at an angle of 30° to the radial direction. The absolute exit velocity, V2, is 90 ft/s. (a) Draw the velocity triangle for the impeller exit flow. (b) Estimate the torque necessary to turn the impeller if the fluid is water. What will the impeller rotation speed become if the shaft breaks? 30° 729 vacuum of 95 mm of mercury and the pressure gage at (2) indicates a pressure of 80 kPa. The diameter of the pipe at the inlet is 110 mm and at the exit it is 55 mm. If z2 − z1 = 0.5 m, what is the actual head rise across the pump? Explain how you would estimate the pump motor power requirement. 12.13 The performance characteristics of a certain centrifugal pump having a 9-in.-diameter impeller and operating at 1750 rpm are determined using an experimental setup similar to that shown in Fig. 12.10. The following data were obtained during a series of tests in which z2 − z1 = 0, V2 = V1, and the fluid was water. V2 Q (gpm) 20 40 60 80 100 120 140 p2 − p1 (psi) 40.2 40.1 38.1 36.2 33.5 30.1 25.8 Power input (hp) + ω 1.58 2.27 2.67 2.95 3.19 3.49 4.00 1 ft Based on these data, show or plot how the actual head rise, ha, and the pump efficiency, η, vary with the flowrate. What is the design flowrate for this pump? *12.14 Determine algebraic equations for the pump head rise, power, and efficiency as functions of flow rate based on the data from Problem 12.18. Analytically determine the Best Efficiency Point and the BEP flow, head rise, and power. How do these values compare with those from Problem 12.18? 1 in. ■ Figure P12.9 12.10 A centrifugal radial water pump has the dimensions shown in Fig. P12.10. The volume rate of flow is 0.25 ft3/s, and the absolute inlet velocity is directed radially outward. The angular velocity of the impeller is 960 rpm. The exit velocity as seen from a coordinate system attached to the impeller can be assumed to be tangent to the vane at its trailing edge. The hydraulic efficiency is 82% and the mechanical efficiency is 96%. Calculate the power required to drive the pump. Q = 0.25 ft 3 /s 55° V1 + 11 in. 3 in. 960 rpm 0.75 in. Section 12.4.3 Net Positive Suction Head (NPSH) 12.15 In Example 12.3, how will the maximum height, z1, that the pump can be located above the water surface change if the water temperature is decreased to 40 °F? 12.16 A centrifugal pump with a 7-in.-diameter impeller has the performance characteristics shown in Fig. 12.12. The pump is used to pump water at 100 °F, and the pump inlet is located 12 ft above the open water surface. When the flowrate is 200 gpm, the head loss between the water surface and the pump inlet is 6 ft of water. Would you expect cavitation in the pump to be a problem? Assume standard atmospheric pressure. Explain how you arrived at your answer. 12.17 Water at 40 °C is pumped from an open tank through 200 m of 50-mm-diameter smooth horizontal pipe as shown in Fig. P12.17 and discharges into the atmosphere with a velocity of 3 m/s. Minor losses are negligible. (a) If the efficiency of the pump is 70%, how much power is being supplied to the pump? (b) What is the NPSHA at the pump inlet? Neglect losses in the short section of pipe connecting the pump to the tank. Assume standard atmospheric pressure. 3m Diameter = 50 mm ■ Figure P12.10 Pump Section 12.4.2 Pump Performance Characteristics 12.11 Water is pumped with a centrifugal pump, and measurements made on the pump indicate that for a flowrate of 240 gpm the required input power is 6 hp. For a pump efficiency of 62%, what is the actual head rise of the water being pumped? 12.12 The performance characteristics of a certain centrifugal pump are determined from an experimental setup similar to that shown in Fig. 12.10. When the flowrate of a liquid (SG = 0.9) through the pump is 120 gpm, the pressure gage at (1) indicates a Length = 200 m ■ Figure P12.17 12.18 The centrifugal pump shown in Fig. P12.18 is not selfpriming. That is, if the water is drained from the pump and pipe as shown in Fig. P12.18(a), the pump will not draw the water into the pump and start pumping when the pump is turned on. However, if the pump is primed [i.e., filled with water as in Fig. P12.18(b)], the pump does start pumping water when turned on. Explain this behavior. 730 Chapter 12 ■ Turbomachines Pump Pump (a) (b) as a function of the valve setting is shown in Fig. P12.23(b). The pump head-capacity relationship is given by the equation ha = 52.0 − 1.01 × 103 Q2 with ha in meters when Q is in m3/s. Assume the friction factor f = 0.02 for the pipe, and all minor losses, except for the valve, are negligible. The fluid levels in the two tanks can be assumed to remain constant. (a) Determine the flowrate with the valve wide open. (b) Determine the required valve setting (percent open) to reduce the flowrate by 50%. Open ■ Figure P12.18 3m Section 12.4.4 System Characteristics and Pump Selection Valve 30 D = 0.1 m 30 m KL 20 Pump 12.20 A centrifugal pump having a 6-in.-diameter impeller and the characteristics shown in Fig. 12.7 is to be used to pump gasoline through 4000 ft of commercial steel 3-in.-diameter pipe. The pipe connects two reservoirs having open surfaces at the same elevation. Determine the flowrate. Do you think this pump is a good choice? Explain. 12.22 Both the suction and discharge piping for the pump shown in Fig. P12.22 consist of 4-in. I.D. 40-ft-long plastic pipe. Find the volume flow rate of 60 °F water through the pump. The connections are glued (equivalent to soldered connections). d1 = 4 in. d2 = 4 in. Pump head hp(ft) 0 0 20 (Closed) 40 5 10 (b) 12.24 Two of the pumps in Fig. P12.24 are operated in series to supply water through the piping system. Determine the flow rate through the piping system for 10 °C water and screwed connections. Then find the total power input to the two pumps. 80 hp η 30 60 20 40 10 20 0 10 60 m of 0.10-m inside diameter, commercial steel pipe, one swing check valve, four 90° standard elbows 20 30 40 Volume flow rate, Q(l3/s) 50 ■ Figure P12.24 60 240 m of 0.10-m inside diameter, commercial steel pipe, four gate valves, one fully open globe valve, twelve 90° standard elbows, four tees with flow through lines 40 m ■ Figure P12.22 12.23 In a chemical processing plant a liquid is pumped from an open tank, through a 0.1-m-diameter vertical pipe, and into another open tank as shown in Fig. P12.23(a). A valve is located in the pipe, and the minor loss coefficient for the valve 80 ■ Figure P12.23 0 Volume flow rate Q(ft3/s) 60 % valve setting (a) Q 50 ft 150 100 50 0 10 0 Head, hp(m) 12.21 A centrifugal pump having the characteristics shown in Example 12.4 is used to pump water between two large open tanks through 100 ft of 8-in.-diameter pipe. The pipeline contains four regular flanged 90° elbows, a check valve, and a fully open globe valve. Other minor losses are negligible. Assume the friction factor f = 0.02 for the 100-ft section of pipe. If the static head (difference in height of fluid surfaces in the two tanks) is 30 ft, what is the expected flowrate? Do you think this pump is a good choice? Explain. 40 Efficiency η (%) 12.19 A centrifugal pump having a head-capacity relationship given by the equation ha = 180 − 6.10 × 10−4 Q2, with ha in feet when Q is in gpm, is to be used with a system similar to that shown in Fig. 12.14. For z2 − z1 = 50 ft, what is the expected flowrate if the total length of constant diameter pipe is 600 ft and the fluid is water? Assume the pipe diameter to be 4 in. and the friction factor to be equal to 0.02. Neglect all minor losses. 100 (Open) Problems †12.25 Water is pumped between the two tanks described in Example 12.4 once a day, 365 days a year, with each pumping period lasting two hours. The water levels in the two tanks remain essentially constant. Estimate the annual cost of the electrical power needed to operate the pump if it were located in your city. You will have to make a reasonable estimate for the efficiency of the motor used to drive the pump. Due to aging, it can be expected that the overall resistance of the system will increase with time. If the operating point shown in Fig. E12.4c changes to a point where the flowrate has been reduced to 1000 gpm, what will be the new annual cost of operating the pump? Assume that the cost of electrical power remains the same. Section 12.5 Dimensionless Parameters and Similarity Laws 12.26 A centrifugal pump having an impeller diameter of 1 m is to be constructed so that it will supply a head rise of 200 m at a flowrate of 4.1 m3/s of water when operating at a speed of 1200 rpm. To study the characteristics of this pump, a 1/5 scale, geometrically similar model operated at the same speed is to be tested in the laboratory. Determine the required model discharge and head rise. Assume that both model and prototype operate with the same efficiency (and therefore the same flow coefficient). 12.27 Do the head–flowrate data shown in Fig. 12.7 appear to follow the similarity laws as expressed by Eqs. 12.39 and 12.40? Explain. 12.28 A centrifugal fan operating in a duct has the dimensionless parameters CQ = Q ωD 3  and CH = Δp 2 ρω D 2 , where CQ is a flow coefficient, CH is a head coefficient, Q is the volume flow rate, ω is the fan speed, D is the fan diameter, ρ is the fluid density, and Δp is the fan pressure rise. Figure P12.28 shows this fan’s performance curve in dimensional form for a fan speed of ω15 = 1500 rpm. Find the fan operating points (Q and Δp) for ω30 = 3000 rpm and corresponding to points 1, 3, and 5 at ω15 = 1500 rpm. Assume similarity between 1500 rpm and 3000 rpm. Fan pressure rise, Δ p (kPa) 4 2 0 an inlet power of 1.24 hp, a flow rate of 220 ft3/min, and a speed of 1800 rpm. Find the corresponding input power and flow rate of the prototype fan. Neglect Reynolds number effects. 12.31 Use the data given in Problem 12.18 and plot the dimensionless coefficients CH, C𝒫, η versus CQ for this pump. Calculate a meaningful value of specific speed, discuss its usefulness, and compare the result with data of Fig. 12.18. 12.32 In a certain application, a pump is required to deliver 5000 gpm against a 300-ft head when operating at 1200 rpm. What type of pump would you recommend? 12.33 A centrifugal pump operates at 300 rpm to deliver 20 °C lubri1 cating oil. A 5-size, geometrically similar pump delivering 15 °C water is used to model the larger pump. How fast should the smaller pump run? Discuss the accuracy of the result. Section 12.6 Axial-Flow and Mixed-Flow Pumps 12.34 A certain axial-flow pump has a specific speed of Ns = 5.0. If the pump is expected to deliver 3000 gpm when operating against a 15-ft head, at what speed (rpm) should the pump be run? Draw a sketch of the pump impeller (front and side views). 12.35 A certain pump is known to have a capacity of 3 m3/s when operating at a speed of 60 rad/s against a head of 20 m. Based on the information in Fig. 12.18, would you recommend a radialflow, mixed-flow, or axial-flow pump? Draw a sketch of the pump impeller (front and side views). 12.36 The system resistance for a pipeline is given by Δpsys = 2.0Q2, where Δpsys is the pressure rise required of a pump to deliver the flow rate Q through the piping system. A pump has the pressure–rise–flow characteristic given by Δpp = 30.0 − 3.0Q2. In both curves, Δp is in kPa and Q is in m3/s. Find the pump input power if this pump is placed in this piping system and the pump overall efficiency is 90%. 12.37 The axial-flow pump shown in Fig. 12.19 is designed to move 5000 gal/min of water over a head rise of 5 ft of water. Estimate the motor power requirement and the U2Vθ2 needed to achieve this flowrate on a continuous basis. Comment on any cautions associated with where the pump is placed vertically in the pipe. 12.38 A propeller-driven airplane is traveling with a velocity V1  =  200 mph (294 ft/s). The propeller diameter is 10 ft and it rotates at 3000 rpm. Figure P12.38 shows a propeller crosssectional profile and the velocity diagrams for a short section of the propeller at a radius of 3.0 ft. The outlet relative velocity W2 is assumed tangent to the propeller at the outlet, so β′1 = β2 = 50°. The air density is constant as it flows over the propeller. Assume 3 1 731 1 2 3 Δ p = 1 – 0.25 Q2 4 0 1 N15 = 1500 rpm 5 2 3 β 2′ β2 4 3 Volume flow rate, Q (m /s) V1 α1 A1 12.30 A prototype fan has a 20-ft diameter, an inlet pressure of 14.40 psia, an inlet temperature of 70 °F, and a speed of 90 rpm. A 1 10 -scale model of the fan has the same inlet pressure and temperature, U V2 ■ Figure P12.28 12.29 A centrifugal pump has the performance characteristics of the pump with the 6-in.-diameter impeller described in Fig. 12.7. Note that the pump in this figure is operating at 3500 rpm. What is the expected head gained if the speed of this pump is reduced to 2800 rpm while operating at peak efficiency? W2 U α2 W1 β1 ■ Figure P12.38 A2 732 Chapter 12 ■ Turbomachines that the flow area for the mass flow rate interacting with this short section of the propeller is the same upstream and downstream (inlet and outlet) from the propeller (i.e., A1 = A2). Produce the velocity diagram downstream from the propeller by finding U, W2, V2, and α2. Section 12.7 Fans 12.39 For the fan of both Examples 5.19 and 5.28 discuss what fluid flow properties you would need to measure to estimate fan efficiency. 12.40 A lossless motor drives the fan shown in Fig. P12.40 at 40 Hz. The power input to the motor is 40 amps at 440 volts. For the geometry shown, what is the discharge flow rate of air through the fan? Assume that the tangential component of the velocity leaving the impeller is equal to that of the impeller at that point. The exit air temperature T2 = 15 °C. Vθ 2 12.45 An inward-flow radial turbine (see Fig. P12.48) involves a nozzle angle, α1, of 60° and an inlet rotor tip speed, U1, of 3 m/s. The ratio of rotor inlet to outlet diameters is 2.0. The absolute velocity leaving the rotor at section (2) is radial with a magnitude of 6 m/s. Determine the energy transfer per unit mass of fluid flowing through this turbine if the fluid is (a) air, (b) water. V2 = 6 m/s Vr2 ω Q Section (1) α1 1.2 m Vr2 Section 12.8 Turbines (also see Sec. 12.3) r1 r2 + Section (2) 0.3 m ■ Figure P12.40 ■ Figure P12.45 12.41 A centrifugal fan has a power input of 25 kW, an inner radius of R1 = 0.5 m, an outer radius of R2 = 1.0 m, and delivers 100 kg/s of air. There are no friction losses, the air inlet absolute velocity has no tangential component, and the outlet absolute velocity has a tangential component equal to the blade velocity at the outer radius R2. The rotor depth (i.e., blade height) is 1.0 m. What is the required rotational speed of the rotor? 12.42 An axial fan operating at 1000 rpm has the characteristics shown in Fig. P12.42. It delivers 15 °C atmospheric air through a 50-cm I.D., galvanized, sheet-metal, horizontal duct having a length of 175 m and seven 90° long-radius elbows. For constant air density what is the flow rate if the duct discharges to the atmosphere? 12.46 The frictionless converging stationary nozzle of the hydraulic turbine shown in Fig. P12.46 has an inlet pressure p0 = 480 kPa, a negligible inlet velocity V0, and an exit pressure p1 = 101 kPa. The velocity V1 is used to drive the axial flow turbine, which has a rotational speed of 185.7 rpm, an outside radius of R = 1.20 m, and a blade height of h = 0.40 m. Determine velocities V1 and V2 and the power transmitted to the turbine in terms of α1 and α2. The fluid density is constant and the velocities of the fluid relative to the blade are directed as shown in Fig. P12.46. Assume that the velocities are uniform over the flow inlet and outlet areas and use an average blade velocity at the blade midheight (r = 1.00 m) and a water temperature of 20°C. VO, pO 20 Static head (cm H2O) Nozzle 15 α1 V1 U 10 β1 W1 5 α1 W2 β 2′ β 1′ β2 U α2 V2 0 0 50 100 150 Section AA 200 U Air flow rate (m3/min) ■ Figure P12.42 12.43 A model fan with wheel diameter 32 in. is tested at a speed of 1750 rpm. The test fluid is air with density 0.075 lbm/ft3. At its BEP, the fan produces 8000 ft3/min at total pressure rise of 8 in.H2O. A geometrically similar fan is to handle 200,000 ft3/min of flue gas with density 0.050 lbm/ft3 and 30 in.H2O total pressure rise. Determine the required size and speed of the flue gas fan. Note any assumptions and/or limitations. 12.44 A fan is to produce a total pressure rise of 6 in.H2O and a flow of 4000 ft3/min. Two motors are available, 3550 rpm and 1160 rpm. For each motor, specify the best (most efficient) type of fan to use and sketch the impeller. Blades A A ω + ω R h Flow area ■ Figure P12.46 p1 Nozzles p2 = p1 733 Problems 12.47 A water turbine wheel rotates at the rate of 100 rpm in the direction shown in Fig. P12.47. The inner radius, r2, of the blade row is 1 ft, and the outer radius, r1, is 2 ft. The absolute velocity vector at the turbine rotor entrance makes an angle of 20° with the tangential direction. The inlet blade angle is 60° relative to the tangential direction. The blade outlet angle is 120°. The flowrate is 10 ft3/s. For the flow tangent to the rotor blade surface at inlet and outlet, determine an appropriate constant blade height, b, and the corresponding power available at the rotor shaft. Is the shaft power greater or less than the power lost by the fluid? Explain. 20° W1 b V1 60° + A h = 60 ft 2 B Turbine ■ Figure P12.49 12.50 A small Pelton wheel is used to power an oscillating lawn sprinkler as shown in Video V12.4 and Fig. P12.50. The arithmetic mean radius of the turbine is 1 in., and the exit angle of the blade is 135° relative to the blade motion. Water is supplied through a single 0.20-in.-diameter nozzle at a speed of 50 ft/s. Determine the flowrate, the maximum torque developed, and the maximum power developed by this turbine. r1 = 2 ft 100 rpm 120° 1 r2 = 1 ft W2 Q Section (2) Section (1) ■ Figure P12.47 12.48 A sketch of the arithmetic mean radius blade sections of an axial-flow water turbine stage is shown in Fig. P12.48. The rotor speed is 1500 rpm. (a) Sketch and label velocity triangles for the flow entering and leaving the rotor row. Use V for absolute velocity, W for relative velocity, and U for blade velocity. Assume flow enters and leaves each blade row at the blade angles shown. (b) Calculate the work per unit mass delivered at the shaft. rm = 6 in. ■ Figure P12.50 12.51 A windmill has an approach velocity V1 = 24 mph (35.3 ft/s) and a blade diameter of 100 ft. The blade rotates at 15 rpm. Figure P12.51 shows a blade cross-sectional profile and the velocity diagrams for a short section of the blade at a radius of r = 25.0 ft. The outlet relative velocity W2 is tangent to the blade at the outlet so β2 = β′2 = 30°. The air density is constant as it flows over the blade. Assume that the flow area for the mass flow rate that interacts with this short section of the blade is the same upstream and downstream of the blade (i.e., A1 = A2). Find the velocity diagram downstream of the blade by finding U, W2, V2, and 𝛼2. 1500 rpm Rotor Stator 70° U U 45° Blade sections at the arithmetic mean radius β2 W2 β 2′ A1 45° Blade ■ Figure P12.48 12.49 Figure P12.49 shows a piping system with frictional losses of hL1−2 = 4.0Q2, with hL1−2 in ft and Q in gal/min. The turbine performance characteristics are given by ht = 20 + 12.0Q, where ht is the turbine head in ft and Q is in gal/min. Find the flow rate Q. W1 β U 1 α1 V1 ■ Figure P12.51 U α2 V2 A2 734 Chapter 12 ■ Turbomachines 12.52 The single-stage, axial-flow turbomachine shown in Fig. P12.52 involves water flow at a volumetric flowrate of 9 m3/s. The rotor revolves at 600 rpm. The inner and outer radii of the annular flow path through the stage are 0.46 and 0.61 m, and β2 = 60°. The flow entering the rotor row and leaving the stator row is axial when viewed from the stationary casing. Is this device a turbine or a pump? Estimate the amount of power transferred to or from the fluid. Q= 3 9 m /s U1 W1 V1 12.59 A 1-m-diameter Pelton wheel rotates at 300 rpm. Which of the following heads (in meters) would be best suited for this turbine: (a) 2, (b) 5, (c) 40, (d) 70, or (e) 140? Explain. 12.60 Draft tubes as shown in Fig. P12.60 are often installed at the exit of Kaplan and Francis turbines. Explain why such draft tubes are advantageous. W2 β2 V2 U2 r0 = 0.61 m r1 = 0.46 m Draft tube 600 rpm ■ Figure P12.52 12.53 For an air turbine of a dentist’s drill like the one shown in Fig. E12.8 and Video V12.5, calculate the average blade speed associated with a rotational speed of 350,000 rpm. Estimate the air pressure needed to run this turbine. 12.54 Water for a Pelton wheel turbine flows from the headwater and through the penstock as shown in Fig. P12.54. The effective friction factor for the penstock, control valves, and the like is 0.032, and the diameter of the jet is 0.20 m. Determine the maximum power output. Elevation = 975 m ■ Figure P12.60 12.61 Turbines are to be designed to develop 30,000 horsepower while operating under a head of 70 ft and an angular velocity of 60 rpm. What type of turbine is best suited for this purpose? Estimate the flowrate needed. 12.62 Water at 400 psi is available to operate a turbine at 1750 rpm. What type of turbine would you suggest to use if the turbine should have an output of approximately 200 hp? 12.63 It is desired to produce 50,000 hp with a head of 50 ft and an angular velocity of 100 rpm. How many turbines would be needed if the specific speed is to be (a) 50, (b) 100? *12.64 Test data for the small Francis turbine shown in Fig. P12.64 is given in the following table. The test was run at a constant 32.8-ft D = 0.90 m l = 1020 m 1.7 m Elevation = 250 m Brake arm 0.20 m ■ Figure P12.54 Brake cord Q 12.55 A Pelton wheel has a diameter of 2 m and develops 500 kW when rotating 180 rpm. What is the average force of the water against the blades? If the turbine is operating at maximum efficiency, determine the speed of the water jet from the nozzle and the mass flowrate. ω 12.56 Water to run a Pelton wheel is supplied by a penstock of length ℓ and diameter D with a friction factor f. If the only losses associated with the flow in the penstock are due to pipe friction, show that the maximum power output of the turbine occurs when the nozzle diameter, D1, is given by D1 = D/(2f ℓ/D)1/4. 12.57 A Pelton wheel is supplied with water from a lake at an elevation H above the turbine. The penstock that supplies the water to the wheel is of length ℓ, diameter D, and friction factor f. Minor losses are negligible. Show that the power developed by the turbine is maximum when the velocity head at the nozzle exit is 2H/3. Note: The result of Problem 12.56 may be of use. 12.58 Water flows through the Pelton wheel turbine shown in Fig. 12.25. For simplicity we assume that the water is turned 180° by the blade. Show, based on the energy equation (Eq. 5.84), that the maximum power output occurs when the absolute velocity of the fluid exiting the turbine is zero. F 6 38 in. ω (rpm) Q (ft 3/s) F (lb) 0 1000 1500 1870 2170 2350 2580 2710 0.129 0.129 0.129 0.124 0.118 0.0942 0.0766 0.068 2.63 2.40 2.22 1.91 1.49 0.876 0.337 0.089 ■ Figure P12.64 Problems head just upstream of the turbine. The Prony brake on the turbine output shaft was adjusted to give various angular velocities, and the force on the brake arm, F, was recorded. Use the given data to plot curves of torque as a function of angular velocity and turbine efficiency as a function of angular velocity. Section 12.9 Compressible Flow Turbomachines 12.65 Obtain photographs/images of a variety of turbo-compressor rotors and categorize them as axial-flow or radial-flow compressors. Explain briefly how they are used. Note any unusual features. Repeat for compressible flow turbines. 12.66 An axial flow compressor stage shown in Fig. P12.66 has the inlet and outlet velocity diagrams shown. Calculate the work per unit mass. Quantities are U1 = U2 = U = 762 ft/s, V1 = 440 ft/s, W1 = 880 ft/s, α1 = 90°, V2 = 545 ft/s, W2 = 622 ft/s, and α2 = 53.8°. 735 12.68 A centrifugal air compressor has a rotor inner diameter of D1 = 2.0 in., a rotor outer diameter of D2 = 6.5 in., a rotor depth of 10 in., and a rotor rotational speed of 3600 rpm. The fluid relative velocities (W) are purely radial. The compressor delivers an air mass flow rate of 1.0 lbm/s with T1 = 70 °F, p1 = 14.7 psia, T2 = 240 °F, and p2 = 33.1 psia. Find the power transferred by the rotor to the air and the inlet relative velocity W1. 12.69 The axial flow steam turbine rotor shown in Fig. P12.69 has a blade outer radius R0 = 2.40 ft, a blade inner radius Ri = 2.00 ft, a steam inlet pressure p1 = 200 psia, a steam inlet density ρ1 = 0.296 lbm/ft3, and an inlet absolute velocity V1 = 1000 ft/s making an angle of 70° with the axial direction. The steam outlet pressure p2 = 50 psia and outlet density is 0.1014 lbm/ft3. β2 = 40°. The rotor rotates at 3600 rpm. Using a blade velocity at the blade midheight (R = 2.20 ft), find the power transferred from the steam to the rotor. Blade α2 W1 U2 V2 α1 = 20° β2 U2 α1 U V1 W2 β1 V1 U Turbine blade Section AA V2 α2 W1 W2 Section AA A A A A ■ Figure P12.66 Ri Ro 12.67 The axial flow gas turbine stage shown in Fig. P12.67 has a mean blade radius of R = 5.0 in., a rotational speed of 15,000 rpm, a mass flow rate of 10.0 lbm/s, W1 = 972 ft/s, V1 = 1560 ft/s, α1 = 20°, W2 = 874 ft/s, V2 = 483 ft/s, and α2 = 99.4°. Find the power transmitted from the gas to the turbine blade ■ Figure P12.69 α1 U W2 U 12.70 The figure below shows a nozzle vane and a rotor blade for an axial flow gas turbine stage. The blade speed is 800 ft/s. The absolute velocity leaving the stage is identical to the absolute velocity entering the stage and both are purely in the axial direction. (a) Draw and label the velocity diagrams. Show values for all three velocities and the absolute and relative angles on each diagram. (b) Calculate the work per unit mass for this stage. α2 Blade V1 V2 W1 Section AA 30° A R = 5 in. A U = 800 ft/s ■ Figure P12.67 ■ Figure P12.70 40° 736 ■ Chapter 12 ■ Turbomachines Lifelong Learning Problems 12.1 LL What do you think are the major unresolved fluid dynamics problems associated with gas turbine engine compressors? For gas turbine engine high-pressure and low-pressure turbines? For jet engine fans? 12.2 LL Pick a specific day (any day will do). During that day note and record at least ten different instances when the operation of one or more turbomachines makes your life better or easier or more comfortable. (If the machine operates more-or-less continuously, like a furnace fan, note only once.) In each case, record the type of machine (e.g., “axial-flow fan”). If you don’t know, make your best guess. Then, for each machine, estimate its flowrate, pressure change, and power input or output. 12.3 LL What are current efficiencies achieved by the following categories of turbomachines? (a) Wind turbines; (b) hydraulic turbines; (c) power plant steam turbines; (d) aircraft gas turbine engines; (e) natural gas pipeline compressors; (f) home vacuum cleaner blowers; (g) laptop computer cooling fan; (h) irrigation pumps; (i) dentist drill air turbines. What is being done to improve these devices? 12.4 LL Perform an on-line search. Identify at least ten different pump manufacturers. For each, choose at least three different types of pumps in their product line. Find performance curves for these pumps and estimate the specific speed for five of them and the suction specific speed for five different pumps. Do these pumps agree with Fig. 12.18? A A.1 Computational Fluid Dynamics Introduction What is Computational Fluid Dynamics (CFD)? As with many questions related to fluid mechanics, this question is much easier to ask than answer. It is not simply the use of computers to model the behavior of fluid components and systems. Going all the way back to hydrostatics, using a computer to perform the arithmetic required to compute the force on a submerged surface from the hydrostatic pressure distribution is not CFD. Using a numerical method to approximate the ordinary differential equation that results from using the Bernoulli and continuity equations to model tank draining or filling is not CFD. Using a computer to compute the friction factor for pipe flow from the implicit Colebrook formula is not CFD. The common trait shared by these examples is that the general continuum-level model of fluid flow as represented by the Navier-Stokes equations has been greatly simplified using assumptions justified by the application before the computational power of the computer has been brought to bear. The set of nonlinear partial-differential equations has been reduced to an explicit algebraic equation, a fairly simple ordinary differential equation, and an implicit algebraic equation, respectively. In all three instances, a reasonably useful answer could be obtained with a pencil, a piece of paper, a calculator, and a modest amount of effort. In all three instances the specialized physics of the process is already deeply imbedded in the model before a computational method is employed. Computational Fluid Dynamics (CFD) is the simulation of fluid flow using the vast computational power of modern computers to obtain approximate solutions to generally applicable models of fluid behavior (e.g., the Navier-Stokes equations, Eqs. 6.127a,b,c and the continuity equation, Eq. 6-27) for a wide variety of boundary and initial conditions. CFD involves replacing the partial differential equations with discretized algebraic equations that approximate the partial differential equations. These algebraic equations are then numerically solved to obtain flow field values at discrete points in space and/or time. Since the Navier–Stokes equations are valid everywhere in the flow field, an analytical solution to these equations would provide the solution for an infinite number of points in the flow. The CFD simulation solves for the relevant flow variables only at the discrete points, which make up the grid or mesh of the solution (discussed in more detail below). Interpolation schemes must be used to obtain values at non-grid point locations. In general, most applications of CFD take the same basic approach. Some of the differences include problem complexity, available computer resources, available expertise in CFD, and whether a commercially available CFD package is used, or a problem-specific CFD algorithm is developed. In today’s market, there are many commercial CFD codes available to solve a wide variety of problems. However, if the intent is to conduct a thorough investigation of a specific fluid flow problem such as in a research environment, it is possible that taking the time to develop a problem-specific algorithm may be most efficient in the long run. CFD can be thought of as a numerical experiment. In a typical fluids experiment, an experimental model is built, measurements of the flow interacting with that model are made, and the results are analyzed. In CFD, the building of the model is replaced with the formulation of the governing equations and the development of the numerical algorithm. The process of obtaining measurements is replaced with using a computer and software to implement the algorithm to simulate the flow interaction. Of course, the analysis of results is common ground to both techniques. 737 738 Appendix A ■ Computational Fluid Dynamics In a very real sense, CFD represents a “marriage” of the disciplines of Fluid Mechanics, Numerical Analysis, and Computer Science; therefore a thorough coverage of CFD topics is well beyond the scope of this textbook. This appendix highlights some of the more important topics in CFD, but is only intended as a brief introduction. The topics include discretization of the governing equations, grid generation, boundary conditions, turbulence models, solving the equations, application of CFD, and some representative examples. A.2 A Very Simple Example We will begin our consideration of CFD with a very simple example. Although modern CFD analyses are developed very formally, the development of this example will be more intuitive. After developing nearly all of the important concepts of CFD in this example, we will be prepared to discuss the features of the large and powerful CFD (computer) codes currently available. EXAMPLE A.1 Flow in a Converging Channel GIVEN Water flows in the converging channel 6 cm shown in Figure EA.1a. FIND Determine the pressure distribution on 1 cm the upper (flat) wall. 100 cm/s 2 cm 50 cm/s y x 1 cm 4 cm ■ Figure EA.1a A converging channel SOLUTION The flow in the channel is governed by the continuity equation and the Navier-Stokes equations. To keep this example simple, we make the following assumptions: 1. The flow is steady. 2. The flow is repeatable in planes parallel to the paper (i.e., it is two-dimensional). 3. Shear stress is negligible, so the fluid is assumed to have zero viscosity. 4. The fluid is incompressible. With these assumptions the governing equations (Eqs. 6.27, 6.127a,b,c) reduce to ∂u ∂v + =0 ∂x ∂y ∂u ∂u 1 ∂p u +v =− ρ ∂x ∂x dy ∂v ∂v 1 ∂p +v =− u ρ ∂y ∂x dy In so-called primitive variables (u, v, p) there are three equations in three unknowns—and two of the equations are nonlinear partial differential equations. Fortunately, from Sec. 6.4.2, a solution to the two momentum equations is simply Bernoulli’s equation p u2 + v2 = Constant + ρ 2 ∂ψ ∂ψ and v = − dy dx (see Eq. 6.37), the continuity equation is automatically satisfied Introducing a stream function ψ such that u = ∂ 2ψ 2 + ∂ 2ψ = −ζz . Finally, because the flow originates in a ∂x ∂y2 region of uniform velocity where ζz = 0 and the fluid has zero viscosity, ζz = 0 everywhere (see Sec. 6.4.3). We can solve the ∂ 2ψ ∂ 2ψ problem by first solving 2 + 2 = 0, then by finding u and ∂x ∂y v from ψ , and, finally, calculating the requested pressure from Bernoulli’s equation. The task now is to solve the partial differential equation (PDE) and ∂ 2ψ 2 ∂x + ∂ 2ψ ∂y2 = 0. (1) Appendix A ■ Computational Fluid Dynamics e d c b a f g h i j k l m n 1 4 7 10 13 16 18 20 22 2 5 8 11 14 17 19 21 23 3 6 9 12 15 ee dd cc bb aa z y x w o 24 v Of course we must provide a set of boundary conditions in order to specify the geometry of this particular channel. Boundary conditions for ψ are especially simple; we specify values for ψ at all boundaries. Recalling that for two-dimensional flow, the difference between the values of ψ between two streamlines is the volume flow between the streamlines, we can write the boundary conditions as: p r 25 s u t ■ Figure EA.1b Channel with grid the first derivative of ψ by either Along the upper wall: ψ = 100 0 ≤ x ≤ 6; y = 2 Along the bottom wall: ψ = 0 0 ≤ x ≤ 1; y = 0, 1 ≤ x ≤ 5; y = 0.25(x − 1), 5 ≤ x ≤ 6; y = 1 It might be possible to solve the equation analytically using, say, a complex variable transformation. If we were able to solve it analytically, the benefit would be that we could determine the velocities at any point in the channel. The disadvantages would be that we would be in for some complicated mathematics and the operations necessary to calculate values at any point would likely be tedious. Let’s decide that we don’t need the pressure and velocities at every point; values at several specific points will be satisfactory. We will seek a numerical (computational) solution. To get started, we need to do two things immediately: establish the points in the field where we will determine values of ψ , and put the PDE in a form that allows us to use algebraic equations to calculate values of ψ for every point that we define. We start by defining a grid that is overlaid on the flow field. Such a grid is shown in Figure EA.1b. For this case, we use a uniform (Δx = Δy = 0.5 cm) Cartesian (square) grid. The points at which we will compute ψ are the points at the intersection of the grid lines and are numbered from 1 to 25. The grid lines at the far left and the far right are the channel inlet and the channel outlet, respectively. These points and the points on the walls are assigned letters rather than numbers; ψ is known at these points and they will provide boundary conditions. In order to calculate 25 unknown values of ψ we will need 25 equations. Since Eq. (1) applies at every point in the field, we can obtain exactly 25 equations by applying it at each of the 25 points. To do so, we must approximate Eq. (1) using the value of ψ at a point together with values at several surrounding points. Figure EA.1c shows a typical point in the field, labeled P, and its neighboring points, labeled N, E, S, and W.* We approximate *The notation comes from the points on a compass. ∂ψ ψE − ψP ≈ ∂x )P Δx (2) ψP − ψW ∂ψ ≈ . ) ∂x P Δx (3) or Along the inlet: ψ = 50y x = 0; 0 ≤ y ≤ 2 Along the outlet: ψ = 100(y − 1) x = 6; 1 ≤ y ≤ 2 739 These two approximations, called a forward difference and a backward difference, become more accurate as Δx becomes smaller. Although intended as approximations of the derivative at P, they are better approximations for the derivative at point PE midway between P and E and point PW, midway between P and W. We obtain an approximation to the second derivative by ∂2ψ 2 ∂x = ∂ψ∂x) PE − ∂ψ∂x) PW ∂ ∂ψ ≈ . ( ) ∂x ∂x Δx Equations 2 and 3 are good approximations for the derivative at points PE and PW, so ∂2ψ 2 ∂x ≈ ( ψ E − ψ P )Δx − ( ψ P − ψ W )Δx ψ E + ψ W − 2ψ P = . Δx ( Δx) 2 (4) This approximation for the second derivative is most accurate at P, because it is “centered” there. By similar reasoning, the y derivative is approximated by ∂2ψ 2 ∂y ≈ ψ N + ψ S − 2ψ P ( Δy) 2 , Δx N Δy PN P W PW PE E PS S ■ Figure EA.1c A typical point (P) with four surrounding points 740 Appendix A ■ Computational Fluid Dynamics P Δ xB Boundary point Δx B ■ Figure EA.1d Regular point (S) lies outside a boundary. S to solve this system including Gaussian elimination and matrix inversion. Programs for performing these operations are common on computer systems and even hand-held calculators. Here we will solve the equations by Gauss-Sidel Iteration. The method begins by rewriting the equations in a form where the value of ψ at the central point is written in terms of values at the surrounding points, for example ψ1 = (75 + 100 + ψ4 + ψ2 )4 ψ5 = (ψ2 + ψ4 + ψ6 + ψ8 )4 ψ12 = (ψ9 + ψ11 + ψ15 + 0)5 and Eq. (1) is approximated by 1 2 1 2 ( ψ + ψ − 2ψ ) + E W P ( Δy ) ( ψ N + ψ S − 2ψ P ) = 0. ( Δx ) Now we proceed in a series of steps If Δx = Δy (a square grid), we get 1 2 ( ψ + ψ W + ψ N + ψ S − 4ψ P ) = 0, ( Δx ) E ■ Guess values for all unknown ψ s . (For our problem, ψ = 50 is a good guess.) ■ “Visit” each point in turn calculating a new value of ψ P using the most recently calculated values for ψ on the right side. If we move through the grid from top to bottom and left to right, the values for ψ N and ψ E will be “new” values, and the values for ψ W and ψ S will be “old” values. ■ Continue the process until the largest change in ψ P is as small as you want it to be. or, as Δx is finite, ψN + ψE + ψS + ψW − 4ψP = 0 (5) Eq. (5) can be written for any interior point in the field (after we change the PNESW notation to a global point-numbering system). For points adjacent to the boundaries, at least one of the surrounding points is on a boundary and the corresponding value of ψ is known. Sometimes the point located a distance Δx away may lie outside the boundary (Fig. EA.1d ), in which case we find a value for ψ at the point by extrapolation: ψS = ψP + Δx ( ψ − ψ P ). ΔxB B Substituting this expression into Eq. (5) gives the modified equation ψN + ψE + aψB + ψW − (3 + a)ψP = 0 (6) Δx . ΔxB Now we can write either Eq. (5) or Eq. (6) for each of the 25 points in the grid. This will give 25 linear, algebraic equations in 25 unknowns (the 25 ψ ’s). A few of the equations are: where a = 75 + 100 + ψ4 + ψ2 − 4ψ1 = 0 ψ2 + ψ4 + ψ6 + ψ8 − 4ψ5 = 0 ψ9 + ψ11 + ψ15 + 2(0) − (3 + 2)ψ12 = 0 This process can be easily programmed for a computer or programmable calculator. The setup and calculations work especially well on an electronic spreadsheet. Simply type the equation for any ψ in terms of the surrounding ψ ’s in the appropriate cell. The software will object that a circular reference occurs; simply turn on “iterations” or switch to manual recalculation. Figure EA.1e shows the converged solution as calculated on a spreadsheet. These results are mildly interesting but this array of numbers is not particularly informative. One thing we could do is plot several streamlines. This would be done by picking several values for ψ (values of 25, 50, and 75 would be convenient), locating the point in each column where that value of ψ occurs, and connecting the points. Interpolation within a column would be required to determine the location of each point. The stated objective of this problem is to determine the pressure distribution on the upper wall. To do this we must find the velocity along the upper wall and then find the pressure from Bernoulli’s equation. The velocities at the upper wall are given by Solving the 25 equations will give the 25 values of ψ . The equations can be represented as AX = B where A is a 25 × 25 matrix, X is the column vector of 25 ψ ’s, and B is the vector of known values assembled from known boundary values of ψ . The matrix A is sparse; no row contains more than 5 entries (typically, four “1’s” and one “–4”). There are many methods 1 2 3 4 5 6 7 8 A B C D E F G H I J u= ∂ψ ;v=0 ∂y Returning for the moment to the PNESW notation (see Fig. EA.1c), the velocity can be estimated by (see Eqs. (2) and (3)) uP ≈ K L M N ψP − ψS Δy O P Upper Wall of Channel 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 100.000 75.000 74.404 73.579 72.327 70.641 68.538 66.004 62.979 59.419 55.381 51.435 50.359 50.000 Channel 50.000 49.038 47.583 45.090 41.697 37.507 32.501 26.492 19.316 10.671 0.000 0.000 0.000 Inlet 25.000 24.166 22.626 18.754 13.549 7.294 0.000 0.000 0.000 0.000 Channel Outlet ■ Figure EA.1e Converged solution for values of ψ calculated by an electronic spreadsheet. Appendix A ■ Computational Fluid Dynamics 0.5 0 2.5 6.5 120 100 Cp –1 80 –1.5 60 V –2 Velocity [cm/s] Pressure coefficient [Dimensionless] –0.5 4.5 741 40 –2.5 20 –3 0 –3.5 x[cm]-Distance along upper wall where P is a point on the upper wall (P = e, f, g, . . . r) and S is the first point in the field (S = d, 1, 5, 7, . . . 25, s). See Figure EA.1b. The pressure is computed in terms of the pressure coefficient Cp = p − pref 1 2 ρV ref 2 =1− u2 + v2 V 2ref Cp, j representative of the velocity at a point halfway between the wall and the first grid point than it would the velocity at the wall. We could obtain a value that better represents the wall velocity by using a three-point (backward) difference at the wall. The formula uses the point at the wall, the first point away from the wall, and the point beyond that. In PNESW notation u≈ where the reference pressure and velocity are those at the channel inlet. To illustrate, for point j ψj − ψ13 100 − 68.538 = = 62.92 Δy 0.5 2 2 uj + vj 62.922 + 02 =1− =1− = −0.584 2 V in 502 uj = ■ Figure EA.1 f Velocity and Pressure Distribution Along Upper Wall of a Converging Channel The velocity and pressure coefficient on the upper wall are graphed in Figure EA.1f. The computed points have been connected by a smooth curve although we have values only at the discrete points. COMMENTS Our results are approximate. There are two things we could do immediately to improve the accuracy. First, we could use a finer grid, say Δx = Δy = 0.25 cm. This would give about 100 points in the field, requiring us to solve 100 equations in 100 unknowns. The second accuracy improvement is a little less obvious. We used a simple backward difference to compute the velocity. The resulting value would be more 3ψP − 4ψS + ψS+1 * . 2Δy Note that it was necessary to specify boundary conditions that completely surround the flow field. Knowing the inlet and wall boundary conditions is not especially difficult, but the outlet boundary is not so clear. From continuity, we know that the average velocity at the outlet must be 100 cm/s but it would not necessarily be uniform. It may be better to “extend” the channel further downstream to assure that the outlet velocity is truly uniform. Computing the flow field in this example was straightforward and successful for one very important reason: our assumptions about the flow led to a linear PDE for the stream function. This led to a set of linear algebraic equations for calculating ψ . Although we chose an iterative method for solving the equations, iteration was not really necessary; a direct solution via Gaussian elimination or matrix inversion was an option. Because the equations are linear and the boundary conditions are well-posed, a unique solution to the problem was guaranteed. Things become far more difficult when the nonlinear Navier-Stokes equations are to be solved. *This formula can be derived by fitting a parabola to the points P, S, and S+1, and evaluating the derivative at point P (i.e., y = 0). Example A.1 demonstrates the steps that are required to perform a CFD simulation. A review of the example will identify the following steps: ■ Identify the flow system to be simulated and determine the features to be included in the simulation. Stated another way, the capabilities of the CFD code and the flow system model must be matched. Will the simulation be fully three-dimensional? Steady or unsteady? Incompressible? Laminar, turbulent, or inviscid? 742 Appendix A ■ Computational Fluid Dynamics ■ ■ ■ ■ ■ ■ ■ Discretize the partial differential equations, converting them into a set of algebraic equations. Discretize the flow field using an appropriate grid (or mesh). Specify the boundary conditions. Select a turbulence model (if the flow is or is expected to be turbulent). Solve the equations (usually by iteration). Process the solution and display the results in useful form. Called post-processing, this step generates graphs, vector plots, and, often, color graphics visualizations of the flow field, pressure field and/or temperature field. Investigate and verify the accuracy of the solution. A few of these steps are self-explanatory. The following sections discuss some options for the others and, in some cases, their consequences. A.3 Discretization The process of discretization involves developing a set of algebraic equations (based on discrete points in the flow domain) to be used in place of the partial differential equations. Of the various discretization techniques available for the numerical solution of the governing differential equations, the following types are most common: (1) the finite element method, (2) the boundary element method, (3) the finite difference method, and (4) the finite volume method. In each of these methods, the continuous flow field (i.e., velocity and pressure as a function of space and time) is described in terms of discrete (rather than continuous) values at prescribed locations. Through these techniques the differential equations are replaced by a set of (usually nonlinear) algebraic equations that can be solved with the aid of a computer. For the finite element method, the flow field is broken into a set of small fluid elements (usually triangular areas if the flow is two-dimensional, or small volume elements if the flow is three-dimensional). A number of nodes are defined for each element, always one at each corner and sometimes elsewhere such as the center of each side. The field variables (u, v, p, etc.) are approximated within each element by an algebraic function containing the nodal values. The approximating functions are then substituted into the conservation equations (mass, momentum, energy). Because the functions are approximations, the conservation equations are not exactly satisfied; a set of residuals is produced. The residuals are then multiplied by a weighting function and integrated over the element. The computational equations are produced by minimizing the weighted residual for each element and then “hooking up” all of the elements in the field (a process called assembly). This process produces a set of N algebraic equations in N unknowns where N is the number of unknown nodal values. These equations are solved with the aid of a computer. The number, size, and shape of elements are dictated in part by the particular flow geometry and flow conditions for the problem at hand. As the number of elements increases (as is necessary for flows with complex boundaries), the number of simultaneous algebraic equations that must be solved increases rapidly. Problems involving one million to ten million (or more) elements are not uncommon in today’s CFD community, particularly for complex three-dimensional geometries. The finite element method is extremely popular in computational solid mechanics but is not as widely used in CFD. Further information about this method can be found in Refs. 1 and 2. For the boundary element method, the boundary of the flow field (not the entire flow field as in the finite element method) is broken into discrete segments (Ref. 3) and appropriate singularities such as sources, sinks, doublets, and vortices are distributed on these boundary elements. The strengths and type of the singularities are chosen so that the appropriate boundary conditions are obtained on the boundary elements. For points in the flow field not on the boundary, the flow is calculated by adding the contributions from the various singularities on the boundary. Although the details of this method are rather mathematically sophisticated, it may (depending on the particular problem) require less computational time and space than the finite element method. Typical boundary elements and their associated singularities (vortices) for two-dimensional flow past an airfoil are shown in Fig. A.1. Such use of the boundary element method in aerodynamics is often termed 743 Appendix A ■ Computational Fluid Dynamics i th panel Γi – 1 Γi Γi+ 1 U Γi = strength of vortex on i th panel ■ Figure A.1 Panel method for flow past an airfoil. the panel method in recognition of the fact that each element plays the role of a panel on the airfoil surface (Ref. 4). Because the boundary element method simulates the flow field using sources, sinks, and vortices, it can only be used to simulate inviscid (potential) flows. The finite difference method for computational fluid dynamics is perhaps the most easily understood of the four methods listed above. The finite difference method was used somewhat informally in Example AE.1. For this method the flow field is dissected into a set of grid points and the continuous functions (velocity, pressure, etc.) are approximated by discrete values of these functions calculated at the grid points. Derivatives of the functions are approximated using the differences between the function values at local grid points divided by the grid spacing. Using a Taylor series expansion to relate the value of a variable at one point to its value at a neighboring point provides a strong mathematical foundation for developing the required approximations and for understanding the limitations of the approximation. (See Ref. 5.) For example, consider a rectangular grid applied to a flow domain as shown in Fig. A.2. The grid stencil shows five grid points in x–y space with the center point being labeled as i, j. Note that the PNESW notation of Example AE.1 has been replaced by double subscript notation. The index notation is used as subscripts on variables to signify location. For example, ui+1, j is the u component of velocity at the first point to the right of the center point i, j. The grid spacing in the i and j directions is given as Δx and Δy, respectively. To find an algebraic approximation to a first derivative term such as ∂u/∂x at the i, j grid point, consider a Taylor series expansion written for u at i + 1 as (Δx) 2 (Δx) 3 ∂u Δx ∂ 2u ∂ 3u ui+1, j = ui, j + ( ) + ( 2) + ( 3) + … ∂x i, j 1! ∂x i, j 2! ∂x i, j 3! (A.1) Solving for the underlined term in the above equation results in the following: ui+1, j − ui, j ∂u + O(Δx) ( ∂x )i, j = Δx (A.2) where O(Δx) contains higher order terms proportional to Δx, (Δx) 2 , and so forth. Equation A.2 represents a forward difference equation to approximate the first derivative using values at i + 1, j and i, j along with the grid spacing in the x direction. Obviously in solving for the ∂u /∂x term we have ignored higher order terms such as the second and third derivatives present in Eq. A.1. This process is termed truncation of the Taylor series expansion. The lowest order term that was truncated included ( Δx) 2 . Notice that the first derivative term contains Δx. When solving for the first derivative, all terms on the right-hand side were divided by Δx. Therefore, the term O( Δx) signifies that this equation has error of “order ( Δx),” which is due to the neglected terms in the Taylor series and is called truncation error. Hence, the forward difference is termed first-order accurate. y i–1 i i+1 j+1 j Δy j–1 Δx x ■ Figure A.2 Standard rectangular grid. 744 Appendix A ■ Computational Fluid Dynamics Thus, we can transform a partial derivative into an algebraic expression involving values of the variable at neighboring grid points. This method of using the Taylor series expansions to obtain discrete algebraic equations is called the finite difference method. Similar procedures can be used to develop backward difference and central difference approximations of the first derivative. The central difference makes use of both the left and right points (i.e., i − 1, j and i + 1, j) and is second-order accurate. In addition, finite difference equations can be developed for the other spatial directions (i.e., ∂u/∂y) as well as for second derivatives (∂ 2u/∂x2 ), which are also contained in the Navier–Stokes equations (see Ref. 5 for details). Applying this method to all terms in the governing equations approximates the differential equations as a set of algebraic equations involving the physical variables at the grid points (i.e., ui, j , pi, j for i = 1, 2, 3, … and j = 1, 2, 3, … , etc.). This set of equations is then solved by appropriate numerical techniques. The larger the number of grid points used, the larger the number of equations that must be solved. For further information on the finite difference method, see Refs. 5 and 6. The concept behind the finite volume method might be obvious to students of fluid mechanics. In this method, the flow field is divided into a large number of tiny control volumes. In the formal development of the method, the continuity and Navier-Stokes equations are integrated over a volume and a mathematical expression called the Gauss divergence theorem is used to convert the volume integrals of momentum and shear stress into surface integrals. The resulting algebraic equation(s) are identical to those we would obtain by applying the continuity and momentum equations from Chapter 5 using the tiny control volume. Early finite volume codes computed the velocity at the faces of the control volume and the pressure at the center and were more-or-less restricted to rectangular (in 2-D) and rectangular parallelepiped (in 3-D) volumes. More recent codes compute both velocity and pressure at the center of the volume and can use triangular (2-D) and tetrahedral (3-D) volumes to better fit irregular geometries. As in the finite element and finite difference methods, applying the finite volume formulation results in a large number of algebraic equations for the field variables. See Refs. 7 and 8 for further details of the finite volume method. If the flow is unsteady, it is necessary to discretize time as well as space. Time appears in the governing equations only as a first derivative and a simple first-order forward difference is most often used ∂u /∂t ≈ (u n+1 − u ni,j )Δt i, j (A.2) where superscripts identify the time level and Δt is the time step. The most significant issue in unsteady problems is not discretizing the time derivative; it is whether to evaluate the remainder of the terms at the old time level (n) (giving an explicit method) or the new time level (n+1) (giving an implicit method). A student of CFD should realize that the discretization of the governing equations yields algebraic equations that are an approximation to the original partial differential equation. Along with this approximation comes some amount of error. This type of error is termed truncation error because the Taylor series expansion used to represent a derivative is “truncated” at some reasonable point and the higher order terms are ignored. The truncation errors tend to zero as the grid is refined by making Δx and Δy smaller, so grid refinement is one method of reducing this type of error. Another type of unavoidable numerical error is the so-called round-off error. This type of error is due to the limit of the computer on the number of digits it can retain in memory. Engineering students can run into round-off errors from their calculators if they plug values into the equations at an early stage of the solution process. Fortunately, for most CFD cases, if the algorithm is set up properly, round-off errors are usually negligible. A.4 The Computational Grid The result of a CFD simulation is a set of values for variables of interest (e.g., velocity, pressure) at a finite number of points in the flow domain. The process by which the location of these points is defined varies with the discretization being employed. For example, a finite-difference model directly defines the locations of the points in space. It is frequently useful to connect these points with lines to Appendix A ■ Computational Fluid Dynamics (a) 745 (b) ■ Figure A.3 Structured grids. (a) Rectangular grid. (b) Grid around a parabolic surface. VIDEO VA.2 Dynamic grid form a “mesh” or “grid” that covers the flow domain. We often refer to the process of defining the points as “generating” a mesh or grid. A finite-volume discretization requires that the flow field be divided into small volumes. In this case, the boundaries of the volumes define the mesh. Again, the process of defining the volumes is usually called grid or mesh generation. As will be explored in more detail in a following section, points at which the values of variables are computed may, or may not, be at the intersection of grid lines. Discretization of the flow field for a finite element model is again referred to as mesh generation, again the mesh is formed by element boundaries, and again the points at which values of variables are computed may or may not be the intersections of grid lines. The type of grid developed for a given problem can have a significant impact on the numerical simulation, including the accuracy of the solution. The grid must represent the geometry correctly and accurately, since an error in this representation can have a significant effect on the solution. The grid must also have sufficient resolution to capture the relevant flow physics, otherwise they will be lost. This particular requirement is problem dependent. For example, if a flow field has small-scale structures, the grid resolution must be sufficient to capture these structures. It is usually necessary to increase the number of grid points or volumes (i.e., use a finer mesh) where large gradients are to be expected, such as in the boundary layer near a solid surface. The same can also be said for the temporal resolution. The time step, Δt, used for unsteady flows must be smaller than the smallest time scale of the flow features being investigated. Generally, the types of grids fall into two categories: structured and unstructured, depending on whether or not there exists a systematic pattern of connectivity of the grid points with their neighbors. As the name implies, a structured grid has some type of regular, coherent structure to the mesh layout that can be defined mathematically. The simplest structured grid is a uniform rectangular grid, as shown in Fig. A.3a and used in Example A.1. However, structured grids are not restricted to rectangular geometries. Figure A.3b shows a structured grid wrapped around a parabolic surface. Notice that grid points are clustered near the surface (i.e., grid spacing in normal direction decreases as one moves toward the surface) to help capture the steep flow gradients found in the boundary layer region. This type of variable grid spacing is used wherever there is a need to increase grid resolution and is termed grid stretching. For the unstructured grid, the grid cell arrangement is irregular and has no systematic pattern. The grid cell geometry usually consists of various-sized triangles for two-dimensional problems and tetrahedrals for three-dimensional grids. An example of an unstructured grid is shown in Fig. A.4. Unlike structured grids, for an unstructured grid each grid cell and the connection information to neighboring cells is defined separately. This produces an increase in the computer code complexity as well as a significant computer storage requirement. The advantage to an unstructured grid is that it can be applied to complex geometries, where structured grids would have severe difficulty. The finite difference method is usually restricted to structured grids whereas the finite volume and finite element methods can use either structured or unstructured grids. 746 Appendix A ■ Computational Fluid Dynamics ■ Figure A.4 Anisotropic adaptive mesh for flow induced by the rotor of a helicopter in hover, with ground effect. Left: flow; Right: grid. (From Newmerical Technologies International, Montreal, Canada. Used by permission.) Other grids include hybrid, moving, and adaptive grids. A grid that uses a combination of grid elements (rectangles, triangles, etc.) is termed a hybrid grid. As the name implies, the moving grid is helpful for flows involving a time-dependent geometry. If, for example, the problem involves simulating the flow within a pumping heart or the flow around a flapping wing, a mesh that moves with the geometry is desired. The nature of the adaptive grid lies in its ability to literally adapt itself during the simulation. For this type of grid, while the CFD code is working toward a converged solution, the grid will adapt itself to place additional grid resources in regions of steep gradients. Such a grid is particularly useful when a new problem arises and the user is not quite sure where to refine the grid due to steep gradients. A.5 Boundary Conditions The same governing equations, the Navier–Stokes equations (see Eqs. 6.127a,b,c), are valid for all Newtonian fluid flow problems. Thus, if the same equations are solved for all types of problems, how is it possible to achieve different solutions for different types of flows involving different flow geometries? The answer lies in the boundary conditions of the problem. The boundary conditions are what allow the governing equations to differentiate between different flow fields (for example, flow past an automobile and flow past a person running) and produce a solution unique to the given flow geometry. It is critical to specify the correct boundary conditions so that the CFD simulation is a wellposed problem and is an accurate representation of the physical problem. Poorly defined boundary conditions can ultimately affect the accuracy of the solution. One of the most common boundary conditions used for simulation of viscous flow is the no-slip condition, as discussed in Section 1.6. Thus, for example, for two-dimensional external or internal flows, the x and y components of velocity (u and v) are set to zero at the stationary wall to satisfy the no-slip condition. Other boundary conditions that must be appropriately specified involve inlets, outlets, far-field, wall gradients, etc. It is important to not only select the correct physical boundary condition for the problem, but also to correctly implement this boundary condition into the numerical simulation. A.6 Turbulence Models The majority of flows of engineering interest are turbulent. One might think that with the ability to solve the unsteady, three-dimensional Navier-Stokes equations, a modern CFD code would be able to calculate turbulent flows with ease. Unfortunately, this is not the case. Turbulence consists of an indeterminate number of chaotically moving fluid masses, which are loosely termed “eddies” (see Sec. 8.3). Eddies exist in an extremely large range of sizes (called scales) and their motion has an extremely large Appendix A ■ Computational Fluid Dynamics 747 number of time scales. For air flowing at 10 m/s in a 0.1-m-diameter duct, the size of the largest eddies are the order of the duct diameter (0.1 m) and the smallest eddies are the order of 10−5 m. The time scale of the motion varies from about 1 s to about 10−4 s. To fully capture the motion in a 10-m length of this duct would require a computation with 1011 grid points (control volumes) and at least 105 time steps—well beyond the capability of the current largest, fastest computers. Fortunately, in most circumstances engineers have no need for that level of detail; information on the average velocities, pressure, and shear stress is sufficient. As explained in Sec. 8.3, we tackle the problem by breaking the velocity and other flow variables into the sum of an average value and a fluctuating component t+T 1 u = u + u′ where u = T ∫ udt and u′ = 0 t To develop a set of equations for computing the average quantities, we substitute into the continuity and Navier-Stokes equations and average the equations (steady in the mean, incompressible flow is assumed here). The result is ∂u ∂v ∂w + + =0 ∂x ∂y ∂z ∂p ∂u ∂u ∂u ∂ ∂u ∂ ∂u ρ u +v +w =− + μ − ρu′u′ + μ − ρu′v′ + ( ∂x ) ( ) ( ) ∂y ∂z ∂x ∂x ∂x ∂y ∂y ∂ ∂u μ − ρu′w′ ) ( ∂z ∂z ∂p ∂v ∂v ∂u ∂ ∂v ∂ ∂v ρ(u +v + w ) = − + (μ − ρu′v′ ) + (μ − ρv′v′ ) + ∂x ∂y ∂z ∂y ∂x ∂x ∂y ∂y ∂ ∂v μ − ρv′w′ ) ∂z ( ∂z ∂p ∂w ∂w ∂w ∂ ∂w ∂ ∂w ρ(u +v + w )= − + (μ − ρu′w′ ) + (μ − ρv′w′ ) + ∂x ∂y ∂z ∂w ∂x ∂x ∂y ∂y ∂ ∂w μ − ρw′w′ ) ∂z ( ∂z (A.3) (A.4a) (A.4b) (A.4c) The last three equations are called the Reynolds averaged Navier-Stokes (RANS) equations. The nine extra terms on the right side (−ρu′v′, −ρv′ 2, etc.) are called Reynolds Stresses (see Sec. 8.3). At first glance it would seem that the problem of computing average quantities in turbulent flow is straightforward; all we need to do is solve the RANS equations. A second look reveals an enormous problem—we have introduced six new unknowns into the problem (the Reynolds stresses) but we have no new equations! This is the closure problem of turbulent flow. Just like it is beyond current capability to solve the full unsteady Navier-Stokes equations for a turbulent flow, it is beyond current capability to calculate the Reynolds stresses. (A little reflection will reveal both are really the same task.) The only path forward is to construct and use a turbulence model; a mathematical relationship for the estimation of turbulence properties (in this case, the Reynolds stresses) in terms of available characteristics of the flow field. These models are typically based on plausible assumptions about the physics of turbulence and always require input from experimental measurements. Many different turbulence models have been proposed and most CFD codes permit the user to choose among several (typically about 10 choices are available). Many models are based on the idea of an eddy viscosity, η, defined by a relationship of the form (see Sec. 8.3.2 and Eq. 8.27): ∂u ∂v τturb, xy = −ρu′v′ = η( + ) ∂y ∂x Note that this is not a physical law and the eddy viscosity is not a fluid property; it varies strongly with flow conditions. The core assumption in the model is that, at any point, the stress is proportional to the mean strain rate. 748 Appendix A ■ Computational Fluid Dynamics The eddy viscosity is typically constructed from a relationship based on dimensional reasoning η = ρ × length scale × velocity scale Eddy viscosity (and other) turbulence models are characterized by the number of differential equations that must be solved to determine the eddy viscosity or the Reynolds stresses. The most common example of a zero equation model is the mixing length model (see Eq. 8.28) where length scale = ℓm and velocity scale = ℓm | | ∂u ∂y In this model, the mixing length must be prescribed in terms of the dimensions of the problem and/or the characteristics of the mean flow. The most widely used turbulence model in engineering CFD is the two equation k − ε model, 1 where k is the turbulence kinetic energy (k = 2 (u′ 2 + v′ 2 + w′ 2 )) and ε is the dissipation of d(loss) mechanical energy caused by the turbulence, ε = ρ . The eddy viscosity is given by dt η = ρCη k2 ε where Cη is a constant determined from experimental data. k and ε are determined by partial differential equations of the form u ∂ϕ ∂ϕ ∂ϕ +v +w = Diffusion of ϕ + Production of ϕ − Destruction of ϕ ∂x ∂y ∂z where ϕ stands for k or ε. The three terms on the right side involve averages of second- and thirdorder products of fluctuating quantities, which themselves must be modeled. There are a number of other turbulence models available. Another popular two-equation model uses the “specific dissipation,” ω(ω = ε/k), in place of ε. In addition to eddy viscosity models, there are Reynolds stress models that solve equations, both differential and algebraic, for the Reynolds stresses themselves. There are also large eddy simulation models, which actually compute the details of the larger eddies and model the smaller scales. Most CFD codes use a special method for applying boundary conditions for turbulent flow near a wall. In order to apply the no-slip condition at the wall, an extremely fine grid would be required to properly resolve the viscous sublayer (see Sec. 8.3.3). To avoid this requirement, socalled wall functions are used. The first grid point away from the wall is assumed to lie in the overlap region and Eq. 8.30 is used to determine the wall shear stress 1 up 1 τw 2 (ρ) yp τw 2 = 2.44 ln( [ ] ) + 5.0. ν ρ Whatever turbulence model is used, the model’s form is somewhat ad hoc, several constants like Cη require input from experimental data, and the model works well for some cases but not for others. It should be apparent that turbulence modeling keeps CFD from being an exact science and demands both intelligent choices and a certain degree of skepticism from the user of CFD. See (Ref. 9) for further information about turbulence models and wall functions. A.7 Solving the Equations After the grid is defined, the partial differential equations discretized (including any PDEs used to model turbulence), and the boundary conditions specified, a set of algebraic equations remains to be solved. If the flow is modeled as inviscid, with either velocity potential or stream function as the unknown, the algebraic equations will be linear. For viscous flows, modeled by either the Navier-Stokes equations or by the boundary layer approximation, the equations will be nonlinear and, in order to solve them, the equations must be linearized in some way. The Appendix A ■ Computational Fluid Dynamics 749 nonlinearity is typically handled by iteration, requiring that the linearized equations be solved at each iteration step. In most instances, the number of equations to be solved will be quite large, between three and ten equations for each grid point or volume, with thousands of grid points.1 No matter how large, the system of equations can be written in matrix notation as AX = B where A is the n × n matrix of coefficients, n is the number of equations, X is a column vector of unknowns such as u, v, k, and B is a column vector of known values derived from problem geometry, boundary conditions, fluid properties, etc. In general, there are two methods for solving systems of linear algebraic equations: direct methods and iterative methods. A direct method solves the equations in a definite number of steps, which can be determined from the number of equations. Probably the most well-known direct method is Gaussian elimination. In this method, the matrix is converted to an upper triangular matrix (all zeros below the diagonal) by successively adding or subtracting multiples of rows to eliminate the coefficients below the diagonal. After the upper triangular matrix is constructed, the equations can be solved “bottom-to-top” by back substitution. Another well-known direct method is to calculate the inverse of matrix A, A−1. Finding the inverse of a matrix requires several steps. Many “canned” software programs and even spreadsheets can be used to determine the inverse of matrices of moderate size. The equations are then solved by X = A −1 B. There are many other methods for direct solution of systems of equations; consult a linear algebra textbook (Ref. 10) for more information. Direct solution methods are seldom used in CFD codes. Although the deterministic character of a direct method may be appealing, these methods require a very large number of arithmetic operations and are subject to problems such as the propagation of round-off error. The matrices that arise in CFD are generally very large and very sparse, meaning that any row or column contains a large number of zeros.2 A direct method would waste a considerable amount of computer time and storage by manipulating zeros. Iterative methods are generally more efficient in CFD codes. An advantage of iterative methods is that the iterations required because the discretized equations are nonlinear are done simultaneously with the iterations for solving the system of equations. In general, the steps in an iterative method for solving AX = B are: ■ ■ ■ Rearrange the equations so that for each, one element from X appears by itself on the left side of the equation and the right side is populated with other elements of X, elements of A, and elements of B. Different iterative methods will produce different right-side expressions. Compute a new value (a new iterate) for each element of X. If the change between iterates is sufficiently small by some measure, an approximate solution has been computed. If the change is not sufficiently small, repeat the two preceding steps until it is sufficiently small (known as convergence) or the solution process fails. In Example A.1, we used the Gauss-Sidel (iterative) method where the most recently available values of the unknowns are used on the right side. The Jacobi method substitutes new values on the right side only after a complete pass through the entire system of equations. An obvious question is “How do we know when to stop iterating?”. With a small system of equations, such as might be solved in a spreadsheet, the iterations are usually stopped when the largest change in any unknown is less than some value, say 0.01%. With a spreadsheet that visually displays the results of each iteration, the iterations may be continued until no change in any variable is observed, which means the change is smaller than the number of digits being displayed. In a large system, such as occurs in a CFD calculation, a more complex method is used. At any step in the 1 An exception is the boundary element method, where the number of elements, hence equations, is typically the order of tens or hundreds. Again, the boundary element method is an exception. 2 750 Appendix A ■ Computational Fluid Dynamics iteration, the current set of variables, say ui, vi, and ki can be substituted into the discretized equations, in this case, the x-momentum, y-momentum, and turbulence model equations. Because the solution is not exact, each equation at each point will not be satisfied; there will be a residual ri (for, say, the x-momentum equation) rx,i = (right side of x-momentum equation) – (left side of x-momentum n equation). A total residual is calculated from Rx = ∑ ∣rx,i ∣. The total residual for each equation is i=1 tracked and when the largest of them drops to an acceptable level, the iteration is considered to have converged. It is not unusual for this process to require hundreds or even thousands of iterations. Sometimes, it is desirable to speed up an iteration to achieve convergence more rapidly, at other times an iteration must be slowed to avoid instability. This is done using a process called relaxation. At any iteration step, we can write ui = λu new + (1 − λ)u old i i where ui is any variable and λ is a relaxation parameter. If 0 ≤ λ ≤ 1 the calculation is said to be underrelaxed and if 1 ≤ λ ≤ 2 the calculation is said to be overrelaxed. The choice of λ for a given simulation is usually made by trial and error. Finally, some CFD codes use a multigrid method to speed convergence. In this method, iterations are done alternately on finer and coarser grids. For details, see Refs. (7 and 8). A.8 Some Unexpected Complications The process of developing a CFD code requires many choices. The developer must choose a discretization method, a type of grid, turbulence model(s), and more. For example, the seemingly ∂φ simple task of approximating the first derivative, , requires choosing between a forward difference, ∂x a backward difference, a central difference, a three-point forward difference (see Example A.1), a three-point backward difference—the list goes on. None of these approximations is intrinsically correct; if one takes the limit as Δx goes to zero, all of the approximations become the (exact) first derivative. Clearly the best choices are those that produce the most accurate flow predictions. Different choices yield different sets of equations to solve and one should expect to get different answers. Usually this would be acceptable because the answers are approximations anyway. Sometimes different choices have more serious consequences; some seemingly correct approximations may not work at all. “Not working” can take on several forms: failure to converge; instability; producing nonphysical results. Computational Fluid Dynamics has proven to be especially sensitive to the selection of proper formulations, and several “obvious” approaches have proven to be totally unsatisfactory. This section discusses some of the special formulations that were discovered only by trial and error. Conservative Form. The (Reynolds averaged) Navier-Stokes equations (Eqs. 6.127a,b,c; A.4a,b,c) are said to be in nonconservative form. Recall that these equations were derived by applying Newton’s Second Law to a fluid element (see Fig. 6.11). If we add the continuity equation in the form of Eq. 6.31 to each of Eqs. 127a,b,c and gather terms, the equations can be written (shown here for the x-component only) as ∂ (ρu) ∂p ∂ ∂ ∂ ∂ 2u ∂ 2u ∂ 2u + (ρu2 ) + (ρuv) + (ρuw) = − + μ( 2 + 2 + 2 ) ∂t ∂x ∂y ∂z ∂x ∂x ∂y ∂z (A.5) Equation A.5 is the x-momentum equation in conservative form and is the form of the equation we would get by applying the momentum equation to an infinitesimal control volume instead of an infinitesimal fluid element. Experience has shown that in some discretization schemes the conservative form of the equations must be used or the calculations will not converge due to failure to conserve mass, momentum, and energy throughout the flow field. Staggered Grid. In CFD, the purpose of the grid is to define a number of discrete locations where the flow variables (u, p, and so on) are to be computed. In the finite difference method, these locations are typically at the intersection of the grid lines, and in the finite volume method, Appendix A ■ Computational Fluid Dynamics 751 the obvious location is the center of the finite volume. In Example A.1, the grid points were defined by the intersection of the grid lines, and the stream function was calculated at those points. When solving the more complicated Navier-Stokes equations, one’s natural inclination would be to define and calculate all variables (u, v, w, p, k, ε) at the same location. Like many other “obvious” choices in CFD, collocating all variables at a single point can lead to serious difficulties; in this case calculating an unrealistic pressure field. The solution to this difficulty is to use a staggered grid. Consider the two-dimensional grid layout shown in Fig. A.5. The finite difference method approximates the PDEs at point P; the finite volume method considers mass and momentum balances for the control volume surrounding point P. In a staggered grid, scalar quantities such as pressure and turbulence kinetic energy are defined and computed at point P, the “center” of the grid element, but velocities are defined and computed at the “halfway” locations PE and PN (i.e., at the faces of the surrounding control volume). Note that the fluid velocities at points PW and PS are associated with points (volumes) W and S. Many CFD codes use the staggered grid but others choose to use higher-order difference approximations on a collocated grid to avoid challenges that arise for the staggered-grid such as accurate modeling of complex geometries. Upwind Differencing. Consider the transport of some property φ out from the control volume shown in Fig. A.5 (φ could be temperature, momentum, turbulence kinetic energy, or any other property that is carried by the flow). Also, suppose the flow is from left to right. The obvious φP + φE discrete expression for the flow of φ (per unit area) out from the control volume is ρuP( ) 2 where the term in parentheses approximates the value of φ at the control volume face. This formulation is called a central difference. At times this “obvious” formulation leads to difficulties (“wiggles” in the solution). Imagine that the flow is very strong. It stands to reason that the value of φ upstream (upwind) of the control volume face would have a stronger influence on the value of φ φP + φE there than the value downstream. This effect is simulated by using only φP in place of . On 2 the other hand, if the flow is from right to left (uP negative), then only φE would be used. This strategy is called upwind differencing or donor cell differencing.3 It effectively eliminates “wiggles” in the flow solution. Upwind differencing does have a negative side. It tends to smooth out or “smear” sharp changes in the flow, an effect called false diffusion. This can be mostly eliminated by blending an upwind difference and a central difference, based on the strength of the flow transport. Pressure-Velocity Coupling. The (Reynolds averaged) Navier-Stokes equations together with the continuity equation (Eqs. 6.27, 6.127a,b,c, A.3, A.4a,b,c) are coupled partial differential equations; in particular, all components of velocity appear in all of the equations. After the equations are discretized, each variable is computed from a different equation. The obvious approach is N v defined here P Control volume vP PN uP P W p, k, ε defined here PE PW PS E uP defined here S ■ Figure A.5 Staggered grid arrangement 3 One of the early developers of upwind differencing has called it the “pigpen method,” noting that the odor is much stronger downwind from a pig pen than it is upwind! 752 Appendix A ■ Computational Fluid Dynamics to compute u using the x-momentum equation, v using the y-momentum equation, and w using the z-momentum equation. This leaves the continuity equation available to compute p—but pressure does not appear in the continuity equation!4 This knotty problem is usually solved by a pressure correction algorithm. This is an iterative method that generally proceeds as follows. A pressure field is guessed and the momentum equations are solved for the velocities. The calculated velocities do not satisfy the continuity equation. A pressure correction is computed from a pressure correction equation derived from the continuity and momentum equations. The pressure is updated with the pressure correction, and the new pressure field is used to compute a new velocity field, which, hopefully, comes closer to satisfying the continuity equation. The iteration process continues until both the velocity field and the pressure field converge. Most of the algorithms for performing these computations are based on or are similar to the SIMPLE algorithm (an ironic name) of Patankar and Spalding (Ref. 11). One helpful feature of this method is that the sequence of iterations accounts for both the pressure correction and the nonlinearity of the momentum equations simultaneously. A.9 Verification and Validation Verification and validation of the simulation are critical steps in the CFD process. This is a necessary requirement for CFD, particularly since it is possible to have a converged solution that is nonphysical. Figure A.6 shows the streamlines for viscous flow past a circular cylinder at a given instant after it was impulsively started from rest. The lower half of the figure represents the results of a finite difference calculation; the upper half of the figure represents the photograph from an experiment of the same flow situation. It is clear that the numerical and experimental results agree quite well. For any CFD simulation, several levels of testing need to be accomplished before one can have confidence in the solution. The most important verification to be performed is a grid convergence study. In its simplest form, it consists of proving that further refinement of the grid (i.e., increasing the number of grid points) does not alter the final solution. When this has been achieved, you have a grid-independent solution. Other verification factors that need to be investigated include the suitability of the convergence criterion, whether the time step is adequate for the time scale of the problem, and comparison of CFD solutions to existing data, at least for baseline cases. Even when using a commercial CFD code that has been validated on many problems in the past, the CFD practitioner still needs to verify the results through such measures as grid-dependence testing. ■ Figure A.6 Streamlines for flow past a circular cylinder at a short time after the flow was impulsively started. The upper half is a photograph from a flow visualization experiment. The lower half is from a finite difference calculation. (From Ref. 12, used by permission.) 4 If the flow is compressible, density ρ does appear in the continuity equation and p is related to ρ and temperature. The energy equation must be included to calculate temperature, and continuity can be used to compute density and pressure then determined. Here we are considering incompressible flow. Appendix A ■ Computational Fluid Dynamics A.10 753 Application of CFD In the early stages of CFD, research and development was primarily driven by the aerospace and nuclear industries. Today, CFD is still used as a research tool, but it also has found a place in industry as a design tool. There is now a wide variety of industries that make at least some use of CFD, including automotive, power generation, HVAC, naval, civil, chemical, biological, and others. Industries are using CFD as an added engineering tool that complements experimental and theoretical work in fluid dynamics. A.10.1 Advantages of CFD VIDEO VA.3 tornado simulation There are many advantages to using CFD for simulation of fluid flow. One of the most important advantages is the realizable savings in time and cost for engineering design. In the past, coming up with a new engineering design often meant somewhat of a trial-and-error method of building and testing multiple prototypes prior to finalizing the design. With CFD, many of the issues dealing with fluid flow can be flushed out prior to building the actual prototype. This translates to a significant savings in time and cost. It should be noted that CFD is not meant to replace experimental testing, but rather to work in conjunction with it. Experimental testing is still a necessary component of engineering design. Other advantages include the ability of CFD to: (1) provide flow information in regions that would be difficult to test experimentally, (2) simulate real flow conditions, (3) conduct large parametric tests on new designs in a shorter time, and (4) enhance visualization of complex flow phenomena. A good example of the advantages of CFD is shown in Figure A.7. Researchers used a type of CFD approach called “large-eddy simulation” or LES to simulate the fluid dynamics of a ■ Figure A.7 Results from a large-eddy simulation showing the visual appearance of the debris and funnel cloud from a simulated medium swirl F3-F4 tornado. The funnel cloud is translating at 15 m/s and is ingesting 1-mm-diameter “sand” from the surface as it encounters a debris field. Please visit the book web site to access a full animation of this tornado simulation. (Photographs and animation courtesy of Dr. David Lewellen (Ref. 13) and Paul Lewellen, West Virginia University.) 754 Appendix A ■ Computational Fluid Dynamics tornado as it encounters a debris field and begins to pick up sand-sized particles. A full animation of this tornado simulation can be accessed by visiting WileyPLUS. The motivation for this work is to investigate whether there are significant differences in the fluid mechanics when debris particles are present. Historically it has been difficult to obtain comprehensive experimental data throughout a tornado, so CFD is helping to shine some light on the complex fluid dynamics involved in such a flow. A.10.2 A Warning An important point that a beginning CFD user should understand is that one cannot treat the computer as a “magic black box” when performing flow simulations. It is quite possible to obtain a fully converged solution for the CFD simulation, but this is no guarantee that the results are physically correct. This is why it is important to have a good understanding of the flow physics and how they are modeled. Any numerical technique (including those discussed above), no matter how simple in concept, contains many hidden subtleties and potential problems. For example, it may seem reasonable that a finer grid would ensure a more accurate numerical solution. While this may be true, it is not always so straightforward; a variety of stability or convergence problems may occur. In such cases the numerical “solution” obtained may exhibit unreasonable oscillations or the numerical result may “diverge” to an unreasonable (and incorrect) result. Other problems that may arise include (but are not limited to): (1) difficulties in dealing with the nonlinear terms of the Navier– Stokes equations, (2) difficulties in modeling turbulent flows, (3) convergence issues, (4) difficulties in obtaining a quality grid for complex geometries, and (5) managing resources, both time and computational, for complex problems such as unsteady three-dimensional flows. A.11 Summary In CFD, there are many different numerical schemes, grid techniques, etc. Each has advantages and disadvantages. A great deal of care must be used in obtaining approximate numerical solutions to the governing equations of fluid motion. The process is not as simple as the often-heard “just let the computer do it.” Remember that CFD is a tool and as such needs to be used appropriately to produce meaningful results and, in the wrong hands, it may do more harm than good. The general field of computational fluid dynamics, in which computers and numerical methods are combined to solve fluid flow problems, represents an extremely important subject area in advanced fluid mechanics. Considerable progress has been made in the past relatively few years, but much remains to be done. The reader is encouraged to consult some of the available literature. References 1. Baker, A. J., Finite Element Computational Fluid Mechanics, McGraw-Hill, New York, 1983. 2. Carey, G. F., and Oden, J. T., Finite Elements: Fluid Mechanics, Prentice-Hall, Englewood Cliffs, N. J., 1986. 3. Brebbia, C. A., and Dominguez, J., Boundary Elements: An Introductory Course, McGraw-Hill, New York, 1989. 4. Moran, J., An Introduction to Theoretical and Computational Aerodynamics, Wiley, New York, 1984. 5. Anderson, J. D., Computational Fluid Dynamics: The Basics with Applications, McGraw-Hill, New York, 1995. 6. Tannehill, J. C., Anderson, D. A., and Pletcher, R. H., Computational Fluid Mechanics and Heat Transfer, 2nd Ed., Taylor and Francis, Washington, D.C., 1997. 7. Versteeg, H., and Malalasekera, W., An Introduction to Computational Fluid Dynamics: The Finite Volume Method, 2nd Ed., Pearson Prentice Hall, 2007. 8. Tu, J., Yeoh, G. H., and Liu, C., Computational Fluid Dynamics: A Practical Approach, 2nd Ed., Butterworth-Heinmann, Paperback, 2012. 9. Wilcox, D. C., Turbulence Modeling for CFD, DCW Industries, Inc., La Canada, California, 1994. 10. Hoffman, K., and Kunze, R., Linear Algebra, 2nd Ed., Paperback, Pearson Prentice Hall, 2009. Appendix A ■ Computational Fluid Dynamics 755 11. Patankar, S. V., and Spalding, D. B., A Calculation Procedure for Heat, Mass, and Momentum Transfer in Three-Dimensional Parabolic Flows, International Journal of Heat and Mass Transfer, Vol. 15, p. 1787, 1972. 12. Hall, E. J., and Pletcher, R. H., Simulation of Time Dependent, Compressible Viscous Flow Using Central and Upwind-Biased Finite-Difference Techniques, Technical Report HTL-52, CFD-22, College of Engineering, lowa State University, 1990. 13. Lewellen, D. C., Gong, B., and Lewellen, W. S., Effects of Debris on Near-Surface Tornado Dynamics, 22nd Conference on Severe Local Storms, Paper 15.5, American Meteorological Society, 2004. B Physical Properties of Fluids 4.0 2.0 1.0 8 6 Glycerin Castor oil 4 2 1 × 10–1 8 6 SAE 30W oil 4 SAE 10W-30 oil μ , Dynamic viscosity, N • s/m2 2 SAE 10W oil –2 1 × 10 8 6 4 Kerosene 2 Mercury 1 × 10–3 8 6 Octane Carbon tetrachloride Heptane 4 Water 2 1 × 10–4 8 6 4 2 Methane Carbon dioxide Air Helium 1 × 10–5 8 6 –20 Hydrogen 0 20 40 60 80 100 Temperature, °C ■ Figure B.1 Dynamic (absolute) viscosity of common fluids as a function of temperature. To convert to BG units of lb ∙ s/ft2 multiply N ∙ sm2 by 2.089 × 10 −2. (Curves from R. W. Fox and A. T. McDonald, Introduction to Fluid Mechanics, 3rd Ed., Wiley, New York, 1985. Used by permission.) 756 120 Appendix B ■ Physical Properties of Fluids 1 × 10–2 8 6 4 1 × 10–3 8 6 SA E 10 in er yc Gl oil W 30 E SA 2 W 4 –3 0 oi l Helium ν , Kinematic viscosity, m2/s 2 Hydrogen 1 × 10–4 8 6 SA E 4 10 W Methane oil Air 2 1 × 10–5 8 6 oxide Carbon di 4 Kero sene 2 1 × 10–6 8 6 Wat Octa ne er Heptane 4 Carbon te trachlor ide 2 Mercury 1 × 10–7 8 –20 0 20 40 60 80 100 120 Temperature, °C ■ Figure B.2 Kinematic viscosity of common fluids (at atmospheric pressure) as a function of temperature. To convert to BG units of ft2s multiply m2s by 10.76. (Curves from R. W. Fox and A. T. McDonald, Introduction to Fluid Mechanics, 3rd Ed., Wiley, New York, 1985. Used by permission.) 757 758 Appendix B ■ Physical Properties of Fluids ■ Table B.1 Physical Properties of Water (BG/EE Units)a Temperature (°F) Density, ρ (slugsft3 )b Specific Weightc, γ (lbft3 ) 32 40 50 60 70 80 90 100 120 140 160 180 200 212 1.940 1.940 1.940 1.938 1.936 1.934 1.931 1.927 1.918 1.908 1.896 1.883 1.869 1.860 62.42 62.43 62.41 62.37 62.30 62.22 62.11 62.00 61.71 61.38 61.00 60.58 60.12 59.83 Dynamic Viscosity, μ (lb ∙ sft2 ) 3.732 E 3.228 E 2.730 E 2.344 E 2.037 E 1.791 E 1.500 E 1.423 E 1.164 E 9.743 E 8.315 E 7.207 E 6.342 E 5.886 E −5 −5 −5 −5 −5 −5 −5 −5 −5 −6 −6 −6 −6 −6 Kinematic Viscosity, ν (ft2s) 1.924 E 1.664 E 1.407 E 1.210 E 1.052 E 9.262 E 8.233 E 7.383 E 6.067 E 5.106 E 4.385 E 3.827 E 3.393 E 3.165 E −5 −5 −5 −5 −5 −6 −6 −6 −6 −6 −6 −6 −6 −6 Surface Tensiond, σ (lbft) 5.18 E 5.13 E 5.09 E 5.03 E 4.97 E 4.91 E 4.86 E 4.79 E 4.67 E 4.53 E 4.40 E 4.26 E 4.12 E 4.04 E −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 Vapor Pressure, pυ [lbin.2 (abs)] 8.854 E 1.217 E 1.781 E 2.563 E 3.631 E 5.069 E 6.979 E 9.493 E 1.692 E 2.888 E 4.736 E 7.507 E 1.152 E 1.469 E −2 −1 −1 −1 −1 −1 −1 −1 +0 +0 +0 +0 +1 +1 Speed of Sounde, c (fts) 4603 4672 4748 4814 4871 4819 4960 4995 5049 5091 5101 5195 5089 5062 a Based on data from Handbook of Chemistry and Physics, 69th Ed., CRC Press, 1988. Where necessary, values obtained by interpolation. To obtain EE units (lbm/ft3) multiply by 32.174. c Density and specific weight are related through the equation γ = ρg. For this table, g = 32.174 fts2. d In contact with air. e Based on data from R. D. Blevins, Applied Fluid Dynamics Handbook, Van Nostrand Reinhold Co., Inc., New York, 1984. b ■ Table B.2 Physical Properties of Water (SI Units)a Temperature (°C) Density, ρ (kgm3 ) Specific Weightb, γ (kNm3 ) 0 5 10 20 30 40 50 60 70 80 90 100 999.9 1000.0 999.7 998.2 995.7 992.2 988.1 983.2 977.8 971.8 965.3 958.4 9.806 9.807 9.804 9.789 9.765 9.731 9.690 9.642 9.589 9.530 9.467 9.399 a Dynamic Viscosity, μ (N ∙ sm2 ) 1.787 E 1.519 E 1.307 E 1.002 E 7.975 E 6.529 E 5.468 E 4.665 E 4.042 E 3.547 E 3.147 E 2.818 E −3 −3 −3 −3 −4 −4 −4 −4 −4 −4 −4 −4 Kinematic Viscosity, ν (m2s) 1.787 E 1.519 E 1.307 E 1.004 E 8.009 E 6.580 E 5.534 E 4.745 E 4.134 E 3.650 E 3.260 E 2.940 E −6 −6 −6 −6 −7 −7 −7 −7 −7 −7 −7 −7 Surface Tensionc, σ (Nm) 7.56 E 7.49 E 7.42 E 7.28 E 7.12 E 6.96 E 6.79 E 6.62 E 6.44 E 6.26 E 6.08 E 5.89 E Based on data from Handbook of Chemistry and Physics, 69th Ed., CRC Press, 1988. Density and specific weight are related through the equation γ = ρg. For this table, g = 9.807 ms2. c In contact with air. d Based on data from R. D. Blevins, Applied Fluid Dynamics Handbook, Van Nostrand Reinhold Co., Inc., New York, 1984. b −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 Vapor Pressure, pυ [Nm2 (abs)] 6.105 E 8.722 E 1.228 E 2.338 E 4.243 E 7.376 E 1.233 E 1.992 E 3.116 E 4.734 E 7.010 E 1.013 E +2 +2 +3 +3 +3 +3 +4 +4 +4 +4 +4 +5 Speed of Soundd, c (ms) 1403 1427 1447 1481 1507 1526 1541 1552 1555 1555 1550 1543 Appendix B ■ Physical Properties of Fluids 759 ■ Table B.3 Physical Properties of Air at Standard Atmospheric Pressure (BG/EE Units)a Temperature (°F) −40 −20 0 10 20 30 40 50 60 70 80 90 100 120 140 160 180 200 300 400 500 750 1000 1500 a Density, ρ (slugsft3 )b 2.939 E 2.805 E 2.683 E 2.626 E 2.571 E 2.519 E 2.469 E 2.420 E 2.373 E 2.329 E 2.286 E 2.244 E 2.204 E 2.128 E 2.057 E 1.990 E 1.928 E 1.870 E 1.624 E 1.435 E 1.285 E 1.020 E 8.445 E 6.291 E −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −3 −4 −4 Specific Weightc, γ (lbft3 ) 9.456 E 9.026 E 8.633 E 8.449 E 8.273 E 8.104 E 7.942 E 7.786 E 7.636 E 7.492 E 7.353 E 7.219 E 7.090 E 6.846 E 6.617 E 6.404 E 6.204 E 6.016 E 5.224 E 4.616 E 4.135 E 3.280 E 2.717 E 2.024 E −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 −2 Dynamic Viscosity, μ (lb ∙ sft2 ) 3.29 E 3.34 E 3.38 E 3.44 E 3.50 E 3.58 E 3.60 E 3.68 E 3.75 E 3.82 E 3.86 E 3.90 E 3.94 E 4.02 E 4.13 E 4.22 E 4.34 E 4.49 E 4.97 E 5.24 E 5.80 E 6.81 E 7.85 E 9.50 E −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 −7 Kinematic Viscosity, ν (ft2s) 1.12 E 1.19 E 1.26 E 1.31 E 1.36 E 1.42 E 1.46 E 1.52 E 1.58 E 1.64 E 1.69 E 1.74 E 1.79 E 1.89 E 2.01 E 2.12 E 2.25 E 2.40 E 3.06 E 3.65 E 4.51 E 6.68 E 9.30 E 1.51 E Based on data from R. D. Blevins, Applied Fluid Dynamics Handbook, Van Nostrand Reinhold Co., Inc., New York, 1984. To obtain EE units (lbm/ft3) multiply by 32.174. c Density and specific weight are related through the equation γ = ρg. For this table g = 32.174 fts2. b −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −4 −3 Specific Heat Ratio, k (—) Speed of Sound, c (fts) 1.401 1.401 1.401 1.401 1.401 1.401 1.401 1.401 1.401 1.401 1.400 1.400 1.400 1.400 1.399 1.399 1.399 1.398 1.394 1.389 1.383 1.367 1.351 1.329 1004 1028 1051 1062 1074 1085 1096 1106 1117 1128 1138 1149 1159 1180 1200 1220 1239 1258 1348 1431 1509 1685 1839 2114 760 Appendix B ■ Physical Properties of Fluids ■ Table B.4 Physical Properties of Air at Standard Atmospheric Pressure (SI Units)a Temperature (°C) −40 −20 0 5 10 15 20 25 30 40 50 60 70 80 90 100 200 300 400 500 1000 a Density, ρ (kgm3 ) 1.514 1.395 1.292 1.269 1.247 1.225 1.204 1.184 1.165 1.127 1.109 1.060 1.029 0.9996 0.9721 0.9461 0.7461 0.6159 0.5243 0.4565 0.2772 Specific Weightb, γ (Nm3 ) 14.85 13.68 12.67 12.45 12.23 12.01 11.81 11.61 11.43 11.05 10.88 10.40 10.09 9.803 9.533 9.278 7.317 6.040 5.142 4.477 2.719 Dynamic Viscosity, μ (N ∙ sm2 ) 1.57 E 1.63 E 1.71 E 1.73 E 1.76 E 1.80 E 1.82 E 1.85 E 1.86 E 1.87 E 1.95 E 1.97 E 2.03 E 2.07 E 2.14 E 2.17 E 2.53 E 2.98 E 3.32 E 3.64 E 5.04 E −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 Kinematic Viscosity, ν (m2s) 1.04 E 1.17 E 1.32 E 1.36 E 1.41 E 1.47 E 1.51 E 1.56 E 1.60 E 1.66 E 1.76 E 1.86 E 1.97 E 2.07 E 2.20 E 2.29 E 3.39 E 4.84 E 6.34 E 7.97 E 1.82 E −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −5 −4 Specific Heat Ratio, k (—) Speed of Sound, c (ms) 1.401 1.401 1.401 1.401 1.401 1.401 1.401 1.401 1.400 1.400 1.400 1.399 1.399 1.399 1.398 1.397 1.390 1.379 1.368 1.357 1.321 306.2 319.1 331.4 334.4 337.4 340.4 343.3 346.3 349.1 354.7 360.3 365.7 371.2 376.6 381.7 386.9 434.5 476.3 514.1 548.8 694.8 Based on data from R. D. Blevins, Applied Fluid Dynamics Handbook, Van Nostrand Reinhold Co., Inc., New York, 1984. Density and specific weight are related through the equation γ = ρg. For this table g = 9.807 ms2. b C Properties of the U.S. Standard Atmosphere ■ Table C.1 Properties of the U.S. Standard Atmosphere (BG/EE Units)a Altitude (ft) −5,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000 60,000 70,000 80,000 90,000 100,000 150,000 200,000 250,000 Temperature (°F) 76.84 59.00 41.17 23.36 5.55 −12.26 −30.05 −47.83 −65.61 −69.70 −69.70 −69.70 −69.70 −67.42 −61.98 −56.54 −51.10 19.40 −19.78 −88.77 Acceleration of Gravity, g (fts2 ) Pressure, p [lbin.2 (abs)] 32.189 32.174 32.159 32.143 32.128 32.112 32.097 32.082 32.066 32.051 32.036 32.020 31.990 31.959 31.929 31.897 31.868 31.717 31.566 31.415 17.554 14.696 12.228 10.108 8.297 6.759 5.461 4.373 3.468 2.730 2.149 1.692 1.049 0.651 0.406 0.255 0.162 0.020 0.003 0.000 Density, ρ (slugsft3 )b 2.745 E 2.377 E 2.048 E 1.756 E 1.496 E 1.267 E 1.066 E 8.907 E 7.382 E 5.873 E 4.623 E 3.639 E 2.256 E 1.392 E 8.571 E 5.610 E 3.318 E 3.658 E 5.328 E 6.458 E −3 −3 −3 −3 −3 −3 −3 −4 −4 −4 −4 −4 −4 −4 −5 −5 −5 −6 −7 −8 Dynamic Viscosity, μ (lb ∙ sft2 ) 3.836 E − 7 3.737 E − 7 3.637 E − 7 3.534 E − 7 3.430 E − 7 3.324 E − 7 3.217 E − 7 3.107 E − 7 2.995 E − 7 2.969 E − 7 2.969 E − 7 2.969 E − 7 2.969 E − 7 2.984 E − 7 3.018 E − 7 3.052 E − 7 3.087 E − 7 3.511 E − 7 3.279 E − 7 2.846 E − 7 a Data abridged from U.S. Standard Atmosphere, 1976, U.S. Government Printing Office, Washington, D.C. To obtain EE units (lbm/ft3), multiply by 32.174 b 761 762 Appendix C ■ Properties of the U.S. Standard Atmosphere ■ Table C.2 Properties of the U.S. Standard Atmosphere (SI Units)a Altitude (m) −1,000 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 15,000 20,000 25,000 30,000 40,000 50,000 60,000 70,000 80,000 a Temperature (°C) 21.50 15.00 8.50 2.00 −4.49 −10.98 −17.47 −23.96 −30.45 −36.94 −43.42 −49.90 −56.50 −56.50 −51.60 −46.64 −22.80 −2.50 −26.13 −53.57 −74.51 Acceleration of Gravity, g (ms2 ) 9.810 9.807 9.804 9.801 9.797 9.794 9.791 9.788 9.785 9.782 9.779 9.776 9.761 9.745 9.730 9.715 9.684 9.654 9.624 9.594 9.564 Pressure, p [Nm2 (abs)] 1.139 E 1.013 E 8.988 E 7.950 E 7.012 E 6.166 E 5.405 E 4.722 E 4.111 E 3.565 E 3.080 E 2.650 E 1.211 E 5.529 E 2.549 E 1.197 E 2.871 E 7.978 E 2.196 E 5.221 E 1.052 E +5 +5 +4 +4 +4 +4 +4 +4 +4 +4 +4 +4 +4 +3 +3 +3 +2 +1 +1 +0 +0 Density, ρ (kgm3 ) 1.347 E 1.225 E 1.112 E 1.007 E 9.093 E 8.194 E 7.364 E 6.601 E 5.900 E 5.258 E 4.671 E 4.135 E 1.948 E 8.891 E 4.008 E 1.841 E 3.996 E 1.027 E 3.097 E 8.283 E 1.846 E Data abridged from U.S. Standard Atmosphere, 1976, U.S. Government Printing Office, Washington, D.C. +0 +0 +0 +0 −1 −1 −1 −1 −1 −1 −1 −1 −1 −2 −2 −2 −3 −3 −4 −5 −5 Dynamic Viscosity, μ (N ∙ sm2 ) 1.821 E − 5 1.789 E − 5 1.758 E − 5 1.726 E − 5 1.694 E − 5 1.661 E − 5 1.628 E − 5 1.595 E − 5 1.561 E − 5 1.527 E − 5 1.493 E − 5 1.458 E − 5 1.422 E − 5 1.422 E − 5 1.448 E − 5 1.475 E − 5 1.601 E − 5 1.704 E − 5 1.584 E − 5 1.438 E − 5 1.321 E − 5 D 0.1 1.0 Compressible Flow Functions for an Ideal Gas with k = 1.4 0.5 1.0 5.0 9.0 0.9 T __ T0 0.8 p0 0.6 p __ 7.0 6.0 A ___ p0 T __ T0 ρ __ ρ0 8.0 ρ __ ρ0 0.7 p __ 10.0 10.0 A* A ___ 0.5 5.0 0.4 4.0 0.3 3.0 0.2 2.0 0.1 1.0 0.0 0.1 0.5 1.0 Ma 5.0 A* 0.0 10.0 ■ Figure D.I Isentropic flow of an ideal gas with k = 1.4. (Graph provided by Dr. Bruce A. Reichert.) 763 764 Appendix D ■ Compressible Flow Functions for an Ideal Gas with k = 1.4 ■ Table D.I Isentropic Flow Functions for an Ideal Gas with k = 1.4 Ma TT0 pp0 ρρ0 AA* Ma TT0 pp0 ρρ0 0.0000 0.0500 0.1000 0.1500 0.2000 0.2500 0.3000 0.3500 0.4000 0.4500 0.5000 0.5500 0.6000 0.6500 0.7000 0.7500 0.8000 0.8500 0.9000 0.9500 1.0000 1.0500 1.1000 1.1500 1.2000 1.2500 1.3000 1.3500 1.4000 1.4500 1.5000 1.5500 1.6000 1.6500 1.7000 1.7500 1.8000 1.8500 1.9000 1.9500 2.0000 1.0000 0.9995 0.9980 0.9955 0.9921 0.9877 0.9823 0.9761 0.9690 0.9611 0.9524 0.9430 0.9328 0.9221 0.9107 0.8989 0.8865 0.8737 0.8606 0.8471 0.8333 0.8193 0.8052 0.7908 0.7764 0.7619 0.7474 0.7329 0.7184 0.7040 0.6897 0.6754 0.6614 0.6475 0.6337 0.6202 0.6068 0.5936 0.5807 0.5680 0.5556 1.0000 0.9983 0.9930 0.9844 0.9725 0.9575 0.9395 0.9188 0.8956 0.8703 0.8430 0.8142 0.7840 0.7528 0.7209 0.6886 0.6560 0.6235 0.5913 0.5595 0.5283 0.4979 0.4684 0.4398 0.4124 0.3861 0.3609 0.3370 0.3142 0.2927 0.2724 0.2533 0.2353 0.2184 0.2026 0.1878 0.1740 0.1612 0.1492 0.1381 0.1278 1.0000 0.9988 0.9950 0.9888 0.9803 0.9694 0.9564 0.9413 0.9243 0.9055 0.8852 0.8634 0.8405 0.8164 0.7916 0.7660 0.7400 0.7136 0.6870 0.6604 0.6339 0.6077 0.5817 0.5562 0.5311 0.5067 0.4829 0.4598 0.4374 0.4158 0.3950 0.3750 0.3557 0.3373 0.3197 0.3029 0.2868 0.2715 0.2570 0.2432 0.2300 ∞ 11.5914 5.8218 3.9103 2.9635 2.4027 2.0351 1.7780 1.5901 1.4487 1.3398 1.2549 1.1882 1.1356 1.0944 1.0624 1.0382 1.0207 1.0089 1.0021 1.0000 1.0020 1.0079 1.0175 1.0304 1.0468 1.0663 1.0890 1.1149 1.1440 1.1762 1.2116 1.2502 1.2922 1.3376 1.3865 1.4390 1.4952 1.5553 1.6193 1.6875 2.0500 2.1000 2.1500 2.2000 2.2500 2.3000 2.3500 2.4000 2.4500 2.5000 2.5500 2.6000 2.6500 2.7000 2.7500 2.8000 2.8500 2.9000 2.9500 3.0000 3.1000 3.2000 3.3000 3.4000 3.5000 3.6000 3.7000 3.8000 3.9000 4.0000 4.5000 5.0000 5.5000 6.0000 6.5000 7.0000 7.5000 8.0000 8.5000 9.0000 10.0000 0.5433 0.5313 0.5196 0.5081 0.4969 0.4859 0.4752 0.4647 0.4544 0.4444 0.4347 0.4252 0.4159 0.4068 0.3980 0.3894 0.3810 0.3729 0.3649 0.3571 0.3422 0.3281 0.3147 0.3019 0.2899 0.2784 0.2675 0.2572 0.2474 0.2381 0.1980 0.1667 0.1418 0.1220 0.1058 0.0926 0.0816 0.0725 0.0647 0.0581 0.0476 0.1182 0.1094 0.1011 0.0935 0.0865 0.0800 0.0740 0.0684 0.0633 0.0585 0.0542 0.0501 0.0464 0.0430 0.0398 0.0368 0.0341 0.0317 0.0293 0.0272 0.0234 0.0202 0.0175 0.0151 0.0131 0.0114 0.0099 0.0086 0.0075 0.0066 0.0035 0.0019 0.0011 0.0006 0.0004 0.0002 0.0002 0.0001 0.0001 0.0000 0.0000 0.2176 0.2058 0.1946 0.1841 0.1740 0.1646 0.1556 0.1472 0.1392 0.1317 0.1246 0.1179 0.1115 0.1056 0.0999 0.0946 0.0896 0.0849 0.0804 0.0762 0.0685 0.0617 0.0555 0.0501 0.0452 0.0409 0.0370 0.0335 0.0304 0.0277 0.0174 0.0113 0.0076 0.0052 0.0036 0.0026 0.0019 0.0014 0.0011 0.0008 0.0005 AA* 1.7600 1.8369 1.9185 2.0050 2.0964 2.1931 2.2953 2.4031 2.5168 2.6367 2.7630 2.8960 3.0359 3.1830 3.3377 3.5001 3.6707 3.8498 4.0376 4.2346 4.6573 5.1210 5.6286 6.1837 6.7896 7.4501 8.1691 8.9506 9.7990 10.7188 16.5622 25.0000 36.8690 53.1798 75.1343 104.1429 141.8415 190.1094 251.0862 327.1893 535.9375 765 Appendix D ■ Compressible Flow Functions for an Ideal Gas with k = 1.4 1.0 1.0 2.0 3.0 4.0 5.0 40.0 p0,y ____ 0.8 p0,x 30.0 p0,y ____ px May p px y ___ 0.6 T Tx y ___ p px May y ___ p p0, x 20.0 0, y _____ ρy ___ ρx V Vy x ___ 0.4 p px 0,y ____ 10.0 0.2 ρ ___y ρx & V x ___ Vy T y ___ Tx 0.0 1.0 2.0 3.0 Max 4.0 0.0 5.0 ■ Figure D.S Normal shock flow of an ideal gas with k = 1.4. (Graph provided by Dr. Bruce A. Reichert.) 766 Appendix D ■ Compressible Flow Functions for an Ideal Gas with k = 1.4 ■ Table D.S Normal Shock Functions for an Ideal Gas with k = 1.4 Max 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50 1.55 1.60 1.65 1.70 1.75 1.80 1.85 1.90 1.95 2.00 2.05 2.10 2.15 2.20 2.25 2.30 2.35 2.40 2.45 2.50 2.55 2.60 2.65 2.70 2.75 2.80 2.85 2.90 2.95 3.00 May pypx 1.0000 1.0000 0.9531 1.1196 0.9118 1.2450 0.8750 1.3763 0.8422 1.5133 0.8126 1.6563 0.7860 1.8050 0.7618 1.9596 0.7397 2.1200 0.7196 2.2863 0.7011 2.4583 0.6841 2.6363 0.6684 2.8200 0.6540 3.0096 0.6405 3.2050 0.6281 3.4063 0.6165 3.6133 0.6057 3.8263 0.5956 4.0450 0.5862 4.2696 0.5774 4.5000 0.5691 4.7363 0.5613 4.9783 0.5540 5.2263 0.5471 5.4800 0.5406 5.7396 0.5344 6.0050 0.5286 6.2763 0.5231 6.5533 0.5179 6.8363 0.5130 7.1250 0.5083 7.4196 0.5039 7.7200 0.4996 8.0262 0.4956 8.3383 0.4918 8.6562 0.4882 8.9800 0.4847 9.3096 0.4814 9.6450 0.4782 9.9862 0.4752 10.3333 TyTx ρyρx = VxVy p0yp0x Max 1.0000 1.0328 1.0649 1.0966 1.1280 1.1594 1.1909 1.2226 1.2547 1.2872 1.3202 1.3538 1.3880 1.4228 1.4583 1.4946 1.5316 1.5693 1.6079 1.6473 1.6875 1.7285 1.7705 1.8132 1.8569 1.9014 1.9468 1.9931 2.0403 2.0885 2.1375 2.1875 2.2383 2.2902 2.3429 2.3966 2.4512 2.5067 2.5632 2.6206 2.6790 1.0000 1.0840 1.1691 1.2550 1.3416 1.4286 1.5157 1.6028 1.6897 1.7761 1.8621 1.9473 2.0317 2.1152 2.1977 2.2791 2.3592 2.4381 2.5157 2.5919 2.6667 2.7400 2.8119 2.8823 2.9512 3.0186 3.0845 3.1490 3.2119 3.2733 3.3333 3.3919 3.4490 3.5047 3.5590 3.6119 3.6636 3.7139 3.7629 3.8106 3.8571 1.0000 0.9999 0.9989 0.9967 0.9928 0.9871 0.9794 0.9697 0.9582 0.9448 0.9298 0.9132 0.8952 0.8760 0.8557 0.8346 0.8127 0.7902 0.7674 0.7442 0.7209 0.6975 0.6742 0.6511 0.6281 0.6055 0.5833 0.5615 0.5401 0.5193 0.4990 0.4793 0.4601 0.4416 0.4236 0.4062 0.3895 0.3733 0.3577 0.3428 0.3283 3.05 3.10 3.15 3.20 3.25 3.30 3.35 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 3.95 4.00 4.10 4.20 4.30 4.40 4.50 4.60 4.70 4.80 4.90 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.00 ∞ May pypx TyTx ρyρx = VxVy p0yp0x 0.4723 0.4695 0.4669 0.4643 0.4619 0.4596 0.4573 0.4552 0.4531 0.4512 0.4492 0.4474 0.4456 0.4439 0.4423 0.4407 0.4392 0.4377 0.4363 0.4350 0.4324 0.4299 0.4277 0.4255 0.4236 0.4217 0.4199 0.4183 0.4167 0.4152 0.4090 0.4042 0.4004 0.3974 0.3949 0.3929 0.3912 0.3898 0.3886 0.3876 0.3780 10.6863 11.0450 11.4096 11.7800 12.1563 12.5383 12.9263 13.3200 13.7196 14.1250 14.5363 14.9533 15.3763 15.8050 16.2396 16.6800 17.1263 17.5783 18.0363 18.5000 19.4450 20.4133 21.4050 22.4200 23.4583 24.5200 25.6050 26.7133 27.8450 29.0000 35.1250 41.8333 49.1250 57.0000 65.4583 74.5000 84.1250 94.3333 105.1250 116.5000 ∞ 2.7383 2.7986 2.8598 2.9220 2.9851 3.0492 3.1142 3.1802 3.2472 3.3151 3.3839 3.4537 3.5245 3.5962 3.6689 3.7426 3.8172 3.8928 3.9694 4.0469 4.2048 4.3666 4.5322 4.7017 4.8751 5.0523 5.2334 5.4184 5.6073 5.8000 6.8218 7.9406 9.1564 10.4694 11.8795 13.3867 14.9911 16.6927 18.4915 20.3875 ∞ 3.9025 3.9466 3.9896 4.0315 4.0723 4.1120 4.1507 4.1884 4.2251 4.2609 4.2957 4.3296 4.3627 4.3949 4.4262 4.4568 4.4866 4.5156 4.5439 4.5714 4.6245 4.6749 4.7229 4.7685 4.8119 4.8532 4.8926 4.9301 4.9659 5.0000 5.1489 5.2683 5.3651 5.4444 5.5102 5.5652 5.6117 5.6512 5.6850 5.7143 6.0000 0.3145 0.3012 0.2885 0.2762 0.2645 0.2533 0.2425 0.2322 0.2224 0.2129 0.2039 0.1953 0.1871 0.1792 0.1717 0.1645 0.1576 0.1510 0.1448 0.1388 0.1276 0.1173 0.1080 0.0995 0.0917 0.0846 0.0781 0.0721 0.0667 0.0617 0.0424 0.0297 0.0211 0.0154 0.0113 0.0085 0.0064 0.0050 0.0039 0.0030 0.0000 Appendix D ■ Compressible Flow Functions for an Ideal Gas with k = 1.4 0.1 5.0 0.5 1.0 5.0 10.0 5.0 p 0 ___ p0* 4.0 4.0 3.0 3.0 2.0 2.0 V ___ V* 1.0 f(l* – l) _______ 1.0 D p __ p* 0.0 0.1 0.5 1.0 Ma T __ T* 5.0 ■ Figure D.F Fanno flow of an ideal gas with k = 1.4. (Graph provided by Dr. Bruce A. Reichert.) 0.0 10.0 767 768 Appendix D ■ Compressible Flow Functions for an Ideal Gas with k = 1.4 ■ Table D.F Fanno Flow Functions for an Ideal Gas with k = 1.4 Ma f (ℓ *−ℓ )D 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50 1.55 ∞ 280.0203 66.9216 27.9320 14.5333 8.4834 5.2993 3.4525 2.3085 1.5664 1.0691 0.7281 0.4908 0.3246 0.2081 0.1273 0.0723 0.0363 0.0145 0.0033 0.0000 0.0027 0.0099 0.0205 0.0336 0.0486 0.0648 0.0820 0.0997 0.1178 0.1361 0.1543 pp* ∞ 21.9034 10.9435 7.2866 5.4554 4.3546 3.6191 3.0922 2.6958 2.3865 2.1381 1.9341 1.7634 1.6183 1.4935 1.3848 1.2893 1.2047 1.1291 1.0613 1.0000 0.9443 0.8936 0.8471 0.8044 0.7649 0.7285 0.6947 0.6632 0.6339 0.6065 0.5808 TT * p0p*0 1.2000 ∞ 1.1994 11.5914 1.1976 5.8218 1.1946 3.9103 1.1905 2.9635 1.1852 2.4027 1.1788 2.0351 1.1713 1.7780 1.1628 1.5901 1.1533 1.4487 1.1429 1.3398 1.1315 1.2549 1.1194 1.1882 1.1065 1.1356 1.0929 1.0944 1.0787 1.0624 1.0638 1.0382 1.0485 1.0207 1.0327 1.0089 1.0165 1.0021 1.0000 1.0000 0.9832 1.0020 0.9662 1.0079 0.9490 1.0175 0.9317 1.0304 0.9143 1.0468 0.8969 1.0663 0.8794 1.0890 0.8621 1.1149 0.8448 1.1440 0.8276 1.1762 0.8105 1.2116 ρρ* = V*V Ma f (ℓ*−ℓ )D pp* TT * p0p*0 ρρ* = V*V ∞ 18.2620 9.1378 6.0995 4.5826 3.6742 3.0702 2.6400 2.3184 2.0693 1.8708 1.7092 1.5753 1.4626 1.3665 1.2838 1.2119 1.1489 1.0934 1.0440 1.0000 0.9605 0.9249 0.8926 0.8633 0.8367 0.8123 0.7899 0.7693 0.7503 0.7328 0.7166 1.60 1.65 1.70 1.75 1.80 1.85 1.90 1.95 2.00 2.10 2.20 2.30 2.40 2.50 2.60 2.70 2.80 2.90 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 ∞ 0.1724 0.1902 0.2078 0.2250 0.2419 0.2583 0.2743 0.2899 0.3050 0.3339 0.3609 0.3862 0.4099 0.4320 0.4526 0.4718 0.4898 0.5065 0.5222 0.5864 0.6331 0.6676 0.6938 0.7140 0.7299 0.7425 0.7528 0.7612 0.7682 0.7740 0.7790 0.8215 0.5568 0.5342 0.5130 0.4929 0.4741 0.4562 0.4394 0.4234 0.4082 0.3802 0.3549 0.3320 0.3111 0.2921 0.2747 0.2588 0.2441 0.2307 0.2182 0.1685 0.1336 0.1083 0.0894 0.0750 0.0638 0.0548 0.0476 0.0417 0.0369 0.0328 0.0293 0.0000 0.7937 0.7770 0.7605 0.7442 0.7282 0.7124 0.6969 0.6816 0.6667 0.6376 0.6098 0.5831 0.5576 0.5333 0.5102 0.4882 0.4673 0.4474 0.4286 0.3478 0.2857 0.2376 0.2000 0.1702 0.1463 0.1270 0.1111 0.0980 0.0870 0.0777 0.0698 0.0000 1.2502 1.2922 1.3376 1.3865 1.4390 1.4952 1.5553 1.6193 1.6875 1.8369 2.0050 2.1931 2.4031 2.6367 2.8960 3.1830 3.5001 3.8498 4.2346 6.7896 10.7188 16.5622 25.0000 36.8690 53.1798 75.1343 104.1429 141.8415 190.1094 251.0862 327.1893 ∞ 0.7016 0.6876 0.6745 0.6624 0.6511 0.6404 0.6305 0.6211 0.6124 0.5963 0.5821 0.5694 0.5580 0.5477 0.5385 0.5301 0.5225 0.5155 0.5092 0.4845 0.4677 0.4559 0.4472 0.4407 0.4357 0.4317 0.4286 0.4260 0.4239 0.4221 0.4207 0.4082 Appendix D ■ Compressible Flow Functions for an Ideal Gas with k = 1.4 0.1 2.5 0.5 1.0 5.0 10.0 2.5 p0 ____ p0,* 2.0 2.0 * ρ V ___ ___ ρ & * 1.5 1.5 V 1.0 1.0 T0 ____ T0,* 0.5 0.5 T __ T* p __ p* 0.0 0.1 0.5 1.0 Ma 5.0 0.0 10.0 ■ Figure D.R Rayleigh flow of an ideal gas with k = 1.4. (Graph provided by Dr. Bruce A. Reichert.) 769 770 Appendix D ■ Compressible Flow Functions for an Ideal Gas with k = 1.4 ■ Table D.R Raleigh Flow Functions for an Ideal Gas with k = 1.4 Ma T0T *0 pp* TT * p0p*0 ρρ* = V*V Ma T0 T *0 pp* TT * p0p*0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.84515 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50 1.55 0.0000 0.0105 0.0370 0.0741 0.1181 0.1667 0.2181 0.2713 0.3257 0.3807 0.4360 0.4915 0.5471 0.6029 0.6587 0.7148 0.7710 0.8221 0.8276 0.8845 0.9420 1.0000 1.0587 1.1181 1.1784 1.2395 1.3017 1.3649 1.4292 1.4947 1.5615 1.6296 1.6990 2.4000 2.3916 2.3669 2.3267 2.2727 2.2069 2.1314 2.0487 1.9608 1.8699 1.7778 1.6860 1.5957 1.5080 1.4235 1.3427 1.2658 1.2000 1.1931 1.1246 1.0603 1.0000 0.9436 0.8909 0.8417 0.7958 0.7529 0.7130 0.6758 0.6410 0.6086 0.5783 0.5500 0.0000 0.0143 0.0560 0.1218 0.2066 0.3044 0.4089 0.5141 0.6151 0.7080 0.7901 0.8599 0.9167 0.9608 0.9929 1.0140 1.0255 1.0286 1.0285 1.0245 1.0146 1.0000 0.9816 0.9603 0.9369 0.9118 0.8858 0.8592 0.8323 0.8054 0.7787 0.7525 0.7268 1.2679 1.2657 1.2591 1.2486 1.2346 1.2177 1.1985 1.1779 1.1566 1.1351 1.1141 1.0940 1.0753 1.0582 1.0431 1.0301 1.0193 1.0116 1.0109 1.0049 1.0012 1.0000 1.0012 1.0049 1.0109 1.0194 1.0303 1.0437 1.0594 1.0777 1.0983 1.1215 1.1473 ∞ 167.2500 42.2500 19.1019 11.0000 7.2500 5.2130 3.9847 3.1875 2.6409 2.2500 1.9607 1.7407 1.5695 1.4337 1.3241 1.2344 1.1667 1.1600 1.0977 1.0450 1.0000 0.9613 0.9277 0.8984 0.8727 0.8500 0.8299 0.8120 0.7959 0.7815 0.7685 0.7568 1.60 1.65 1.70 1.75 1.80 1.85 1.90 1.95 2.00 2.10 2.20 2.30 2.40 2.50 2.60 2.70 2.80 2.90 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 10 ∞ 1.7699 1.8422 1.9161 1.9915 2.0685 2.1472 2.2275 2.3096 2.3934 2.5663 2.7466 2.9344 3.1299 3.3333 3.5448 3.7644 3.9923 4.2286 4.4734 5.8276 7.4050 9.2111 11.2500 13.5244 16.0362 18.7870 21.7778 25.0095 28.4827 32.1979 36.1557 44.8000 ∞ 0.5236 0.4988 0.4756 0.4539 0.4335 0.4144 0.3964 0.3795 0.3636 0.3345 0.3086 0.2855 0.2648 0.2462 0.2294 0.2142 0.2004 0.1879 0.1765 0.1322 0.1026 0.0818 0.0667 0.0554 0.0467 0.0399 0.0345 0.0301 0.0265 0.0235 0.0210 0.0170 0.0000 0.7017 0.6774 0.6538 0.6310 0.6089 0.5877 0.5673 0.5477 0.5289 0.4936 0.4611 0.4312 0.4038 0.3787 0.3556 0.3344 0.3149 0.2969 0.2803 0.2142 0.1683 0.1354 0.1111 0.0927 0.0785 0.0673 0.0583 0.0509 0.0449 0.0399 0.0356 0.0290 0.0000 1.1756 1.2066 1.2402 1.2767 1.3159 1.3581 1.4033 1.4516 1.5031 1.6162 1.7434 1.8860 2.0451 2.2218 2.4177 2.6343 2.8731 3.1359 3.4245 5.3280 8.2268 12.5023 18.6339 27.2113 38.9459 54.6830 75.4138 102.2875 136.6235 179.9236 233.8840 381.6149 ∞ ρρ* = V*V 0.7461 0.7364 0.7275 0.7194 0.7119 0.7051 0.6988 0.6929 0.6875 0.6778 0.6694 0.6621 0.6557 0.6500 0.6450 0.6405 0.6365 0.6329 0.6296 0.6173 0.6094 0.6039 0.6000 0.5971 0.5949 0.5932 0.5918 0.5907 0.5898 0.5891 0.5885 0.5875 0.5833 mil 1.98E5 rd μ Å 5.029E10 rd 5.029E6 rd Rod (rd) mm 5029 rd mm 1.829E3 fath mm 1000 m mm 914.4 yd mm 304.8 ft mm 25.4 in. mm 10 cm cm 502.9 rd cm 182.9 fath cm 100 m cm 9.144 yd cm 30.48 ft cm 2.54 in. in. 198 rd in. 72 fath in. 39.37 m in. 36 yd in. 12 ft ft 5280 mi ft 16.5 rd ft 6 fath ft 3.281 m ft 3 yd Angstrom (Å) Micron (μ) Mil Millimeter (mm) Centimeter (cm) Inch (in.) Foot (ft) cm Yard (yd) yd 2026 mi yd 1760 mi yd 5.5 rd yd 2 fath yd 1.094 m Meter (m) m 1852 mi m 1609 mi m 5.029 rd m 1.829 fath ( 393.7 mil ) = 1.143 cm. 450.0 mil μ mil ft mm cm in. Int’l nautical Å 7.291E7 mi 1.852E6 mi 1.852E5 mi 7.291E4 mi 6076.12 mi 1.852E13 mi 1.852E9 mile (mi) mi μ cm in. mil mm Statute mile Å 6.336E7 mi 1.609E6 mi 1.609E5 mi 6.336E4 mi 1.609E13 mi 1.609E9 (mi) mi mil 7.2E4 fath μ Å 1.829E10 fath 1.829E6 fath mil 2.54E5 m Fathom (fath) μ 1.0E6 m mil 3.6E5 yd μ Å 9.144E9 yd 9.144E5 yd Yard (yd) Å 1.0E10 m mil 1.2E5 ft μ Å 3.048E9 ft 3.048E5 ft Foot (ft) Meter (m) mil 1000 in. μ 2.54E4 in. Å 2.54E8 in. Inch (in.) mil 393.7 cm Å 1.0E8 cm Centimeter (cm) μ 1.0E4 cm Å 1.0E7 mm Millimeter (mm) μ 25.4 mil mil 39.37 mm Å 2.54E5 mil Mil Solution: Locate the column for the smaller unit (mil) and the rwo for the larger unit (cm). If unsure as to the smaller unit, try the second option if your first guess doesn’t work. The intersection gives the conversion factor. In this case dividing by the conversion factor (equal to unity) cancels the mils, so Example: Convert 450.0 mils to centimeters. μ 1000 mm 4 Å 10 μ Micron (μ) Angstrom (Å) Table E.2 Length conversion factors Å Fathom (fath) fath 1012 mi fath 879.7 mi fath 2.75 rd in cm mm Mil μ Å Hour (hr) Mil Rod (rd) rd 368.3 mi rd 320 mi μ min 60 hr mil 393.7 cm Statute mile Int’l nautical (mi) mile (mi) st mi I.N. mi mm Minute (min) sec 60 min Second (sec) Hour (hr) Second (sec) sec 3600 hr 1.151 Minute (min) Table E.1 Time conversion factors E Comprehensive Table of Conversion Factors 771 Cubic inch (in3) ) mil in3 ml mi3 ) 3.381E4 fl oz m3 fl oz yd3 fl oz ft3 2.585E4 957.5 fl oz 128 gal fl oz 33.81 1 fl oz 32 qt qt 1 1000 qt m3 U.S. fluid ounce (fl oz) U.S. liquid quart (qt) Liter (1) m2 mi2 264.2 202.0 7.48 gal m3 gal yd3 gal ft3 m2 4047 acre 2.589E6 Square meter (m2) yd2 mi2 ft3 m3 ft3 yd3 yd2 m2 35.31 27 1.196 yd2 4840 acre 3.098E6 Square yard (yd2) U.S. liquid gallon (gal) Cubic foot (ft3) 1 gal ft3 1.101E12 3 1.472E11 mi3 mi mi3 1 m3 1 yd3 1 ft3 4.166E12 764.6 qt yd3 1.057E3 807.9 28.3 1 gal acre mi2 3.7854 640 Acre qt ft3 qt gal 29.92 4 1.057 Square statute mile (mi2) Table E.3 Area conversion factors in3 1.409E14 fl oz 4.404E11 qt mi3 mi3 mi3 Cubic inch (in3) 2.544E14 in3 m3 in3 yd3 Table E.4 Volume conversion factors ( Milliliter (ml) cubic centimeter 4.166E15 Cubic statute mile (mi3) 6.102E4 ml m3 1.0E6 Cubic meter (m3) 4.666E4 ml yd3 7.646E5 Cubic yard (yd3) in3 ft3 1728 ml ft3 2.83E4 Cubic foot (ft3) in3 gal 231 in3 1 in3 qt ml gal 3785.4 U.S. liquid gallon (gal) 61.02 1000 Liter (1) ml 1 946.4 57.75 in3 1.805 fl oz ml qt mil 29.574 fl oz 16.39 U.S. fluid quart (qt) U.S. fluid ounce (fl oz) ( Milliliter (ml) cubic centimeter ft2 mi2 ft2 m2 yd3 mi3 Cubic yard (yd3) 5.452E9 yd3 m3 ft2 yd2 1.308 9 10.76 ft2 4.356E4 acre 2.788E7 Square foot (ft2) 929.0 in2 ft2 Cubic meter (m3) m3 mi3 8361 in2 yd2 Square foot (ft2) Square inch (in2) cm2 cm2 in2 ft2 yd2 Square centimeter (cm2) Square yard (yd2) cm2 m2 Square meter (m2) Acre cm2 Cubic statute mile (mi3) 6.452 1.0E4 cm2 4.047E7 acre in2 m2 4.166E9 144 1296 1550 Square centimeter (cm2) in2 cm2 Square statute 2 2 2.589E10 mi mi2 mile (mi ) in2 6.272E6 acre 4.014E9 Square inch (in2) 772 Appendix E ■ Comprehensive Table of Conversion Factors pdl s ton 9.964E3 Long ton pdl l ton Poundal (pdl) 7.207E4 Table E.6 Force conversion factors Dyne 9.807E3 Metric ton N s ton N lb Newton (N) N l ton N m ton 8896 4.448 9964 6.435E4 dyne s ton pdl lb dyne l ton 8.896E3 Short ton 32.174 dyne lb pdl N 9808 4.448E5 Pound (lb) (avdp.) 7.233 dyne pdl 7.094E4 m ton m ton E5 dyne N dyne 1.383E4 pdl Newton (N) Poundal (pdl) Dyne lb s ton Pound (lb) (avdp.) lb l ton lb m ton 2240 2205 2000 Slug s ton l ton s ton m ton kg slug Short ton 1.12 1.102 14.594 Kilogram (kg) m ton l ton lbm kg lbm slug Metric ton 1.016 2.2046 32.174 Pound (lbm) (avdp.) Table E.5 Mass conversion factors oz lbm oz kg oz slug Long ton 16 35.27 514.8 Ounce (oz) (avdp.) g lbm g kg g slug g 28.35 oz 453.6 1000 1.459E4 Gram (g) gr lbm gr kg gr slug gr 15.43 g gr 437.5 oz 7000 1.543E4 2.252E5 Grain (gr) Grain (gr) Gram (g) Ounce (oz) (avdp.) Pound (lbm) (avdp.) Kilogram (kg) Slug Appendix E ■ Comprehensive Table of Conversion Factors 773 fpm 5 in./sec fpm 54.68 km/hr in. sec km hr (fpm) ft min cm s cm/s 100 m/s fpm 196.9 m/s 101.3 cm/s 51.44 knot cm/s 44.7 mph fpm knot fpm 88 mph cm/s 30.48 fps cm/s 27.78 km/hr cm/s 2.54 in./sec cm/s 1.667 m/min psi 14.696 atm bar 1.0132 atm m min m/min 60 m/s m/min 30.87 knot m/min 26.82 mph m/min 18.29 fps m/min 16.67 km/hr m/min 1.524 in./sec in. H2O (39.2°F) mm Hg (0°C) in. sec in./sec 39.37 m/s in./sec 20.26 knot in./sec 17.6 mph in./sec 12 fps km hr km/hr 3.6 m/s km/hr 1.852 knot km/hr 1.609 mph km/hr 1.097 fps statute mile hr (mph) ft sec (fps) mph 2.237 m/s Torr mm Hg Knots knot 1.944 m/s 1.00 Pa Torr 1.00 m s Pa mm Hg Pa in H2O Pa ft H2O Pa in Hg 133.3 249.1 3386 mm Hg Torr 1868 in. H2O in. H2O fps 3.281 m/s fps 1.688 knot 1.868 Torr in. Hg Pa bar 6.895E3 Pa psi 1.0E5 1.013E5 Pa atm 2989 25.4 Torr psi Torr bar Torr atm ( ) Pascal N m2 mm Hg Torr 304.8 ft H2O ft H2O mm Hg in. Hg 22.42 25.4 mph 1.151 knot in. H2O ft H2O fps 1.47 mph 12 ft H2O in. H2O 13.60 in. Hg in. Hg 51.71 mm Hg in. H2O ft H O 51.71 27.68 2.307 psi2 psi psi in. Hg 2.036 psi 1.133 750.1 760 Torr in. H2O mm Hg ft H O 401.5 750.1 33.46 bar2 bar bar in. H2O mm Hg ft H2O 760 406.8 33.90 atm atm atm ft H2O (39.2°F) in. Hg 29.53 bar in. Hg 29.92 atm in. Hg (32°F) in./sec 10.94 km/hr psi 14.504 bar psi Bar Table E.8 Velocity conversion factors m s Knots statute mile hr (mph) (fps) fpm 60 fps fpm 3.281 m/min m min ft sec fpm 1.969 cm/s cm s (fpm) ft min atm Table E.7 Pressure conversion factors 2 2 2 2 2 2 2 2 10 dyne/cm Pa dyne/cm Torr 1000 dyne/cm mm Hg 1333 dyne/cm in. H2O 2491 dyne/cm ft H2O 2.989E4 dyne/cm in. Hg 3.386E4 dyne/cm psi 6.895E4 dyne/cm bar 1.0E6 dyne/cm atm 1.032E6 dyne cm2 2 dyne cm2 ( ) Pascal N m2 Torr mm Hg (0°C) in. H2O (39.2°F) ft H2O (39.2°F) in. Hg (32°F) psi Bar atm 774 Appendix E ■ Comprehensive Table of Conversion Factors 775 Appendix E ■ Comprehensive Table of Conversion Factors Table E.9 Absolute viscosity conversion factors lb . sec in2 lb . sec ft2 lbm in. · sec in2 sec ft2 sec m2 s N.s m2 kg m. s ( cm · s ) Centipoise 386.1 6895 6895 6.895E4 6.895E6 lb · sec/ft2 lb · sec/in2 lbm/in. · sec lbm/ft · sec lb · sec/in2 lb · sec/in2 N · s/m2 lb · sec/in2 kg/m · s lb · sec/in2 poise lb · sec/in2 centipoise lb · sec/in2 2.681 32.174 47.88 47.88 478.8 4.788E4 lbm/in. · sec lb · sec/ft2 lbm/ft · sec lb · sec/ft2 N · s/m2 lb · sec/ft2 kg/m · s lb · sec/ft2 poise lb · sec/ft2 centipoise lb · sec/ft2 12 17.86 17.86 178.6 1.786E4 4633 poise lbm/ft · sec kg/m · s N · s/m2 centipoise lbm/in. · sec lbm/in. · sec lbm/in. · sec lbm/in. · sec lbm/in. · sec 100 centistoke stoke 645.2 6.452 centistoke in2/sec stoke in2/sec 9.29E4 929 centistoke ft2/sec stoke ft2/sec 1.0E6 10,000 929 10.76 centistoke m2/s stoke m2/s in2/sec m2/s ft2/sec m2/s Centistoke Stoke (cm2/s) in2 sec ft2 sec 1.488 1.488 14.88 1488 N ·s/m2 lbm/ft · sec kg/m · s lbm/ft · sec poise lbm/ft · sec centipoise lbm/ft ·sec 1 in2/sec 144 2 ft /sec kg/m · s N · s/m2 10 poise N · s/m2 10 poise kg/m·s 1000 centipoise N · s/m2 1000 centipoise kg/m ·s 100 m2 s centipoise poise lb · sec in2 lb · sec ft2 lbm in. · sec lbm ft. · sec N·s m 2 kg m·s Poise g ( cm · s ) Centipoise Table E.10 Kinematic viscosity conversion factors Table E.11 Temperature Conversions Convert from Poise g 144 Centistoke Stoke (cm2/s) lbm ft · sec to Celsius (°C) Fahrenheit (°F) TF = 1.8 TC + 32° Fahrenheit (°F) Celsius (°C) TC = (5/9) TF – 32° TK = (5.556E – 1) TR Rankine (°R) Kelvin (K) Kelvin (K) Rankine (°R) TR = (1.8) TK Celsius (°C) Kelvin (K) TK = TC + 273.15 Fahrenheit (°F) Rankine (°R) TR = TF + 459.67 Btu/hr kW 3600 Btu/hr Btu/sec 60 g · cal/min gm · cal/s 252.0 g · cal/min Btu/min 1.069E4 g · cal/min hp 1.432E4 g · cal/min kW 1.512E4 g · cal/min Btu/sec g · cal min ft· lb/min g · cal/s 778 ft· lb/min Btu/min 3.3E4 ft· lb/min hp 4.425E4 ft· lb/min kW 4.666E4 ft· lb/min Btu/sec ft· lb min Btu/hr Btu/min Btu hr 3413.4 Btu/hr hp 2542 60 Btu/hr g · cal/s 14.29 Btu/hr ft · lb/sec 185.2 4.629 Btu/hr W 3.4134 19.44 g · cal/min W ft · lb/min W g · cal/min ft · lb/sec 14.34 44.25 ft· lb/min ft· lb/sec g · cal/min Btu/hr ft· lb/min Btu/hr m3 s 60 4.20 12.96 ft· lb/min g · cal/min 3.086 Table E.13 Power conversion factors Btu sec Kilowatt (kW) Horsepower (hp) Btu min g · cal s ft· lb sec Watt (W) Btu hr g · cal min ft · lb min W g · cal/s W ft · lb/sec W Btu/sec W kW Watt (W) 1054 1000 W 745.7 hp 17.57 W Btu/min 4.184 1.356 35.61 ft3/sec m3/s ft3 sec ft · lb sec ft · lb/sec Btu/sec 778 ft · lb/sec kW 737.6 ft · lb/sec hp 550 ft · lb/sec Btu/min 12.96 ft · lb/sec g · cal/s l/s l/s m3/s ft3/sec 3.086 28.32 1000 Liter s ft3/min l/s g · cal s g · cal/s Btu/sec 252.0 g · cal/s kW 238.7 g · cal/s hp 6.416E5 g · cal/s Btu/min 4.200 2.119 60 ft3/min ft3/sec 2119 ft3/min m3/s ft3 min Table E.12 Volume flow rate conversion factors gpm ft3/min gpm l/s 60 Btu min Btu/min Btu/sec Btu/min kW 56.83 Btu/min hp 42.44 7.481 15.85 449 gpm ft3/sec 15,850 gpm m3/s gal min (gpm) hp cm3 s gal min (gpm) ft3 min Liter s ft3 sec m3 s Kilowatt (kW) hp kW 1.054 Btu/sec Btu/sec kW Horsepower (hp) 1.414 1.341 63.09 cm3/s gpm 471.9 cm3/s ft3/min 1000 cm3/s l/s 2.832E4 cm3/s ft3/sec 60E6 cm3/s m3/s cm3 s Btu sec 776 Appendix E ■ Comprehensive Table of Conversion Factors 1400 ft· lb/lbm kcal/kg 4184 J/kg kcal/kg ft· lb/lbm Btu/slug 2.5020E4 ft· lb/slug Btu/lbm 4.504E4 ft· lb/slug g · cal/g 4.504E4 ft· lb/slug kcal/kg ft · lb slug erg/g Btu/slug 2.324E7 erg/g Btu/lbm 4.184E7 erg/g g· cal/g 4.184E7 erg/g kcal/kg erg g Joule kg J/kg g · cal/g 4184 J/kg Btu/lbm 2324 J/kg Btu/slug ft · lb lbm ft· lb/slug g · cal/g 1400 ft· lb/lbm Btu/lbm 778 ft· lb/lbm Btu/slug 778 7.224E5 24.17 J/kg ft · lb/lbm ft· lb/slug ft · lb/lbm erg/g ft· lb/lbm 72.23 2.990 32.174 ft · lb/slug J/kg erg/g J/kg 2116 6855 Btu slug Btu lbm Btu/lbm kcal/kg 1.800 57.90 Btu/slug kcal/kg Btu/lbm g · cal/g 1.800 24.2 g · cal liter · atm Btu/slug g · cal/g 57.90 Btu/slug Btu/lbm 32.174 10.41 g ·cal g g ·cal/g kcal/kg 1.0 ft·lb g ·cal 3.086 74.74 ft·lb liter · atm 778 ft·lb Btu ft·lb kcal g · cal kcal 252.0 g · cal Btu 3086 1000 liter · atm Btu 2.989E4 10.76 1.0E4 erg/g ft · lb/slug 929.4 41.29 liter · atm kcal ft·lb ft3 ·atm Btu kcal ft3 · atm kcal ft·lb hp · hr 1.98E6 g· cal hp · hr 6.416E5 g· cal ft3 · atm 3.966 1.459 2.649E4 liter · atm hp · hr ft·lb kWh 2.655E6 28.32 Btu hp · hr ft3 · atm hp · hr kcal hp · hr g· cal kWh 8.606E5 ft · lb liter · atm ft3 · atm 2546 935.9 641.6 liter · atm kWh 3.553E4 g · cal Btu ft3 · atm Btu kWh ft3 · atm kWh kcal kWh Liter · atm 2.721 3414 1255 860.6 hp · hr kWh Btu 1.341 ft3 · atm kcal hp · hr Table E.15 Specific energy conversion factors kcal kg g · cal g Btu lbm Btu slug ft · lb lbm Joule kg ft · lb slug erg g kWh Table E.14 Energy conversion factors J kcal J Btu 896.8 in.·lb liter · atm 9332 in.·lb Btu 2.539E4 in.·lb ft3 ·atm 3.703E4 in.·lb kcal in.·lb hp · hr 2.376E7 in.·lb kWh 3.186E7 in.·lb J ft·lb in.·lb ft·lb 8.850 in.·lb J 12 in.·lb J 37.03 g·cal g·cal kcal kg 1.356 4.184 101.3 J liter · atm 1054 J ft3 ·atm 2869 4184 J hp · hr 2.685E6 J kWh 3.6E6 Joule (J) (N ·m) erg ft · poundal 4.214E5 erg ft · poundal 1.130 in.·lb in.·lb 2.681 erg J 1.356E7 erg ft·lb 4.184E7 erg g ·cal 3.578E7 erg liter · atm 1.054E10 erg Btu ft · poundal 1.0E7 J 23.73 ft · poundal ft·lb 32.174 ft · poundal g ·cal 99.29 ft · poundal liter · atm 2405 ft · poundal Btu 25020 ft · poundal ft3 ·atm 6.809E4 1.0133E9 erg ft3 ·atm 4.184E10 erg kcal 9.929E4 ft · poundal kcal erg hp · hr 2.684E13 erg kWh 3.6E13 erg (dyne · cm) ft · poundal hp · hr 6.370E7 ft · poundal kWh 8.543E7 ft · poundal erg (dyne ·cm) ft · poundal in.·lb Joule (J) (N·m) ft·lb g · cal Liter ·atm Btu ft3 · atm kcal hp · hr kWh Appendix E ■ Comprehensive Table of Conversion Factors 777 Answers to Selected Homework Problems Chapter 1 1.6 1.8 1.10 1.12 1.16 1.20 1.22 1.28 1.30 1.32 1.34 1.36 1.38 1.40 1.42 1.44 1.46 1.48 1.50 1.52 1.56 1.58 1.60 1.62 1.64 1.66 1.68 1.70 1.72 1.74 1.76 1.78 1.80 1.82 1.84 1.86 1.88 1.94 1.96 1.98 1.100 1.102 1.104 1.106 1.108 (a) LT−2; (b) T−2; (c) L2 T−2 (a) L2 T−2; (b) M2 L−3 T−3; (c) M0 L0 T 0 (a) is not; (b) is; (c) is not; (d) is not (in H2O)(in1.22)ft yes (a) 4.32 mms; (b) 70.2 kg; (c) 13.4 N; (d) 22.3 ms2; (e) 1.12 N ∙ sm2 (a) 6.47 × 105 m2; (b) 56.8 × 10−2 m3; (c) 3.86 × 105 m; (d) 5.90 × 104 W; (e) 289 K 0.250 m3 (a) 0.240 mi3; (b) 4.41 × 105 lb 30.6 kg; 37.3 N ρeg = 1,130 kgm3 Δh = 0.0059 ft; no 0.9971 lb 991.5 kgm3 D = 0.729 in = 18.5 mm m = 9.38 × 10−4 slug (a) 0.0214 kgm3; (b) ρMarsρearth = 1.75% madded = 0.0185 kg, pfinal = 273 kPa 98.7 psia 668 lb 0.6 N ∙ sm2; 1.3 × 10−2 lb ∙ sft2 31% 0.277 N ∙ sm2 2.05 × 10−5 N ∙ sm2 (a) No; (b) Not correct 184 C = 1.43 × 10−6 kg(m ∙ s ∙ K12); S = 107 K D = 1.767 × 10−6 N ∙ sm2; B = 1.870 × 103 K; 5.76 × 10−4 N ∙ sm2 0.300 N ∙ sm2 V2 = 0.201 m/s, left 11.7 × 10−4 ft 377 N 0.0883 ms τ = 0.45 lb/ft2 3.43 × 10−4 lb 9.53 × 10−4 ft ∙ lb 0.072 ft-lb 2.03 × 103 psi 0.0163 m3 decrease 104 psi (gage) 2.88 kgm3 23 tires Ev(water) Ev(air) = 1.52 × 104 % change = 1.69%, incompressible OK A = 102 ft2 1.110 1.112 1.114 1.116 1.118 1.120 1.122 1.124 1.126 1.128 1.130 14.1% 4.74 psi (abs) 13 kPa (abs) 5.81 kPa (abs); 0.842 psi (abs) 0.060 Nm 538 Pa σ = 5.97 × 10−2 Nm (a) 24.5 deg; (b) sinks because W > T 0.0614 in.; Yes 1.80 × 10−2 ft 7.49 mm Chapter 2 2.2 2.4 2.6 2.8 2.10 2.12 2.16 2.18 2.20 2.24 2.26 2.28 2.32 2.34 2.36 2.38 2.40 2.42 2.44 2.46 2.48 2.50 2.54 2.56 2.58 2.60 2.61 2.64 2.66 2.68 2.70 2.74 2.76 2.78 2.82 2.83 p = 111 MPa gage p = 209 kPa SG = 1.53 10.4 m 114 MPa gage 101 kPa gage 1m 464 mm 62.9% 4,250 ft 543 m 60 kPa 14.4 psia; 99.3 kPa (abs) (a) 18.2 ft; (b) 8.73 psi; 21.7 psia 2 cm pA = 15.0 psia; pB = 15.0 psia; pC = 15.5 psia −3.32 kPa (a) 4.00 ft; (b) 2.08 ft 1.45 ft (a) 26.9 kPa; (b) 0.202 m S = 0.667 (a) h1 − h2 = 0; (b) h1 − h2 = h 21.6 ft 0.100 m error = 4% 27.8° 0.212 m 0.304 ft ℓ = [d + 11.31 − (d 2 + 18.61d + 128)12]2 36,000 lb p = 0.416 θ 2.90 lb 149 kN 2,460 lb 107 kPa 665 lb ANS-1 ANS-2 2.84 2.88 2.89 2.92 2.94 2.96 2.98 2.100 2.102 2.103 2.104 2.106 2.108 2.110 2.114 2.116 2.118 2.120 2.122 2.124 2.126 2.128 2.130 2.134 2.136 2.138 2.142 2.144 2.146 2.148 2.150 2.152 2.154 2.156 2.158 2.160 2.162 Answers to Selected Homework Problems 7.11 m; No FA = 89,900 lb; FB = 22,500 lb; FC = 75,400 lb FB = γπR34 F = 2.0 × 107 lb; LOA @ 21.7 ft above base FHEAD = 312,000 lb; yR,HEAD = 66.7 ft; FTAIL = 34,300 lb; y′R,TAIL = 24.4 ft No No FV = 98,100 downward; LOAV 1.00 m from left edge of gate h = 16.2 m; No 1680 lb 5.24 ft 3.55 m 3.63 ft 102 lb-ft Alike - same magnitude & LOA; Different opposite directions No 581 lb; 0.289 psi 7.77 × 109 lb acting 406 ft above base 64.4 kN FH = 1.92 × 105 lb toward front, yR = 30.3 ft FH = 12,300 N; yR = 0.167 m; FV = 82,400 N; LOA 0.525*D from left edge of log. 3370 lb; 882 lb 485 kN downward along centerline A = 93.4 ft2 786,000 lb; 315,000 lb 54,600 lb 127 lb bottom 13.3 ft submerged; stable 681 lb at 6.22 ft to right of A W = 3.88 × 107 lb Vol = 4,700 ft3 front (a) 68.9 lbft2; (b) 57.4 lb T = 0.062 N h = 5.22 cm 35.1 rpm pB = 14.732 psia; pC = 8.10 psia 28.8 kPa Chapter 3 3.2 3.4 3.6 3.8 3.12 3.14 3.21 3.25 3.29 3.34 3.36 13.7 ms −30.0 kPam −0.838 psift; 0.0292 psift (a) −2ρa2V 02 [1 − (ax)2]x3 (b) p0 + ρV 02 [(ax)2 − (ax)42] 12.0 kPa; −20.1 kPa (a) 4.97 lbft3; 0.681 lbft3 3.21 Nm2 203 fts 46.5 ms 20.9 fts 45.7 ft 3.38 3.43 3.44 3.45 3.48 3.51 3.58 3.59 3.60 3.64 3.66 3.68 3.70 3.76 3.78 3.80 3.82 3.91 3.93 3.94 3.95 3.97 3.99 3.109 3.111 3.113 3.116 3.118 3.123 3.131 3.133 1248 lbft2 20.8 lbs 43.0 psi 0.0119 ft3s 0.624 cm 0.223 ft; 9.22 lbft2 3.19 ft 0.0156 m3s for any D 6.59 ft; B above A D = D1[1 − (π2gD14 z8Q2]14 3.13 ft 2.43 m 0.0758 ft3s; 499 lbft2; 488 lbft2; −11.7 lbft2 (a) 1.06 × 10−3; (b) 3.02 × 10−3; 0.118 29.5 Pa 0 (gage); 0.146 m (a) 0.0696 m3s; (b) 0.574 m 4 13.0 psi 1.30 ft3s 0.607 ft3s; (b) 21.3 ft 0.123 ft3s 0.400 ft 404.5 kPa 252 kPa; 114 kPa 9.10 × 10−3 m3s; 57.9 kPa 3120 ft3s; 2080 ft3s; 2200 ft3s 155 Nm2 0.630 ft; 4.48 ft 1.73 m2s 25.4 m; 6.51 m; −9.59 m Chapter 4 4.2 4.4 4.8 4.10 4.14 4.18 4.20 4.22 4.26 4.30 4.34 4.37 4.39 4.44 4.46 4.49 4.51 V= V0 (V0 + ΔV) V0 + ΔVe−α(V0 + ΔV)t x = y + y22 x = −2, y = 2 Everywhere; x2 + y2 = constant x = −h (u0υ0) ln (1 − yh) 10, 9, 6, 4 fts Steady, Two-Dimensional, Plot Two-Dimensional; Two-Dimensional; Two-Dimensional; Two-Dimensional; Three-Dimensional; Two-Dimensional; Three-Dimensional (a) 4fts2; 2 fts2; (b) negative −1.65 × 1011 fts2; −5.12 × 109 720 ms2; 1440 ms2; 2160 ms2 −33.8 fts2; 1.05 (225x + 150) ms2; 0; 150 ms2; 375 ms2 8.02 fts 8860 ms2; 7440 ms2 (a) 10 fts2; (b) 20 fts2, 10 fts2; (c) 22.4 fts2 −225 fts2; −28.1 fts2; −8.33 fts2 Answers to Selected Homework Problems 5.41 5.43 5.46 5.48 5.50 5.52 5.58 5.60 5.64 5.68 5.70 5.72 5.74 5.76 5.78 5.80 5.82 5.84 5.86 5.89 5.97 5.102 5.104 5.106 5.116 5.118 5.122 5.124 5.126 5.128 5.130 5.132 5.135 5.137 5.141 5.151 5.155 5.157 5.159 2.66 × 10−4 m3s 15.5 lb 7.01 ft3s; 674 lb 352 lb 2.69 lb (a) 6.26 ms; (b) 392 N; (c) 675 N 17,900 lb 28.7 ft3s (a) W1 − 9.8 lb; (b) W2 + 14.7 lb −185 kN; 45.8 kN 1290 lbft; 3120 lbft 3580 lb 2.96 lb 66.6 N 13.3 lb non-uniform = (43) uniform (a) 181 lb; (b) 146 lb (b) 234 lb 9170 lb 78.5 kN (a) 231 N ∙ m, 185 rads; (b) 200 N ∙ m, 160 rads; (c) 116 N ∙ m, 92.5 rads (a) 43 deg; (b) 53.4 ft ∙ lbs −36.8 ft ∙ lbslug −3130 ft ∙ lbslug 0.042 m3s 4.58 × 10−3 m3s 926 ft ∙ lbslug; 200 ft ∙ lbslug 734 ft 2.22 × 106 W 5.48 hp Will work for 2 atm., not for 3 atm. 0.052 m3s 1610 ft ∙ lbslug; 1110 ft ∙ lbs (a) 4.08 hp; (b) 9.94 ft 1.78 kW (a) 17.2 deg, 4.29 ms; 558 N ∙ ms (a) 3.43 ft ∙ lblbm; (b) 3.36 ft ∙ lblbm 11.6 N ∙ mkg; 0.796 148 ft ∙ lbslug; 0.875 Chapter 6 6.1 6.4 6.5 6.7 6.9 6.11 6.15 2x, 4xt2; −2y, 4yt2; V = 0; a = 4iˆ − 4jˆ ft/s2 ; a = 5.66 fts2 (a) 0; (b) ω = −(y2 + z)iˆ + (5z2)jˆ − (y2)kˆ; No ζ = 3xy2 kˆ ; No 3 x3 u = − x2 + + f (y) 2 3 −r0ω γ˙ = r0 − ri No (a) 2.24 ms, −63.4 deg, x = y = 0; (b) 4.12 ms, 76.0 deg, none 6.16 No 6.18 υ= 6.23 6.24 6.26 6.27 6.30 6.31 6.34 6.36 6.37 6.38 6.39 6.40 6.42 6.46 6.51 6.55 6.59 6.60 6.64 6.66 6.67 6.69 6.72 6.78 6.79 6.80 6.86 6.90 6.96 6.98 −y 2 + f (x) 2x 19.6y − 100 kPa, where y is m Derivation −Ar 2 ψ= + C; ψ = −A ln r + C 2 ϕ = A ln r + Br cos θ + C; θ = π, r = AB 60.5 psi (a) Yes; (b) Yes, ϕ = 2(x + y) + C; (c) 0 −Kr2/2 + C; No −7.10 kPa ψ = 5x2y − 5y33 + C 92.3 ft 3.10 ft (a) 1.60 × 103r 3 kPam; (b) 184 kPa 0.00796 ft to right of slit; 0.0250 ft 80 mph; 53.3 mph; 40 mph h2 = m2πA 8.49 × 10−4 ms (23)12 10 34,900 lb; 105,000 lb 55.56 rad/s CW τxy = 30μxy; τyz = τxz = 0 (a) ∂υ/∂y = −2x; (b) a = 2x3 ˆi ; (c) ∂p/∂x = 2μ − 2ρx3 43.2 lbft3; 0.75 fts (a) u = Uyb; (b) Ub2 u = [(U1 + U2)b]y − U2 U(1 + μ2μ1) (a) 𝒯 = 2πr03 μωℓ(r0 − ri); (b) γ˙ = −U/b (a) Proven; (b) 2.08 Pa V0 = (Δpℓ)[2r 2i ln (ri r0 ) − (r 2i − r 20 )]4μ (a) 8.79 kPa; (b) No Chapter 7 7.6 7.10 7.11 7.13 7.14 7.17 7.20 7.27 7.29 7.31 7.37 7.39 7.40 7.41 ANS-3 (a) 103 ms; (b) 444 ms ωℓ12g12 = ϕ(hℓ) Qd 52 g12 = ϕ(ρgρa, hd) δd = ϕ (ℓd, ρV2E, μVdE) hℓ = ϕ(σℓ2 gρ) Δp ∝ 1D2 (for a given velocity) HP; LP t√gd = ϕ(Vd 3, ρ √g d 3/2 μ) (a) VD √ρ𝒲 = ϕ(bd, dD); (b) V = √2𝒲bπρdD2 5.58 × 107; 1.88; 1.44; 4.17 × 10−3 ΔpρV12 = −1.10 (A1A2)2 + 1.07 (A1A2) − 0.0103; 6.26 lbft2 Δp(ρV 2) = 0.505 (Dd )3.99 (a) Hb = ϕ(hb, ℓb); (b) Hb = 0.0833 (hb)−1.00 (a) Correct; (b) t = 1.36 μ ANS-4 7.42 7.44 7.48 7.50 7.51 7.53 7.54 7.55 7.57 7.58 7.60 7.61 7.63 7.66 7.70 7.71 7.73 7.73 7.81 Answers to Selected Homework Problems (a) −12; (b) Yes CD; Re; CD = 0.549 Re−0.08; 0.0497 N 129 ms 3750 mph; No 1170 kmhr (a) 4.50 mph; (b) 8.00 ft 0.125; 0.0625 (a) hH = ϕ(dH, b/H, ρgγs, V √gH); (b) 21.2 mph 50.2 kPa (abs) [[See answer for 7.69 in 7th ed. solution manual]] (a) Δpρc2 = ϕ(Eρc2d3) (a) V0m tmDm = V0 tD, ωm tm = ωt, ρm V0m Dmμm = ρV0 Dμ, D ΔpℓρV 02 = Dm Δpℓmρm V 0m2; (b) 160 rads 100 mph to 450 mph 60 mph; M30 = M604 (a) Vℓ2Q = ϕ(ℓiℓ, Q2ℓ5g, ρQℓμ); no (b) 0.410 gpm; 2.46 in. (a) nDV = ϕ(VDν); (b) Yes; (c) 7.54 fts (a) VmUm = VU, Vm Dmνsm = VDνs, V m2 gm Dm = V 2gD, ( ρ − ρs)mρm = (ρ − ρs)ρ; (b) No 0.440 ft; 2.80 ft3s (a) d 2y*dx*2 = (Pℓ2 EI)(x* − 1) 8.61 8.65 8.67 8.69 8.71 8.73 8.78 8.81 8.83 8.86 8.88 8.89 8.93 8.96 8.99 8.103 8.107 8.111 8.117 8.119 8.123 8.126 8.129 8.131 8.132 1.54 kPa 0.899 lbft2 (a) 135 ft; (b) 137 ft 84.0 ft 0.750 psi 24.4 hp 16.5 ft 48.0 psi 379 kW 5.46 × 10−4 m3s 0.0615 ft3s 0.0289 m3s 155 hp 0.0494 ft3s 74.8 min 0.535 ft 0.0445 m 0.500 ft3s 0.0180 m3s 0.0284 m3s; 0.0143 m3s; 0.0141 m3s 0.0936 ft3s 32.4 kPa 0.0221 m3s 5.69 ft3s 11.1 in. Chapter 8 Chapter 9 8.2 8.4 8.6 8.7 8.9 8.10 8.12 8.14 8.15 8.17 8.18 8.20 8.22 8.28 8.31 8.32 8.34 8.35 8.37 8.40 8.42 8.44 8.50 8.53 8.55 8.57 8.58 8.59 9.2 9.5 9.8 9.11 9.13 9.15 9.19 9.21 9.25 9.27 9.31 9.36 9.38 9.44 9.46 9.48 9.57 9.61 9.63 9.64 9.66 9.69 9.71 9.74 9.78 9.80 9.82 9.84 Turbulent 3.36 ft 4.95 × 10−6 m3s at 20 °C 3 m; 8.83 Nm2 2 ms; 1 ms; 1.26 × 10−3 m3s 1.89 ms; 7.42 × 10−3 m3s 0.15 lbft2; 0.06 lbft2; 0 lbft2 (b) to (a) 0.354 D 3.43 m; 166 kPa 5 z = 1 + e−(t2.8 × 10 ) ft, where t is seconds (a) 4.69 × 10−3 ft3s; 3.30 × 10−3 ft3s 8.88 × 10−8 m3s (a) 0.707 R; (b) 0.750 R 1.02 × 10−4 ft B to A; A to pump (a) 0.266 psi; (b) 1.13 psi; (c) −0.601 psi 0.649 Pa; 2.04 × 10−3 W 0.0162 0.0404 19.6 m 0.0132 ft (a) 0.556 ft; (b) 3.70 ft; (c) 0.926 ft 39.7 kPa; 39.7 kPa, 93.0 kPa 13 0.223 lbft2 9 Disagree (a) 0.06 (ρU22); (b) 2.40 5.43 kN 0.142 7.54 × 10−3 fts; 0.471 fts; 2.62 × 10−5 fts 1.12 m; 7.92 × 10−3 m 6.65 × 10−6 m2s d = 1 + 0.0304√x ft where x is ft (a) 27.6 fts; (b) 20.5 mph (a) 0.171 ft; (b) 0.134 ft (b) δ/x = 5.48 √Rex (b) δ = 5.83√νxU 0.707 48.8 lb 0.10; streamlined (a) 1303 W; (b) 3900 W 1.88 × 104 lb; 33.3% 2.8 s (a) 567 fts; (b) 118 fts; (c) 13.5 fts 133 N 77.0 fts (a) 7080 N ∙ m; 8950 N ∙ m 30.6 fts 378 lb 58.4 hp 0.946 fts in water; 30.2 fts in air (a) 2.44 ms; (b) 2.13 ms 5.31 ms 1.42 lb; 142 lb Answers to Selected Homework Problems 9.86 9.88 9.92 9.93 9.96 9.97 9.99 9.101 9.103 9.106 9.107 9.110 9.114 9.115 9.117 9.122 7220 ms2 9.65 lbft2 (a) 4.31 MN; (b) 4.17 MN (a) 53.5 kW; (b) 4.46 kW; (d) 42.0 kW 4.35% 41.6 fts (a) 0.0166 hp; (b) 32.4 hp 0.851; 0.301 65.9 hp 0.225 1.72 U 0.206 Yes 0.480; 0.409; 0.451; 0.482 5.72% 28.4% Chapter 10 10.3 10.6 10.8 10.10 10.13 10.15 10.17 10.19 10.27 10.33 10.37 10.38 10.41 10.44 10.46 10.48 10.49 10.52 10.57 10.58 10.60 10.62 10.66 10.67 10.70 10.74 10.76 10.81 10.82 Subcritical; 1.04 m 5.66 ft 2.80 ms 175 fts 5.82 ft or 1.129 ft 1.774 ft or 1.974 ft 1.828 ft 0.694 ft 0.000532 (a) 1.80 lbft2; (b) 0.0469; (c) 0.636 23.7% 48,500 ft3s Yes 0.355 ft3s; Re = 53 18.2 m3s 1148 m3s 119 ft3s 0.841 m Each requires the same. 2.74 ft 5.16 ft 10.66 m 0.000664 0.000269 0.452 in. 1.68 m 401 kW 0.652 m; 2.61 m (a) 37.7 ft2s; (b) 1.74 ft, 6.29 ft; (c) 4.33 fts, 5.99 fts; (d) 9.2 hpft 10.84 4.70 ft Chapter 11 11.2 (a) 71,700 Jkg; (b) 100,000 Jkg; (c) −1.58 kgm3; (d) 396 Jkg ∙ K ANS-5 905 °R; −4.17 × 106 ft ∙ lbslug; 877 ft ∙ lbslug 588.6 °F; 28450 Btu°R*hr 5300 ms; 1470 ms; 340 ms (a) 210 ms (b) 204 ms (a) 0.0328; (b) 0.0722; (c) 0.131 (a) 732 ms; (b) 2400 fts; (c) 1636 mph (a) 590 mph; (b) 866 fts; (c) 264 ms From 0.47 to 0.75 92 kPa (abs); 82 kPa (abs) No (a) 259 ms, 584 K, 600 kPa (c) 586 K, 1040 kPa (a) 13.8; (b) 13.7 283 ms; 0.89 1.8; 1580 fts 0.335, 461 psia, 498 psia 0.9608, 0.4348, 0.0640; 1.854, 0.605 0.36 m2, 23 kPa, 113 K 1.35 m 27.7 kgs 17 psia; 868 fts; 1390 Btus at duct entrance T = 130 K; p = 6.0 kPa (abs); T0 = 293 K; p0 = 101 kPa (abs); V = 571 ms; just upstream of shock T = 296 K; p = 17 kPa (abs); T0 = 396 K; p0 = 47.7 kPa (abs); V = 448 ms; just downstream of shock T = 354 K; p = 30.8 kPa (abs); T0 = 396 K; p0 = 46.4 kPa (abs); V = 299 ms; at duct exit T = 351 K; p = 26.6 kPa (abs); T0 = 407 K; p0 = 45.9 kPa (abs); V = 337 ms 11.70 2090 kJs; 660.7 °C; 57.1 kPa 11.7 11.10 11.16 11.18 11.21 11.23 11.26 11.27 11.31 11.32 11.34 11.38 11.42 11.45 11.50 11.52 11.54 11.59 11.64 11.66 11.69 Chapter 12 12.5 12.7 12.9 12.11 12.13 12.15 12.18 12.21 12.22 12.26 12.27 12.32 12.34 12.44 12.46 12.49 12.51 12.53 12.54 12.58 12.61 −71 ft ∙ lbslug (b) fan; (c) 1780 ft2s2 379 ft ∙ lblbm (b) 891 ft ∙ lb 61.3 ft 107 gpm (a) 2.07 kW; (b) 12.6 m 158 galmin; Yes 1740 galmin; Yes 0.0328 m3s; 8.00 m Yes Centrifugal 1900 rpm (a) −15.6 m2s2; (b) −15.6 m2s2 −19.8 hp 0.0109 ft3s; 0.0409 hp 816 kW pump 23,200 kW 26,600 N; 37.6 ms; 707 kgs 70 m impulse Index Absolute pressure, 13, 50 Absolute temperature in British Gravitational (BG) System, 7 in English Engineering (EE) System, 8 in ideal gas law, 12 in International System (SI), 7 Absolute velocity in centrifugal pumps, 688–689 defined, 206, 681 in rotating sprinkler example, 228–229 in turbomachine example, 681–684 vs. relative velocity, 185–186 Absolute viscosity of air, 759–760 of common gases, 756 of common liquids, 756 defined, 15 of U.S. Standard Atmosphere, 761–762 of water, 758 Acceleration along streamline, 167–168 centrifugal, 173, video V3.4 convective, 169, 173, 323 defined, 165 as field, 165, 279 gravitational, 7, 8, 761–762 local, 168 normal, 103, 171–173, video V3.1, video V3.4 particle, 103, 165, 279 role of material derivative, 165–168, 279 streamwise, 103, 173, video V3.1, video V5.3 Accessibility of flow, 569 Acoustic velocity, 22 Actual head rise, 690–692 Adaptive grids, 746 Adiabatic flow, 235, 607, 632, 652–659 with friction .see Fanno flow reversible .see Isentropic flow Adverse pressure gradient, 314, 511 Aerodynamics, in history of fluid mechanics, 28 Affinity laws, pump, 703 Air, table of properties, 759–760 Air bridges, 163 Airfoils, 537, 542, video V9.21, video V9.23, video V9.24 Alternate depths, 568 Andrade’s equation, 18 Aneroid barometers, 56 Angle of attack, 362, 513, 541, 543 Angle of contact, 25 Angular deformation, 278, 280–283, video V6.3 Angular momentum. see Moment-ofmomentum Angular momentum flow, 226, video V5.12 Annulus, flow in, 330–331 Apparent viscosity, 16 Aqueducts, 572 Archimedes, 27, 28, 71 Archimedes’ principle, 70–73, video V2.7 Area centroid, 60–62 compressible flow with change in, 632–647 isentropic flow of ideal gas with change in, 635–641 and Mach number, 636–638 moment of inertia, 61 Aspect ratio airfoil, 542 plate, 355 Assemblies, 742 Atmosphere, standard. see Standard atmosphere Atmospheric pressure, 13, 48–49, 52, video V2.2 Available energy, 242–243, 247, 607–608 Available net positive suction head, 694–695 Average velocity, in mass flowrate, 199–200 Axial-flow impulse turbines, 715 Axial-flow machines compressors, 720–722 defined, 680, 681 pumps, 704–707 Axisymmetric flow, 484 Backward curved blades, 691, 692 Backward difference, 739, 741 Bar (unit), 52 Barometers, 52 aneroid, 56 mercury, 51 Basic dimensions, 4, 349 Bats, 538 Bellows meters, 466 Bends, 439 Bernoulli, D., 27, 28, 107 Bernoulli equation cautions on use of, 107 derivation from Euler’s equations, 294–295 derivation from first law of thermodynamics, 241–247 derivation from F = ma, 107, 293–294, video V3.1 examples of use, 117–131 extended, 243 for incompressible flow, 630–631 irrotational flow, 137, 297 physical interpretation, 110–113 restrictions on use of, 134–138, 295, video V3.12 unsteady flow, 135–136 use of, video V3.2 Best efficiency points, 694 Best hydraulic cross section, 578–579 BG units. see British Gravitational (BG) System of units Bingham plastic, 16–17 Birds, wings of, 536 Blades curved and radia, 691, 692 radial, 691 rotor, 705, 721 stator, 705, 721 turbomachine, 679 Blade velocity, 681–684, 688–689 Blasius, H., 29, 369, 496 Blasius formula, 369, 433 Blasius solution, 497–499 Blood pressure, measuring, 2, 45, 52, video V2.3 Blowers, 679, 707 Blunt bodies, 484, video V9.2 Body forces, 42, 212, 290 Boiling, 23–24, 124–125 Boosters, 707 Bottom slope, 566–567, 572 Boundary conditions in Bernoulli equation, 107 inviscid flow, 309, 310 no-slip, 14, 311, 324, video V1.4, video V6.12, video V6.13 in solving CFD simulation problems, 746 Boundary element method, 742–743 Boundary layer, 491–515 concept in history of fluid mechanics, 27–28 concept of, 296–297, 489, video V1.4, video V6.12, video V9.12 displacement thickness, 493–494 effect of surface roughness, 527 effects of pressure gradient, 510–514, video V9.10 equations, 495–499 flat plate, 491–495 and flow separation, 489 laminar flow approximate solution, 503–504 Blasius solution, 497–499 Prandtl solution, 495–497 momentum integral equation for, 499–504, 514–515 momentum thickness, 495 pipe, 405 thickness, 492–495, video V9.6 transition from laminar to turbulent flow, 492, 504–505, video V9.7 turbulent flow, 492, 506–509 velocity profiles, 505, 511, 512, video V9.4 vorticity, 492 Boundary layer separation, 511–512, video V6.8 Bound vortex, 542, video V4.6 Bourdon pressure gage, 56, video V2.4 Boussinesq, J., 423 Bow thrusters, 222 Brake horsepower, 693 British Gravitational (BG) System of units, 7–8 Broad-crested weirs, 590–592, video V10.14 Buckingham, E., 29, 350 Buckingham pi theorem, 349–350 Buick Dynaflow, 687 Bulk modulus, video V1.7 characterizing compressibility, 20, 21 defined, 20 Bulk modulus of elasticity, 614 Buoyant force, 70–73, video V2.7, video V2.9, video V11.5 Canals, 570 Capacity, pump, 693 Capillary action, 25, video V1.10 I-1 I-2 Index Capillary tube viscometer, 33, video V1.5 Casings, 687, 688 Cauchy, A., 28, 364 Cauchy number, 362, 364 Cavitation, 718 defined, 24, 124 producing in flowing liquid, 125 Cavitation number, 364, 378, 387 Ceiling fans, hi-tech, 708 Center of buoyancy, 71–72, video V2.12 Center of pressure, 62 Centrifugal acceleration, 173 Centrifugal pumps, 687–700, 707, video V12.3 net positive suction head, 694–696 performance characteristics, 692–694 performance curves, 692, 697–700 system curve, 696–697 theoretical considerations, 688–692 Centroid, 60–62 CFD. see Computational fluid dynamics (CFD) Chaos, 427 Chezy, A., 28, 572 Chezy coefficient, 572 Chezy equation, 571–572 Choked flow Fanno, 650, 655–657 Rayleigh, 662 in variable area duct, 639–640, 643 Chord length, 537, 542 Circulation flow past airfoils, 541–543 flow past cylinders, 543 flow past spheres, 543–545 and vortex motion, 306 Closure problem, 747 Coefficients, 365. see also specific coefficients Colebrook formula, 432 Combustion chamber, 665–666 Completely turbulent flow, 430–431 Composite body drag, 530–534 Compressibility of fluids, 20–23, 46–48, 526–527 Compressible flow, 604–672, video V11.2, video V11.5 analogy with open-channel flow, 667–668 effect on use of Bernoulli equation, 134–135 with friction, 647–659 graphs of, 763, 765, 767, 769 with heating or cooling, 660–666 ideal gas thermodynamics, 605–610 introduction, 604–605 isentropic, 626–632 Mach number and speed of sound, 612–617 regimes, 617–621 as reversible process, 134–135 shock waves, 621–626 stagnation properties, 610–612 two-dimensional supersonic, 668–669 in variable area duct, 632–647, video V11.6, video V11.7 Compressible flow passage, 640–641 Compressible flow turbines, 723–725 Compressible flow turbomachines, 719–725 Compressors, 719–723, video V12.6 axial-flow, 720–722 defined, 707, 719 multi-stage, 719–720 as pumps, 679 radial-flow, 720 Compressor stall, 721 Compressor surge, 721 Computational fluid dynamics (CFD), 427, 737–754, video V2.11, video V6.2, video V6.16, video V11.6 advantages for flow simulation studies, 753 boundary conditions, 746 computational grid, 744–746 converging channel example, 738–742 difficulties in using, 750–752, 754 discretization, 742–744 overview, 278, 332, 737–738 solving equations, 748–750 turbulence models, 746–748 verification and validation, 752 Computational grids. see Grids Conservation of energy. see First law of thermodynamics Conservation of linear momentum, 289–293, video V5.4, video V5.5, video V5.6 Conservation of mass, 120, 197–210, 283–289 Conservative form, 750 Contact angle, 25, video V1.11 Continuity equation, 198–210. see also Conservation of mass cylindrical polar coordinates, 286 defined, 284–285 derivation of, 198–200 differential form, 283–285 finite control volume, 120–121 incompressible flow, 205, 206, video V5.1, video V5.2 unsteady flow, 199, 285 rectangular coordinates, 283–284 Continuum hypothesis, 4 Contraction coefficient, 119, 120, 129, 593 Control surfaces, 174 Control volume, 173–175, video V4.14 defined, 174 deforming, 175, 210–212 in derivation of continuity equation, 198–199 and first law of thermodynamics, 234–235 fixed, 174–175, 200–206 guidelines for selection, 186–187, 208–210 infinitesimal, 277 moving, 185–186, 206–208, 224–226 velocity, 206 Control volume statement, 251–252 Convective acceleration, 169, 173, 323 Convective derivative, 168–169 Converging channel, 738–742 Converging-diverging duct flow, 635, 637–638, video V11.6, video V11.7 Converging-diverging nozzles, 634, 643–647, video V11.6, video V11.7 Converging duct flow, 633–634 Converging nozzles, 641–643, 646–647 Conversion factors, for units of measure, 9 Cooling, 660–666 Corrected compressor mass flowrate, 723 Cotton candy, 119 Couette flow, 325–327 Creeping flow, 332, 363 Critical depth, 568 Critical flow, open-channel, 561, 565 Critical pressure ratios, 645, 646 Critical properties, 631 Critical Reynolds number, 492, 504 Critical state, compressible flow, 631–632 Cross, H., 460 Crystal Rapid, 584 Curl operator, 282 Curved shock waves, 621, 622 Curved surfaces, forces on, 68–70 Cylinders circular, 314–317 drag coefficient for, 523–524, video V9.13 inviscid flow around, 315, video V6.6 large vs. low Reynolds number flow around, 489–490, video V6.8, video V6.16 pressure distribution on surface, 315, 513–514 rotating, 317–319, 543 separation location, 489, 511–512 Cylindrical coordinates, 77–78 Cylindrical polar coordinates, 286, 298, 322 D’Alembert, J., 27, 28, 316 D’Alembert’s paradox, 316, 511, 512 Darcy, H., 28, 414 Darcy friction factor, 414 Darcy-Weisbach equation, 430 DaVinci, L., 27, 28 Deepwater pipelines, 456 Deepwater waves, 564 Deformation angular, 278, 280–283, video V6.3 linear, 278, 279–280 rate of, 280 Deforming control volume, 175, 208–210 Degree of reaction, turbine, 708 Del (gradient) operator, 43 Density of air, 759–760 as fluid characteristic, 11, video V2.10 ideal gas, 608–609 mass flow, 648–649, 660 stagnation, 611 of U.S. Standard Atmosphere, 761–762 of water, 11, 758 Dental drills, 714–715 Depth ratio, 583–584 Derivative convective, 168–169 local, 168 material, 165–168, 170–171, 182–183, 279 substantial, 165–168, 279 Design flowrate, 694 Differential analysis, 277–334 and conservation of linear momentum, 289–293 and conservation of mass, 283–289 determination of pi terms by example, 357–358 determination of pi terms by inspection, 359–360 and fluid element kinematics, 278–283 introduction, 277–278 and inviscid flow, 293–300 and plane potential flows, 300–319 Index Differential equations. see Differential analysis Diffuser pumps, 687, video V8.13 Diffusers, 438 Dimensional analysis, 346–369 and Buckingham pi theorem, 349–355 common dimensionless groups, 361–366 concept of dimensions, 4–5 correlation of experimental data in, 366–369 defined, 346–347 determination of pi terms by inspection, 359–360 determination of reference dimensions, 357–358 of ducts, 443–446 for fully developed laminar pipe flow, 413–415 overview, 347–349 of pipe flow, 427–446 selection of variables, 355–357 Dimensional homogeneity, 4–5, 350 Dimensionless groups common in fluid mechanics, 361–366 defined, 348–349 Dimensionless products defined, 348–349 pi terms as, 350 Dimensions basic, 4, 349 of common physical quantities, 5 reference, 350, 357–358 Discharge coefficient nozzle, 462 orifice, 461 underflow gates, 593 Venturi meter, 463 Discretization, 742–744 Displacement thickness, 493–494 Distorted models, 374–375, video V7.8, video V7.11, video V7.12, video V7.17, video V7.18 Diverging-converging duct flow, 635 Diverging duct flow, 633–634 Donor cell differencing, 751 Doppler effect, 619 Double-suction impellers, 688 Doublet, 307–309 Drag, 515–534 on circular cylinder, 316 defined, 484–485 example, 485–486 form, 517–519, video V7.2, video V9.11 friction, 516–517 friction coefficient, 508–509 overview, 515–516 pressure, 517–519 on rotating cylinder, 318 Drag coefficient airfoil, 542 automobile, 531–532 composite body, 530–534 cylinder, 523–524, video V9.13 defined, 487 as dimensionless group in fluid mechanics, 362 effect of compressibility on, 526–527, 617 effect of Froude number on, 529–530, video V9.18 effect of Mach number on, 526–527 effect of shape on, 519–520, 525, video V9.11, video V9.19 effect of surface roughness on, 527–529 ellipse, 519–520 flat plate normal to flow, 519, 521, 522, 525, video V7.2 flat plate parallel to flow, 508–509, 519–522, 525 form, 517–519 friction, 503, 508, 516–517 in model studies of flow around bodies, 379 pressure, 517–519 Reynolds number dependence, 521–526 selected three-dimensional objects, 532 selected two-dimensional objects, 532–534 sphere, 381, 521–524, 527–528, 543, 544, video V9.17 streamlined body, 520 Drag friction coefficient, 503 Drooping head curve, 693 Ducts dimensional analysis, 443–446 vs. pipes, 401 Dynamic pressure, 113 Dynamic similarity, 372 Dynamic viscosity of air, 759–760 of common gases, 756 of common liquids, 756 defined, 15 of U.S. Standard Atmosphere, 761–762 of water, 758 Eddies, 421, 746 Eddy viscosity, 423, 747–748, video V8.7 EE units. see English Engineering (EE) System of units Efficiency fan performance example, 250 hydraulic, 693 isentropic, 722 mechanical, 693 overall, pump, 693 polytropic, 722 pumps vs. turbines, 717–718 volumetric, 693 Elbows, losses in, 439–440 Elevation head, 111, 132 Enclosed impellers, 688 Energy available, 242–243 internal, 234, 605–606, 608–609 kinetic, 111, 234 per unit mass, 234 per unit volume, 245 per unit weight .see Head potential, 111, 234 specific, 567–570 stored, 234 Energy balance, 566–567 Energy equation, 234–250 application to nonuniform flows, 247–250 application to one-dimensional flows, 237–241 combined with moment-of-momentum equation, 250 I-3 comparison with Bernoulli equation, 241–247 defined, 237 derivation, 234–237 semi-infinitesimal control volume statement of, 251 Energy line, 448–449, 566–567 Energy line (EL), 131–133 English Engineering (EE) System of units, 8–9 Enthalpy, 238, 606, 608–609, 611 Entrance length, 406, video V8.12 Entrance loss, 435, 436 Entrance region, 297, 405, 435 Entropy, 251–252, 607–609 and Fanno flow, 649–650 and normal shock, 625 of Rayleigh flow, 661 T–s diagrams, 630 Equation of state, 12–13, 605 Equations. see also specific equations general homogeneous, 5 of motion, 292–293 restricted homogeneous, 5 Equipotential lines, 301 Equivalent length, 434, 435 Euler, L., 27, 28, 293, 364 Eulerian description of flow, 159–160 Eulerian description of motion, video V6.2 Euler number, 362, 364, 387 Euler’s equations, 102 Euler’s equations of motion, 293 Euler turbomachine equation, 685 Exhausters, 707 Exit pipe loss, 436, video V8.12 Expansion waves, 668, video V11.9 Experimental data, 366–369 Explicit method, 744 Extended Bernoulli equation, 243 Extensive property, 175–176 External flow, 482–546 background, 482–483 boundary layer characteristics, 491–515 drag, 515–534 general characteristics, 483–491 lift, 533–545 Eyes, in pump casings, 687 Eyes, pressure of, 114 Falling (drooping) head curve, 693 Fan laws, 708 Fanning friction factor, 414 Fanno flow, 648–659 basic equations of, 648–651 choked vs. non-choked, 653, 655–657 defined, 648 duct sizing in, 657–659 for ideal gas, 767–768 of ideal gas, 652–654 supersonic, 659 Fanno line, 648–651, 654–655, 666 Fans, 679, 682, 707–708 Favorable pressure gradient, 511 Field representation, 156–157, video V4.2 Finite control volume analysis, 197–255, 251–253 Finite difference method, 743 Finite element method, 742–743 I-4 Index Finite volume method, 744 First law of thermodynamics, 197, 234–250, 252–253, video V5.15, video V5.16 First moment of area, 60 First Tds equation, 251 Fittings, losses in, 439–440 Fixed control volume, 174–175, 200–206 Flaps, airfoil, 539, video V9.23, video V9.24 Flat plate drag coefficient, 508–509 Flat plates Blasius solution, 497–499, 498 drag coefficient normal to flow, 519, 521, 522, 525, video V7.2 parallel to flow, 519–520, 521, 522, 525 momentum integral equation, 499–504 laminar flow, 500 turbulent flow, 506 relative roughness on, 508 Flettner, A., 319 Floating ball water fountains, 325 Floating bodies, video V2.7 Archimedes’ principle, 70–73 stability, 73–74 Flow, in converging channel, 738–742 Flow coefficient, 701, 717 Flow measurement, 126–131 external flow Pitot tube, 114–115, 132, video V3.8 internal flow bellows meter, 466 magnetic flowmeter, 465 nozzle meter, 127, 460–464 nutating disk meter, 465–466, video V8.16 orifice meter, 127, 460, 461 rotameter, 464, video V8.15 turbine meter, 464–465 ultrasonic flowmeter, 465 Venturi meter, 127, 128, 460, 462–463, video V3.10 open-channel flow sluice gate, 128–130, 593–594 uniform, 570–572 weir, 130–131, 587–592, video V10.13 Flow meters. see Flow measurement Flow net, 301 Flowrate mass, 120, 199 normal (design), 694 of toilets, 206 volume, 120, 199 Flow separation, 489, 513–514, video V9.9, video V9.19 Flow visualization, 160, 161–165, video V4.1, video V4.6, video V4.9, video V7.16, video V9.19, video V11.5 Flow work, 236 Fluid dynamics vs. fluid statics, 11 Fluid machines, videos V12.1–V12.5 Fluid mechanics defined, 1 flow parameters, 1–2 history of, 27–29 Fluids defined, 3 describing characteristics, 3–4 flow parameters, 1–2 ideal, 295 inviscid, 293–300, 296 molecular structure, 3–4 Newtonian, 15 non-Newtonian, 16, video V1.6 properties, 11, 14, 755–760 viscous, 15 Fluid speed in Fanno flow, 650 of Rayleigh flow, 661 Fluid statics equilibrium in, 40 forces in, 59–64 pressure height relation, 43–44, 46–50 vs. fluid dynamics, 11 Force body, 42, 212, 290 buoyant, 70–73, video V2.7, video V2.9 compressibility, 364 drag, 316, 484–487, 515–534 hydrostatic on curved surfaces, 68–70, video V2.6 on plane surfaces, 59–64, video V2.5 lift, 316, 484–487 pressure, 55, 66 shear, 14 surface, 42–43, 212, 290–292 surface tension, 24–26, video V1.9 as unit of measure, 7–9 viscous, 15 Forced vortex, 305 Form drag, 517–519 Forward curved blades, 691 Forward difference, 739 Fountains, 446 Francis, J., 716 Francis turbine, 716, 717 Free jets, 118–120, video V3.9 Free-stream velocity, 510 Free surface, 59 Free-surface flow, 382–383 Free vortex, 109, 305, 307, video V3.6, video V6.4 Frequency parameter, 365 Friction, 647–659 Friction coefficient, 362, 502–503 Friction drag. see Drag Friction drag coefficient, 508 Friction factor data correlation for, 431 in dimensional analysis of laminar pipe flow, 413–415 for Fanno flow, 657–659 Moody chart, 430–432 noncircular conduits, 444 smooth pipe, 431 in wholly turbulent flow, 430–431 Frictionless flow, 632 compressible adiabatic .see Isentropic flow compressible with heat transfer .see Rayleigh flow with heating or cooling, 660–666 incompressible, 107, 295 open channel .see Open-channel flow in variable area duct, 632–647 Friction slope, 566, 580–581 Fricton velocity, 424 Frontal area, 487, 520 Frontinus, S., 27, 28 Froude, W., 28, 363 Froude number, 362, 363, 383–384, 387, 561, 564–566, video V7.4, video V10.5 defined, 32 and drag coefficient, 529–530 and Mach number, 667 Fuel pumps, 697 Fully developed flow applying Navier-Stokes equations, 412–413 laminar, 407–416, video V8.10 in pipes, 405–406 turbulent, 417–427 Fully developed region, 297 Gc, 9 Gage fluid, 53 Gage pressure, 13, 50 Galilei, G., 27, 28 “Galloping Gertie,” 374 Gas constants defined, 14 in ideal gas equation of state, 12, 605 universal, 605 Gas dynamics. see Compressible flow Gas turbine engines, 723–725 Gas turbines, 620 Gauss-Sidel iteration, 740 General homogeneous equations, 5 Geometric similarity, 372 Glass wool, 119 Grade line energy, 131–133, 566–567 hydraulic, 131–133 Gradient, pressure. see Pressure gradient Gradient operator, 166, 279, 298 Gradually varied flow (GVF), 560, 579–581 Grand Canyon, 584 Gravity, acceleration of, 7, 8, 761–762 Grids, 739, 743–746, 750–751, video VA.2 Guide vanes, 439, 705, 721 GVF. see Gradually varied flow Hagen, G., 28, 327, 410–411 Hagen-Poiseuille flow, 327, 410–411, video V6.14 Half-body, 309–312, video V6.5 Head elevation, 111, 132 in energy line concept, 131–132 loss .see Head loss piezometric, 132 pressure, 45, 111, 112, 132 pump, 690–692, 694–696 shaft, 245 shaft work head, 245 shutoff, 693 total, 111, 132 velocity, 111, 132 weir, 587–591 Head coefficient, 700, 717 Head curve falling (drooping), 693 rising, 693 Index Head loss defined, 245 in diffusers, 438 in enlargements and contractions, 436–439 in entrances, 435, 436, video V8.12 in exits, 436, video V8.12 in gradual contractions, 438 in hydraulic jumps, 583–584, video V10.11, video V10.12 major, 428–433 minor, 428, 433–443, video V8.14 in mitered bends, 439 in nozzles, 438–439 in open-channel flow, 567 in pipe bends, 439–440 in pipe entrances, 435 in pipes, 415, 438–440 in sudden area changes, 435 in valves and fittings, 439–440 Head loss coefficient, 433–436 Head rise coefficient, 700 Heat exchangers, 419 Heating, 660–666 Heat transfer net rate, 234 rate, expressed in first law of thermodynamics, 234 Hilsch, Rudolph, 610 Hilsch tube, 610 History of fluid mechanics, 27–29 Homogeneous flow, 560 Horseshoe vortex, 542, video V4.6, video V9.1, video V9.25 Housings, pump, 687 Hurricanes, 305 Hybrid grids, 746 Hydraulically smooth wall, 431 Hydraulic diameter, 330, 444 Hydraulic efficiency, 693 Hydraulic grade line (HGL), 131–133 Hydraulic jumps, 582–587, video V10.11, video V10.12 classification of, 586–587 depth ratio across, 583–584 as example of rapidly varied flow, 582 head loss across, 583–584 Hydraulic loss, 693 Hydraulic radius, 572 Hydraulics, in history of fluid mechanics, 27 Hydrodynamics, in history of fluid mechanics, 27 Hydrometers, 32, 72, video V2.8 Hydrostatic force on curved surfaces, 68–70, video V2.6 on plane surfaces, 59–64, video V2.5 Hydrostatic pressure, 113 Hydrostatic pressure distribution, 44–47 Hydrostatic pressure forces, 63 Hypersonic flow, 620 Ideal fluid, 295 Ideal gases compressible flow functions for, 763–770 isentropic flow of, 635–641 speed of sound, 614 thermodynamics of, 605–610 Ideal gas flow, 605, video V11.1 with area change, 635–641 Fanno, 652–659 isentropic, 627–630 Rayleigh, 662–666 Ideal gas law, 12–13, 605, video V11.1 Ideal head rise, 690–692 Impellers, 679, 687, 688 Implicit method, 744 Impulse function, 660 Impulse turbines, video V12.5 axial-flow, 715 defined, 708 dental drills as, 714–715 Pelton wheel example, 708–714 vs. reaction turbines, 708 Inclined tube manometer, 55–56 Incompressible flow, 107, 285, 295 Bernoulli’s equation for, 630–631 constant-area duct flow with friction vs., 647–648 Incompressible fluids, 20, 21, 44–48, video V1.7 Induced drag, 542 Inertial control volume, 212 Infinitesimal control volume, 277 Inhalers, 123 Inlet guide vanes, 721 Inlet velocity, 691 Intensity of turbulence, 420 Intensive property, 175–176 Internal energy, 234, 605–606, 608–609 International System (SI) of units, 7 Inviscid core, 405 Inviscid flow applying differential analysis, 293–300 applying Newton’s second law, 102–103 defined, 102, 293 Inviscid fluids, 105, 296 Irreversible flow, 251–253 Irrotational flow, 137, 282–283, 295–300, video V3.6, video V6.4 Irrotational vortex, 304–305 Isentropic efficiency, 722 Isentropic flow, 626–632 with area change, 635–641 basic equations for, 627–628 and Bernoulli’s equation, 630–631 and compressible flow condition, 134–135 critical state, 631–632 defined, 610 ideal gas, 627–630 for ideal gas, 763–764 Mach number in, 615 Isentropic process, 21, 607, 610 Isentropic stagnation properties, 612 Isothermal atmosphere, 47–48 Isothermal process, 21 Jacobi method, 749 Jet exit pressure, 118–119 Joule, 7 Jumps, hydraulic. see Hydraulic jumps Kaplan, V., 716 Kaplan turbines, 716–717 I-5 Kármán vortex street, 365 Kármán vortex trail, video V4.12, video V6.16, video V9.12 Kilogram, 7 Kinematic similarity, 372 Kinematics of fluid flow, 156–188, 278–283 Kinematic viscosity of air, 759–760 of common gases, 757 of common liquids, 757 defined, 20 of water, 758 Kinetic energy in first law of thermodynamics, 234 turbulence, 748 in work-energy principle, 111 Kinetic energy coefficient, 247–250 Knudsen number, 366 Kutta-Joukowski law, 319 Lagrangian description of flow, 159–160, video V4.4, video V4.5 Lake Nyos, 73 Laminar boundary layer Blasius solution, 497–499 description of, 492, video V9.4 effect of pressure gradient, 510–514 friction drag coefficient for, 503 Prandtl solution, 495–497 thickness, video V9.4 Laminar flow, video V4.7 in an annulus, 330–331 applying dimensional analysis, 413–415 in boundary layers, 492 between parallel plates, 323–327 past cylinders, 512 past spheres, 523–524 in pipes, 327–329, 403–405, 407–416, video V6.14, video V8.2 vs. turbulent flow, 161 Laplace’s equation, 298, 300 Laplacian operator, 298 Lapse rate, 49 Large eddy simulation models, 748, 753–754 Law of the wall, 424 Length ratio, 362 Length scale, 374 Lift, 316, 318–319, 484–487, 533–545 Lift coefficient airfoil, 537 defined, 487 as dimensionless group in fluid mechanics, 362 rotating cylinder or sphere, 543–545 writing, 535 Lift-drag polar diagram, 538 Lift to drag ratio, 538 Linear deformation, 278–280 Linear momentum. see Newton’s second law of motion Linear momentum equation, 211–226 application, 212–226 derivation, 211–212 Linear momentum flow, 211, video V5.4, video V5.5, video V5.6, video V5.8, video V5.9 I-6 Index Linear partial differential equations. see Laplace’s equation Liquid-jet scalpels, 161 Liquid knife, 641 Liquid mirror telescope (LMT), 79 Local acceleration, 168 Local derivative, 168 Local friction coefficient, 502–503 Logarithmic velocity profile, 424 Loss, major and minor. see Head loss Loss coefficient, 362, 433–436 Loss of available energy, 242–243, 247, 607–608 Low Reynolds number flow, 487–491, 521–523, video V1.3, video V7.7, video V8.6 Mach, E., 28, 364 Mach cone, 618–621, video V1.8, video V11.3, video V11.4 Machine Mach number, 723 Mach number, 23, 362, 364, 387, 612–617 and area for isentropic flow, 636–638 defined, 37 and drag coefficient, 526–527, 617 for Fanno flow, 652–654 and Froude number, 667 in isentropic flow, 628 and isentropic flow pressure, 631 and Mach cone angle, 619–620 and normal shock, 623–625 and stagnation properties, 614–617 Mach number tables, 616–617 Mach waves, 619, 668–669, video V1.8, video V11.3 Magnetic flowmeter, 465 Magnus, H., 543 Magnus effect, 319, 543 Major loss. see Head loss Manning, R., 28, 572 Manning coefficient, 572–574 Manning equation, 572–574 Manometer rule, 53, 54 Manometers defined, 52 differential, 54–55 inclined tube, 55–56 piezometer tube, 52–53, 113, video V2.3 U-tube, 53–55 Mass. see Conservation of mass Mass, units of in British Gravitational (BG) System, 8 in English Engineering (EE) System, 8 in International System (SI), 7 Mass flow density, 648–649, 660 Mass flow function, dimensionless, 639, 640, 643 Mass flowrate, 120, 199 Mass flow rate for converging nozzle, 642–643 in duct, 639–640 Mass vs. weight, 7, 8, 9 Material derivative, 165–168, 170–171, 182–183, 279 Maximum flowrate, 572 Measurement, flow. see Flow measurement Mechanical efficiency, 693 Mechanical energy equation, 243 Mechanical heart assist devices, 707 Mechanical loss, 693 Meniscus, 52 Mesh. see Grids Meters, flow. see Flow measurement Method of repeating variables, 350–352, 354–355 Method of superposition, video V6.5 Minor loss, 428, 433–443 Minor loss coefficient, 433–436 Mixed-flow machines, 680 Mixed-flow pumps, 704–707 Mixing length, 423 Model design conditions, 370 Modeling laws, 370 Models defined, 370 distorted, 374–375, video V7.8, video V7.12, video V7.17, video V7.18 practical use of, 347 scales for, 373–374 similitude concept, 346 theory of, 370–373 true, 374, 375 typical studies with, 376–385, video V7.1, video V7.9, video V7.18, video V7.19, video V7.20 Moment of inertia, 61 Moment-of-momentum, 685–687 application of equation, 227–234, video V5.13 derivation, 226–227 turbomachine considerations, 231, 685–687 Moment-of-momentum equation, 250 Momentum angular .see Moment-of-momentum linear .see Newton’s second law of motion Momentum coefficient, 221 Momentum equation. see also Linear momentum equation for control volume, 211, 223, 290 for inviscid flow, 293–300 and wheel brake design, 213 Momentum flux, 221, 222, video V5.7 Momentum integral equation boundary layer flow on flat plate, 499–504 boundary layer with nonzero pressure gradient, 514–515 boundary layer with zero pressure gradient, 500 Momentum thickness, 495 Moody, L., 29, 430 Moody chart, 430–432 Motorized surfboard, 217 Moving grids, 746 Multigrid methods, 750 Multiple pipe system. see Pipe systems Multi-stage compressors, 719–720 Multi-stage pumps, 688 Multistage turbines, 697 Nanoscale, 404 Nappe, 590 Navier, L., 28, 322 Navier-Stokes equation applying to laminar flow, 323–325, 412–413 cautions on use of, 332 cylindrical polar coordinates, 322 defined, 321–322, 332 rectangular coordinates, 321–322 Net positive suction head, 694–696 Newton (unit), 7 Newton, I., 27, 28 Newtonian fluids, 15, 320–322 Newton’s law of viscosity, 421 Newton’s second law of motion, 41, 101–110, 197, 211–234, 290 along a streamline, 104–108 and Bernoulli equation, 241 normal to a streamline, 108–110 Nikuradse, J., 430 Nodes, 742 Noncircular conduits, 443–446 Nonconservative form, 750 Noninertial reference frame, 212 Non-Newtonian fluids, 16, video V1.6 Nonuniform flow, 246–250, 560 Normal acceleration, 103, 171–173, video V3.1, video V3.4 Normal depth, 574 Normal flowrate, 694 Normal shock flow for ideal gas, 765–767 Normal shock waves, 621–625, video V11.7 basic equations, 622–625 defined, 621 examples, 625–626 and Rayleigh/Fanno lines, 649, 666 Normal stress in conservation of linear momentum, 290 in Newtonian fluids, 15, 320, 321 and work transfer, 235–236 No-slip condition, 14, 311, 324, video V1.4, video V6.12, video V6.13 Nozzle discharge coefficient, 462 Nozzle meters, 127, 460–464 Nozzles converging, 641–643, 646–647 converging-diverging, 634, 643–647, video V11.6, video V11.7 defined, 118 for inhalers, 123 loss coefficients, 438–439 overexpanded flow, 645, video V11.6, video V11.7 rocket, 646 thrust, video V5.11 underexpanded flow, 645, video V11.7 Numerical methods, 332, video V6.15 Nutating disk meters, 465–466, video V8.16 Oblique shock waves, 621, 669, video V1.8, video V11.6, video V11.7, videos V11.9–V11.11 One-dimensional compressible flow in constant-area duct with friction, 647–659 in constant-area duct with heating or cooling, 660–666 in variable area duct, 632–647 One-dimensional flow, 160–161 One-equation model, 427 Index Open-channel flow, 559–595 analogy with compressible flow, 667–668 critical flow, 561, 565 defined, 559 energy considerations with, 566–570 gradually varied flow, 579–581 gradually varying, 560 hydraulic jump, 582–587, video V10.11, video V10.12 laminar, 560 Manning equation for, 572–574 normal depth, 574 rapidly varied flow, 581–594 rapidly varying, 560 surface wave propagation, 561–566 transitional, 560 turbulent, 560 uniform flow, 560, 570–579 varied flow, video V10.6, video V10.8 Open channels best hydraulic cross sections, 578–579 characteristics of, 559–561, video V10.6 geometric properties of, 570–574, video V10.7 shear stress in, 571–572 velocity contours in, 571 Open impellers, 688 Operating point, 697 Orifice discharge coefficient, 461 Orifice meters, 127, 460, 461 Orifices, flow through, 119 Outer layer, turbulent pipe flow, 422, 423 Outlet velocity, 691 Overall efficiency, pump, 693 Overexpanded flow, 645, video V11.6, video V11.7 Overlap layer, turbulent pipe flow, 422, 423 Panel method, 743 Parachutes, 372 Parallel plates Couette flow, 325–327 steady laminar flow between, 323–325 Particle image velocimeter (PIV), 157 Pascal (unit), 13 Pascal, B., 28, 41 Pascal’s law, 41 Pathlines, 163–165, video V4.11 Pelton wheels, 708–714, video V5.6, video V12.2, video V12.4 Perfect gas law. see Ideal gas law Perfect guiding assumption, 681 Perfectly expanded flow, 645 Physical properties of fluids of air, 759–760 of common gases, 756–757, 759–760 of common liquids, 756–758 of U.S. Standard Atmosphere, 761–762 of water, 758 Piezometers, 52–53, 113, video V2.3 Piezometric head, 132, video V2.3 Pipes, 401–467 aging, 430–431 compressible adiabatic flow in .see Fanno flow deepwater, 456 dimensional analysis of flow in, 427–446 fittings, 439–440 flow characteristics, 402–407 flow meters, 460–465 flow rate, 460–465 fully developed flow in, 327–329, 405–416 head loss .see Head loss heat transfer in .see Rayleigh flow hydraulically smooth, 431 laminar flow in, 327–329, 403–405, video V6.14 laminar vs. turbulent flow in, 405–406 noncircular, 443–446 pressure and shear stress in, 406–407 relative roughness of, 429–432 smooth, 431 standard components, 439–441 transitional flow in, 403, 417–419, video V8.2, video V8.3, video V8.9 turbulent flow in, 403–405, 417–427, video V8.2, video V8.3 typical system components, 401, 402 vs. ducts, 401 Pipe systems, video V6.15 loop, 457–458 multiple pipes, 456–460 networks, 459–460 parallel, 456, 457 series, 456–457 single pipes, 446–455 Pi terms defined, 350 determination by inspection, 359–360 determination by method of repeating variables, 350–355 determination example, 357–358 problems with, 366–369 uniqueness of, 358–359 Pi theorem, 349–350 Pitot, H., 28 Pitot tubes, 114–117, 132, 626, video V3.8 Plane Poiseuille flow, 329 Planform area, 487, 520, 537 Poise, 20 Poiseuille, J., 28, 327, 410–411 Poiseuille flow, 327, 329, 410–411, video V6.14 Poiseuille’s law, 329, 411 Polar coordinates, 77 Polar diagram, lift-drag, 538 Polytropic efficiency, 722 Position vector, 157 Potential energy, 111, 234 Potential flow theory, video V6.9, video V6.10 basic plane flows, 300–301, 309, video V6.6, video V6.7, video V6.11 doublet, 307–309 sink, 302–304 source, 302–303 uniform flow, 301–302 vortex motion, 304–307 defined, 298 finding solutions to problems, 319–320 singularity, 303 superposition of basic plane flows, 309–319, video V6.5 flow past circular cylinder, 314–317 flow past half-body, 309–312, video V6.5 flow past Rankine oval, 312–314 flow past rotating cylinder, 317–319 Potential function, 298 Pound, 8 Power as expression of work transfer rate, 235 and lift for flight, 536–537 shaft, 686, 689 units of, 7 Power coefficient, 700, 717 Power-law velocity profile, 424 Power specific speed, 717–719 Prandtl, L., 27–29, 423, 489 Prandtl boundary layer equations, 495–497 Prediction equation, 370 Pressure absolute, 13, 50 atmospheric, 13, 52, 761–762 center of, 62 defined, 12–13, 40 and drag, 485–486 dynamic, 113 gage, 13, 50 hydrostatic, 44–47, 113 and ideal gas law, 12–13 in isentropic flow, 627, 628, 630–631 isentropic stagnation .see Isentropic stagnation properties and lift, 536–537 measurement of, 50–52, 114, 117, video V2.3, video V2.4 and normal shock, 623–624 at a point, 40–41, 103, video V3.7 and raindrops, 106 and shear stress in pipes, 406–407 stagnation .see Stagnation pressure static, 113 suction, 50 thermodynamic, 113 of tires, 58 total, 114 vacuum, 50 vapor, 23–24 Pressure at a jet exit, 118–119 Pressure coefficient, 362, 364, 518 Pressure correction algorithm, 752 Pressure distribution airfoils, 483 automobile, 535, video V2.1 cylinder, inviscid flow, 315 cylinder, viscous effects, 513–514 entrance of pipe, 406, 435 hydrostatic, 44–47 in rigid-body motion, 74–79 in rotating fluids, 77–78 standard atmosphere, 48–50 Pressure drag. see Drag Pressure drag coefficient, 517–519 Pressure forces, 41–43 Pressure gradient adverse, 314, 511 effect on surface forces, 43 effects on boundary layer, 510–514, video V9.10 favorable, 511 Pressure head, 45, 111, 112, 132 Pressure prism, 65–68 Pressure recovery coefficient, 438 I-7 I-8 Index Pressure-sensitive paint, 485 Pressure tap, 116, 117, 132 Pressure transducers, 57–59 defined, 57 illustrated, 57, 59 role of strain gages, 58–59 Pressure-velocity coupling, 751, 752 Pressure vessels, 70 Pressure waves, 618–620 Primary quantities, 4 Primitive variable equation, for mass flow rate, 639 Primitive variables, 738 Product of inertia, 61 Profile, velocity. see Velocity profile Propeller pumps, 705 Properties of air, 759–760 of common gases, 756–757, 759–760 of common liquids, 756–758 of U.S. Standard Atmosphere, 49, 761–762 of water, 758 Prototype, 370 Pump affinity laws, 703 Pump performance curves, 692, 697–700 Pumps axial-flow, 704–707 centrifugal, 687–700, 707, video V12.3 defined, 679 dimensionless parameters, 700–701 efficiency, 246 geometrically similar, 700–701 head, 690, 694–696 ideal vs. actual head rise, 690–692 mixed-flow, 704–707 parallel, 699–700 performance characteristics, 691–694 propeller, 705 scaling laws, 701–703 series, 699 similarity laws, 701–703 specific speed, 703–704 stages in, 705 system equation, 696 vs. fans, 707–708 vs. turbines, 684–685 Quasi-steady flow, 136 Radial blades, 691 Radial-flow compressors, 720 Radial-flow machines, defined, 680, 681 Raindrops, 106 Rankine ovals, 312–314 Ranque, George, 610 Ranque vortex tube, 610 Rapidly varied flow (RVF), 560, 581–594, video V10.9, video V10.10 Rate of angular deformation, 283, video V6.3 Rate of heat transfer, 234 Rate of shearing strain, 15, 283 Rayleigh, Lord, 29 Rayleigh flow, 660–666 basic equations for, 660–662 in combustion chamber, 665–666 defined, 660 for ideal gas, 769–770 of ideal gas, 662–666 normal shock wave in, 666 Rayleigh line, 660–662, 664–666 Reaction turbines, 708, 716–718 Rectangular weirs, 130–131, 589 Reentrant entrance, 435 Reference dimensions, 350, 357–358 Relation parameters, 750 Relative roughness, 362, 429–432, 508 Relative velocity in centrifugal pumps, 688–689 defined, 206, 681 and Reynolds transport theorem, 186 in rotating sprinkler example, 229 as turbomachine energy consideration, 681–684 vs. absolute velocity, 185–186 Relaxation, 750 Repeating variable method, 350–352, 354–355 Required net positive suction head, 694, 695 Residuals, 742, 750 Resistance coefficient, Manning, 572–574 Restricted homogeneous equations, 5 Reversible procesess, 607 Reynolds, O., 29, 363, 403, 422 Reynolds averaged Navier-Stokes equations, 747 Reynolds number critical .see Transition to turbulence defined, 18, 363 as dimensionless group in fluid mechanics, 362 and drag coefficient, 617 large vs. low, properties of flows, 487–491, video V1.1, video V1.2, video V7.3, video V7.7, video V9.15 similarity, 387 Reynolds pipe flow experiment, 403, video V8.2 Reynolds stress, 422, 423 Reynolds transport theorem, 176–187 defined, 175 derivation, 177–182 example of use, 179 general form, 182 obtaining for moving control volume, 185–186 physical interpretation, 182 relationship to material derivative, 182–183 role in selection of control volume, 186–187 steady effects, 183 unsteady effects, 184–185 Rheology, 3 Right-hand rule, 229 Rigid-body motion of fluids, 74–79 Ripple tank, comparison of compressible and open-channel flows, 667 Rising head curve, 693 Rocket nozzles, 646 Rogue waves, 561 Rotameters, 464, video V8.15 Rotating cylinder viscometer, 36 Rotation, 137, 278, 280–283 Rotor blades, 705, 721 Rotors, turbomachine, 228, 232, 680, 682 Roughness effect on models, 376, 382, 383, video V7.13 on flat plates, 508 in open channels, 572–574, 577–578 in pipes, 429–432 relative, 429–432 typical values of, 430 Round-off errors, 744 Runners, 679 Scales length, 374 types for models, 373–374 Scales, turbulence model, 746 Scaling laws for pumps, 701–703 Schlieren visualization, of flow, 619–620 Secondary flow, in pipes, 439 Secondary quantities, 4 Second law of thermodynamics, 197, 251–253 combination of first and, 252–253 semi-infinitesimal control volume statement of, 251–252 Second moment of area, 61 Semi-infinitesimal control volume statement of energy equation, 251 of second law of thermodynamics, 251–252 Separation, video V6.8, video V9.9, video V9.16, video V9.19 in boundary layers, 297, 489, 511–513 in pipes, 435, 439 Separation location on airfoils, 513–514 on cylinders, 489, 511–512 defined, 511 illustrated, 490 Shaft power, 230, 686, 689 Shaft torque, 230, 685, 689 Shaft work, 230, 237, 243, 245, 686–687 Shaft work head, 245, 690, 692 Shallow water waves, 564 Sharp-crested weirs, 130–131, 587–590, video V10.13 Shearing strain defined, 14 rate of, 15, 283, video V6.3 Shearing stress defined, 3 distribution in pipes, 409, 415 drag from, 485–486 effect on fluids, 3, 14, 15–17, 19 as force acting on differential element, 291 in laminar flow, 420–421 lift from, 536–537 in Newtonian fluid, 321 in open-channel flow, 571–572 and pressure in pipes, 406–407 in turbulent flow, 419–423 in work transfer, 236 Shear thickening fluids, 16 Shear thinning fluids, 16 Ships, 319 Shock absorbers, 247 Shock regime, 645 Shock waves, 621–626, videos V11.6–V11.11 curved, 621, 622 in gas turbines, 620 Index normal, 621–625 .see Normal shock waves oblique, 621, 669, video V11.6, video V11.7, videos V11.9–V11.11 Shrouded impellers, 688 Shutoff head, 693 Similarity laws, pump, 701–703 Similarity requirements, 370–372 Similarity variable, 497, 498 Similitude based on differential equations, 385–388 as concept in experimentation, 346 Single-stage pumps, 688 Single-suction impellers, 688 Sink flow defined, 303 example, 303–304 strength of, 303 Sinusoidal waves, 563–564 Siphon, 125–126 SI (International System) units, 7 Slip, in experimental data, 366 Slope of channel bottom, 566–567, 572 energy line, 131–133, 566–567 friction, 566, 580–581 Slug, 8 Sluice gates discharge coefficient, 593 flow rate under, 128–130, 593–594 force on, 222–223 free outflow, 593 submerged, 593–594 Smooth pipe, 431 Sonar systems, 421 Sonic flow, 619 Sonification, 614 Sound, speed of. see Speed of sound Sound waves, patterns in, 614 Source flow, 302–303, video V6.5 Space shuttle, fuel pump of, 697 Specific energy, 567–570 Specific energy diagram, 567–568 Specific gravity, 12, 26 Specific head. see Specific energy Specific heat constant pressure, 21, 606–607, 610 constant volume, 21, 605–607, 610 Specific heat ratio of air, 759–760 defined, 21, 607 Specific speed for pumps, 703–704 and turbine efficiency, 717–719 Specific volume, 11 Specific weight of air, 759–760 of water, 758 Speed of sound, video V11.3 defined, 22–23 ideal gas, 22, 614 and Mach number, 612–617 table of values for air at standard atmospheric pressure, 759–760 table of values for water at standard atmospheric pressure, 758 Spheres drag coefficient for, 381, 521–524, 527–528, 543, 544, video V9.17 effect of surface roughness, 527–528 low Reynolds number flow past, 521 rotating, 543–545 Spillways, 584, video V10.15, video V10.16 Stability, 73–74, video V2.11, video V2.12 Staggered grid, 750–751 Stagnation density, 611 Stagnation enthalpy, 611 Stagnation point, 113, video V3.7 Stagnation pressure, 114, video V3.8 across normal shock waves, 625 and critical state, 632 defined, 611 in Fanno flow, 654 isentropic .see Isentropic stagnation properties in isentropic flow, 627 and Mach number, 615 in Rayleigh flow, 664 Stagnation pressure probe. see Pitot tubes Stagnation properties, 610–612. see Isentropic stagnation properties in isentropic flow, 627 and Mach number, 614–617 mass flow rate in terms of, 639 and normal shock, 624 Stagnation streamline, 114, video V3.7, video V4.9 Stagnation temperature across normal shock waves, 624 and critical state, 631 defined, 611 in isentropic flow, 627 and Mach number, 614–615 in Rayleigh flow, 663–664 Stall condition, 514, 542, video V9.21 Standard atmosphere defined, 49 pressure variation, 48–50 properties of, 49, 761–762 State, equation of, 12–13, 605 Static fluids, pressure variation in, 43–48 Static pressure, 113 Static pressure tap, 116, 117, 132 Static properties, 610–612, 624, 639 Stator blades, 705, 721 Steady flow, video V4.7 defined, 161 fluid particles, 102 and Reynolds transport theorem, 183 streaklines in, 163 streamlines in, 162 Steam tables, 239 Steel wool, 119 Stoke, 20 Stokes, G., 28, 322, 522 Stokes’ law, 367 Stratified flow, 560 Stratosphere, 49 Streaklines, 163–165, video V4.10, video V7.14 Stream function, 286–289 Streamline bodies, video V2.1, video V9.2, video V9.20 Streamline coordinates, video V4.13 I-9 Streamline curvature, 102–103 Streamlined bodies, 484 Streamlines, video V3.1, video V4.9 defined, 102 equation, 287 and Newton’s second law of motion, 104–110 use of coordinate systems, 171–173 in visualization and analysis of flow fields, 162–165 Streamtube, 237–238 Streamwise acceleration, 103 Stress Newtonian fluid, 15, 320–321 normal, 235–236, 290, 321 notation, 291 shearing .see Shearing stress sign convention, 291 tangential, 236 turbulent, 419–423 Strouhal, V., 29, 365 Strouhal number, 362, 364–365, 386, video V7.5 Structured grids, 745 Strutt, J. W., 660 Subcritical flow, 561, 565–566, 568, video V10.5 Submerged area, forces on, 59–62 Subsonic flow, 23, 619 in coverging/diverging duct, 633–634 Fanno, 650 Rayleigh, 661, 663 Substantial derivative, 165–168, 279 Suction, 50 Suction specific speed, 704–705 Sudden contraction, 437 Sudden expansion, 437 Supercritical flow, 561, 565–566, 568, video V10.5 Superposition, 309–319, video V6.5 Supersonic flow, 23, 619, video V1.8, video V11.6, video V11.7 in coverging/diverging duct, 633–634 Fanno, 650, 659 in gas turbines, 620 in pitot tube, 626 Rayleigh, 663 two-dimensional, 668–669 Surface forces, 42–43, 212, 290–292 Surface pressure, 485 Surface tension, video V1.9, video V7.6, video V10.3 defined, 24–26 effect on models, 382–383 and oil spills, 26 of water, 758 Surface waves, 561–566, video V7.4, video V9.18, videos V10.1–V10.5 Sutherland equation, 18 System curve, 696–697 System equation, 696 Systems, 173–175 Systems of units, 7–9 Tacoma Narrows Bridge, 374 Tangential stress work, 236 T ds equations, 608 I-10 Index Temperature absolute, 7, 8 Celsius, 7 centigrade, 7 Fahrenheit, 7 in isentropic flow, 627, 628 Kelvin, 7 and normal shock, 623–624 Rankine, 7 of Rayleigh flow, 662 stagnation .see Stagnation temperature of U.S. Standard Atmosphere, 761–762 Temperature–entropy (T–s) diagrams, 630, 649 Terminal velocity, of falling object, 525–526 Thermal choking, 662 Thermodynamic pressure, 113. see Pressure Thermodynamics for Fanno flow, 652–654 first law, 197, 234–250 of ideal gases, 605–610 second law, 197, 251–253 static properties, 610 Three-dimensional flow, 160 Three-dimensional objects, 484, 532–534 Three Gorges Dam, 63 Three-point difference, 741 Three-reservoir problem, 458–459 Throat, duct, 634, 635 Thrust, 221–222, video V5.11 Tidal waves. see Tsunamis Time-averaged equations, 427 Time-averaged velocity, 418–420 Tire pressure, 58 Toilets, 206 Torque in moment-of-momentum equation, 226–232 shaft, 230, 685, 689 Torque converter, 687 Torricelli, E., 28, 51, 118 Torricelli’s formula, 118 Total head, 111, 132 Total pressure, 114 Total pressure ratio, 719 Total properties, 612 Trailing vortex, 542, video V4.6, video V9.1, video V9.25 Tranquil flow, 561 Transition to turbulence, video V9.5 in boundary layer flows, 492, 504–505, video V9.7 in fully developed flows, 417–419 in open-channels, 560 in pipe flows, 403, 417–419, video V8.2, video V8.3 Translation, in fluid element flow, 278–380 Transonic flow, 620 Triangular weirs, 131, 589–590, video V10.13 Troposphere, 49–50 True models, 374, 375 Truncation, 743 Truncation errors, 744 Tsunamis, 564 Turbine flow meters, 464–465 Turbines, 708–719 axial-flow, 681 compressible flow, 719, 723–725 defined, 708 Francis, 716, 717 gas, 723–725 impulse, 708–715, video V12.4, video V12.5 Kaplan, 716–717 multistage, 697 Pelton wheel, 708–714, video V5.6, video V5.14, video V12.4 performance maps, 725 radial-flow, 681 reaction, 708, 716–718 vs. pumps, 684–685 wind, 683 Turbomachines, 679–726 angular momentum considerations, 685–687 axial-flow, 680, 705–707, 720–722 basic energy considerations, 681–685 centrifugal pumps, 687–700, video V12.3 compressible flow, 719–725 defined, 679–680 dimensionless parameters, 700–701 fans, 707–708 mixed-flow, 680, 705–707 pump scaling laws, 701–703 radial-flow, 680, 720 rotary sprinkler, 228–231, video V4.11, video V5.12, video V12.2 suction specific speed, 704–705 turbines, 708–719 Turbopumps, 697 Turbulence chaos, 427 characteristics of, 161, 419–421, video V1.1, video V8.1, video V8.7, video V8.8, video V8.11 intensity of, 420 mixing in, 419, 421, video V1.1, video V8.4, video V8.5, video V8.7 modeling, 427 Reynolds stress in, 422 sensitivity of bats to, 538 spots in, 504 Turbulence kinetic energy, 748 Turbulence models, 746–748 Turbulent boundary layer approximate solution for flat plate, 506–508 effect on separation, 513–514 properties, 506–508, video V9.8 transition from laminar boundary layer, 492 velocity profiles in, 505 Turbulent jet, video V8.1 Turbulent pipe flow, video V8.2, video V8.3, video V8.9 fully developed, 417–427 general characteristics, 403–405 logarithmic, 424 mean vs. fluctuating velocity, 419–420 overlap layer, 422, 423 power law, 424 shear stress distribution, 409, 420, 422 velocity profile, 423–425, video V8.2 viscous sublayer, 422–423 Turbulent stresses, 419–423 Two-dimensional flow, 160 Two-dimensional objects, 483–484, 532 Two-dimensional supersonic flow, 668–669 Two-equation model, 427 Ultrasonic flowmeter, 465 Unchoked flow, 643, 655–657 Underexpanded flow, 645, video V11.7 Underflow gates, 593–594 Underrelaxed calculations, 750 Uniform flow in open channels, 560, 570–579 potential, 301–302 in sections, 200 Uniform flow field, 295 Units conversion factors, 9 defined, 4 need for consistency, 9–11 systems of, 6–9 Universal gas constant, 605 Unsteady Bernoulli equation, 135 Unsteady flow, 135–136, 161–163, 168, 184–185, 204–205, video V3.11, video V4.7, video V5.10 Unstructured grids, 745 Upstream velocity, 483 Upwind differencing, 751 U.S. standard atmosphere. see Standard atmosphere Useful energy. see Available energy U-tube manometers, 53–55 Vacuum pressure, 50 Valve losses, 440 Valves, 434, 440–441 Vanes, 687. see also Guide vanes Vapor pressure in barometer tubes, 51 boiling, 23–24 defined, 23–24 effect on cavitation, 24 of water, 758 Variable area duct one-dimensional compressible flow in, 632–647 Varied flow. see Gradually varied flow (GVF); Rapidly varied flow (RVF) Velocity absolute, 206, 228–229, 681–684, 688–689 absolute vs. relative, 185–186 average, 199–200 blade, 681–684, 688–689 friction, 424 inlet, 691 outlet, 691 particle, 102, 157 Pressure-velocity coupling, 751, 752 relative, 206, 229, 681–684, 688–689 time-averaged, 418–420 Velocity diagrams, 682–683, 686, 689, 715 Velocity distribution. see Velocity profile Velocity field, 156–165, 279, video V4.2, video V4.3, video V6.1, video V6.2 Velocity gradient, 14 Velocity head, 111, 132 Velocity measurement. see Flow measurement Velocity potential, 297–298 Index Velocity profile in boundary layers effect of pressure gradient on, 511, 512 laminar flow, 503 in transition from laminar to turbulent flow, 505 turbulent flow, 506–508 between flat plates, 323–325 in open channels, 571 in pipes laminar flow, 328, 409–410, video V8.10 power law, 424 turbulent flow, 423–425, video V8.10 Velocity ratio as dimensionless group in fluid mechanics, 362 in Fanno flow, 653 Velocity vector, 157 Vena contracta, 119, 129, 435 Venturi, G., 28 Venturi discharge coefficient, 463 Venturi meters, 127, 128, 460, 462–463, video V3.10 Venturi regime, 645 Venturi tube, 635 Viscometers capillary tube, 33, video V1.5 rotating cylinder, 36 Viscosity, video V1.3 absolute or dynamic, 15 of air, 759–760 apparent, 16 of common gases, 756–757 of common liquids, 756–757 defined, 14 eddy, 423, 747–748 effect of temperature on, 17–18 kinematic, 20 and oil spills, 26 of U.S. Standard Atmosphere, 761–762 of water, 758 Viscous dissipation, 415 Viscous flow, 320–322 Viscous fluids, 15 Viscous stresses, 321 Viscous sublayer, turbulent pipe flow, 422–423 Visualization flow, 160, 161–165, video V4.6, video V4.9, video V7.16, video V9.19 Volume flow meters, 465–466 Volume flowrate, 120 Volumetric dilatation, 280 Volumetric efficiency, 693 Volumetric loss, 693 Volutes, 687 Von Kármán, T., 29, 365, 500 Vortex, 304–307 bound, 542, video V4.6, video V9.25 combined, 305, video V4.8 defined, 304–305 forced, 305 free, 109, 305, 307, video V3.6, video V6.4 horseshoe, 542 irrotational, 304–305 Kármán trail, video V6.16, video V9.14 ring, video V4.1 rotational, 305 strength of, 305 tornado, video VA.3 trailing, 542, video V3.5, video V4.6, video V9.1, video V9.25 Vortex shedding, 365, 372, video V4.12, video V7.5, video V7.10, video V9.14 Vortex street, video V6.16, video V7.5, video V9.12 Vorticity, 282–283, 295, 298, 492 Wake, video V9.3, video V9.9 and boundary layer, 297 streamlined object, 520 vortex street, video V6.16, video V9.14 Wake region defined, 489 flat plate, 489 Wall friction factor, 651 Wall functions, 748 Wall shear stress, 409, 415, 499 Water, table of properties, 758 Water horsepower, 693 Water jet, 20, 112 Water table, comparison of compressible and open-channel flows, 667 Water tunnels, 372 Watt, 7 Wave motion, 563–564 Waves calculating speed, 561–564 deepwater, 564 expansion, 668, video V11.9 gravity, video V10.4, video V11.4 mach, 619, 668–669, video V1.8, video V11.3 pressure, 618–620 rogue, 561 role of gravity, 562–563 shallow water, 564 I-11 shock .see Shock waves sinusoidal, 563–564 sound, 614 surface, 561–566, video V9.18, video V10.3, video V10.4, video V10.5, video V11.4 Wave speed, 561–564, 668, video V11.3 Weather, 52 Weber, M., 29, 365 Weber number, 362, 365, 387, video V7.6 Weight vs. mass, 7, 8, 9 Weir coefficient, 589–591 Weir head, 587–591 Weirs broad-crested, 590–592, video V10.14 defined, 130, 587 rectangular, 130–131, 589 sharp-crested, 130–131, 587–590, video V10.13 triangular, 131, 589–590, video V10.13 Weisbach, J., 28 Wetted perimeter, 330, 444, 570–571 Wheel brake design, 213 Wheels, 679 Wholly turbulent flow, 430–431 Wicket gates, 716 Windmills, 683, video V12.1 Wind tunnels, 442, video V3.3, video V7.15, video V9.11, video V9.22 Wind turbines, 683 Winglets, 543 Wing loading, 537 Wing tip vortices, 282, 542–543, video V3.5 Work expressed in first law of thermodynamics, 235 flow, 236 normal stress, 235–236 rate of, 230, 235 shaft, 230, 237, 243, 245, 686–687 shear stress, 236 tangential stress, 236 and turbomachines, 685–687 unit of, 7 Work-energy equation, 110–111 Zero-equation model, 427 Zero equation model, 748 Zone of action, 619 Zone of silence, 619 68 68 60 68 68 60 60 60 Temperature (°F) 3.09 1.53 1.32 2.44 26.3 1.77 1.99 1.94 99.5 49.3 42.5 78.6 847 57.0 64.0 62.4 Specific Weight, γ (lbft3 ) 20 20 15.6 20 20 15.6 15.6 15.6 Temperature (°C) b In contact with air. Isentropic bulk modulus calculated from speed of sound. c Typical values. Properties of petroleum products vary. a Carbon tetrachloride Ethyl alcohol Gasolinec Glycerin Mercury SAE 30 oilc Seawater Water Liquid 1,590 789 680 1,260 13,600 912 1,030 999 Density, ρ (kgm3 ) 15.6 7.74 6.67 12.4 133 8.95 10.1 9.80 Specific Weight, γ (kNm3 ) ■ Table 1.6 Approximate Physical Properties of Some Common Liquids (SI Units) b In contact with air. Isentropic bulk modulus calculated from speed of sound. c Typical values. Properties of petroleum products vary. a Carbon tetrachloride Ethyl alcohol Gasolinec Glycerin Mercury SAE 30 oilc Seawater Water Liquid Density, ρ (slugsft3 ) ■ Table 1.5 Approximate Physical Properties of Some Common Liquids (BG Units) Kinematic Viscosity, ν (m2s) 6.03 E − 7 1.51 E − 6 4.6 E − 7 1.19 E − 3 1.15 E − 7 4.2 E − 4 1.17 E − 6 1.12 E − 6 9.58 E − 4 1.19 E − 3 3.1 E − 4 1.50 E + 0 1.57 E − 3 3.8 E − 1 1.20 E − 3 1.12 E − 3 6.47 E − 6 1.63 E − 5 4.9 E − 6 1.28 E − 2 1.25 E − 6 4.5 E − 3 1.26 E − 5 1.21 E − 5 2.00 E − 5 2.49 E − 5 6.5 E − 6 3.13 E − 2 3.28 E − 5 8.0 E − 3 2.51 E − 5 2.34 E − 5 Dynamic Viscosity, μ (N ∙ sm2 ) Kinematic Viscosity, ν (ft2s) Dynamic Viscosity, μ (lb ∙ sft2 ) 2.69 E − 2 2.28 E − 2 2.2 E − 2 6.33 E − 2 4.66 E − 1 3.6 E − 2 7.34 E − 2 7.34 E − 2 Surface Tension,a σ (Nm) 1.84 E − 3 1.56 E − 3 1.5 E − 3 4.34 E − 3 3.19 E − 2 2.5 E − 3 5.03 E − 3 5.03 E − 3 Surface Tension,a σ (lbft) E+4 E+3 E+4 E−2 E−1 — 1.77 E + 3 1.77 E + 3 1.3 5.9 5.5 1.4 1.6 Vapor Pressure, pυ [Nm2 (abs)] E+0 E−1 E+0 E−6 E−5 — 2.56 E − 1 2.56 E − 1 1.9 8.5 8.0 2.0 2.3 Vapor Pressure, pυ [lbin.2 (abs)] 1.31 E + 9 1.06 E + 9 1.3 E + 9 4.52 E + 9 2.85 E + 10 1.5 E + 9 2.34 E + 9 2.15 E + 9 Bulk Modulus,b Eυ (Nm2 ) 1.91 E + 5 1.54 E + 5 1.9 E + 5 6.56 E + 5 4.14 E + 6 2.2 E + 5 3.39 E + 5 3.12 E + 5 Bulk Modulus,b Eυ (lbin.2 ) 59 68 68 68 68 68 68 2.38 E − 3 3.55 E − 3 3.23 E − 4 1.63 E − 4 1.29 E − 3 2.26 E − 3 2.58 E − 3 7.65 E 1.14 E 1.04 E 5.25 E 4.15 E 7.28 E 8.31 E −2 −1 −2 −3 −2 −2 −2 Specific Weight, γ (lbft3 ) 3.74 E 3.07 E 4.09 E 1.85 E 2.29 E 3.68 E 4.25 E −7 −7 −7 −7 −7 −7 −7 Dynamic Viscosity, μ (lb ∙ sft2 ) 15 20 20 20 20 20 20 b Values of the gas constant are independent of temperature. Values of the specific heat ratio depend only slightly on temperature. a Air (standard) Carbon dioxide Helium Hydrogen Methane (natural gas) Nitrogen Oxygen Gas Temperature (°C) 1.23 E + 0 1.83 E + 0 1.66 E − 1 8.38 E − 2 6.67 E − 1 1.16 E + 0 1.33 E + 0 Density, ρ (kgm3 ) 1.20 E 1.80 E 1.63 E 8.22 E 6.54 E 1.14 E 1.30 E +1 +1 +0 −1 +0 +1 +1 Specific Weight, γ (Nm3 ) 1.79 E 1.47 E 1.94 E 8.84 E 1.10 E 1.76 E 2.04 E −5 −5 −5 −6 −5 −5 −5 Dynamic Viscosity, μ (N ∙ sm2 ) ■ Table 1.8 Approximate Physical Properties of Some Common Gases at Standard Atmospheric Pressure (SI Units) b Values of the gas constant are independent of temperature. Values of the specific heat ratio depend only slightly on temperature. a Air (standard) Carbon dioxide Helium Hydrogen Methane (natural gas) Nitrogen Oxygen Gas Temperature (°F) Density, ρ (slugsft3 ) ■ Table 1.7 Approximate Physical Properties of Some Common Gases at Standard Atmospheric Pressure (BG Units) −4 −5 −3 −3 −4 −4 −4 1.46 E 8.03 E 1.15 E 1.05 E 1.65 E 1.52 E 1.53 E −5 −6 −4 −4 −5 −5 −5 Kinematic Viscosity, ν (m2s) 1.57 E 8.65 E 1.27 E 1.13 E 1.78 E 1.63 E 1.65 E Kinematic Viscosity, ν (ft2s) 2.869 E + 2 1.889 E + 2 2.077 E + 3 4.124 E + 3 5.183 E + 2 2.968 E + 2 2.598 E + 2 Gas Constant,a R (Jkg ∙ K) 1.716 E + 3 1.130 E + 3 1.242 E + 4 2.466 E + 4 3.099 E + 3 1.775 E + 3 1.554 E + 3 Gas Constant,a R (ft ∙ lbslug ∙ °R) 1.40 1.30 1.66 1.41 1.31 1.40 1.40 Specific Heat Ratio,b k 1.40 1.30 1.66 1.41 1.31 1.40 1.40 Specific Heat Ratio,b k ■ Table 1.3 Conversion Factors from BG and EE Units to SI Unitsa Acceleration Area Density Energy Force Length Mass Power Pressure Specific weight Temperature Velocity Viscosity (dynamic) Viscosity (kinematic) Volume flowrate To Convert from to Multiply by fts2 ft2 lbmft3 slugsft3 Btu ft ∙ lb lb ft in. mile lbm slug ft ∙ lbs hp in. Hg (60 °F) lbft2 (psf) lbin.2 (psi) lbft3 °F °R fts mihr (mph) lb ∙ sft2 ft2s ft3s galmin (gpm) ms2 m2 kgm3 kgm3 J J N m m m kg kg W W Nm2 Nm2 Nm2 Nm3 °C K ms ms N ∙ sm2 m2s m3s m3s 3.048 E − 1 9.290 E − 2 1.602 E + 1 5.154 E + 2 1.055 E + 3 1.356 4.448 3.048 E − 1 2.540 E − 2 1.609 E + 3 4.536 E − 1 1.459 E + 1 1.356 7.457 E + 2 3.377 E + 3 4.788 E + 1 6.895 E + 3 1.571 E + 2 TC = (59) (TF − 32°) 5.556 E − 1 3.048 E − 1 4.470 E − 1 4.788 E + 1 9.290 E − 2 2.832 E − 2 6.309 E − 5 a For more unit conversions, refer to Appendix E. ■ Table 1.4 Conversion Factors from SI Units to BG and EE Unitsa To Convert from Acceleration Area Density Energy Force Length Mass Power Pressure Specific weight Temperature Velocity Viscosity (dynamic) Viscosity (kinematic) Volume flowrate a 2 ms m2 kgm3 kgm3 J J N m m m kg kg W W Nm2 Nm2 Nm2 Nm3 °C K ms ms N ∙ sm2 m2s m3s m3s For more unit conversions, refer to Appendix E. to Multiply by 2 fts ft2 lbmft3 slugsft3 Btu ft ∙ lb lb ft in. mile lbm slug ft ∙ lbs hp in. Hg (60 °F) lbft2 (psf) lbin.2 (psi) lbft3 °F °R fts mihr (mph) lb ∙ sft2 ft2s ft3s galmin (gpm) 3.281 1.076 E + 1 6.243 E − 2 1.940 E − 3 9.478 E − 4 7.376 E − 1 2.248 E − 1 3.281 3.937 E + 1 6.214 E − 4 2.205 6.852 E − 2 7.376 E − 1 1.341 E − 3 2.961 E − 4 2.089 E − 2 1.450 E − 4 6.366 E − 3 TF = 1.8 TC + 32° 1.800 3.281 2.237 2.089 E − 2 1.076 E + 1 3.531 E + 1 1.585 E + 4 0.008 0.009 0.01 0.015 0.02 0.025 103 2(103) 0.003–0.03 0.001–0.01 0.0006–0.003 0.00085 0.0005 Riveted steel Concrete Wood stave Cast iron Galvanized iron Commercial steel or wrought iron Drawn tubing Plastic, glass 4 6 8 0.00015 0.000005 0.0 (smooth) Feet Pipe 104 2(104) 0.045 0.0015 0.0 (smooth) 0.9–9.0 0.3–3.0 0.18–0.9 0.26 0.15 Millimeters Equivalent Roughness, 𝜺 4 ■ Table 8.1 Equivalent Roughness for New Pipes [Adapted from Moody (Ref. 7) and Colebrook (Ref. 8)] Transition range Laminar flow 6 8 105 Smooth Wholly turbulent flow ρ VD Re = _____ μ 2(105) 4 6 8 106 2(106) ■ Figure 8.20 Friction factor as a function of Reynolds number and relative roughness for round pipes—the Moody chart. (Data from Ref. 7 with permission.) f 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 4 6 8 107 2(107) 4 6 8 0.00001 0.00005 0.0001 0.0002 0.0004 0.0006 0.001 0.0008 0.002 0.004 0.006 0.008 0.01 0.015 0.02 0.03 0.04 0.05 D ε __ WILEY END USER LICENSE AGREEMENT Go to www.wiley.com/go/eula to access Wiley’s ebook EULA.