The Third Edition of Basic Heat and Mass Transfer offers complete coverage for introductory engineering courses on heat
148 25 40MB
English Pages 1026 [1025] Year 2015
BASIC HEAT AND MASS TRANSFER Third Edition A. F. Mills Professor o f M echanical and Aerospace Engineering, Emeritus The University o f California at Los Angeles, Los Angeles, CA
C. F. M. Coimbra Professor o f M echanical and Aerospace Engineering The University o f California at San Diego, La Jolla, CA
Temporal Publishing, LLC  San Diego, CA 92130
Library of Congress CataloginginPublication Data Mills, A. F. and Coimbra, C. F. M. Basic Heat and Mass Transfer 3/E by Anthony F. Mills & Carlos F. M. Coimbra p. cm. Includes bibliographical references and index. ISBN 9780996305303 CIP data available. © 2015 by A.F. Mills and C.F.M. Coimbra. All Rights Reserved. Exclusive Publishing Rights to Temporal Publishing, LLC  San Diego, CA 92130
The authors and the publisher have used their best efforts in preparing this book. These efforts include the development, research, and (when applicable) testing of the theories and programs to determine their effectiveness. The authors and publisher make no warranty of any kind, expressed or implied, with regard to these programs or the documentation contained in this book and in the solutions manual. The authors and publisher shall not be liable in any event for incidental or consequential damages in connection with, or arising out of, the furnishing, performance, or use of the theory, results and/or programs. » All rights reserved. No part of this book may be reproduced, translated or stored in any storage retrieval device in any form or by any means, without permission in writing from Temporal Publishing, LLC. There are no authorized electronic or international hardcopy versions of this book. If the book you are reading is not a hardcopy published by Temporal Publishing LLC, you are infringing on U.S. and international copyright laws. Address inquiries or comments to: [email protected] www.temporalpublishing.com Printed in the United States of America 10987
ISBN 9780996305303
To Brigid For your patience and understanding.
To Kaori For your loving support.
PREFACE
For this third edition of Basic Heat and Mass Transfer Anthony Mills is joined by Carlos Coimbra as a coauthor. Professor Coimbra brings to this venture the perspec tive and skills of a younger generation of heat transfer educators and his own special expertise in areas of heat transfer research. Fifteen years after the second edition was published, a new edition is perhaps overdue, but in a mature field such as heat transfer, it is not at all clear what topics should be introduced, and then what topics should be removed to retain an acceptable length for an introductory text. As a result, our main motivation in publishing a third edition has been a different consideration. Our concern was the excessive prices of college textbooks, which in recent years have destroyed the established role played by these texts in the education of engi neering students. Traditionally, students bought a required textbook, became familiar with it in taking the course, and then retained the book as a tool for subsequent courses and an engineering career. Nowadays the pattern is for a student to sell the textbooks back to the university bookstore at the end of the course in order to obtain funds for buying textbooks for the next term. Alternatively, electronic versions of portions of the text are used during the course, or course readers containing selected material from the text may be used. It is particularly frustrating to instructors of subsequent design and laboratory courses to find that the students no longer have appropriate textbooks. Also, the traditional role formerly played by textbooks as professional manuals for engineering practice has been significantly affected. Basic methodology and data are more easily and reliably obtained from a familiar text than from an internet search. In an attempt to mitigate these problems and improve the experience of our engi neering students we decided to retain creative and publishing rights over the content of this book for this and future editions. A company called Temporal Publishing LLC was created to publish quality engineering textbooks at more reasonable prices. 1 1 Books can be ordered directly at discounted prices at www.temporalpublishing.com
V
PREFACE
This entailed first converting the previous edition of Basic Heat and Mass Transfer to LaTeX, which we could then modify efficiently. Since the conversion proved to be a major project in itself, our objective with this third edition is rather modest. We have focused on corrections, clarifications, minor updates and the production of a dedicated companion website.2 We envisage this website to be an integral part of the project and hope to make it a really useful adjunct to the text, for both students and instructors. The website contains links to the dedicated software BHMT that automates most of the calculations done in the text, instructor aides (such as com plete solutions manual for adoptees of the text, additional examples and exercises, presentations, etc.) and a compilation of answers to oddnumbered exercises to assist selfstudy by students. We will be continuously adding new technical content to the website while we work on future editions of the textbook. Also, given our closer association with the printondemand process, it will be easy for the authors to im plement small improvements in subsequent printings of this edition. We certainly welcome input and suggestions from users to improve our product. In preparing this new edition we have had valuable assistance from: Marius Andronie Kuang Chao Kaori Yoshida Coimbra We would like to dedicate the collaborative effort of bringing a new edition of Basic Heat and Mass Transfer to the memory of Prof. Donald K. Edwards, our teacher. A. F. Mills Santa Barbara, CA [email protected] C. F. M. Coimbra La Jolla, CA [email protected]
2 www.temporalpublishing.com/bhrntstudents
PREFACE TO THE SECOND EDITION
Basic Heat and Mass Transfer has been written for undergraduate students in me chanical engineering programs. Apart from the usual lowerdivision mathematics and science courses, the preparation required of the student includes introductory courses in fluid mechanics and thermodynamics, and preferably the usual juniorlevel engineering mathematics course. The ordering of the material and the pace at which it is presented have been carefully chosen so that the beginning student can proceed from the most elementary concepts to those that are more difficult. As a result, the book should prove to be quite versatile. It can be used as the text for an introductory course during the junior or senior year, although the coverage is sufficiently comprehensive for use as a reference work in undergraduate laboratory and design courses, and by the practicing engineer. Throughout the text, the emphasis is on engineering calculations, and each topic is developed to a point that will provide students with the tools needed to practice the art of design. The worked examples not only illustrate the use of relevant equations but also teach modeling as both an art and science. A supporting feature of Basic Heat and Mass Transfer is the fully integrated software available from the author’s website3. The software is intended to serve primarily as a tool for the student, both at college and after graduation in engineering practice. The programs are designed to reduce the effort required to obtain reliable numerical results and thereby increase the efficiency and effectiveness of the engineer. I have found the impact of the software on the educational process to be encouraging. It is now possible to assign more meaningful and interesting problems, because the students need not get bogged down in lengthy calculations. Parametric studies, which are the essence of engineer ing design, are relatively easily performed. Of course, computer programs are not a substitute for a proper understanding. The instructor is free to choose the extent to ? http://www.mae.ucla.edu/people/faculty/anthonymills
vii
viii
PREFACE TO THE SECOND EDITION
which the software is used by students because of the unique exact correspondence between the software and the text material. My practice has been to initially require students to perform various hand calculations, using the computer to give immediate feedback. For example, they do not have to wait a week or two until homework is returned to find that a calculated convective heat transfer coefficient was incorrect because a property table was misread. The extent to which engineering design should be introduced in a heat transfer course is a controversial subject. It is my experience that students can be best in troduced to design methodology through an increased focus on equipment such as heat and mass exchangers: Basic Heat and Mass Transfer presents more extensive coverage of exchanger design than do comparable texts. In the context of such equip ment one can conveniently introduce topics such as synthesis, parametric studies, tradeoffs, optimization, economics, and material or health constraints. The com puter programs HEX2 and CTOWER assist the student to explore the consequences of changing the many parameters involved in the design process. If an appropriate selection of this material is taught, I am confident that Accreditation Board for Engi neering and Technology guidelines for design content will be met. More important, I believe that engineering undergraduates are well served by being exposed to this material, even if it means studying somewhat less heat transfer science. More than 300 new exercises have been added for this edition. They fall into two categories: ( 1 ) relatively straightforward exercises designed to help students understand fundamental concepts, and (2 ) exercises that introduce new technology and that have a practical flavor. The latter play a very important role in motivating students; considerable care has been taken to ensure that they are realistic in terms of parameter values and focus on significant aspects of real engineering problems. The practical exercises are first steps in the engineering design process and many have substantial design content. Since environmental considerations have required the phasing out of CFC refrigerants, R 12 and Rl 13 property data, worked examples and exercises, have been replaced with corresponding material for R22 and Rl34a. Basic Heat and Mass Transfer complements Heat Transfer, which is published concurrently. Basic Heat and Mass Transfer was developed by omitting some of the more advanced heat transfer material from Heat Transfer and adding a chapter on mass transfer. As a result, Basic Heat and Mass Transfer contains the following chapters and appendixes: 1. Introduction and Elementary Heat Transfer 2. Steady OneDimensional Heat Conduction 3. Multidimensional and Unsteady Conduction 4. Convection Fundamentals and Correlations 5. Convection Analysis 6.
Thermal Radiation
7. Condensation, Evaporation, and Boiling 8.
Heat Exchangers
PREFACE TO THE SECOND EDITION
ix
9. Mass Transfer A. Property Data B. Units, Conversion Factors, and Mathematics C. Charts In a first course, the focus is always on the key topics of conduction, convection, radiation, and heat exchangers. Particular care has been taken to order the material on these topics from simpler to more difficult concepts. In Chapter 2 onedimensional conduction and fins are treated before deriving the general partial differential heat conduction equation in Chapter 3. In Chapter 4 the student is taught how to use con vection correlations before encountering the partial differential equations governing momentum and energy conservation in Chapter 5. In Chapter 6 radiation properties are introduced on a total energy basis and the shape factor is introduced as a geo metrical concept to allow engineering problem solving before having to deal with the directional and spectral aspects of radiation. Also, wherever possible, advanced topics are located at the ends of chapters, and thus can be easily omitted in a first course. Chapter 1 is a brief but selfcontained introduction to heat transfer. Students are given an overview of the subject and some material needed in subsequent chapters. Interesting and relevant engineering problems can then be introduced at the earli est opportunity, thereby motivating student interest. All the exercises can be solved without accessing the property data in Appendix A. Chapters 2 and 3 present a relatively conventional treatment of heat conduction, though the outdated and approximate Heissler and Grober charts are replaced by exact charts and the computer program COND2. The treatment of finitedifference numerical methods for conduction has been kept concise and is based on finitevolume energy balances. Students are encouraged to solve the difference equations by writing their own computer programs, or by using standard mathematics software such as Mathcad or MATLAB. In keeping with the overall philosophy of the book, the objective of Chapter 4 is to develop the students’ ability to calculate convective heat transfer coefficients. The physics of convection is explained in a brief introduction, and the heat transfer coefficient is defined. Dimensional analysis using the Buckingham pi theorem is used to introduce the required dimensional groups and to allow a discussion of the importance of laboratory experiments. A large number of correlation formulas fol low; instructors can discuss selected geometrical configurations as class time allows, and students can use the associated computer program CONV to reliably calculate heat transfer coefficients and skin friction coefficients or pressure drop for a wide range of configurations. Being able to do parametric studies with a wide variety of correlations enhances the students’ understanding more than can be accomplished by hand calculations. Design alternatives can also be explored using CONV. Analysis of convection is deferred to Chapter 5: simple laminar flows are consid ered, and highspeed flows are treated first in Section 5.2, since an understanding of
PREFACE TO THE SECOND EDITION
the recovery temperature concept enhances the students’ problemsolving capabili ties. Mixing length turbulence models are briefly discussed, and the chapter closes with a development of the general conservation equations. Chapter 6 focuses on thermal radiation. Radiation properties are initially defined on a total energy basis, and the shape factor is introduced as a simple geometrical concept. This approach allows students to immediately begin solving engineering radiation exchange problems. Only subsequently need they tackle the more difficult directional and spectral aspects of radiation. For gas radiation, the ubiquitous Hottel charts have been replaced by the more accurate methods developed by Edwards; the accompanying computer program RAD3 makes their use particularly simple. The treatment of condensation and evaporation heat transfer in Chapter 7 has novel features, while the treatment of pool boiling is quite conventional. Heatpipes are dealt with in some detail, enabling students to calculate the wicking limit and to analyze the performance of simple gascontrolled heatpipes. Chapter 8 expands the presentation of the thermal analysis of heat exchangers beyond the customary and includes the calculation of exchanger pressure drop, thermalhydraulic design, heat transfer surface selection for compact heat exchang ers, and economic analysis leading to the calculation of the benefitcost differential associated with heat recovery operations. The computer program HEX2 serves to introduce students to computeraided design of heat exchangers. Chapter 9 is an introduction to mass transfer. The focus is on diffusion in a sta tionary medium and low masstransfer rate convection. As was the case with heat convection in Chapter 4, mass convection is introduced using dimensional analysis and the Buckingham pi theorem. Of particular importance to mechanical engineers is simultaneous heat and mass transfer, and this topic is given detailed consideration with a focus on problems involving water evaporation into air. The author and publisher appreciate theefforts of all those who provided input that helped develop and improve the text. We remain dedicated to further refining the text in future editions, and encourage you to contact us with any suggestions or comments you might have. A. F. Mills [email protected] Bill Stenquist Executive Editor william_stenquist @prenhall.com
ACKNOWLEDGEMENTS TO THE FIRST AND SECOND EDITIONS
Reviewers commissioned for the first edition, published by Richard D. Irwin, Inc., provided helpful feedback. The author would like to thank the following for their contributions to the first edition. Martin Crawford, University of Alabama—Birmingham Lea Der Chen, University of Iowa Prakash R. Damshala, University of Tennessee— Chattanooga Tom Diller, Virginia Polytechnic Institute and State University Abraham Engeda, Michigan State University Glenn Gebert, Utah State University Clark E. Hermance, University of Vermont Harold R. Jacobs, Pennsylvania State University— University Park John H. Lienhard V, Massachusetts Institute of Technology Jennifer Linderman, University of Michigan—Ann Arbor Vincent P. Mano, Tufts University Robert J. Ribando, University of Virginia Jamal SeyedYagoobi, Texas A&M University—College Station The publisher would also like to acknowledge the excellent editorial efforts on the first edition. Elizabeth Jones was the sponsoring editor, and Kelley Butcher was the senior developmental editor.
xii
ACKNOWLEDGEMENTS TO THE FIRST AND SECOND EDITIONS
Some of the material in Basic Heat and Mass Transfer, in the form of examples and exercises, has been adapted from an earlier text by my former colleagues at UCLA, D. K. Edwards and V. E. Denny {Transfer Processes 1/e, Holt, Rinehart & Winston, 1973; 2/e HemisphereMcGrawHill, 1979). I have also drawn on material in radiation heat transfer from a more recent text by D. K. Edwards (Radiation Heat Transfer Notes, Hemisphere, 1981). I gratefully acknowledge the contributions of these gentlemen, both to this book and to my professional career. The late D. N. Bennion provided a chemical engineering perspective to some of the material on mass exchangers. The computer software was ably written by Baek Youn, HaeJin Choi, and Benjamin Tan. I would also like to thank former students S. W. Hiebert, R. Tsai, B. Cowan, E. Myhre, B. H. Chang, D. C. Weatherly, A. Gopinath, J. I. Rodriguez, B. R Dooher, M. A. Friedman, and C. Yuen. In preparing the second edition, I have had useful input from a number of peo ple, including Professor F. Forster, University of Washington; Professor N. Shamsundar, University of Houston; Professor S. Kim, Kukmin University; and Professor A. Lavine, UCLA. Students who have helped include P. Hwang, M. Tari, B. Tan, J. Sigler, M. Fabbri, F. Chao, and A. NaNakompanom. My special thanks to the secretarial staff at UCLA and the University of Auck land: Phyllis Gilbert, Joy Wallace, and Julie Austin provided enthusiastic and expert typing of the manuscript. Mrs. Gilbert also provided expert typing of the solutions manual.
NOTES TO THE INSTRUCTOR AND STUDENT
These notes have been prepared to assist the instructor and student and should be read before the text is used. Topics covered include conventions for artwork and mathematics, the format for example problems, organization of the exercises, com ments on the thermophysical property data in Appendix A, and a guide for use of the accompanying computer software.
ARTWORK Conventions used in the figures are as follows. —► ^ >
Conduction or convection heat flow Radiation heat flow Fluid flow Species flow
MATHEMATICAL SYMBOLS Symbols that may need clarification are as follows. ~ Nearly equal ~ Of the same order of magnitude  All quantities in the term to the left of the bar are evaluated at x xiii
xiv
NOTES TO THE INSTRUCTOR AND STUDENT
EXAMPLES Use of standard format for presenting the solutions of engineering problems is a good practice. The format used for the examples in Basic Heat and Mass Transfer, which is but one possible approach, is as follows. Problem statement Solution Given: Required: Assumptions: 1. 2.
etc.
Sketch (when appropriate) Analysis (diagrams when appropriate) Properties evaluation Calculations Results (tables or graphs when appropriate) Comments 1. 2.
etc.
It is always assumed that the problem statement precedes the solution (as in the text) or that it is readily available (as in the Solutions Manual). Thus, the Given and Required statements are concise and focus on the essential features of the problem. Under Assumptions, the main assumptions required to solve the problem are listed; when appropriate, they are discussed further in the body of the solution. A sketch of the physical system is included when the geometry requires clarification; also, ex pected temperature and concentration profiles are given when appropriate. (Schemat ics that simply repeat the information in the problem statements are used sparingly. We know that many instructors always require a schematic. Our view is that students need to develop an appreciation of when a figure or graph is necessary, because artwork is usually an expensive component of engineering reports. For example, we see little use for a schematic that shows a 10 m length of straight 2 cmO.D. tube.) The analysis may consist simply of listing some formulas from the text, or it may require setting up a differential equation and its solution. Strictly speaking, a property should not be evaluated until its need is identified by the analysis. However, in routine calculations, such as evaluation of convective heat transfer coefficients, it
NOTES TO THE INSTRUCTOR AND STUDENT
XV
is often convenient to list all the property values taken from an Appendix A table in one place. The calculations then follow with results listed, tabulated, or graphed as appropriate. Under Comments, the significance of the results can be discussed, the validity of assumptions further evaluated, or the broader implications of the problem noted. In presenting calculations for the examples in Basic Heat and Mass Transfer, we have rounded off results at each stage of the calculation. If additional figures are retained for the complete calculations, discrepancies in the last figure will be observed. Since many of the example calculations are quite lengthy, we believe our policy will facilitate checking a particular calculation step of concern. As is common practice, we have generally given results to more significant figures than is justified, so that these results can be conveniently used in further calculations. It is safe to say that no engineering heat transfer calculation will be accurate to within 1 %, and that most experienced engineers will be pleased with results accurate to within 1 0 % or 20%. Thus, preoccupation with a third or fourth significant figure is misplaced (unless required to prevent error magnification in operations such as subtraction). Fundamental constants are rounded off to no more than five significant figures.
EXERCISES The diskette logo next to an exercise statement indicates that it can be solved using the Basic Heat and Mass Transfer software, and that the sample solution provided to the instructor has been prepared accordingly. There are many additional exercises that can be solved using the software but that do not have the logo designation. These exercises are intended to give the student practice in hand calculations, and thus the sample solutions were also prepared manually. The exercises have been ordered to correspond with the order in which the mate rial is presented in the text, rather than in some increasing degree of difficulty. Since the range of difficulty of the exercises is considerable, the instructor is urged to give students guidance in selecting exercises for selfstudy. Answers to all exercises are listed in the Solutions Manual provided to instructors. Odd and evennumbered exercises are listed separately; answers to oddnumbered exercises are available to students on the book website.
PROPERTY DATA A considerable quantity of property data has been assembled in Appendix A. Key sources are given as references or are listed in the bibliography. Since Basic Heat and Mass Transfer is a textbook, our primary objective in preparing Appendix A was to provide the student with a wide range of data in an easily used form. Whenever possible, we have used the most accurate data that we could obtain, but accuracy was not always the primary concern. For example, the need to have consistent data over a wide range of temperature often dictated the choice of source. All the tables are in SI units, with temperature in kelvins. The computer program UNITS can be used
xvi
NOTES TO THE INSTRUCTOR AND STUDENT
for conversions to other systems of units. Appendix A should serve most needs of the student, as well as of the practicing engineer, for doing routine calculations. If a heat transfer research project requires accurate and reliable thermophysical property data, the prudent researcher should carefully check relevant primary data sources.
SOFTWARE The Basic Heat and Mass Transfer software has a menu that describes the content of each program. The programs are also described at appropriate locations in the text. The input format and program use are demonstrated in example problems in the text. Use of the text index is recommended for locating the program descriptions and examples. There is a onetoone correspondence between the text and the software. In principle, all numbers generated by the software can be calculated manually from formulas, graphs, and data given in the text. Small discrepancies may be seen when interpolation in graphs or property tables is required, since some of the data are stored in the software as polynomial curve fits. The software facilitates selfstudy by the student. Practice hand calculations can be immediately checked using the software. When programs such as CONV, PHASE, and BOIL are used, properties evaluation and intermediate calculation steps can also be checked when the final results do not agree. Since there is a large thermophysical property database stored in the software package, the programs can also be conveniently used to evaluate these properties for other purposes. For example, in CONV both the wall and fluid temperatures can be set equal to the desired temperature to obtain property values required for convection calculations. We can even go one step further when evaluating a con vective heat transfer coefficient from a new correlation not contained in CONV: if a corresponding item is chosen, the values of relevant dimensionless groups can also be obtained from CONV, further simplifying the calculations. Presently the BHMT software is only available in a DOS version, which runs on both Mac OS X and Windows platforms with DOS emulators. We are preparing a Windows version of the BHMT software and will announce its availability on the website. However, all the heat transfer components (excluding material relevant to Chapter 9 only) are already available in the Windows HT (Heat Transfer) package on the book companion website at www.temporalpublishing.com/bhmt. Some ex amples in the text show sample inputs from the DOS version of the software. For the Windows version the inputs are essentially the same.
CONTENTS
CHAPTER
INTRODUCTION AND ELEMENTARY HEAT TRANSFER 1.1 1.2 1.3
1.4
1.5 1.6
1.7 1.8
2
Introduction 2 Heat Transfer and Its Relation to Thermodynamics 3 Modes of Heat Transfer 7 1.3.1 Heat Conduction 8 1.3.2 Thermal Radiation 13 1.3.3 Heat Convection 17 Combined Modes of Heat Transfer 24 1.4.1 Thermal Circuits 24 1.4.2 Surface Energy Balances 27 Transient Thermal Response 29 1.5.1 The Lumped Thermal Capacity Model 29 Mass Transfer and Its Relation to Heat Transfer 34 1.6 .1 Modes of Mass Transfer 36 1.6.2 A Strategy for Mass Transfer 37 Dimensions and Units 37 Closure 39 Exercises 39
STEADY ONEDIMENSIONAL HEAT CONDUCTION 2.1 2.2
2.3
2.4
1
57
Introduction 58 Fourier’s Law of Heat Conduction 58 2.2.1 Thermal Conductivity 59 2.2.2 Contact Resistance 61 Conduction Across Cylindrical and Spherical Shells 63 2.3.1 Conduction across a Cylindrical Shell 63 2.3.2 Critical Thickness of Insulation on a Cylinder 67 2.3.3 Conduction across a Spherical Shell 70 2.3.4 Conduction with Internal Heat Generation 72 Fins 76
xvii
xviii
CONTENTS
2.5
2.4.1 The Pin Fin 76 2.4.2 Fin Resistance and Surface Efficiency 84 2.4.3 Other Fin Type Analyses 85 2.4.4 Fins of Varying CrossSectional Area 90 2.4.5 The Similarity Principle and Dimensional Analysis Closure 101 References 102 Exercises 102
MULTIDIMENSIONAL AND UNSTEADY CONDUCTION 3.1 3.2
3.3
3.4
3.5
3.6
4
4.3
4.4
133
Introduction 134 The Heat Conduction Equation 134 3.2.1 Fourier’s Law as a Vector Equation 135 3.2.2 Derivation of the Heat Conduction Equation 135 3.2.3 Boundary and Initial Conditions 140 3.2.4 Solution Methods 143 Multidimensional Steady Conduction 144 3.3.1 Steady Conduction in a Rectangular Plate 144 3.3.2 Steady Conduction in a Rectangular Block 151 3.3.3 Conduction Shape Factors 154 Unsteady Conduction 157 3.4.1 The Slab with Negligible Surface Resistance 158 3.4.2 The SemiInfinite Solid 165 3.4.3 Convective Cooling of Slabs, Cylinders, and Spheres 177 3.4.4 Product Solutions for Multidimensional Unsteady Conduction Numerical Solution Methods 193 3.5.1 A FiniteDifference Method for TwoDimensional Steady Conduction 194 3.5.2 FiniteDifference Methods for OneDimensional Unsteady Conduction 202 Closure 211 References 212 Exercises 213
CONVECTION FUNDAMENTALS AND CORRELATIONS 4.1 4.2
98
Introduction 244 Fundamentals 244 4.2.1 The Convective Heat Transfer Coefficient 4.2.2 Dimensional Analysis 251 4.2.3 Correlation of Experimental Data 263 4.2.4 Evaluation of Fluid Properties 267 Forced Convection 269 4.3.1 Forced Flow in Tubes and Ducts 269 4.3.2 External Forced Flows 280 Natural Convection 293 4.4.1 External Natural Flows 293
245
243
188
xix
CONTENTS
4.5
4.6 4.7 4.8 4.9
5
CONVECTION ANALYSIS 5.1 5.2
5.3
5.4
5.5
5.6
5.7
6
4.4.2 Internal Natural Flows 301 4.4.3 Mixed Forced and Natural Flows 308 Tube Banks and Packed Beds 315 4.5.1 Flow through Tube Banks 316 4.5.2 Flow through Packed Beds 323 Rotating Surfaces 330 4.6.1 Rotating Disks, Spheres, and Cylinders Rough Surfaces 333 4.7.1 Effect of Surface Roughness 334 The Computer Program CONV 343 Closure 343 References 352 Exercises 355 381
Introduction 382 HighSpeed Flows 383 5.2.1 A Couette Flow Model 383 5.2.2 The Recovery Factor Concept 388 Laminar Flow in a Tube 390 5.3.1 Momentum Transfer in Hydrodynamically Fully Developed Flow 391 5.3.2 Fully Developed Heat Transfer for a Uniform Wall Heat Flux 394 Laminar Boundary Layers 400 5.4.1 The Governing Equations for Forced Flow along a Flat Plate 401 5.4.2 The Plug Flow Model 403 5.4.3 Integral Solution Method 405 5.4.4 Natural Convection on an Isothermal Vertical Wall 414 Turbulent Flows 420 5.5.1 The Prandtl Mixing Length and the Eddy Diffusivity Model 421 5.5.2 Forced Flow along a Flat Plate 424 5.5.3 More Advanced Turbulence Models 427 The General Conservation Equations 428 5.6.1 Conservation of Mass 428 5.6.2 Conservation of Momentum 430 5.6.3 Conservation of Energy 434 5.6.4 Use of the Conservation Equations 438 Closure 439 References 439 Exercises 440
THERMAL RADIATION 6.1 6.2
330
449
Introduction 450 The Physics of Radiation 450 6.2.1 The Electromagnetic Spectrum 6.2.2 The Black Surface 452
451
CONTENTS
6.3
6.4
6.5
6.6
6.7
6.8
6.2.3 Real Surfaces 454 Radiation Exchange Between Surfaces 456 6.3.1 Radiation Exchange between Black Surfaces 456 6.3.2 Shape Factors and Shape Factor Algebra 458 6.3.3 Electrical Network Analogy for Black Surfaces 465 6.3.4 Radiation Exchange between Two Diffuse Gray Surfaces 6.3.5 Radiation Exchange between Many Diffuse Gray Surfaces 6.3.6 Radiation Transfer through Passages 483 Solar Radiation 486 6.4.1 Solar Irradiation 486 6.4.2 Atmospheric Radiation 488 6.4.3 Solar Absorptance and Transmittance 490 Directional Characteristics of Surface Radiation 495 6.5.1 Radiation Intensity and Lambert’s Law 496 6.5.2 Shape Factor Determination 499 6.5.3_ Directional Properties of Real Surfaces 502 Spectral Characteristics of Surface Radiation 508 6.6.1 Planck’s Law and Fractional Functions 508 6.6.2 Spectral Properties 511 Radiation Transfer Through Gases 517 6.7.1 The Equation of Transfer 518 6.7.2 Gas Radiation Properties 519 6.7.3 Effective Beam Lengths for an Isothermal Gas 527 6.7.4 Radiation Exchange between an Isothermal Gas and a Black Enclosure 532 6.7.5 Radiation Exchange between an Isothermal Gray Gas and a Gray Enclosure 533 6.7.6 Radiation Exchange between an Isothermal Nongray Gas and a SingleGraySurface Enclosure 537 Closure 539 References 540 Exercises 541
CONDENSATION, EVAPORATION, AND BOILING 7.1 7.2
7.3 7.4
569
Introduction 570 Film Condensation 570 7.2.1 Laminar Film Condensation on a Vertical Wall 572 7.2.2 Wavy Laminar and Turbulent Film Condensation on a Vertical Wall 580 7.2.3 Laminar Film Condensation on Horizontal Tubes 586 7.2.4 Effects of Vapor Velocity and Vapor Superheat 592 Film Evaporation 599 7.3.1 Falling Film Evaporation on a Vertical Wall 599 Pool Boiling 603 7.4.1 Regimes of Pool Boiling 603 7.4.2 Boiling Inception 606 7.4.3 Nucleate Boiling 609
468 475
CONTENTS
7.5
7.6
7.4.4 The Peak Heat Flux 611 7.4.5 Film Boiling 614 Heatpipes 620 7.5.1 Capillary Pumping 623 7.5.2 Sonic, Entrainment, and Boiling Limitations 7.5.3 GasLoaded Heatpipes 630 Closure 634 References 635 Exercises 637
HEAT EXCHANGERS 8.1 8.2
8.3
8.4 8.5
8.6
8.7
9.3
649
Introduction 650 Types of Heat Exchangers 650 8.2.1 Geometric Flow Configurations 652 8.2.2 Fluid Temperature Behavior 655 8.2.3 Heat Transfer Surfaces 657 8.2.4 DirectContact Exchangers 657 Energy Balances and Overall Heat Transfer Coefficients 658 8.3.1 Exchanger Energy Balances 658 8.3.2 Overall Heat Transfer Coefficients 660 Singlestream Steadyflow Heat Exchangers 665 8.4.1 Analysis of an Evaporator 6 6 6 Twostream Steadyflow Heat Exchangers 669 8.5.1 The Logarithmic Mean Temperature Difference 669 8.5.2 Effectiveness and Number of Transfer Units 674 8.5.3 BalancedFlow Exchangers 682 Elements of Heat Exchanger Design 685 8.6.1 Exchanger Pressure Drop 687 8.6.2 ThermalHydraulic Exchanger Design 694 8.6.3 Surface Selection for Compact Heat Exchangers 701 8.6.4 Economic Analysis 704 8.6.5 ComputerAided Heat Exchanger Design: HEX2 709 Closure 720 References 721 Exercises 721
MASS TRANSFER 9.1 9.2
628
745
Introduction 746 Concentrations and Fick’s Law of Diffusion 749 9.2.1 Definitions of Concentration 749 9.2.2 Concentrations at Interfaces 752 9.2.3 Fick’s Law of Diffusion 754 9.2.4 Other Diffusion Phenomena 756 Mass diffusion 758 9.3.1 Steady Diffusion through a Plane Wall 758 9.3.2 Transient Diffusion 765
xxii
CONTENTS
9.4
9.5
9.6
9.7
9.8
9.9
9.3.3 Heterogeneous Catalysis 772 Mass convection 111 9.4.1 The Mass Transfer Conductance 111 9.4.2 Low Mass Transfer Rate Theory 778 9.4.3 Dimensional Analysis 779 9.4.4 The Analogy between Convective Heat and Mass Transfer 9.4.5 The Equivalent Stagnant Film Model 789 Simultaneous Heat and Mass Transfer 792 9.5.1 Surface Energy Balances 793 9.5.2 The Wet and DryBulb Psychrometer 798 9.5.3 Heterogeneous Combustion 806 Mass Transfer in Porous Catalysts 810 9.6.1 Diffusion Mechanisms 810 9.6.2 Effectiveness of a Catalyst Pellet 812 9.6.3 Mass Transfer in a Pellet Bed 817 Diffusion in a Moving Medium 820 9.7.1 Definitions of Fluxes and Velocities 821 9.7.2 The General Species Conservation Equation 824 9.7.3 A More Precise Statement of Fick’s Law 827 9.7.4 Diffusion with One Component Stationary 828 9.7.5 High Mass Transfer Rate Convection 833 Mass Exchangers 836 9.8.1 Catalytic Reactors 837 9.8.2 Adiabatic Humidifiers 842 9.8.3 Counterflow Cooling Towers 847 9.8.4 CrossFlow Cooling Towers 855 9.8.5 ThermalHydraulic Design of Cooling Towers 858 Closure 871 References 872 Exercises 873
782
APPENDIX
PROPERTY DATA Table A .la Table A .l£ Table A .lc Table A.2 Table A.3 Table A.4 Table A.5a
903
Solid metals: Melting point and thermal properties at 300 K 905 Solid metals: Temperature dependence of thermal conductivity 907 Solid metals: Temperature dependence of specific heat capacity 908 Solid dielectrics: Thermal properties 909 Insulators and building materials: Thermal properties 911 Thermal conductivity of selected materials at cryogenic temperatures 913 Total hemispherical emittance at Ts ~ 300 K, and solar absorptance 914
CONTENTS
Table A.5b Table A,6a Table A.6b Table A.7 Table A.8 Table A.9 Table A.IOa Table A.10& Table A .ll Table A.12a Table AA2b Table A. 12c TableA.12d Table A.12e Table A.12/ Table A.13a Table A.13b Table A.14a Table A. 14b Table A.14c Table A.14d Table A .\4e Table A .14/ Table A.15 Table A.16 Table A.17« Table A.17b Table A.18 Table A.19 Table A.20 Table A.21 Table A.22a Table A.22b Table A.23 Table A.24 Table A.25 Table A.26
xxiii
Temperature variation of total hemispherical emittance for selected surfaces 917 Spectral and total absorptances of metals for normal incidence 918 Spectral absorptances at room temperature and an angle of incidence of 25° from the normal 919 Gases: Thermal properties 920 Dielectric liquids: Thermal properties 924 Liquid metals: Thermal properties 927 Volume expansion coefficients for liquids 928 Density and volume expansion coefficient of water 929 Surface tensions in contact with air 930 Thermodynamic properties of saturated steam 931 Thermodynamic properties of saturated ammonia 934 Thermodynamic properties of saturated nitrogen 935 Thermodynamic properties of saturated mercury 936 Thermodynamic properties of saturated refrigerant22 937 Thermodynamic properties of saturated refrigerant134a 938 Aqueous ethylene glycol solutions: Thermal properties 939 Aqueous sodium chloride solutions: Thermal properties 940 Dimensions of commercial pipes [mm] (ASA standard) 941 Dimensions of commercial tubes [mm] (ASTM standard) 942 Dimensions of seamless steel tubes for tubular heat exchangers [mm] (DIN 28 180) 943 Dimensions of wrought copper and copper alloy tubes for condensers and heat exchangers [mm] (DIN 178583) 943 Dimensions of seamless cold drawn stainless steel tubes [mm] (LN 9398) 944 Dimensions of seamless drawn wrought aluminum alloy tubes [mm] (LN 9223) 944 U.S. standard atmosphere 945 Selected physical constants 946 Diffusion coefficients in air at 1 atm 947 Schmidt number for vapors in dilute mixture in air at normal temperature, 948 Schmidt numbers for solution in water at 300 K 949 Diffusion coefficients in solids 950 Selected atomic weights 951 Henry constants for dilute aqueous solutions at moderate pressures 952 Equilibrium compositions for the NH 3 water system 953 Equilibrium compositions for the S0 2 water system 953 Solubility and permeability of gases in solids 954 Solubility of inorganic compounds in water 956 Combustion data 957 Thermodynamic properties of water vaporair mixtures at 1 atm 958
XXIV
CONTENTS
B
UNITS, CONVERSION FACTORS, AND MATHEMATICS 959 Table B .la Table B .lb Table B .lc TableB.lt/ Table B.2 Table B.3 Table B.3a Table B.3b Table B.4 CHARTS
Base and supplementary SI units 960 Derived SI units 960 Recognized nonSI units 961 Multiples of SI units 961 Conversion factors 962 Bessel functions 963 Bessel functions of the first and second kinds Modified Bessel functions of the first and second kinds The complementary error function 968 969
Figure C .la Figure CAb Figure C .lc Figure C.2a Figure C.2b Figure C.2c Figure Figure Figure Figure
C.3a C.3b C.3c C.4a
Figure CAb Figure C.4c Figure C.4d
Bibliography Nomenclature Index
993
964 966
Centerplane temperature response for a convectively cooled slab; Bi = hcL /k , where L is the slab halfwidth 970 Centerline temperature response for a convectively cooled cylinder; Bi = hcR /k 971 Center temperature response for a convectively cooled sphere; Bi = hcR /k 971 Fractional energy loss for a convectively cooled slab; Bi = hcL /k , where L is the slab halfwidth 972 Fractional energy loss for a convectively cooled cylinder; Bi = hcR /k 972 Fractional energy loss for a convectively cooled sphere; Bi = hcR /k 973 Shape (view) factor for coaxial parallel disks 973 Shape (view) factor for opposite rectangles 974 Shape (view) factor for adjacent rectangles 974 LMTD correction factor for a heat exchanger with one shell pass and 2, 4, 6 , . . . tube passes 975 LMTD correction factor for a crossflow heat exchanger with both fluids unmixed 975 LMTD correction factor for a crossflow heat exchanger with both fluids mixed 976 LMTD correction factor for a crossflow heat exchanger with two tube passes (unmixed) and one shell pass (mixed) 976
977 987
CHAPTER
1
INTRODUCTION AND ELEMENTARY HEAT TRANSFER CONTENTS
1.1
INTRODUCTION
1.2
HEAT TRANSFER AND ITS RELATION TO THERMODYNAMICS
1.3
MODES OF HEAT TRANSFER
1.4
COMBINED MODES OF HEAT TRANSFER
1.5
TRANSIENT THERMAL RESPONSE
1.6
MASS TRANSFER AND ITS RELATION TO HEAT TRANSFER
1.7
DIMENSIONS AND UNITS
1.8
CLOSURE
1
CHAPTER 1 INTRODUCTION AND ELEMENTARY HEAT TRANSFER
INTRODUCTION The process of heat transfer is familiar to us all. On a cold day we put on more clothing to reduce heat transfer from our warm body to cold surroundings. To make a cup of coffee we may plug in a kettle, inside which heat is transferred from an elec trical resistance element to the water, heating the water until it boils. The engineering discipline of heat transfer is concerned with methods of calculating rates of heat transfer. These methods are used by engineers to design components and systems in which heat transfer occurs. Heat transfer considerations are important in almost all areas of technology. Traditionally, however, the discipline that has been most concerned with heat transfer is mechanical engineering because of the importance of heat transfer in energy conversion systems, from coalfired power plants to solar water heaters. Many thermal design problems require reducing heat transfer rates by providing suitable insulation. The insulation of buildings in extreme climates is a familiar ex ample, but there are many others. The space shuttle has thermal tiles to insulate the vehicle from hightemperature air behind the bow shock wave during reentry into the atmosphere. Cryostats, which maintain the cryogenic temperatures required for the use of superconductors, must be effectively insulated to reduce the cooling load on the refrigeration system. Often, the only way to ensure protection from severe heating is to provide a fluid flow as a heat “sink”. Nozzles of liquidfueled rocket motors are cooled by pumping the cold fuel through passages in the nozzle wall be fore injection into the combustion chamber. A critical component in a fusion reactor is the “first wall” of the containment vessel, which must withstand intense heating from the hot plasma. Such walls may be cooled by a flow of helium gas or liquid lithium. A common thermal design problem is the transfer of heat from one fluid to an other. Devices for this purpose are called heat exchangers. A familiar example is the automobile radiator, in which heat is transferred from the hot engine coolant to cold air blowing through the radiator core. Heat exchangers of many different types are required for power production and by the process industries. A power plant, whether the fuel be fossil or nuclear, has a boiler in which water is evaporated to produce steam to drive the turbines, and a condenser in which the steam is con densed to provide a low back pressure on the turbines and for water recovery. The condenser patented by James Watt in 1769 more than doubled the efficiency of steam engines then being used and set the Industrial Revolution in motion. The common vapor cycle refrigeration or airconditioning system has an evaporator where heat is absorbed at low temperature and a condenser where heat is rejected at a higher tem perature. On a domestic refrigerator, the condenser is usually in the form of a tube coil with cooling fins to assist transfer of heat to the surroundings. An oil refinery has a great variety of heat transfer equipment, including rectification columns and thermal crackers. Many heat exchangers are used to transfer heat from one process stream to another, to reduce the total energy consumption by the refinery. Often the design problem is one of thermal control, that is, maintaining the op erating temperature of temperaturesensitive components within a specified range.
1.2 HEAT TRANSFER AND ITS RELATION TO THERMODYNAMICS
3
Cooling of all kinds of electronic gear is an example of thermal control. The develop ment of faster computers is now severely constrained by the difficulty of controlling the temperature of very small components, which dissipate large amounts of heat. Thermal control of temperaturesensitive components in a communications satellite orbiting the Earth is a particularly difficult problem. Transistors and diodes must not overheat, batteries must not freeze, telescope optics must not lose alignment due to thermal expansion, and photographs must be processed at the proper temperature to ensure high resolution. Thermal control of space stations present even greater problems, since reliable lifesupport systems are also necessary. From the foregoing examples, it is clear that heat transfer involves a great variety of physical phenomena and engineering systems. The phenomena must first be un derstood and quantified before a methodology for the thermal design of an engineer ing system can be developed. Chapter 1 is an overview of the subject and introduces key topics at an elementary level. In Section 1.2, the distinction between the subjects of heat transfer and thermodynamics is explained. The first law of thermodynamics is reviewed, and closed and opensystem forms required for heat transfer analysis are developed. Section 1.3 introduces the three important modes of heat transfer: heat conduction, thermal radiation, and heat convection. Some formulas are developed that allow elementary heat transfer calculations to be made. In practical engineering problems, these modes of heat transfer usually occur simultaneously. Thus, in Sec tion 1.4, the analysis of heat transfer by combined modes is introduced. Engineers are concerned with the changes heat transfer processes effect in engineering systems and, in Section 1.5, an example is given in which the first law is applied to a simple model closed system to determine the temperature response of the system with time. In Section 1.6, the subject of mass transfer is briefly introduced and its relation to heat transfer explained. Finally, in Section 1.7, the International System of units (SI) is reviewed, and the units policy that is followed in the text is discussed.
HEAT TRANSFER AND ITS RELATION TO THERMODYNAMICS When a hot object is placed in cold surroundings, it cools: the object loses internal energy, while the surroundings gain internal energy. We commonly describe this interaction as a transfer o f heat from the object to the surrounding region. Since the caloric theory of heat has been long discredited, we do not imagine a “heat substance” flowing from the object to the surroundings. Rather, we understand that internal energy has been transferred by complex interactions on an atomic or sub atomic scale. Nevertheless, it remains common practice to describe these interactions as transfer, transport, or flow, of heat. The engineering discipline of heat transfer is concerned with calculation of the rate at which heat flows within a medium, across an interface, or from one surface to another, as well as with the calculation of associated temperatures. It is important to understand the essential difference between the engineering dis cipline of heat transfer and what is commonly called thermodynamics. Classical thermodynamics deals with systems in equilibrium. Its methodology may be used
CHAPTER 1 INTRODUCTION AND ELEMENTARY HEAT TRANSFER
to calculate the energy required to change a system from one equilibrium state to another, but it cannot be used to calculate the rate at which the change may occur. For example, if a 1 kg ingot of iron is quenched from 1000°C to 100°C in an oil bath, thermodynamics tells us that the loss in internal energy of the ingot is mass (1 kg) x specific heat capacity (~450 J/kg K) x temperature change (900 K), or approximately 405 kJ. But thermodynamics cannot tell us how long we will have to wait for the temperature to drop to 100°C. The time depends on the temperature of the oil bath, physical properties of the oil, motion of the oil, and other factors. An appropriate heat transfer analysis will consider all of these. Analysis of heat transfer processes does require using some thermodynamics concepts. In particular, the first law of thermodynamics is used, generally in par ticularly simple forms since work effects can often be ignored. The first law is a statement of the principle o f consemcition o f energy, which is a basic law of physics. This principle can be formulated in many ways by excluding forms of energy that are irrelevant to the problem under consideration, or by simply redefining what is meant by energy. In heat transfer, it is common practice to refer to the first law as the energy conservation principle or simply as an energy or heat balance when no work is done. However, as in thermodynamics, it is essential that the correct form of the first law be used. The student must be able to define an appropriate system, recognize whether the system is open or closed, and decide whether a steady state can be assumed. Some simple forms of the energy conservation principle, which find frequent use in this text, follow. A closed system containing a fixed mass of a solid is shown in Fig. 1.1. The system has a volume V[m3], and the solid has a density p [kg/m3]. There is net heat transfer into the system at a rate of Q [J/s or W], and heat may be generated within the solid, for example, by nuclear fission or by an electrical current, at a rate Qv [W]. Solids may be taken to be incompressible, so no work is done by or on the system. The principle of conservation of energy requires that over a time interval A t [s], Change in internal energy _ Net heat transferred within the system ~ into the system
AU = QAt + QvAt
Heat generated within the system (1.1)
Dividing by At and letting At go to zero gives dU i f 
.
•
Q+Qv
System boundary
Application of the energy conservation principle to a closed system. Figure 1.1
5
1.2 HEAT TRANSFER AND ITS RELATION TO THERMODYNAMICS
The system contains a fixed mass (p V ); thus, we can write dU = pVdu, where u is the specific internal energy [J/kg]. Also, for an incompressible solid, du = cvdT, where cv is the constantvolume specific heat1 [J/kg K], and T [K] is tempera ture. Since the solid has been taken to be incompressible, the constantvolume and constantpressure specific heats are equal, so we simply write du = cdT to obtain p V c — = Q + Qv dt
(1.2)
Equation (1.2) is a special form of the first law of thermodynamics that will be used often in this text. It is written on a rate basis; that is, it gives the rate of change of temperature with time. For some purposes, however, it will prove convenient to return to Eq. (1.1) as a statement of the first law. w
Figure 1.2
Application of the energy conservation principle to a steadyflow open system.
Figure 1.2 shows an open system (or control volume), for which a useful form of the first law is the steadyflow energy equation. It is used widely in the thermody namic analysis of equipment such as turbines and compressors. Then
riiA
+ y + g zj = Q + W
(1.3)
where m [kg/s] is the mass flow rate, h [J/kg] is the specific enthalpy, V [m/s] is velocity, g [m/s2] is the gravitational acceleration, z is elevation [m], Q [W] is the net rate of heat transfer, as before, and W [W] is the rate at which external (shaft) work is done on the system .2 Notice that the sign convention here is that external work done on the system is positive; the opposite sign convention is also widely used. The symbol AX means Xout —Xjn, or the change in X. Equation (1.3) applies to a pure 1 The terras specific heat capacity and specific heat are equivalent and interchangeable in the heat transfer literature. 2 Equation (1.3) has been written as if h, V, and z are uniform in the streams crossing the control volume boundary. Often such an assumption can be made; if not, an integration across each stream is required to give appropriate average values.
CHAPTER 1 INTRODUCTION AND ELEMENTARY HEAT TRANSFER
substance when conditions within the system, such as temperature and velocity, are unchanging over some appropriate time interval. Heat generation within the system has not been included. Inmany types of heat transferequipment, no external work is done, and changes in kinetic and potential energy arenegligible; Eq.(1.3) then reduces to mAh = Q
(1.4)
The specific enthalpy h is related to the specific internal energy u as h = u + Pv
(1*5)
where P [N/m2 or Pa] is pressure, and v is specific volume [m3 /kg]. Two limit forms of Ah are useful. If the fluid enters the system at state 1 and leaves at state 2: 1. For ideal gases with Pv = RT, Ah = [ 2 CpdT
(1.6a)
Ji\
where R [J/kg K] is the gas constant and cp [J/kg K] is the constantpressure specific heat. 2. For incompressible liquids with p = 1/v — constant rT2
/ h
p2  A
cdT +  
(1.6b)
P
where c = cv = cp. The second term in Eq. (1.6b) is often negligible as will be assumed throughout this text. Equation (1.4) is the usual starting point for the heat transfer analysis of steadystate open systems. The second law o f thermodynamics tells us that if two objects at temperatures T\ and 72 are connected, and if T\ > 7 2 , then heat will flow spontaneously and irre versibly from object 1 to object 2. Also, there is an entropy increase associated with this heat flow. As 72 approaches T\ , the process approaches a reversible process, but simultaneously the rate of heat transfer approaches zero, so the process is of little practical interest. All heat transfer processes encountered in engineering are irre versible and generate entropy. With the increasing realization that energy supplies should be conserved, efficient use of available energy is becoming an important con sideration in thermal design. Thus, the engineer should be aware of the irreversible processes occurring in the system under development and understand that the opti mal design may be one that minimizes entropy generation due to heat transfer and fluid flow. Most often, however, energy conservation is simply a consideration in the overall economic evaluation of the design. Usually there is an important tradeoff between energy costs associated with the operation of the system and the capital costs required to construct the equipment.
1.3 MODES OF HEAT TRANSFER
7
MODES OF HEAT TRANSFER In thermodynamics, heat is defined as energy transfer due to temperature gradients or differences. Consistent with this viewpoint, thermodynamics recognizes only two modes of heat transfer: conduction and radiation. For example, heat transfer across a steel pipe wall is by conduction, whereas heat transfer from the Sun to the Earth or to a spacecraft is by thermal radiation. These modes of heat transfer occur on a molecular or subatomic scale. In air at normal pressure, conduction is by molecules that travel a very short distance (~ 0.065jim) before colliding with another molecule and exchanging energy. On the other hand, radiation is by photons, which travel almost unimpeded through the air from one surface to another. Thus, an important distinction between conduction and radiation is that the energy carriers for conduc tion have a short mean free path, whereas for radiation the carriers have a long mean free path. However, in air at the very low pressures characteristic of highvacuum equipment, the mean free path of molecules can be much longer than the equipment dimensions, so the molecules travel unimpeded from one surface to another. Then heat transfer by molecules is governed by laws analogous to those for radiation. A fluid, by virtue of its mass and velocity, can transport momentum. In addition, by virtue of its temperature, it can transport energy. Strictly speaking, convection is the transport of energy by bulk motion of a medium (a moving solid can also convect energy in this sense). In the steadyflow energy equation, Eq. (1.3), convec tion of internal energy is contained in the term thAh, which is on the lefthand side of the equation, and heat transfer by conduction and radiation is on the righthand side, as Q. However, it is common engineering practice to use the term convection more broadly and describe heat transfer from a surface to a moving fluid also as convection, or convective heat transfer, even though conduction and radiation play a dominant role close to the surface, where the fluid is stationary. In this sense, convec tion is usually regarded as a distinct mode of heat transfer. Examples of convective heat transfer include heat transfer from the radiator of an automobile or to the skin of a hypersonic vehicle. Convection is often associated with a change of phase, for example, when water boils in a kettle or when steam condenses in a power plant condenser. Owing to the complexity of such processes, boiling and condensation are often regarded as distinct heat transfer processes. The hot water home heating system shown in Fig. 1.3 illustrates the modes of heat transfer. Hot water from the furnace in the basement flows along pipes to radiators located in individual rooms. Transport of energy by the hot water from the basement is true convection as defined above; we do not call this a heat transfer process. Inside the radiators, there is convective heat transfer from the hot water to the radiator shell, conduction across the radiator shell, and both convective and radiative heat transfer from the hot outer surface of the radiator shell into the room. The convection is natural convection: the heated air adjacent to the radiator surface rises due to its buoyancy, and cooler air flows in to take its place. The radiators are heat exchang ers. Although commonly used, the term radiator is misleading since heat transfer
CHAPTER 1 INTRODUCTION AND ELEMENTARY HEAT TRANSFER
Figure 1.3
A hotwater home heating system illustrating the modes of heat transfer.
from the shell surface can be predominantly by convection rather than by radiation (see Exercise 120). Heaters that transfer heat predominantly by radiation are, for example, electrical resistance wire units. Each of the three important subject areas of heat transfer will now be introduced: conduction, in Section 1.3.1; radiation, in Section 1.3.2; and convection, in Section 1.3.3. 1.3.1 Heat Conduction On a microscopic level, the physical mechanisms of conduction are complex, en compassing such varied phenomena as molecular collisions in gases, lattice vibra tions in crystals, and flow of free electrons in metals. However, if at all possible, the engineer avoids considering processes at the microscopic level, preferring to use phenomenological laws, at a macroscopic level. The phenomenological law gov erning heat conduction was proposed by the French mathematical physicist J. B. Fourier in 1822. This law will be introduced here by considering the simple problem of onedimensional heat flow across a plane wall—for example, a layer of insu lation .3 Figure 1.4 shows a plane wall of surface area A and thickness L, with its face at x = 0 maintained at temperature T\ and the face at jc = L maintained at 7 2 . The heat flow Q through the wall is in the direction of decreasing temperature: if 3 In thermodynamics, the term insulated is often used to refer to a perfectly insulated (zeroheatflow or adiabatic) surface. In practice, insulation is used to reduce heat flow and seldom can be regarded as perfect.
1.3 MODES OF HEAT TRANSFER
9
Figure 1.4 Steady onedimensional conduction across a plane wall, showing the application of the energy conservation principle to an elemental volume Ax' thick.
T\ > 7 2 , Q is in the positive x direction .4 The phenomenological law governing this heat flow is Fourier’s law of heat conduction, which states that in a homogeneous substance, the local heat flux is proportional to the negative of the local temperature gradient: (1.7) where q is the heat flux, or heat flow per unit area perpendicular to the flow direction [W/m2], T is the local temperature [K or °C], and x is the coordinate in the flow direction [m]. When dT jd x is negative, the minus sign in Eq. (1.7) gives a positive q in the positive x direction. Introducing a constant of proportionality k, ( 1 .8) where k is the thermal conductivity of the substance and, by inspection of the equa tion, must have units [W/m K]. Notice that temperature can be given in kelvins or degrees Celsius in Eq. (1.8): the temperature gradient does not depend on which of these units is used since one kelvin equals one degree Celsius (1 K = 1°C). Thus, the units of thermal conductivity could also be written [W/m°C], but this is not the recommended practice when using the SI system of units. The magnitude of the thermal conductivity k for a given substance very much depends on its microscopic structure and also tends to vary somewhat with temperature; Table 1.1 gives some selected values of k. 4 Notice that this Q is the heat flow in the * direction, whereas in the first law. Eqs. (1.1)(1.4), Q = Qm — Q0ut is the net heat transfer into the whole system. In linking thermodynamics to heat transfer, some ambiguity in notation arises when common practice in both subjects is followed.
C H A PT E R 1
Table 1.1
IN T R O D U C T IO N A N D ELEM EN TA RY HEAT T R A N SF E R
Selected values of thermal conductivity at 300 K (~ 25°C). k
Material
W/m K
Copper Aluminum Brass (70% Cu, 30% Zn) Mild steel Stainless steel, 188 Mercury Concrete Pyrex glass Water Neoprene rubber Engine oil, SAE 50 White pine, perpendicular to grain Polyvinyl chloride (PVC) Freon 12 Cork Fiberglass (medium density) Polystyrene Air
386 204 111
64 15 8.4 1.4 1.09 0.611 0.19 0.145 0.10 0.092 0.071 0.043 0.038 0.028 0.027
Note: Appendix A contains more comprehensive data.
Figure 1.4 shows an elemental volume AV located between x and x\Ax', AV is a closed system, and the energy conservation principle in the form of Eq. (1.2) applies. If we consider a steady state, then temperatures are unchanging in time and d T /d t = 0; also, if there is no heat generated within the volume, Qv = 0. Then Eq. (1.2) states that the net heat flow into the system is zero. Because the same amount of heat is flowing into AV across the face at x, and out of AV across the face at x + Ax, Q\x — Q x+Ax
Since the rate of heat transfer is constant for all x, we simplify the notation by drop ping the I* and 1*+^ subscripts (see the footnote on page 9), and write Q = Constant But from Fourier’s law, Eq. (1.8 ), Q = qA — —kA —— dx The variables are separable: rearranging and integrating across the wall,
where Q and A have been taken outside the integral signs since both are constants. If the small variation of k with temperature is ignored for the present we obtain
11
1.3 MODES OF HEAT TRANSFER
e = T
a ,)
Comparison of Eq. (1.9) with Ohm’s law, I = E /R , suggeststhatAT = T\ — Tz can be viewed as a driving potential for flow ofheat, analogous to voltagebeing the driving potential for current. Then R = L/kA can be viewed as a thermal resistance analogous to electrical resistance. If we have a composite wall of two slabs of material, as shown in Fig. 1.5, the heat flow through each layer is the same:
Hot wali
id)
(c)
(e)
(A
Figure 1.10 Some commonly encountered flows, (a) Forced flow in a pipe, Re# ~ 50,000. The flow is initially laminar because of the “bellmouth” entrance but becomes turbulent downstream, (b) Laminar forced flow over a cylinder, Re# ~ 25. (c) Forced flow through a tube bank as found in a shellandtube heat exchanger, (d) Laminar and turbulent natural convection boundary layers on vertical walls, (e) Laminar natural convection about a heated horizontal plate, (f) Cellular natural convection in a horizontal enclosed fluid layer.
CHAPTER 1 INTRODUCTION AND ELEMENTARY HEAT TRANSFER
approximately by the following, rather complicated correlation of experimental data: V0Sk06(pcp)0A hc = 0.023£ )0 .2 v 0.4
(1.22)
In contrast to laminar flow, hc is now strongly dependent on velocity, V, but only weakly dependent on diameter. In addition to thermal conductivity, other fluid prop erties involved are the kinematic viscosity, v; density, p; and constantpressure spe cific heat, cp. In Chapter 4 we will see how Eq. (1.22) can be rearranged in a more compact form by introducing appropriate dimensionless groups. Equations (1.21) and ( 1 .2 2 ) are only valid at some distance from the pipe entrance and indicate that the heat transfer coefficient is then independent of position along the pipe. Near the pipe entrance, heat transfer coefficients tend to be higher, due to the generation of largescale vortices by upstream bends or sharp comers and the effect of suddenly heating the fluid. Figure 1.11 shows a natural convection flow on a heated vertical surface, as well as a schematic of the associated variation of hc along the surface. Transition from a laminar to a turbulent boundary layer is shown. In gases, the location of the transition is determined by a critical value of a dimensionless group called the Grashof num ber. The Grashof number is defined as GrA= (/3AT)gx* / v2, where AT = Ts — Te, g is the gravitational acceleration [m/s2], x is the distance from the bottom of the surface where the boundary layer starts, and /3 is the volumetric coefficient of expansion, which for an ideal gas is simply l/T , where T is absolute temperature [K]. On a vertical wall, transition occurs at GrA~ 109. For air, at normal temperatures, experi ments show that the heat transfer coefficient for natural convection on a vertical wall can be approximated by the following formulas:
Figure 1.11 A naturalconvection boundary layer on a vertical wall, showing the variation of local heat transfer coefficient. For gases, transition from a laminar to turbulent flow occurs at a Grashof number of approximately 109; hence xtr — [109v2//3ATg]1/ 3.
21
1.3 MODES OF HEAT TRANSFER
Laminar flow:
hc = 1.07(AT/.v)1/4 W/m2K
Turbulent flow: hc = 1.3(AT) 1''3 W/m2K
104 < Gr* < 109
(1.23a)
109 < Grx < 1( ) 12
(1.23b)
Since these are dimensional equations, it is necessary to specify the units of hc, AT, and x, which are [W/m2 K], [K], and [m], respectively. Notice that hc varies as x ~ 1/4 in the laminar region but is independent of x in the turbulent region. Usually the engineer requires the total heat transfer from a surface and is not too interested in the actual variation of heat flux along the surface. For this purpose, it is convenient to define an average heat transfer coefficient hc for an isothermal surface of area A by the relation 0 becomes .m dQ = 2nrLQv dr
Introducing Fourier’s law, Q = A q = 2nrL\—k(dT/dr)], and assuming k constant gives ± ( r* L \ = _ & r dr \ dr ) k
(2.26)
which is a secondorder linear ordinary differential equation for T(r). Two boundary conditions are required; the first comes from symmetry: r = 0:
dT_ dr
=
0
(2.27a)
To obtain a result of some generality, we will take as the second boundary condition a specified temperature on the outer surface of the cylinder: n;
T = T{
(2.27b)
In a typical engineering problem, T\ might not be specified; however, we shall see that the result will be in a form suitable for problem solving. Integrating Eq. (2.26) once gives
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
or
^ = dr
2 k
Cl r
Applying the first boundary condition, Eq. (2.27a), 0
= 0+^,
or Ci = 0
Integrating again, r =  i % r 2+ c 2 4 k Applying the second boundary condition, Eq. (2.27b) allows C2 to be evaluated: 1d" ? 7'' =  ^ T r >+C2 Substituting back gives the desired temperature distribution, T(r): 1 ti" T  h =  ^  { r ]  r 2)
(2.28)
The maximum temperature is at the centerline of the cylinder. Setting r = 0 in Eq. (2.28) gives ■ffi 2 rma*  7 i =  % ^ (2.29) 4 k The use of this result is illustrated in the following example.
EXAMPLE 2.4 Temperature Distribution in a Nuclear Reactor Fuel Rod Uranium oxide fuel is contained inside 0.825 cmI.D., 0.970 cmO.D. Zircaloy4 tubes. The tubes have a 1.75 cm pitch in a square array. The power averaged over the volume including the space between the fuel rods is 152.4 W/cm3. At a specific location along the bundle the coolant water is at 400 K and the convective heat transfer coefficient hc is 1.0 x 104 W/m2 K. If the interfacial conductance between the fuel and the tube, /i;, is 6000 W/m2K, determine the maximum temperature in the fuel rods. Solution Given: Nuclear reactor fuel rod. Required: Maximum rod temperature at location where the cool water is at Te = 400 K. Assumptions: Steady onedimensional heat flow. We cannot immediately use Eq. (2.29) to obtain Tmsx because the surface temperature of the fuel rod is unknown. We proceed as follows: first we calculate Qv in the fuel itself, (a)
75
2.3 CONDUCTION ACROSS CYLINDRICAL AND SPHERICAL SHELLS
Qv = 152.4
Volume of array Volume of fuel
(1.75)2 = 152 4 ' (tt/4)(0.825)2 = 873 W/cm3 = 8.73 x 108 W/m3 Next we find the temperature of the outer surface of the fuel rod. For unit length of rod, the heat flow across the Zircaloy tube wall is Q — Qv [nD1j 4)(1): Q = (8.73 x 108)(tt/4)(0.825 x 10~2)2(1) Tube
= 46,700 W/m
Q""*........................... O.................... O*..."”"WV 1 (yV,) __ 2nrjhj 2nrjt^
Water
o
From the thermal circuit, as shown, Tu = Te + Q Z R = 400 + 46,700
1  ln(0.485/0.413)  1 (2tt) (0.00413) (6000) (2*r) (0.00485) (104)_
= 400 + 46,700(0.00642 + 0.0256/&Zr + 0.00328) As a guess, we take the mean temperature of the tube to be 600 K; from Table A. 16, the conductivity of Zircaloy4 is 17.2 W/m K, and Tu = 400 + 46,700(0.00642 + 0.00149 + 0.00328) = 923K Now Eq. (2.29) can be used to obtain Tmax. If we guess a mean temperature of 1500 K for the uranium oxide, Table A.2 gives &uo2 —2.6 W/m K, and
T
1Q'"rf ___ (8.73 x 108)(0.00413)2 „ „ „ „ = T“+ 4 J ^ = 923+  ( 4 ) f e L = 2 3 5 5 K
To check if our guessed mean temperatures are appropriate, we first determine the mean temperature of the tube. From the thermal circuit, (0.00328 + 0.00149/2) 7 tube — 4 0 0 + ( 9

)^
„
0 0 6 4 2 + 0 0 0 [ 49 + 0 0 0 3 2 8 ) 
which is close enough to our guess of 600 K. The mean temperature of the fuel rod is _
923 + 2355
^
T u o 2  2 =
and at this temperature, &uo2 = 2.5 W/m K. The new value of Tmax is (8.73 x 108)(0.00413)2 .......... = 923 +  (4)(25) = 2412 ~ 2400 K
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
Comments 1. Since the kvalue of UO 2 is given to only two significant figures, no further iteration is warranted. 2. Notice that the conductivity of Zircaloy4 is lower than that of pure zirconium. 3. The largest thermal resistance in the circuit is at the fuelcladding interface. The accuracy of the result depends primarily on our ability to obtain a reliable value of hj. In fact, it could be argued that the second iteration for Tmax was unwarranted due to uncertainty in the value of hj.
4. Notice that we cannot extend the thermal circuit into the rod because Q is not constant when there is internal heat generation.
FINS Heat transfer from a system can be increased by extending the surface area through the addition of fins. Fins are used when the convective heat transfer coefficient hc is low, as is often the case for gases such as air, particularly under naturalconvection conditions. Common examples are the cooling fins on electronics components, on the cylinders of aircooled motorcycles and lawnmowers, and on the condenser tubes of a home refrigerator. Figure 2.8 shows a variety of fin configurations. A careful examination of an automobile radiator will show how it is designed to provide a large exterior surface. Fins are added to increase the hcA product and hence decrease the convective thermal resistance 1/h cA. But the added area is not as efficient as the original surface area since there must be a temperature gradient along the fin to conduct the heat. Thus, for cooling, the average temperature difference (Ts — Te) is lower on a finned surface compared with the unfinned surface, and an appropriate thermal resistance for a fin is \ / h cA r\f, where A is the surface area of the fin and 77/ is the efficiency of the fin (0 < r\f < 1). For short fins of high thermal conductivity, 77/ is large, but as the fin length increases, rff decreases. Our objective here is to analyze heat flow in a fin to determine the temperature variation along the fin and, hence, to evaluate its efficiency r)f. Because fins are thin in one direction, it can be assumed that the temperature variation in this direction is negligible; this key assumption allows the conduction along the fin to be treated as if it were onedimensional, which greatly simplifies the analysis. 2.4.1 The Pin Fin Simple pin fins, such as those used to cool electronic components, will be analyzed to develop the essential concepts of fin theory. The first law is used to derive the governing differential equation, which, when solved subject to appropriate boundary conditions, gives the temperature distribution along the fin. The heat loss from the fin is then obtained and put in dimensionless form as the fin efficiency.
77
2.4 FINS
Figure 2.8
Some heat sinks incorporating fins for cooling of standard packages for integrated circuits. (Photograph courtesy of EG&G Wakefield Engineering, Wakefield, Mass.)
Governing Equation and Boundary Conditions Consider the pin fin shown in Fig. 2.9. The crosssectional area is Ac = tzR2 where R is the radius of the pin, and the perimeter & — 2 k R. Both A c and R are uniform, that is, they do not vary along the fin in the x direction. The energy conservation principle, Eq. (1.2), is applied to an element of the fin located between x and x + Ax. Heat can enter and leave the element by conduction along the fin and can also be lost by convection from the surface of the element to the ambient fluid at temperature Te. The surface area of the element is & Ax\ thus, qAc \x  qAc ^+Av  hc0>Ax{T  T e) = 0 Dividing by Ax and letting A*
0 gives
 ±  ( q A e)  h c& { J  T e) = 0
(2.30)
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
A c ~ nR2 R
3
) hci*>Lx{T  Te)
Figure 2.9 A pin fin showing the coordinate system, and an energy balance on a fin element. For the pin fin, A c is independent of x; using Fourier’s law q — —k d T / d x with k constant gives (2.31)
kAc ^  h c9 > ( T  T e) = 0
which is a secondorder ordinary differential equation for T — T[x). Notice that modeling of the conduction along the fin as onedimensional has caused the con vective heat loss from the sides of the fin to appear in the differential equation, in contrast to the problems dealt with in Section 2.3, where convection became involved as a boundary condition. Next, boundary conditions for Eq. (2.31) must be specified. Since we wish to examine the performance of the fin itself, it is appropriate to take its base temperature as known; that is, (2.32)
7 l* = o = Tb
At the other end, the fin loses heat by Newton’s law of cooling: dT Ack—
= Achc{T\x=L
Te)
(2.33a)
x —L
where the convective heat transfer coefficient here is, in general, different from the one for the sides of the fin because the geometry is different. However, because the area of the end, Ac, is small compared to the side area, the heat loss from the end is correspondingly small and usually can be ignored. Then Eq. (2.33a) becomes (2.33b) and this boundary condition is simpler to use than Eq. (2.33a). An even simpler result
79
2.4 FINS
can be obtained if the temperature distribution along the fin is assumed identical to that for an infinitely long fin, for which the appropriate boundary condition is lim T = Te
(2.33c)
.v—>°o
Figure 2.10 illustrates these boundary conditions.
71ix —►oo
(a)
0)
*e
(c)
Figure 2.10 Three tip boundary conditions for the pin fin analysis, (a) Heat loss by convection, (b) Insulated tip. (c) Infinitely long fin.
Temperature Distribution We will use Eq. (2.33b) for the second boundary condition as a compromise between accuracy and simplicity of the result. For mathematical convenience, let 0 — T — Te and /32 = hc& / k A c\ then Eq. (2.31) becomes 0  i 3 20 = O
(2.34)
For P a constant, Eq. (2.34) has the solution e = Cie^I + C2e~px or 0 = B i sinh fix + B2 cosh fix The second form proves more convenient; thus, we have T — Te = B\ sinh fix + B2 cosh fix
(2.35)
Using the two boundary conditions, Eqs. (2.32) and (2.33b) give two algebraic equa tions for the unknown constants B\ and B2, TB  T e = B\ sinh(0) + B2 cosh(0); dT dx
B2 = Tb  Te
= pBi cosh fiL + PB2 sinh j3L = 0; x —L
B\ = Z^tanhjSL
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
Substituting B\ and B2 in Eq. (2.35) and rearranging gives the temperature distribu tion as T T e cosh p (L —x) T ^ fe= cosh/31 ’
u Wh6re
( h c& > \ ^ 2 (Z36)
T  Te
TB~Te
Figure 2.11
Fin temperature distributions calculated from Eq. (2.36).
Figure 2.11 shows a plot of Eq. (2.36). When /3 is small— for example, if the fin is made of aluminum and has a high thermal conductivity—the temperature T does not drop much below the base temperature 7#. For large /3, T approaches the fluid temperature at the tip of the fin.2 Heat Loss The heat dissipated from the fin can be found by integrating the heat loss over the side surface of the fin (there is no heat loss from the fin tip): Q = [ L hc0 > ( T  T e)dx Jo
(2.37)
with T obtained from Eq. (2.36). Substituting gives
hc&{TB  T e) [L
f i = .."cos'hjSZ
uo,r
w
Jo
To simplify the integration, let § = /3(L  * ) ; then dx =  d £ //3 and 2 A c in /3 = (hc& / k A cy / 2 is the crosssectional area of the fin. The subscript c denotes “cross section” and not “con vection” as in the heat transfer coefficient hc. The area for convective heat loss is the surface area o f the fin, & L .
2.4 FINS
\
Q =
coshj3L hc&
IfSL
(TB  T e)
p
hc& T ~
81
sinh 0 —sinh/3L cosh fiL
(Tb — r e)tanh/3L
(2.38)
A less obvious alternative, but usually a more convenient way to find the heat dissipation, is to apply Fourier’s law at the base of the fin: • , dT Q = —kAc— dx
(2.39)
Substituting from Eq. (2.36), 2 =  k A c(TB 
kAc(Tg
cosh PL Te)
CQsh/3L
= kAcf3(TB  T e) tanh/3L
(2.40)
Since /32 = hc& / k A c, Eqs. (2.38) and (2.40) give the same result, which is to be expected since there is no heat loss from the end of the fin. Fin Efficiency Let us now put Eq. (2.38) in dimensionless form by dividing through by hc0>L(TB  T e):
Q hc&»L(TB  Te)
1 tanh PL PL
(2.41)
The dimensions of the lefthand side of this equation are [W]/[W/m2 K][m][m][K] = 1, as desired. The righthand side must also be dimensionless since /3 has dimensions [m1 ] and the group /3L has dimensions [m_1][m] = 1. (Of course, /3L must be dimensionless to be the argument of the tanh function.) Now hc£PL(TB — Te) is the rate at which heat would be dissipated if the entire fin surface were at the base temperature TB\ in reality, there is a decrease in temperature along the fin, and the actual heat loss is less. Thus, the lefthand side of Eq. (2.41) can be viewed as the ratio of the actual heat loss to the maximum possible and is termed the fin efficiency, r]f. The righthand side is a function of the dimensionless parameter /3L only; we will set fiL = x as a fin parameter, and then Eq. (2.41) can be written in the compact form = ta n h X
X
(2t, the ratio gPjAc is simply \ / t , and /3 = (hc/ k t ) ]/ 2.
Figure 2.13
A straight rectangular fin.
83
2.4 FINS
Computer Program FIN1 The program FIN1 calculates the temperature distribution, fin efficiency, and base heat flow of straight rectangular fins. There are three options for the tip boundary condition: (1) infinitely long fin, (2) insulated, and (3) convective heat loss. The analysis for option 2 was given above; the analyses for options 1 and 3 are given as Exercises 258 and 259, respectively. For all three options, rjf is defined in terms of an isothermal fin heat loss of Q = hc& L(7# — Te). Use of FIN1 is illustrated in the example that follows.
EXAMPLE 2.5 Fins to Cool a Transistor An array of eight aluminum alloy fins, each 3 mm wide, 0.4 mm thick, and 40 mm long, is used to cool a transistor. When the base is at 340 K andthe ambient air is at 300 K, how much power do they dissipate if the combined convection and radiation heat transfer coefficient is estimated to be 8 W/m2 K? The alloy has a conductivity of 175 W/m K. Solution Given: Aluminum fins to cool a transistor. Required: Power dissipated by 8 fins. Assumptions: 1. Heat transfer coefficient constant along fin. 2. Heat loss from fin tip negligible. For one fin, Ac = (0.003) (0.0004) = 1.2 x 10~6m2 IP = 2(0.003 + 0.0004) = 6.8 x 103 m r
hi.V
= kAc
(8.0 W/m2 K) (6.8 x 10“3 m) ~ (175 W/m 1C)(12 x 106 m2) = 259 m2 P = 16.1 m1 % = fiL= (16.1 m“ ')(0.040m) = 0.644 Substituting in Eq. (2.42) 1
________
rif = 0.644 tanh (0.644)
1
*2(0.644) _ !
0.644 *2(0.644) + 1
= 0.881
The side surface area of one fin is £?L = (6.8 x 10 3)(0.040) = 2.72 x 10 4 m2. If each fin were 100% efficient, it would be dissipate h(9>L)(JB  Te) = (8)(2.72 x 10~4)(340  300) = 8.70 x 102 W
84
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
Since the fins are only 88.1% efficient, Q = (0.881)(8.70 x 10“2) = 7.67 x 10'2 W For 8 fins, Qtotal = (8)(7.67 x 10"2) = 0.613 W. Solution using FIN1
The required input is: Boundary condition = 2 Halfthickness, length, and width = 0.0002, 0.040, 0.003 Thermal conductivity = 175 Heat transfer coefficient = 8 Base temperature and ambient temperature = 340, 300 *range for plot = 0.0, 0.04 FIN 1 gives the output: rjf = 0.881 Q
= 7.67 x 10'2 (watts)
Comments
1. Any consistent system of units can be used with FIN1. Since SI units were used here, the heat flow is in watts. 2. Notice the use of h = hc + hr to account for radiation.
2.4.2 Fin Resistance and Surface Efficiency It is useful to have an expression for the thermal resistance of a pin fin for use in thermal circuits. Equation (2.38) can be rewritten as q
Q
— ______ ~
_________
\/[{hc0>/p) tanh pL]
•
(2 4 3 )
{}
Thus, the thermal resistance of a pin fin is Rfl" = (hc'P /P ) tank p L = hc 3? Li) f
(2‘44)
Notice that this thermal resistance accounts for both conduction along the fin and convection into the fluid. There are two parallel paths for heat loss from a finned surface— one through the fins and one through the area between the fins,as shown in Fig. 2.14. The respective conductances are thus additive; however, quite often the heat loss through the area between the fins is negligible. The total surface efficiency r\t of a surface with fins of fin efficiency ?]/ is obtained by adding the unfinned portion of the surface area at 100% efficiency to the surface area of the fins at efficiency rjf. A t1, = (A —Af ) + TjjAf
(2.45)
85
2.4 FINS
(4Af ) Ac(y4^)
MMMMM^^IyMMMHi
^nn = hcAf rjf
Figure 2.14
A finned surface showing the parallel paths for heat loss.
where A/ is the surface area of the fins and A is the total heat transfer surface area, including the fins and exposed tube or other surface. Solving for r\t, Af . . n ,= i  j  ( i  v f )
(2.46)
The corresponding thermal resistance of the finned surface is then R=
1 hcArft
(2.47)
Design calculations for the finned surfaces used in heat exchangers, such as automo bile radiators, are conveniently made using Eq. (2.47). 2.4.3 Other Fin Type Analyses The key feature of the fin analysis presented in Section 2.4.1 was that the thinness of the fin allowed us to ignore the temperature variation across the fin and, hence, to account for the convective loss from the surface directly in the differential equation for T(x). The same assumption is valid for extended surfaces unrelated to cooling fins, and the results obtained in Section 2.4.1 are directly or indirectly applicable to these surfaces. Sometimes it is quite obvious that the situation is similar to that for a cooling fin. For example, Fig. 2.15 shows a thermocouple installation used to measure the temperature of a hot air stream. The thermocouple junction is at a lower temperature than the air since the conduction heat flow along the thermocouple wires to the colder wall must be balanced by convection from the air. The temperature variation along
TV
Duel Hot gases
Thermocouple
Figure 2.15
A thermocouple immersed in a fluid system.
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
Insulati
Figure 2.16 An element of a perforatedplate heat exchanger showing heat conduction along the plates. the thermocouple is identical to that for a pin fin, so Eq. (2.36), with appropriate choices for the kAc product, can be used to determine the error expected in the thermocouple reading. Sometimes it is not obvious that the situation resembles that for a cooling fin, yet the assumption of negligible temperature variation in the thin direction of a wire or plate gives a differential equation similar to Eq. (2.30). The perforated plates in the heat exchanger shown in Fig. 2.16 can be treated as fins since the temperature varia tion across the plates is small compared to the temperature variation along the plates between the hot and cold streams. The copper conductors on the circuit board shown in Fig. 2.17 can be treated as fins, as can the circuit board between the conductors. The examples that follow relate to Figs. 2.15 and 2.16, and Exercise 288 is based on Fig. 2.17.
Phen< clotl
Copper conductor
Figure 2.17
Copper conductors on a circuit board.
EXAMPLE 2.6 Error in Thermocouple Readings Duplex thermocouple leads have both wires embedded in polyvinyl electrical insulation. One available size has wires of 0.25 mm diameter in insulation with an outside perimeter of 1.5 mm and is to be used in the situation depicted in Fig. 2.15. The air temperature is 350 K,
87
2.4 FINS
and the wall temperature is 300 K. What length of immersion is required for the error in the thermocouple reading to be 0.1 K when the heat transfer coefficient on the perimeter is approximately 30 W/m2 K? The wires are (i) copper and constantan (type T), (ii) iron and constantan (type J), and (iii) chromel and alumel (type K). Solution Given: Duplex thermocouple leads for types T, J, and K thermocouples. Required: Length of immersion for a specified error. Assumptions: Temperature variation across lead is small compared to the variation along the lead. This is a fintype problem since the temperature variation across the lead is small compared to the 50 K variation along the lead. The effective /32 is
Insulation
hc& P2 = LkAc where hc = 30 W/m2 K, & = 1.5 x 103 m, and £A:AC must be evaluated for the thermal resistances of the two wires and the insulation in parallel. For each wire, Ac = (7t/4)(0.25 x 103 )2 = 4.91 x 10~8 m2, and for the insulation Ac ~ 10 x 10“ 8m2. Thermal conductivity and kAc values are given in the following table.
Wire Material
k W/mK
Copper Constantan (55% Cu, 45% Ni) Iron ChromelP (90% Ni, 10% Cr) Alumel (95% Ni, 2% Mn, 2% Al) Insulation
385 23 73 17 48 0.1
kAc Wm/K 19 x106 1.1 x 10"6 3.6 x 10"6 0.83 x 10~6 2.36 x 106 0.01 x 10“6
The contribution of the insulation to is seen to be negligible; hence its precise shape or composition is unimportant. The temperature of the thermocouple junction (located at x = L) is given by Eq. (2.36) as TL T e TB T e
0.1 300350
1 cosh/3L
Solving, cosh/3 L = 500 pL = 6.91 from a calculator, or use coshx = (1 /2)(ex f e~x) L = 6.91//3 Evaluating j3 for each thermocouple pair gives the following results:
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
Thermocouple Type
ZkAc Wm/K
P
m1
L cm
T J K
20.1 x 10~6 4.7 x 10~6 3.2 x 10~6
47.3 97.8 119
14.6 7.1 5.8
Comments 1. Type T thermocouples are to be avoided when conduction along the wires may cause a significant error. 2. There are other criteria for choosing thermocouple pairs, including operating temperature range, emf output, and corrosion resistance. 3. Type T thermocouples are widely used because the component wires are relatively free from inhomogeneities; hence, calibration charts are reliable.
EXAMPLE 2.7 A Perforated Plate Heat Exchanger Perforatedplate heat exchangers are used in cryogenic refrigeration systems. The fluids flow through highconductivity perforated plates separated by insulating spacers. Heat is trans ferred from the hot stream to the cold stream by conduction along the plates, as shown in Fig. 2.16. A counterflow helium to helium unit has 0.5 mmthick rectangular aluminum plates with 0.9 mmdiameter holes in a square array of pitch 1.3 mm. The plate length exposed to each stream is 20 mm, and the plate width is 80 mm. The spacer is 4 mm wide and 0.86 mm thick. If the heat transfer coefficient is 400 W/m2 K for both streams, calculate the overall heat transfer coefficient. Take k = 200 W/m K for the aluminum. Solution Given: A perforatedplate heat exchanger. Required: Overall heat transfer coefficient U. Assumptions: 1. The plates are thin enough for a fintype analysis to be valid. 2. Heat flow along the plates is onedimensional.
1! L\ ” 20 mm
^ — L i  4 mm
OOOO OOOO O O O O O O O O O O O O O O O O O O O O O O O O O O
oooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooooooo iO O O O O O O O O O O O O O Q Q O O O O O O O O Q O O Q
5.
Table 2.2 Fins of various shapes: Efficiency, surface area per unit width (S')a and profile area (Ap) for straight fins; surface area (S)b and volume (V) for annular fins and spines: (3 = (hc/ k t ) 1/2. Straight Fins
1. Rectangular y=t
T)f — — tanh (5L S'
2. Parabolic y = f(l —x / L ) 1/2
2L,
S’ = ,B + (t2/2L)]n(2L/t + B)
P L I  1/3 ( j ^ )
B = V + 4 Ljt2, Ap — jtL
nt
1 /, (2jSL)
s ' = 2Vrr + I ?
*?/ =
S' = L B + (L2/2t)\n(2t/L + B)
4. Parabolic y = t{ 1  x / L ) 2
A p = 2tL
_ 1 /2/3(f/3L)
3. Triangular y = r(l —x/L)
=
pL
y/4(pL)2 + 1 + 1
5
= ^ /1 + 4 /2 /L 2,
Ap = \ tL
Annular Fins
5. Rectangular ir2 ~ r\) Ko(Pri )h (Pr2) +Io(pr\)K\ (pr2) v= t S = 2n{r\  r]),
V = 2n[r\  r\)t
2ri/(3 (>2 6.
Hyperbolic y = t(n/r)
Jr
I  2/3 (  Pr2y / r2/ n ^  I 2/3 ( /3 r 2 x / ^ / n ) 12/3 (  0 n )
+ n ) //3 ( 2 p r,) /2/3 ^ p r 2y/ r 2/ r ^  / _ 2/ 3 ( f /3r2v/r 2//*i) /  1/3
5  2 ^ { c  B + (r /2)lnJc + (j j f i ^ }
+
V = t o r i (r2  r i )
C = \ / ( r  / n )2 + 12
Spines (Circular Cross Section)
7. Pin
Tjf =
S = 2ntL,
tanh(V2/3L),
V = 7tt2L
y=t 8. Parabolic
2
y = t ( 1 —x / L ) 1/2
/1
(%y/2PL)
(fv/2/3Z .)/o ( f \/2 0 Z ,) ’
s = (t47t/6L2){(4L2/ t 2 + l) 3/2 —1} V = (n/2)t2L
9. Triangular y = r(]  x / L ) 10. Parabolic y = r(l  x / L ) 2
nf(2V2mh{2 M Tjf =
r
■. =  , 1+ 1
v/8/9(/5L)2 +
s  * , V l 2 + ' 2' V = (it/5)t2L
5 = (roi3/16f){i4B(L/4f)ln[(4/B/L)+A]} M = 1+ (8 r/L 2),
B = \ j \ + (4'2/ i 2)
a,bNote that the formulas for rjy were derived assuming dS' = dx (straight fins) and dS —& d x (spines). Use of the exact expressions for S' and S when calculating Q from r \f gives a result in better agreement with exact numerical solutions for twodimensional conduction in short fins (for long fins the difference is negligible).
96
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
EXAMPLE 2.8 Cooling Fin for a Transistor An aluminum annular fin is used to cool a transistor. The inner and outer radii are 5 mm and 20 mm, respectively, and the thickness is 0.2 mm. Calculate its efficiency and the heat dissipated when its base is at 380 K, the ambient air temperature is 300 K, and the estimated heat transfer coefficient is 8.2 W/m2K. Take the conductivity of aluminum as 205 W/m K. Solution Given: Aluminum annular fin. Required: Efficiency and heat dissipated. Assumptions: 1. Heat transfer coefficient constant over the fin surface. 2. Heat loss from tip negligible. For an annular fin, /32 = hc/kt\ (8.2) P = (205)(0.1 x 103) = 400m~
j3 = 20m1
20 mm
The fin effectiveness is given by Eq. (2.52): Vf ISri
5 mm
(2n /P) K\ (Pn)I\ (Pr2)  1\ (j3r\)K\ (pr2) (fj  r\) Ko(Pr\)/) (Pr2)+Io(pr\)K] (Pr2) (20) (0.005) = 0.1;
jSr2 = (20)(0.020) = 0.4
From Appendix B, Table B.3b, the required values of Bessel functions are: Pr
/o
h
Ko
Ki
0.1 0.4
1.0025
0.0501 0.2040
2.4271
9.8538 2.1843
Substituting in Eq. (2.52), (2)(0.005)/(20) (9.8538)(0.2040)  (0.0501 )(2.1843) n f ~ (0.0202 —0.0052) (2.4271)(0.2040) + (1.0025)(2.1843) _
'
The heat dissipation is the efficiency times the dissipation for an isothermal fin: Q = nAI*c)(Mx)(r22 r i) ( T B T e) = (0.944)(8.2)(2)(tz) (0.0202  0.0052) (380 —300) = 1.46W
97
2.4 FINS
Solution using FIN2 The required input in SI units is: Item number = 5 Thermal conductivity and density of the fin = 205, 2700 Heat transfer coefficient = 8.2 Base temperature and ambient temperature = 380, 300 t = 0.0001 n and r2 = 0.005, 0.020 FIN2 gives the following output: Fin efficiency = 0.944 Base heat flow = 1.459 (watts) Mass of fin = 6.362 x 10~4 (kilograms) Comments 1. Notice that Table B.3b gives e~x times /o(.*) and I\ (*) to simplify the tabulation. 2. The high efficiency suggests that the thickness of such fins is determined by rigidity rather than by heat transfer considerations. 3. Any consistent system of units can be used in FIN2. Since SI units were used here, the base heat flow is in watts, and the mass of the fin is in kilograms.
EXAMPLE 2.9 Heat Loss from a Parabolic Fin A straight Duralumin fin has a parabolic profile y = /(I —x/L)2, with t = 3 mm and L = 20 mm. Determine the heat dissipation by the fin when its base temperature is 500 K and it is exposed to fluid at 300 K with a heat transfer coefficient of 2800 W/m2 K. Also calculate the fin mass. Solution Given: Straight fin with a parabolic profile. Required: Heat dissipation, mass. Assumptions: Constant heat transfer coefficient over fin surface. From Table A.lb, the conductivity of Duralumin at a guessed average fin temperature of (1/2) (500 + 300) = 400 K is 187 W/m K. Using Table 2.2, item 4, 1/2
p
\kt J
2800 [ ( 187)(0.003)
/3Z. = (70.65)(0.02) = 1.413
1/2
= 70.65 ml
CHAPTER 2
V
STEADY ONEDIMENSIONAL HEAT CONDUCTION
2
2
[4(/3L)2 + l]1/2 + 1
[4(1.413)2 + l]1/2 + 1
0.500
21 1/2 1.044
= (0.02) (1.044) + = 0.0406 m For a unit width of fin, Q = hcS/(TB T e)vf = (2800)(0.0406)(500300)(0.500) = 11,370 W/m From Table A. la the density of Duralumin is 2770 kg/m3; thus, Fin mass = App — ~^tLp = Q ) (0.003)(0.02)(2770) =0.1108kg/m Solution using FIN2 The required input in SI units is: Item number = 4 Thermal conductivity and density of the fin = 187, 2770 Heat transfer coefficient = 2800 Base temperature and ambient temperature = 500, 300 t = 0.003 L = 0.02 FIN2 gives the following output: Fin efficiency = 0.500 Base heat flow = 11,370 (watts/meter) Mass of fin = 0.1108 (kilograms/meter) Comments Exercise 2113 shows that this fin profile gives the maximum heat loss for a given weight of any profile. 2.4.5 The Similarity Principle and Dimensional Analysis To conclude our analysis of fins we use the pin fin problem of Section 2.4.1 to illustrate the similarity principle and dimensional analysis. These are important concepts used in the analysis of more complex heat transfer problems. Equation (2.42) showed that the fin performance could be expressed as a relation between just
99
2.4 FINS
two dimensionless parameters: the fin efficiency, ?]/ = Q jhc@*L{jB — Te), and a fin parameter % = (hc& L?/ M c)*/2. The similarity principle for this problem is simply the statement that r\f is a function of £ only. Thus, for example, if & and A c are both doubled, rjf remains the same. We say that all pin fins with the same value of X are similar, even though their sizes, materials, or heat transfer coefficients may be quite different. The dimensionless groups relevant to a given problem are required for use of the similarity principle. We can deduce these dimensionless groups without actually solving the governing equations, as was done in Section 2.4.1. For this purpose, we use dimensional analysis, for which a number of methods are available. The pin fin problem will be used to demonstrate a method that requires a transformation of variables to make the governing equation and boundary conditions dimensionless. The first step is to choose dimensionless forms of the independent variable x and the dependent variable T. For x, an obvious choice is § = x /L , where L is the length of the fin; § then varies from zero to unity as x varies from zero to L. For T we will choose 6 = (T — Te)/(T B — Te)\ 0 has a value of unity at the fin base and will approach zero at the tip of an infinitely long fin. Next, we transform the problem statement into the new variables. The rules of the transformation are X = L$
T = (TB  T e)0 + Te
dx = Ld%
d T = {TB  T e)dO
and Eq. (2.31) becomes ( 7 W ^ ) 0 _ hcg?{TB _ Tt)e = o or _ /? L2 d § 2
=o c
The boundary condition Eq. (2.32) becomes £  0:
0 = 1(2.53a)
and the boundary condition Eq. (2.33b) becomes cTB  T e) d O _ L dt, or df)
= 0
(2.53b)
The differential equation is now put in dimensionless form. Dividing by kAc/ L , ^ dV
^
9 iA c
= 0
CHAPTER 2
STEADY ONEDIMENSIONAL HEAT CONDUCTION
or
(2.54) and we see that the fin parameter % appears quite naturally in the dimensionless form of the governing equation. The boundary conditions are already dimensionless. Equation (2.54) is a differential equation for 0 as a function of § and contains one dimensionless parameter, the boundary conditions contain no further parameters. Thus, the solution must be of the form
or (2.55) For this simple case, the fin efficiency appears quite naturally as the dimensionless form of the heat dissipation rate. Although G = (§ ,^ ), the definite integral in Eq. (2.55) is not a function of §, so that Vf
= Vf ( x )
(2.56)
which corresponds to the analytical solution, Eq. (2.42). More complex heat transfer problems are often governed by differential equations that are difficult or impossible to solve analytically. Use of the above procedure allows the most concise form of the solution to be determined, which can be used as a basis for correlating experimental data or the results of numerical solutions. Also, when properly used, dimensional analysis facilitates the estimation of errors incurred in making simplifying assumptions. Such an approach is especially important for the analysis of convective heat transfer, as is shown in advanced texts. As a rather simple example of error estimation, consider the pin fin problem when the tip heat loss is not neglected. The appropriate second boundary condition is then Eq. (2.33a), which transforms into
2.5 CLOSURE
101
The Biot number Bi = hcL /k was discussed in Section 1.5.1. This boundary con dition introduces a second parameter into the problem; the temperature distribution must now be of the form e = e(