HEAT TRANSFER Sixth Edition J. P. HOLMAN Professor of Mechanical Engineering ‘Southern Methodist University oe l McGraw-Hill Book Company New York St, Louis San Francisco Auckland ‘Bogoté Hamburg Johannesburg London Madrid Mexico Montreal New Delhi Panama Paris Sao Paulo Singapore Sydney Tokyo TorontoHEAT TRANSFER, INTERNATIONAL STUDENT EDITION Copyright © 1986 Exclusive rights by McGraw-Hill Book Co. - Singapore for manufacture and export. This book cannot be #€-exported from the country to which itis consigned by MeGraw Fil Ist printing 1986 oF retrieval system, without the prior written permission of the publisher. ‘This book was set in Times Roman by General Graphic Services ‘The editors were Anne Murphy, Madelaine Eichberg, and Steven Tenney; ‘The designer was Elliot Epstein; ‘The production supervisor was Phil Galea New drawings were done by J & R Services, Ine. rary of Congress Cataloging in Publication Date Holman, J. P. (Jack Philip) Heat Transfer Includes bibliographical references and index. 1. Heat~Transmission, I. Title Qecs20.HGs 1986 5362 85-13783 ISBN 0-07-029620-0 When ordering this title use ISBN 0-07-Y66459. ned by KIN KEONG PRINTING CO, TE LED ~ Rep SgtHEAT a aera eee 72eho Stet ab: gil Sol SoS ssibigle Feoe : jhe WPA aikel sole Shs VAde : Gaeé Sslos Bl eSylie | Samer obitl sle tee gb lee su yT AYaaaF Gab —yABOUT THE AUTHOR Dr. JACK P. HOLMAN received the Ph.D. degree in mechanical engineering from Oklahoma State University in 1958. After two years active duty as a research scientist in the Air Force Aerospace Research Laboratory, he joined the faculty of Southern Methodist University, where he is presently Professor of Me- chanical Engineering. During his tenure at Southern Methodist University he has eight times been voted the Outstanding Engineering Faculty Member by the student body in a poll conducted annually. His planning activities have included membership on the University Master Plan committees for engineering and for graduate edu- cation, and Chairman of the faculty planning committees for the Doctor of Engineering and Environmental Systems degrees. He has been active as a faculty representative to the Board of Trustees, member of Tenure and Ethics committees, Trustee of the SMU Retirement Plan, and Councils on Continuing. Education and Management Review. He has held administrative positions as Director of Thermal and Fluid Sciences Center. Head of the Civil and Me- chanical Engineering Department, and Assistant Provost for Instructional Media, |AS a principal investigator for research sponsored by the Atomic Energy Commission, National Science Foundation, NASA, and the Environmental Protection Agency, he has published extensively in such journals as Industrial and Engineering Chemisty, International Journal of Heat and Mass Transfer, Journal of the Aerospace Sciences, and others He is also the autho of three widely used textbooks: Heat Transfer, 1963 (6th edition 1986), Expe. mental Methods for Engineers, 1966 (4th edition 1984). and Thermodynamics, 1969 (3rd edition 1980), all published by the McGraw: Hill Book Company. These books have been translated into Spanish, Chinese, Japanese, Korean, and Portugese and have received world-wide distribution through International Student Editions Printed in Japan. Dr. Holman is theAbout he authors consulting editor for the McGraw-Hill Series in Mechanical Engineering, and Senior Consulting Editor for the McGraw-Hill/Hemisphere Advanced Series in Thermal/Fluids Engineering, Dr. Holman also consults widely for industry in the fields of energy conservation and energy systems. A member of the American Society of Engineering Education, he is past Chairman of the National Mechanical Engineering Division and past Chairman of the A.S.M.E. Region X Mechanical Engineering Department Heads. Dr. Holman is a registered professional engineer in the state of Texas, and received the “Mechanical Engineer of the Year'’ award by the North Texas Section of the American Society of Mechanical Engineers in 1971 In 1972 Dr. Holman was recipient of the George Westinghouse Award from the American Society of Engineering Education for distinguished contributions to Engineering Education. r1 12 13 14 18 16 17 24 22 23 24 25 26 27 28 29 210 an Preface List of Symbols Introduction Conduction Heat Transfer ‘Thermal Conductivity Convection Heat Transfer Radiation Heat Transfer Dimensions and Units Computer Solution of Heat-Transfer Problems Summary Steady-State Conduction—One Dimension Introduction ‘The Plane Wall Insulation and K Values Radial Systems—Cylinders The Overall Heat-Transfer Coefficient Critical Thickness of Insulation Heat-Source Systems Cylinder with Heat Sources Conduction-Convection Systems Fins ‘Thermal Contact Resistance CONTENTSse conten at 32 a4 35 36 38 39 310 at 42 43 44 45 46 47 48 49 St 52 53 Bs 55 56 87 58 59 510 S11 5412 513 et 62 Steady-State Conduction—Multiple Dimensions Introduction Mathematical Analysis of Two-Dimensional Heat Conduction Graphical Analysis, ‘The Conduction Shape Factor ‘Numerical Method of Analysis, Numerical Formulation in Terms of Resistance Elements Gauss-Seidel Iteration ‘Accuracy Considerations Electrical Analogy for Two-Dimensional Conduction Summary Unsteady-State Conduction Introduction Lumped-Heat-Capacity System Transient Heat Flow in a Semi-Infinite Solid Convection Boundary Conditions Multidimensional Systems Transient Numerical Method ‘Thermal Resistance and Capacity Formulation Graphical Analysis—The Schmidt Plot ‘Summary Principles of Convection Introduction Viscous Flow Inviscid Flow Laminar Boundary Layer on a Flat Plate Energy Equation of the Boundary Layer ‘The Thermal Boundary Layer The Relation between Fluid Friction and Heat Transfer Turbulent-Boundary-Layer Heat Transfer ‘Turbulent-Boundary-Layer Thickness Heat Transfer in Laminar Tube Flow ‘Turbulent Flow in a Tube Heat Transfer in High-Speed Flow Summary Empirical and Practical Relations for Forced-Convection Heat Transfer Introduction Empirical Relations for Pipe and Tube Flow n n n 6 8 98 100 16 n7 BE br 13 136 39 132 158 166, 186, 189 207 207 21 25 m 25 25 27 246 6 250 253 259 aT am 2B63 ea 65 66 70 mo 742 743 Flow across Cylinders and Spheres Flow across Tube Banks Liquid-Metal Heat Transfer Summary Remarks Natural-Convection Systems Introduction Free-Convection Heat Transfer on a Vertical Flat Plate Empirical Relations for Free Convection Free Convection from Vertical Planes and Cylinders, Free Convection from Horizontal Cylinders Free Convection from Horizontal Plates Free Convection from Inclined Surfaces Nonnewtonian Fluids ‘Simplified Equations for Air Free Convection from Spheres Free Convection in Enclosed Spaces Combined Free and Forced Convection ‘Summary Radiation Heat Transfer Introduction Physical Mechanism Radiation Properties Radiation Shape Factor Relations between Shape Factors Heat Exchange between Nonblackbodies Infinite Parallel Planes Radiation Shields Gas Radiation Radiation Network for an Absorbing and Transmitting Medium Radiation Exchange with Specular Surfaces Radiation Exchange with Transmitting, Reflecting, and Absorbing Media Formulation for Numerical Solution Solar Radiation Radiation Properties of the Environment Effect of Radiation on Temperature Measurement ‘The Radiation Heat-Transfer Coefficient ‘Summary Ccononts be 288 299 30s 310 323 323 323 330 339 339 342 33 Ms 34s 346 MT 354 358 373 373 373 375 384 393 400 407 409 4B 24 430 436 492 439 470 a ax Contents ot o2 Sete o7 10 101 10-2 103 10-4 105 10-6 10-7 10-8 10-9 "1 4 112 143, 14 115 116 42 124 122 123 124 125 126 Condensation and Boiling Heat Transfer Introduction Condensation Heat-Transfer Phenomena ‘The Condensation Number Film Condensation inside Horizontal Tubes Boiling Heat Transfer ‘Simplified Relations for Boiling Heat Transfer with Water ‘Summary and Design Information Heat Exchangers Introduction ‘The Overall Heat-Transfer Coefficient Fouling Factors ‘Types of Heat Exchangers ‘The Log Mean Temperature Difference Effectiveness-NTU Method ‘Compact Heat Exchangers Analysis for Variable Properties ‘Heat-Exchanger Design Considerations Mass Transfer Introduction Fick's Law of Diffusion Diffusion in Gases Diffusion in Liquids and Solids ‘The Mass-Transfer Coefficient Evaporation Processes in the Atmosphere ‘Special Topics in Heat Transter Introduction Heat Transfer in Magnetoftuidynamic (MED) Systems Transpiration Cooling Low-Density Heat Transfer Ablation ‘The Heat Pipe Appendixes Tables 1 41 491 498 S13 si 525 525 526 331 $32 613 2 04contents B_ Exact Solutions of Laminar-Boundary-Layer Equations 653 © Analytical Relations for the Heisler Charts 659 Index 667PREFACE This book presents an elementary treatment of the principles of heat transfer. AS a text it contains sufficient material for a one-semester course which may be presented at the junior level. or higher, depending on individual course objectives. A background in ordinary differential equations is helpful for proper understanding of the material. Although some familiarity with fluid mechanics will aid in the convection discussions, it is not essential. The concepts of thermodynamic energy balances are also useful in the various analytical de- velopments. Presentation of the subject follows classical lines of separate discussions for conduction, convection, and radiation, although itis emphasized that the phys- ical mechanism of convection heat transfer is one of conduction through the stationary fluid layer near the heat transfer surface. Throughout the book em- phasis has been placed on physical understanding while, at the same time, relying on meaningful experimental data in those circumstances which do not permit a simple analytical solution. Conduction is treated from both the analytical and the numerical viewpoint, ‘so that the reader is afforded the insight which is gained from analytical solutions, as well as the important tools of numerical analysis which must often be used in practice. A similar procedure is followed in the presentation of convection, heat transfer. An integral analysis of both free- and forced-convection boundary layers is used to present a physical picture of the convection process. From this physical description inferences may be drawn which naturally lead to the presentation of empirical and practical relations for calculating convection heat- transfer coefficients. Because it provides an easier instruction vehicle than other methods, the radiation-network method is used extensively in the introduction of analysis of radiation systems, while a more generalized formulation is given later. Systems of nonlinear equations requiring iterative solutions are also dis- cussed in the conduction and radiation chapters.tv Preace ‘The log-mean-temperature-difference and effectiveness approaches are pre- sented in heat-exchanger analysis since both are in wide use and each offers its own advantages to the designer. A brief introduction to diffusion and mass transfer is presented in order to acquaint the reader with these processes and to establish more firmly the important analogies between heat, mass, and mo- ‘mentum transfer. ‘A number of special topics are discussed in Chapter 12 which give added flavor to the basic material of the preceding chapters. Problems are included at the end of each chapter. Some of these problems are of a routine nature to familiarize the student with the numerical manipu- lations and orders of magnitude of various parameters which occur in the subject of heat transfer. Other problems extend the subject matter by requiring students, to apply the basic principles to new situations and develop their own equations, Both types of problems are important. The subject of heat transfer is not static. New developments occur quite regularly, and better analytical solutions and empirical data are continuously made available to the professional in the field. Because of the huge amount of information which is available in the research literature, the beginning student could easily be overwhelmed if too many of the nuances of the subject were displayed and expanded. The book is designed to serve as an elementary text, 30 the author has assumed a role of interpreter of the literature with those findings and equations being presented which can be of immediate utility to the reader. It is hoped that the student's attention is called to more extensive works in a sufficient number of instances to emphasize the greater depth which is available on most of the subjects of heat transfer. For the serious student, then, the end-of-chapter references offer an open door to the literature of heat transfer which can pyramid upon further investigation. A ‘textbook in its sixth edition obviously reflects many compromises and evolutionary processes over the years. This book is no exception. While the basic physical mechanisms of heat transfer have not changed, analytical tech- niques and experimental data are constantly being revised and improved. One objective of this new edition is to keep the exposition up to date with recent information while still retaining a simple approach which can be understood by the beginning student. ‘The computer is now the preferred vehicle for solution of many heat-transfer problems. Personal computers with either local software or communication links offer the engineer ample power for the solution of most problems. Despite the ready availability of this computing power I have resisted the temptation to include specific computer programs for two reasons: (I) each computer installation is somewhat different inits input-output capability and (2) a number of programs for microcomputers in a menu-driven format are already on the scene or soon to be available. The central issue here has been directed toward problem setup which can be adapted to any computational facility. For those persons wishing to exploit the convenience and utility of the microcomputer a separate software package, developed by Professor Alan D.Kraus of the Naval Postgraduate School, is available from McGraw-Hill. This package contains a diskette with programs as well as documentation illustrating their use. “The SI (metric) system of units is the primary one for the text. Because the Btu-ft-pound system is still in wide use, answers and intermediate steps to ‘examples are occasionally stated in these units. A few examples and problems fare completely in English units. Some figures have dual coordinates that show both systems of units. These displays will enable the student to develop a “pilingual” capability during the period before full metric conversion is achieved. For this edition examples and problems oriented toward numerical (com- puter-generated) solutions have been expanded for both steady state and tran- vient conduction in Chapters 3 and 4. New convection correlations have been added in Chapters 5, 6, and 7, and summary tables have been provided for convenience of the reader. New examples have also been provided in the radiation, convection, and heat exchanger material and over 250 new problems have been added throughout the book. Over 200 of the previous problems have been restated so that they are “*new’” for student work. In addition, all problems have been reorganized to follow the sequence of chapter topics. A total of over 850 problems is provided. With a book at this stage of revision the list of people who have been generous with their comments and suggestions has grown very long indeed. Rather than risk omission of a single name, I hope that a grateful general acknowledgment ‘will express my sincere gratitude for these persons” help and encouragement. J.P. HolmanLocal velocity of sound ‘Attenuation coefficient (Chap. 8) ‘Area Albedo (Chap. 8) = Fin profile area (Chap. 2) “Magnetic field strength © Specific heat, usually ki/kg * °C © Concentration (Chap. 11) Co Drag coefficient, defined by Eq. GB) G__ Friction coefficient, defined by Eq. (5-52) Specific heat at constant pres- sure, usually kW/kg * °C pase > cc, Specific heat at constant volume, usually Kkg * °C 4 Diameter Depth or diameter D__ Diffusion coefficient (Chap. 11) Dy Hydraulic diameter, defined by Eq. (6-14) € Internal energy per unit mass, usually kJ/kg E Internal energy, usually ky E _Emissive power, usually W/m? (Chap. 8) Es, Solar constant (Chap. 8) LIST OF SYMBOLS En =a a als. ho Blackbody emmissive power per unit wave- Tength, defined by Ea. @12) Electric field vector Friction factor, defined by Eq, (5-107) or Eq, (10-29) Force, usually N Radiation shape factor for radiation from sur- face m to surface ‘Acceleration of gravity Conversion factor, de- fined by Ea. (1-14) Mass velocity Irradiation (Chap. 8) Heat-transfer coefficient, usually Wim? «°C Mass-transfer coeffi: cient, usually nvh Enthalpy of vaporiza- tion, Ki/kg, Radiation heat-ransfer coefficient (Chap. 8)vst o symbols Sn Zz Ts esaper 2 uy ao eee ae atmosphere ree ae ee co a ei eal aac eee aa Corrected fin length (Chap. 2) es errs Molecular weight (Chap. 11) ‘Turbidity factor, defined by Eq. ie ; eee on aon aes ee ee i a es a Maecenas Jaxation method (Chaps. 3,4) nae eee imran) ome ree ee eee ee ae on ¥ rr) ae eee oe 5 Ro onewceees Ble ae ‘Thickness, applied to fin problems (Chap. 2) ‘Temperature Velocity Velocity Specific volume usually m'kg Velocity Molecular volume (Chap. 11) Weight, usually N Space coordinates in carte- sian system ‘Thermal diffusivity, usually mis Absorptivity (Chap. 8) ‘Accommodation coefficient (Chap. 12) Solar altitude angle, deg (Chap. 8) Volume coefficient of expan- sion, 17K, ‘Temperature coefficient of thermal conductivity, °C Isentropic exponent, dimen- sionless Condensate mass flow per unit depth of plate (Chap. 9) Hydrodynamic-boundary- layer thickness ‘Thermal-boundary-layer thickness, Heat-exchanger effectiveness Emissivity Eddy diffusivity of heat and momentum (Chap. 5) Ratio of thermal-boundary- layer thickness to hydrody- rnamic-boundary-layer thick- Similarity variable, defined by Eq. (B-6) Fin efficiency, dimensionless ‘Angle in spherical or cylin- drical coordinate system Temperature difference, TratwanceFo Gr The reference temperature is chosen differently for different sys- tems (see Chaps. 2004) Wavelength ‘Mean free path (Chap. 12) Dynamic viscosity Kinematic viscosity Frequency of ra diation (Chap. 8) Density, usually kgm" Reflectivity (Chap. 8) Charge density Electrical conduc- tivity Stefan-Boltzmann constant Surface tension of liquid-vapor inter- face (Chap. 9) Time Shear stress be- tween fluid layers ‘Transmissivity (Chap. 8) ‘Angle in spherical ‘or cylindrical co- ordinate system Stream function Dimensionless Groups Biot number Eckert number a Fourier number — BAT. — To Grashof ~ number: stot symbols abe Gr = GrNu Modified Grashot num ber for con- stant heat fux 4 wz = Re Pr raetz number ox = Re pr or wi Knudsen. Ka=y number « Lewis number reap (Chap. 1) Mach number Bx Magnetic: n = Bes influence num- ~ ber he Nusselt Nene number fx Average Nus- Noa ye selt number Peclet number Prandtl number Rayleigh number Reynolds umber Schmidt num: ber (Chap. 11) Sherwood umber (Chap. 11) Stanton umber Average Stan- ton number Subseripts Adiabatic wall conditions Refers to blackbody conditions (Chap. 8) Evaluated Based on diameter bulk conditions1 Ustof symbols Jf Evaluated at film conditions 1 At specified radial position 8 Saturated vapor conditions Evaluated at condition of sur- (Chap. 9) roundings {Initial or inlet conditions x Denotes some local position ’ L Based on length of plate ‘with respect to x coordinate na m Mean flow conditions w Evaluated at wall conditions ‘m,n Denotes nodal positions in nu. *-~—_(Superscript) Properties evalu- ‘merical solution (see Chaps. ated at reference temperature, 3.4) given by Eq. (5-124) © Denotes stagnation flow condi- Evaluation at free-stream con- tions (Chap. 5) or some initial ditions condition at time zeroHEAT TRANSFERINTRODUCTION Heat transfer is that science which seeks to predict the energy transfer which may take place between material bodies as a result of a temperature difference. ‘Thermodynamics teaches that this energy transfer is defined as heat. The sci- ence of heat transfer seeks not merely to explain how heat energy may be transferred, but also to predict the rate at which the exchange will take place under certain specified conditions. The fact that a heat-ransfer rate is the desired objective of an analysis points out the difference between heat transfer and thermodynamics. Thermodynamics deals with systems in equilibrium; it may be used to predict the amount of energy required to change @ system from ‘one equilibrium state to another; it may not be used to predict how fast a change will take place since the system is not in equilibrium during the process. Heat transfer supplements the first and second principles of thermodynamics by providing additional experimental rules which may be used to establish energy- transfer rates. As in the science of thermodynamics, the experimental rules used as a basis of the subject of heat transfer are rather simple and easily expanded to encompass a variety of practical situations ‘As an example of the different kinds of problems which are treated by thermodynamics and heat transfer, consider the cooling of a hot steel bar which is placed in a pail of water. Thermodynamics may be used to predict the final equilibrium temperature of the steel bar-water combination. Thermodynamics will not tell us how long it takes to reach this equilibrium condition or what the temperature of the bar will be after a certain length of time before the equilibrium condition is attained. Heat transfer may be used to predict the ‘temperature of both the bar and the water as a function of time. Most readers will be familiar with the terms used to denote the three modes of heat transfer: conduction, convection, and radiation. In this chapter we seek to explain the mechanism of these modes qualitatively so that each may be considered in its proper perspective. Subsequent chapters treat the three types of heat transfer in detail 12 Invedcton 1-1. CONDUCTION HEAT TRANSFER When a temperature gradient exists in a body, experience has shown that there is an energy transfer from the high-temperature region to the low-temperature region. We say that the energy is transferred by conduction and that the heat- transfer rate per unit area is proportional to the normal temperature gradient: apes Ay ax ‘When the proportionality constant is inserted, ar = at 1 q--uZ an where q is the heat-transfer rate and T/ax is the temperature gradient in the direction of the heat flow. The positive constant k is called the thermal con- ductivity of the material, and the minus sign is inserted so that the second principle of thermodynamics will be satisfied; i.e., heat must flow downhill on the temperature scale, as indicated in the coordinate system of Fig. 1-1. Equa- tion (1-1) is called Fourier’s law of heat conduction after the French mathe- ‘matical physicist Joseph Fourier, who made very significant contributions to the analytical treatment of conduction heat transfer. It is important to note that Eq, (1-1) is the defining equation for the thermal conductivity and that k has the units of watts per meter per Celsius degree in a typical system of units in which the heat flow is expressed in watts. ‘We now set ourselves the problem of determining the basic equation which ‘governs the transfer of heat in a solid, using Eq. (I-1) as a starting point. Consider the one-dimensional system shown in Fig. 1-2. If the system is in a steady state, i.e., if the temperature does not change with time, then the problem is a simple one, and we need only integrate Eq. (1-1) and substitute the appropriate values to solve for the desired quantity. However, if the tem- perature of the solid is changing with time, or if there are heat sources or sinks within the solid, the situation is more complex. We consider the general case where the temperature may be changing with time and heat sources may be 1 ‘Temperature ‘oie Tig. 14 Sketch showing direction of heat fow.CConcicton heat ansler 3 « lee Fig. 1-2 Elemental volume or one-dimensional heat foxes ceonduetion analy present within the body. For the element of thickness dx the following energy balance may be made: Energy conducted in left face + heat generated within element = change in internal energy + energy conducted out right face ‘These energy quantities are given as follows: Energy in left face = 9, = Energy generated within element = 4A dx CChange in internal energy = pea % dx ar Energy out right face = qryae = —kA 2] ® wales 2 (22 [e+e 8) +] energy eeerated per nt yume, Win? specific heat of material, 1/kg-°C p= density, kg/m? where Combining the relations above gives aT, ar ar ka + GA dx = pA ds a[eze a (,a or S (12) eee (2) +] g1§ Ble 2)This is the one-dimensional heat-conduction equation. To treat more than one- dimensional heat flow, we need consider only the heat conducted in and out of a unit volume in all three coordinate directions, as shown in Fig. 1-3a. The energy balance yields aE Invite + Iyody + Geode + @ Gy vay + 4 a Ge ty + de + Gee and the energy quantities are given by a= ~k dy do Qevas = - [e262 (1S) a] owe o* wk de de ew =~ [+ 2 (eZ) 9] ara ar 9, = ~kdx dys dere = («+2 (2) ae] ds dy duen = G de dy de dE a 0 that the general three-dimensional heat-conduction equation is 2 (,T) , 2 (a0), a(,2 ar ROS) +a (5) ta (G) tere eo For constant thermal conductivity Eq. (1-3) is written ar eT eT ¢ la ae ae tad ka ae where the quantity a = kipc is called the thermal diffusivity of the material. ‘The larger the value of a, the faster heat will diffuse through the material. This may be seen by examining the quantities which make up a. A high value of a could result either from a high value of thermal conductivity, which would indicate a rapid energy-transfer rate, or from a low value of the thermal heat capacity pe. A low value of the heat capacity would mean that less of the energy moving through the material would be absorbed and used to raise the temper- ature of the material; thus more energy would be available for further transfer. ‘Thermal diffusivity a has units of square meters per second. ar pe dx dy dz == (13a)Conduction nest raneer 8 few dacarae « o Fig. 1-3. Elemental volume for three-dimensional heal-conduction analysis. (a) cartesian coor K1)(T. — 110) = 3960 W and. T. = USC (41°F) ‘The heat generated per unit volume g is calculated from P w Gar 3960, = SATO ~ 5-2 MW Im (541 > 10" Bru Finally, the center temperature of the wire is calculated from Eq, (2-26) 2 S.602 x 1ONLLS x 10-9)? +215 = 231.6 [445 Tas 21 (arr) ot, nearConductoncomecton systems 43 2.9 CONDUCTION-CONVECTION SYSTEMS ‘The heat which is conducted through a body must frequently be removed (or delivered) by some convection process. For example, the heat lost by con- duction through a furnace wall must be dissipated to the surroundings through convection. In heat-exchanger applications a finned-tube arrangement might be used to remove fieat from a hot liquid. The heat transfer from the liquid to the finned tube is by convection. The heat is conducted through the material and finally dissipated to the surroundings by convection. Obviously, an analysis, of combined conduction-convection systems is very important from a practical standpoint. We shall defer part of our analysis of conduction-convection systems to Chap. 10 on heat exchangers. For the present we wish to examine some simple extended-surface problems. Consider the one-dimensional fin exposed to a surrounding fluid at a temperature T.. as shown in Fig. 2-9. The temperature Of the base of the fin is Ty. We approach the problem by making an energy balance on an element of the fin of thickness dx as shown in the figure. Thus Energy in left face = energy out right face + energy lost by convection The defining equation for the convection hez = bAT, ~ T) 229) where the area in this equation is the surface area for convection. Let the cross- sectional area of the fin be A and the perimeter be P. Then the energy quantities are (transfer coefficient is recalled as Energy in left face = gy HPT) Fig. 29 Sketch ilustraing one. ‘dimensional conduction and con: ‘ection through a rectangular fn44, Siendy-siate corducton—one cimension Energy out right face Energy lost by convection AP dx(T - Ta) Here it is noted that the differential surface area for convection is the product of the perimeter of the fin and the differential length dx. When we combine the s, the energy balance yields eT AP Bat T=0 (2-304) Let 9 = T ~ T.. Then Eq. (2-30a) becomes 20 WP mmo (2-306) One boundary condition is O=%=To-Te atx=0 The other boundary condition depends on the physical situation, Several cases may be considered: case 1 The fin is very long, and the temperature at the end of the fin is es- sentially that of the surrounding fluid, case2 The fin is of finite length and loses heat by convection from its end. ©ASES The end of the fin is insulated so that dT idx = O atx = L. Tf we let m? = AP/KA, the general solution for Eq. (2-306) may be written 0 = Cem + Crem 30 For case | the boundary conditions are = % atx=0 O=0 atram and the solution becomes fomeeTieit 4 TT. For case 3 the boundary conditions are % atx =0 em 232) a Gito ateakConcctoncarvecton systems 48 Thus - 0 = m(—Cye-m! + Cye™*y Solving for the constants C, and C3, we obtain aaa a Treen + Ta eine oe cosh [m(L = x1 = cosh [miL ~ x)] 2335) cosh mL ‘The hyperbolic functions are defined as cosh. x = sinhx _ e* tanh x = cosh et tee ‘The solution for case 2 is more involved algebraically, and the result is plea q=T. All of the heat lost by the fin must be conducted into the base atx = 0. Using the equations for the temperature distribution, we can compute the heat loss trom ‘An alternative method of integrating the convection heat loss could be used: APT — Taydx = f° hPOds In most cases. however. the first equation is easier to apply. For case 1, 4 = KAU mite") = VRPEA ty (235) For case 3, = =kAdym ‘ er 2:36) = VAPEA 6 tanh mL The heat flow for case 2 is 4 = VAPIA (To ~ 1.) Shee + th 37 cosh mL. + (himk) sin mL. In the above development it has been assumed that the substantial temperature gradients occur only in the x direction. This assumption will be satisfied if the448 Steady-state condvetion—one dinenson fin is sufficiently thin, For most fins of practical interest the error introduced by this assumption is less than I percent. The overall accuracy of practical fin calculations will usually be limited by uncertainties in values of the convection coefficient h. It is worthwhile to note that the convection coefficient is seldom uniform over the entire surface, as has been assumed above. If severe non- uniform behavior is encountered, numerical finite-difference techniques must bbe employed to solve the problem. Such techniques are discussed in Chap. 3. 210 FINS In the foregoing development we derived relations for the heat transfer from a rod or fin of uniform cross-sectional area protruding from a flat wall. In practical applications. fins may have varying cross-sectional areas and may be attached to circular surfaces. In either case the area must be considered as a variable in the derivation, and solution of the basic differential equation and the mathematical techniques become more tedious. We present only the results for these more complex situations. The reader is referred to Refs. | and 8 for details on the mathematical methods used to obtain the solutions. To indicate the effectiveness of a fin in transferring a given quantity of heat, a new parameter called fin efficiency is defined by actual heat transferred _ heat which would be transferred — if entire fin area were at base temperature For case 3 above, the fin efficiency becomes _ VARA Q tanh ml. _ tanh mL _ APL 6 ~ mb The fins discussed above were assumed to be sufficiently deep that the heat flow could be considered one-dimensional. The expression for mL. may be written mL = |aP, [Qe +20, oy VO kee where zis the depth of the fin and ris the thickness. Now, if the fin is sufficiently deep, the term 2z will be large compared with 21, and oO [2h 2h, ee cen eg Multiplying numerator and denominator by L'? gives [2h so mL = Vie’ Lr is the profile area of the fin, which we define as Fin efficiency = v y (238)Fee 47 An = Lt [Ryn so that mb = Va 239) We may therefore use the expression in Eq. (2-39) to compute the efficiency of a fin with insulated tip as given by Eq. 2-38). Harper and Brown [2] have shown that the solution in case 2 above may be ‘expressed in the same form as Eq. (2-38) when the length of the fin is extended by one-half the thickness of the fin. A corrected length L. is then used in all the equations which apply for the case of the fin with an insulated tip. Thus t ees Lek+s (2-40) The error which results from this approximation will be less than 8 percent when myer! 241 i) A 41) If a cylindrical spine extends from a wall as shown in Fig. 2-10g, the corrected fin length is calculated from dla ad L+dis (2-42) Examples of other types of fins are shown in Fig. 2-10 according to Ref. 8. Figure 2-11 presents a comparison of the efficiencies of a triangular fin and a « o @ @ w w o o Fig. 2410 Ditieren ypes of tinea surlaces, according to Kem and Kraus (8). (a) Longitudinal tin of rectangular proie, (0) cylndcal tube equipped wit ins of rectangular profie;(c) long tudinal fin oftrapezoidal prolie (d) long tudinal in of parabolic prot, (e) ylinccal tube equipped ‘with racal in of rectangular prove; cyindtica tube equipped wih radial fin of truncated conical Prot, () eyinariea! spine; (H) tuncated conical spre. () parabolic spine48. Sindy tale conducten—one cmension t + £ rectanguiar fin b+ $ rectang Langa tn he rectangular tin twangular fin Fm ftsemy my pecent 2A1 Eticiencies of rectan Lema! ular ang tangula ins straight rectangular fin corresponding to case 2. Figure 2-12 shows the effi ciencies of circumferential fins of rectangular cross-sectional area. Notice that the corrected fin lengths L, and profile area A., have been used in Figs 2-11 and 2-12. We may note that as r:,/ri > 1.0, the efficiency of the circumferential fin becomes identical to that of the straight fin of rectangular profile, Itis interesting to note that the fin efficiency reaches its maximum value for the trivial cave of L = 0, or no fin at all. Therefore. we should not expect te be able 10 maximize fin performance with respect to fin length. It 1s possible. however, to maximize the efficiency with respect to the quantity of fin materi! (mass, volume, or cost). and such a maximization process has rather obvious economic significance. We have not discussed the subject of rudiation heat transfer from fins, The radiant transfer is an important consideration in a number of applications. and the interested reader should consult Siege! and Howell [9 for information on this subject In some cases a valid method of evaluating fin performance is to compare the heat transfer with the fin to that which would be obtained without the finFre 49 Fig. 212. Eticiencies of cvcunferenia ins of rectangular piotle, according to fet 3 ‘The ratio of these quantities is —auvith fin _ mAphdy qwithout fin hA,6> where A, is the total surface area of the fin and A, is the base area. For the insulated tip fin described by Eq. (2-36), Ar= PL AA and the heat ratio would become q.with fin _ tanh mk 4q without fin ~ VRAIKP | Conditions When Fins Do Not Help At this point we should remark that the installation of fins on a heat-transfer surface will not necessarily increase the heat-transfer rate. If the value of h, the convection coefficient, is large, as it is with high-velocity fluids or boiling liquids, the fin may produce a reduction in heat transfer because the conduction resistance then represents a larger impediment to the heat flow than the con- vection resistance. To illustrate the point, consider a stainless-steel pin fin which has k = 16 Wim -°C, L = 10cm, d = 1 em and which is exposed to|80 Stendysateconducton—one mension a boiling-water convection situation with k = 5000 Wim? - °C. From Eq. (2-36) we can compute qwith fin. _ tanh mL q without 244)" ES ee FO) uo x 10-9} [some x 10732)" 6a x 10-5), 13 ‘Thus, this rather large pin produces an increase of only 13 percent in the heat transfer. Still another method of evaluating fin performance is discussed in Prob. 2- 68. Ken and Kraus [8] give a very complete discussion of extended-surface heat transfer. Some photographs of different fin shapes used in electronic cool- ing applications are shown in Fig. 2-13. 1» EXAMPLE 2-5 influence of thermal conductivity on fn temperature profes Compare the temperature distributions in a straight fin of rectangular profile having a thickness of 2 cm and a length of 10 em and exposed to a convection environment with 25 Wim? « °C, for three fin materials: copper {k = 385 Wim - °C], stainless steel [k= 17 Wim - °C), and glass [k = 0.8 Wim - °C]. Also compare the relative heat flows and fin efficiencies. Solution We have AP _ (25)n(0.02) _ 5000 1A * “kx(0.00? ‘The terms of interest are therefore ak bs Material a n mL Copper 12.99 3.604 0,3604 Stainless steel 294.1 17.5 7s Glass 6250 9.067.906Fee 84 Fig. 2.18. Some fin arrangements used in electronic cooling applications. (Courtesy Wakatield Engineering Inc., Wakefield, Mass.) ‘These values may be inserted into Eq. (2-332) to calculate the temperatures at different «locations along the rod, and the results are shown in the accompanying figure. We notice that the glass behaves as a “‘very long’” fin, and its behavior could be calculated from Eq. (2-32). The fin efficiencies are calculated from Eq. (2-38) by using the corrected length approximation of Eq. (2-40). We have L=L+5 oe 3 = Mem 430i‘The parameters of interest for the heat flow and efficiency comparisons are now tabu- lated as Material —_—_‘APKA mL, Copper 0.190 0.3964 Stainless steel 0.0084 1.8865 Glass 39% 10 8.697 To compare the heat flows we could either calculate the values from Eq, (2-36) for a iS lunit value of 6, or observe that the fin efficiency gives a relative heat-low comparison : because the maximum heat transfer is the same for all three cases: i.e,, we are dealing with the same fin size, shape, and value of h. We thus calculate the values of 1y from Eq, (2-38) and the above values of mL, relative (9 Materiat copper, Fe Copper 0.951 100 Stainless steel 0.506 36 Glass ons mI ‘The temperature profiles in the accompanying figure can be somewhat misleading. The lass has the steepest temperature gradient at the base, but its much lower value of k Produces a lower heat-transfer rate Fig. Ex. 25Fes sa = EXAMPLE 2-6 ‘An aluminum fin [k = 200 Wim - °C] 3.0 mm thick and 7.5 cm long protrudes from a wall, asin Fig, 2-9. The base is maintained at 300°C, and tue ambient temperature is 50°C with k= 10 Wim? - °C, Calculate the heat loss from the fin per unit depth of material, ‘Solution ‘We may use the approximate method of solution by extending the fin a fictitious length 112 and then computing the heat transfer from a fin with insulated tip as given by Eq. (2:36). We have L402 = 75 + 0.15 = 7.65 em [3.01 in} (iB [ase any” moa” ae | wen the fin depth «> 1. So exw)” m+ lawoa'ewrs) -57 From Eq. (2-36), for an insulated-tip fin @ = (tanh mL.) VRPIA 8 tanh mL, For a 1-m depth A= (13. 10°) = 3 10-7 me (4.65 int} and 4 = (5.774)200\3. x 10-300 — 50) tanh {(5.774N'0.0765)), 359 Wim [373.5 Btwih » ft} = EXAMPLE 27 ‘Aluminum fins 1.5 cm wide and 1,0 mm thick are placed on a 2.5-cm-diameter tube to dissipate the heat. The tube surface temperature is 170°C. and the ambient-Auid tem- perature is 25°C. Calculate the heat loss per fin for h = 130 Wim? ~°C. Assume k = 200 Wim - °C for aluminum. ‘Soltion For this example we can compute the heat transfer by using the fin-efficiency curves in Fig. 2-12. The parameters needed are Low L +e = 1S +005 = 1.55 em 2.52 = 1.25em rn = n+ Ly = 125 + LSS = 2.80.em rade = 28OV1.25 = 2.24 Ag = tr. ~ 1) = @.0012.8 ~ 1.25(10- 9) = 1.55 x 10° me Bo = 0135 [apes 0) [om wa)164 Steady-state conducien-—cne amonsion From Fig. 2-12 ry = 82 percent. The heat which would be transferred if the entire fin were at the base temperature is (both sides of fin exchanging heat) ou = Petra? — H7VM(To ~ To) (2.8? ~ 1.25°)10-Y(130K170 — 25) 4.35 W [253.7 Buh] ‘The actual heat transfer is then the product of the heat flow and the fn efficiency: duc: = (0.82\74.35) = 60.97 W (208 Btu/b} EXAMPLE 2.8 Rod with heat sources. ‘A rod containing uniform heat sources per unit volume 4 is connected to two temper- atures as shown in the accompanying figure. The rod is also exposed to an environment ‘with convection coefficient A and temperature T.. Obtain an expression for the tem- perature distribution in the rod. Solution We first must make an energy balance on the element of the rod shown, similar to that used to derive Eq. (2-30). We have Energy in left face + heat generated in element = energy out right face + energy lost by convection we eaten ta (Ee Eas) oar 9 ‘Simplifying, we have or, with @ = T o‘We can make a further variable substitution as 0 = 0 alkent ‘0 that our differential equation becomes (dns oo - mo =0 © which has the general solution O = cerm + eens @ ‘The two end temperatures are used to establish the boundary conditions: T, ~ Te ~ lhe = C, + Cy 0 = Ob = Ty ~ To ~ dik? = Cert + Cent Solving for the constants C, and C: gives © 241 THERMAL CONTACT RESISTANCE Imagine two solid bars brought into contact as indicated in Fig. 2-14, with the sides of the bars insulated so that heat flows only in the axial direction. The materials may have different thermal conductivities, but if the sides are insu- lated, the heat flux must be the same through both materials under steady-state conditions. Experience shows that the actual temperature profile through the ‘two materials varies approximately as shown in Fig. 2-14b. The temperature a Fig. 2-14 tlustratons of therm: * contactresistance effect: (a) physical stuation;(b) temperature 0 roti{80 Steady-state conducton—one dimeneon drop at plane 2, the contact plane between the two materials, is said to be the result of a thermal contact resistance. Performing an energy balance on the two materials, we obtain Ti = Tu | Tun - Tas Tos = Ty 90 be ha ~ KA T-% or q (2-43) * KaglksA + Wh-A + BxalkyA where the quantity 1/h.A is called the thermal contact resistance and A, is called the contact coefficient. This factor can be extremely important in a number of applications because of the many heat-transfer situations which involve me- chanical joining of two materials. The physical mechanism of contact resistance may be better understood by examining a joint in more detail, as shown in Fig. 2-15. The actual surface roughness is exaggerated to implement the discussion. No real surface is per- fectly smooth, and the actual surface roughness is believed to play a central role in determining the contact resistance. There are two principal contributions to the heat transfer at the joint: 1. The solid-to-solid conduction at the spots of contact 2. The conduction through entrapped gases in the void spaces created by the contact The second factor is believed to represent the major resistance to heat flow. because the thermal conductivity of the gas is quite small in comparison to that of the solids. Designating the contact area by A, and the void area by A., we may write for the heat flow across the joint Tx -T gy Ta - Ta - TikyA, + Lyftkes, * AT ha where L, is the thickness of the void space and ky is the thermal conductivity of the fluid which fills the void space. The total cross-sectional area of the bars is A. Solving for h., the contact coefficient, we obtain q Fig. 218 Jontoughness model for analyse of thermal contact resistance,Terma! contact esitance 87 Ac thu ; (& peer 2-44) Akt ke A In most instances, air is the fluid filling the void space and kis small compared with k, and ky. If the contact area is small, the major thermal resistance results from the void space. The main problem with this simple theory is that it extremely difficult to determine effective values of A., Ay, and L, for surfaces in contact. From the above physical model, we may tentatively conclude: L. The contact resistance should increase with a decrease in the ambient gas pressure when the pressure is decreased below the value where the mean free path of the molecules is large compared with a characteristic dimension of the void space, since the effective thermal conductance of the entrapped ‘gas will be decreased for this condition. 2. The contact resistance should be decreased for an increase in the joint pres- sure since this results in a deformation of the high spots of the contact surfaces, thereby creating a greater contact area between the solids. A very complete survey of the contact-resistance problem is presented in Refs. 4, 6, 7, and 10. Unfortunately, there is no satisfactory theory which will, predict thermal contact resistance for all types of engineering materials, nor have experimental studies yielded completely reliable empirical correlations. Table 2-2. Contact Conductance of Typical Surfaces Pres —_ sure, PER) mi °C aim Btu X10 416 Stainless, ground, air 100254 90-200 3-25 0.0015 264 304 Stainless, ground, air 45.4620 40-700.008, 5.28 416 Stainless, i 254 30.2007 0.002 332 with 0.00 shim, sir ‘Aluminum, ground, air 1002.54 150 12-28——0.0005 088 10 025 1302-28 0.0001, os ‘Aluminum, ground, with 100284150, 12-200 0.0007 123 (0.001-in brass shim, Copper, ground, air so 127 12-200 0,00008 oar Copper, milled-ir 10381 20 10-S0—0.0001 os Copper, milled, vacuum 100.250 7-70—_—0.0005 0.88{68 Sieagystateconducton one dimension This is understandable because of the many complex surface conditions which i may be encountered in practice. ! Radiation heat transfer across the joint can also be important when high temperatures are encountered. This energy transfer may be calculated by the methods discussed in Chap. 8. For design purposes the contact conductance values given in Table 2-2 may be used in the absence of more specific information. Thermal contact resistance can be reduced markedly, perhaps as much as 75 percent, by the use of a “thermal grease" like Dow 340, EXAMPLE 20 ‘Two 3.0-cm-diameter 304 stainless-steel bars, 10.em long, have ground surfaces and are exposed to air with a surface roughness of about 1 jum. If the surfaces are pressed. together with a pressure of 50 atm and the two-bar combination is exposed to an overall, temperature difference of 100°C, calculate the axial heat flow and temperature drop across the contact surface. solution ‘The overall heat flow is subject to three thermal resistances, one conduction resistance for each bar, and the contact resistance. For the bars ar (CoO | Ro KA © (16.3)m3 x 10-7)? assed From Table 2-2 the contact resistance is : “A | 6.28 x 10-48) ; Ty SRO | onrew is ‘The total thermal resistance is therefore ER = (2X8.679) + 0.747 andthe overall heat flow is AT _ _100 ER ~ 79.108 “The temperature drop across the contact is found by taking the ratio ofthe contact resistance to the total thermal resistance: Re gp - (O747K100) ~ ER 18.105 18.105 @ 5-52 W [18.83 Bru/h} aT. = 413°C [39.43°F] i VIEW QUESTIONS ‘What is meant by the term one-dimensional when applied to conduction problems? What is meant by thermal resistance? Why is the one-dimensional heat-flow assumption important in the analysis of fins? Define fin efliciency. |Problems 59 ‘5 Why is the insulated-tip solution important for the fin problems? 6 What is meant by thermal contact resistance? Upon what parameters does this resistance depend? PROBLEMS 241 A wall 2cm thick is to be constructed from material which has an average thermal conductivity of 1.3 Wim - °C. The wall is to be insulated with material having an average thermal conductivity of 0.35 WimC, so that the heat loss per square ‘meter will not exceed 1830 W. Assuming that the inner and outer surface tem: peratures of the insulated wall are 1300 and 30°C, calculate the thickness of in sulation required. 22 A certain material 2.5 em thick, with a cross-sectional area of 0.1 m?, has one side ‘maintained at 35°C and the other at 95°C. The temperature at the center plane of the material is 62°C, and the heat flow through the material is | kW. Obtain an expression for the thermal conductivity of the material as a function of temperature. 23. A composite wall is formed of a 2.5-em copper plate, a 3.2-mm layer of asbestos, and a S-cm layer of fiber glass. The wall is subjected to an overall temperature difference of 560°C. Calculate the heat flow per unit area through the composite structure. ‘24 Find the heat transfer per unit area through the composite wall sketched. Assume ‘one-dimensional heat flow. ky = (So Wine G0 =m Sem—vfe 50cm Fig. P24 25 One side of a copper block $ cm thick is maintained at 260°C. The other side is covered with a layer of fiber glass 2.5 em thick. The outside of the fiber glass is ‘maintained at 38°C. and the total heat flow through the copper-fiber-glass combi nation is 44 kW. What is the area of the slab? 26 An outside wall for a building consists of @ 10-cm layer of common brick and a 2.S-cm layer of fiber glass (k = 0.05 Wim » °C]. Calculate the heat flow through the wall for a 45°C temperature differential(60, Steady-state conductonone dimension 27 2.10 aa 2a 213, 216 a7 One side of a copper block 4 cm thick is maintained at 175°C. The other side is covered with a layer of fiber glass 1.5 cm thick. The outside of the fiber glass is ‘maintained at 80°C, and the total heat flow through the composite slab is 300 W. ‘What is the area of the slab? ‘A plane wall is constructed of a material having a thermal conductivity that varies a the square of temperature according to the relation k = ky(I + BT°). Derive ‘an expression for the heat transfer in such a wall. A certain material has a thickness of 30 cm and a thermal conductivity of 0.04 ‘Wim - °C. At a particular instant in time the temperature distribution with x, the distance from the left face, is T = 150x? - 30x, where x is in meters. Calculate the heat flow rates at x = 0 and x = 30 cm, Is the solid heating up or cooling down? A wall is constructed of 2.0 em of copper, 3.0 mm of asbestos sheet [k = 0.166 ‘Wim - °C}, and 6.0 em of fiber glass. Calculate the heat low per unit area for an overall temperature difference of 500°C. A certain building wall consists of 6.0 in of concrete {k = 1.2 Wim : °C], 2.0 in ‘of fiber-glass insulation, and dn of gypsum board [k = 0.05 Wim - °C]. The inside ‘and outside convection coefficients are 2.0 and 7.0 Btuh - ft - °F, respectively. ‘The outside air temperature is 20°F, and the inside temperature is 72°F. Calculate the overall heat-ransfer coefficient for the wall, the R value, and the heat loss per unit area A wall is constructed of a section of stainless steel [ = 16 Wim -°C] 4.0 mm thick with identical layers of plastic on both sides of the steel. The overall heat- transfer coefficient, considering convection on both sides of the plastic, is 120 ‘Wim? - °C. If the overall temperature difference across the arrangement is 60°C, calculate the temperature difference across the stainless steel. ‘An ice chest is constructed of styrofoam (k = 0.033 Wim - °C] with inside di ‘mensions of 25 by 40 by 100 cm. The wall thickness is 5.0 cm. The outside of the chest is exposed to airat 25°C with h = 10 Wim? - °C. Ifthe chest is completely filled with ice, calculate the time for the ice to completely melt. State your as- sumptions. The heat of fusion for water is 330 ki/kg, A spherical tank, 1-m in diameter, is maintained at a temperature of 120°C and exposed to a convection environment. With k= 25 Wim? ~°C and T. =15°C, ‘what thickness of urethane foam should be added to ensure that the outer tem- perature of the insulation does not exceed 40°C? What percentage reduction in heat loss results from installing this insulation? A hollow sphere is constructed of aluminum with an inner diameter of 4 cm and fan outer diameter of 8 cm. The inside temperature is 100°C and the outer tem- perature is SO°C. Calculate the heat transfer. ‘Suppose the sphere in problem 2-15 is covered with a 1-cm layer of an insulating ‘material having k = 50 mWim -°C and the outside of the insulation is exposed toanenvironment with h = 20 Wim? » °Cand T. = 10°C. The inside of the sphere remains at 100°C. Calculate the heat transfer under these conditions. In Appendix A, dimensions of standard steel pipe are given. Suppose a3-in sched2418 219 22 223 228 227 ule 80 pipe is covered with 1 in of an insulation having k = 60 mW/m - *C and the outside ofthe insulation is exposed to an environment having h = 10 Wim? « °C and 7. = 20°C. The temperature ofthe inside ofthe pipe is 250°C. For unit length of the pipe calculate (a) overall thermal resistance and (b) heat loss. A stee! pipe witha 5-cm OD is covered with a6.4-mm asbestos insulation {k = 0.096 Buwh - ft “F] followed by a 2.5-cm layer of fber-glass insulation [k = 0.028 Btwh - ft - °F]. The pipe-wall temperature is 315°C, and the outside insulation temperature is 38°C. Calculate the interface temperature between the asbestos and fiber glass. Derive an expression for the thermal resistance through a hollow spherical shell of inside radius r, and outside radius r, having a thermal conductivity k. ‘A LO-mm-diameter wire is maintained at a temperature of 400°C and exposed to a convection environment at 40°C with h = 120 Wim? «°C. Calculate the thermal conductivity which will just cause an insulation thickness of 0.2 mm to produce “critical radius.” How much of this insulation must be added to reduce the at transfer by 75 percent from that which would be experienced by the bare ‘A.2.0-in schedule 40 steel pipe (see Appendix A) has k= 27 Btuhh ft - °F. The fluid inside the pipe has h = 30 Btu/h - ft °F, and the outer surface of the pipe is covered with 0.5-in fber-glass insulation with k = 0.023 Btwh - ft °F. The convection coefficient on the outer insulation surface is 2.0 Btwh = ft °F. The inner fluid temperature is 320°F and the ambient temperature is 70°F. Calculate the heat loss per foot of length. Derive a relation for the critical radius of insulation for a sphere. A cylindrical tank 80 cm in diameter and 2.0 m high contains water at 80°C. The tank is 90 percent full, and insulation is to be added so that the water temperature will not drop more than 2°C per hour. Using the information given in this chapter, specify an insulating material and calculate the thickness required for the specified cooling rate. ‘A hot steam pipe having an inside surface temperature of 250°C has an inside diameter of 8 cm and a wall thickness of 5.5 mm. It is covered with a 9-cm layer of insulation having k = 0.5 Wim + °C, followed by a 4-cm layer of insulation having k = 0.25 Wim» °C. The outside temperature of the insulation is 20°C. Calculate the heat lost per meter of length. Assume k = 47 Wim - °C for the pipe. ‘A house wall may be approximated as two 1.2-cm layers of fiber insulating board, 1 8.0-cm layer of loosely packed asbestos, and a 10-cm layer of common brick. ‘Assuming convection heat-transfer coefficients of 15 Wim? « °C on both sides of the wall, calculate the overall heat-transfer coefficient for this arrangement. Calculate the R value for the following insulations: (a) urethane foam, (b) fiber- alass mats, (c) mineral wool blocks, (d) calcium silicate blocks. ‘An insulation system is to be selected for a furnace wall at 1000°C using first a layer of mineral wool blocks followed by fiber-glass boards. The outside of the ulation is exposed to an environment with A= 15 Wim? - °C and 7. = 40°C. Using the data of Table 2-1 calculate the thickness of each insulating material(62 s1030)sateconsucion—ane denon 290 a a such that the interface temperature is not greater than 400°C and the outside temperature is not greater than 55°C. Use mean values for the thermal conduc- tivities. What is the heat loss in this wall in watts per square meter? Derive an expression for the temperature distribution in a plane wall having uniformly distributed heat sources and one face maintained at a temperature T, while the other face is maintained at a temperature 7;. The thickness of the wall ‘may be taken as 2L. Derive an expression for the temperature distribution in a plane-wall in which distributed heat sources vary according to the linear relation, 4 = 4 + BT - TO) where Gu is a constant and equal to the heat generated per unit volume at the wall temperature T... Both sides of the plate are maintained at 7, and the plate thickness is 2L. A plane wall 6.0 cm thick generates heat internally atthe rate of 0.3 MW/m?. One side of the wall is insulated, and the other side is exposed to an environment at 98°C. The convection heat-transfer coefficient between the wall and the environ- ‘ment is 570 Wim? - °C. The thermal conductivity of the wall is 21 Wim - °C. Calculate the maximum temperature in the wall Consider a shielding wall for a nuclear reactor. The wall receives a gamma-ray flux such that heat is generated within the wall according to the relation a= he where gois the heat generation at the inner face of the wall exposed to the gamms- ray flux and a is a constant. Using this relation for heat generation, derive an expression for the temperature distribution in a wall of thickness L, where the inside and outside temperatures are maintained at T, and T,, respectively. Also ‘obtain an expression for the maximum temperature in the wall Repeat Prob. 2-31, assuming that the outer surface is adiabatic while the inner surface temperature is maintained at 7; Rework Prob. 2-29 assuming that the plate is subjected to a convection environ- ‘ment on both sides of temperature T. with a heat-transfer coefficient h. T, is now ‘some reference temperature not necessarily the same as the surface temperature. Heat is generated in a 2.5-cm-square copper rod at the rate of 35.3 MW/m?. The rod is exposed to a convection environment at 20°C, and the heat-transfer coef- ficient is 4000 Wim? - °C. Calculate the surface temperature of the rod. ‘Aplane wall of thickness 2L has an intemal heat generation which varies according to q = de cos ax, where dy is the heat generated per unit volume at the center of the wall (= 0) and a is a constant. If both sides of the wall are maintained ata constant temperature of 7... derive an expression forthe total heat loss from the wall per unit surface area A certain semiconductor material has a conductivity of 0.0124 Wem - °C. A rectangular bar of the material has a cross-sectional area of 1 cm? and a length of 3. cm, One end is maintained at a 300°C and the other end at 100°C, and the2a7 238 299 241 243 244 bar carries a current of 50 A. Assuming the longitudinal surface is insulated, calculate the midpoint temperature in the bar. Take the resistivity as 15 x 10-90 om, ‘The temperature distribution in a certain plane wall is where T; and T; are the temperatures on each side of the wal, If the thermal conductivity of the wall is constant and the wall thickness is L, derive an expres- sion for the heat generation per unit volume as a function of x, the distance from the plane where T = T,, Let the heat generation rate be gy al x = Electric heater wires are installed in a solid wall having a thickness of & em and k = 2.5 Wim-°C. The right face is exposed to an environment with h = 50 ‘Wim? - °C and T. = 30°C, while the left face is exposed to h = 75 Wim? » °C and T= 50°C. What is the maximum allowable heat generation rate such that the ‘maximum temperature in the solid does not exceed 300°C? ‘A 2.0-cmthick plate has heat generated uniformly at the rate of S x 10* Wim’. ‘One side of the plate is maintained at 200°C and the other side at S0°C. Calculate ‘the temperature at the center of the plate for k = 20 Wim « Heat is generated uniformly in a stainless steel plate having k = 20 Wim - °C. ‘The thickness of the plate is 1.0 cm and the heat generation rate is 500 MWim. Ifthe two sides ofthe plate are maintained at 100 and 200°C respectively, calculate the temperature at the center of the plate A plate having a thickness of 4.0 mm has an internal heat generation of 200 MW/m* ‘and a thermal conductivity of 25 Wim °C. One side of the plate is insulated and the other side is maintained at 100°C. Calculate the maximum temperature in the plat. A3.2-mm-diameter stainless-steel wire 30 cm long has a voltage of 10 V impressed. fn it, The outer surface temperature of the wire is maintained at 93°C. Calculate the center temperature of the wire. Take the resistivity of the wire as 70 wf - em and the thermal conductivity as 22.5 Wim: °C. The heater wire of EX. 2-4 is submerged in a fluid maintained at 93°C. The convection heat-transfer coefficient is 5.7 kWim? - °C. Calculate the center tem- perature of the wire. ‘An electric current is used to heat a tube through which a suitable cooling fluid flows. The outside of the tube is covered with insulation to minimize heat loss to the surroundings, and thermocouples are attached to the outer surface of the tube to measure the temperature. Assuming uniform heat generation in the tube, derive an expression for the convection heat-transfer coefficient on the inside of the tube in terms of the measured variables: voltage E, current I, outside tube wall temperature To, inside and outside radii r, and r., tube length L, and fluid temperature 7. Derive an expression for the temperature distribution in a sphere of radius r with uniform heat generation q and constant surface temperature T..(64, Steed-sate conduction one dimension A stainless-steel sphere {k = 16 Wim » °C] having a diameter of 4 cm is exposed to a convection environment at 20°C, h = 15 Wim? - °C. Heat is generated uni- formly in the sphere at the rate of 1.0 MWim?. Calculate the steady-state tem- perature for the center of the sphere ‘An aluminunvalloy electrical cable has k = 190 Wim - °C, a diameter of 30 mm, and carries an electric current of 230 A. The resistivity ofthe cable is 2.9 ft - cm, and the outside surface temperature of the cable is 180°C. Calculate the maximum temperature in the cable if the surrounding air temperature is 15°C. Derive an expression for the temperature distribution in a hollow cylinder with heat sources which vary according to the linear relation gaat br with g, the generation rate per unit volume at r= r,. The inside and outside temperatures are T= T,atr = rand T= T, atr ‘The outside of a copper wire having a diameter of 2 mm is exposed to a convection ‘environment with k = $000 Wim? - °C and T. = 100°C. What current must be passed through the wire to produce a center temperature of 150°C? Repeat for an aluminum wire of the same diameter. ‘A hollow tube having an inside diameter of 2.5 cm and a wall thickness of 0.4 mm is exposed to an environment at h °C and 7. heat generation rate in the tube will produce a maximum tube temperature of 250°C for k = 24 Wim «°C? ‘Water flows on the inside ofa steel pipe with an ID of 2.5 cm. The wall thickness is 2 mm, and the convection coefficient on the inside is 500 Wim? - °C. The ‘convection coefficient on the outside is 12 W/m? - °C. Calculate the overall heat- transfer coefficient, What is the main determining factor for U? ‘The pipe in Prob. 2-51 is covered with a layer of asbestos [k = 0.18 Wim - °C] while still surrounded by a convection environment with = 12 Wim? - °C. Cal- culate the critical insulation radius. Will the heat transfer be increased or de- creased by adding an insulation thickness of (a) 0.5 mm, (6) 10 mm? Calculate the overall heat-transfer coefficient for Prob. 2-4. Calculate the overall heat-transfer coefficient for Prob. 2-5. ‘Ai at 120°C in a thin-wall stainless-steel tube with h = 65 Wim? - °C. The inside diameter of the tube is 2.5 em and the wall thickness is 0.4 mm. k = 18 W/m - °C for the steel, The tube is exposed to an environment with k= 6.5 Wim? - *C and T. = 13°C. Calculate the overall heat transfer coefficient and the heat loss per ‘meter of length. What thickness of an insulation having k = 40 mWim - °C should bbe added to reduce the heat loss by 90 percent? ‘An insulating glass window is constructed of two S-mm glass plates separated by ‘an air layer having a thickness of 4 mm. The air layer may be considered stagnant ‘0 that pure conduction is involved. The convection coefficients forthe inner and ‘outer surfaces are 12 and 50 Wim? «°C respectively. Calculate the overall heat transfer coefficient for this arrangement, and the R value. Repeat the calculation for a single glass plate S mm thick. ‘A wall consists ofa I-mm layer of copper, a 4-mm layer of 1 percent carbon288 261 262 Probieme 68 steel, a L-cm layer of asbestos sheet, and 10 em of iber-glass blanket. Calculate the overall heat-transfer coefficient for this arrangement. If the two outside sur- faces are at 10 and 150°C, calculate each of the interface temperatures. ‘A thin rod of length £ has its two ends connected to wo walls which are maintained ‘at temperatures 7, and T:, respectively. The rod loses heat to the environment at T, by convection. Derive an expression (a) for the temperature distribution in the rod and (b) for the total heat lost by the rod. A rod of length L has one end maintained at temperature Ty and is exposed 10 fan environment of temperature T... An electrical heating element is placed in the rod $0 that heat is generated uniformly along the length at a rate 4. Derive an expression (a) for the temperature distribution in the rod and (b) for the total heat transferred to the environment. Obtain an expression for the value of q which will make the heat transfer zero at the end which is maintained at Ty. One end of a copper rod 300 cm long is firmly connected to a wall which is ‘maintained at 200°C. The other end is firmly connected to a wall which is main- tained at 93°C, Air is blown across the rod so that a heat-transfer coefficient of 17 Wim? - °C is maintained, The diameter of the rod is 12.5 mm. The temperature of the air is 38°C. What is the net heat lost to the air in watts? Verify the temperature distribution for case 2 in Sec. 2-9 ie., that T= Te _ cosh iL ~ x) + (himk) sinh miL ~ x) ToT. cosh mb + (almk) sinh mL Subsequently show that the heat transfer is sinh mL. + (himk) cosh mL. = NRPKA (To ~ 7) cos ml + (himk) sinh mL. ‘An aluminum rod 2.5 em in diameter and 15 cm long protrudes from a wall which is maintained at 260°C. The rod is exposed to an environment at 16°C. The convection heat-transfer coefficient is 15 Wim? - C. Calculate the heat lost by the rod, Derive Eq. (2-35) by integrating the convection heat loss from the rod of case | in Sec. 29, Derive Eq. (2-36) by integrating the convection heat loss from the rod of case 3 in See. 2.9 ‘Along, thin copper rod 6.4 mm in diameter is egposed to an environment at 20°C. ‘The base temperature of the rod is 150°C. The heat-transfer coefficient between the rod and the environment is 24 Wim? - °C. Calculate the heat given up by the rod. ‘A very long copper rod [k = 372 Wim °C] 2.5 cm in diameter has one end maintained at 90°C. The rod is exposed to a fluid whose temperature is 40°C. The heat-transfer coefficient is 3.5 Wim? °C. How much heat is lost by the rod? ‘An aluminum fin 1.6 mm thick is placed om a circular tube with 2.5-em OD. The fin is 6.4 mm long. The tube wall is maintained at 150°C, the environment tem- perature is 15°C, and the convection heat-transfer coefficient is 23 Wim? C. Calculate the heat lost by the fin.(00 steady-state conduction—one dnension an an an 273 2m 278 ‘The total efficiency for a finned surface may be defined as the ratio of the total heat transfer of the combined area of the surface and fins to the heat which would bbe transferred if this total area were maintained at the base temperature Ty. Show that this efficiency can be calculated from fae ” fa-w where 9, = total efficiency ‘A;™ surface area of all fins A = total heat-ransfer area, including fins and exposed tube or otter surface ty = fin efficiency A triangular fin of stainless steel (18% Cr, 8% Ni) is attached to a plane wall maintained at 460°C. The fin thickness is 6.4 mm, and the length is 2.5 cm. The ‘environment is at 93°C, and the convection heat-transfer coefficient is 28 Wim? = °C. Calculate the heat lost from the fin, ‘A 2.5-cm-diameter tube has circumferential fins of rectangular profile spaced at 9.5-mm increments along its length. The fins are constructed of aluminum and ‘are 0.8 mm thick and 12.5 mm long. The tube wall temperature is maintained at 200°C, and the environment temperature is 93°C. The heat-transfer coefficient is 110 Wim? - °C. Calculate the heat loss from the tube per meter of length. ‘A circumferential fin of rectangular cross section surrounds a 2.S.cm-diameter tube. The length of the fin is 6.4 mm, and the thickness is 3.2 mm. The fin is constructed of mild steel. [fair blows over the fin so that a heat-transfer coefficient of 28 Wim? - °C is experienced and the temperatures of the base and air are 260 ‘and 93°C, respectively, calculate the heat transfer from the fin AA straight rectangular fin 2.0 cm thick and 14 cm long is constructed of steel and placed on the outside of a wall maintained at 200°C. The environment temperature is 15°C, and the heat-transfer coefficient for convection is 20 Wim? «°C. Calculate the heat lost from the fin per unit depth. ‘An aluminum fin 1.6 mm thick surrounds a tube 2.5 em in diameter. The length of the fin is 12.5 mm. The tube-wall temperature is 200°C, and the environment temperature is 20°C. The heat-transfer coefficient is 60 Wim? - °C. What is the hheat lost by the fin? Obtain an expression for the optimum thickness of a straight rectangular fin for a given profile area. Use the simplified insulated-tip solution. Derive a differential equation (do not solve) for the teinperature distribution in a straight triangular fin. For convenience take the coordinate axis as shown and ‘assume one-dimensional heat flow. TT276 am 270 200 292 283 2.04 285 288 Probleme 67 A long stainless-steel rod {k = 16 W/m » °C] has a square cross section 12.5 by 12,5 mm and has one end maintained at 250°C. The heat-transfer coefficient is 40, ‘Wim? - °C, and the environment temperature is 90°C. Calculate the heat lost by the rod, AA straight fin of rectangular profile is constructed of duralumin (94% Al, 3% Cu) with a thickness of 2.4 mm. The fin is 19 mm long, and it is subjected to a convection environment with = 85 Winn?» °C. Ifthe base temperature is 90°C ‘and the environment is at 25°C, calculate the heat transfer per unt length of fin A certain internal-combustion engine is air-cooled and has a cylinder constructed of cast iron [k = 35 Btu/h «ft °F]. The fins on the cylinder have a length of {in and thickness of bin, The convection coefficient is 12 Btwhh ft? °F. The cylinder diameter is 4 in, Calculate the heat loss per fin for a base temperature of 450°F and environment temperature of 100°F. A L.6-mm-diameter stainless-steel rod (k = 22 Wim - °C] protrudes from a wall ‘maintained at 49°C, The rod is 12.5 mm long, and the convection coefficient is, 570 Wim’ « °C. The environment temperature is 25°C. Calculate the temperature of the tip of the rod, Repeat the calculation for h = 200 and 1200 Wim? = °C. A 2eemediameter glass rod 6 em long [k= 0.8 Wim - °C] has a base temperature (of 100°C and is exposed to an air convection environment at 20°C. The temperature at the tip of the rod is measured as 35°C, What is the convection heat-transfer coefficient?” How much heat is lost by the rod? A straight rectangular fin has a length of 2.0 em and a thickness of 1.5 mm. The thermal conductivity is $5 Wim - °C. and it is exposed to a convection environment at 20°C and ft = 500 Wim? °C. Catculate the maximum possible heat loss for a base temperature of 200°C. What is the actual heat loss? A straight rectangular fin has a length of 3.5 em and a thickness of 1.4 mm. The thermal conductivity is $5 Wim - °C. The fin is exposed to a convection environ- ment at 20°C and h = 500 Wim? - °C. Calculate the maximum possible heat loss for a base temperature of 150°C. What is the actual heat loss for this base tem- perature? A circumferential fin of rectangular profile is constructed of 1 percent carbon Steel and attached to a circular tube maintained at 150°C, The diameter ofthe fin is 5 em, and the length is also S em with a thickness of 2 mm. The surrounding. air is maintained at 20°C and the convection heat-transfer coefficient may be taken as 100 Wim? «°C. Calculate the heat lost from the fn A circumferential fin of rectangular profile is constructed of aluminum and sur- rounds a 3-cm-diameter tube. The fin is 2 em long and I mm thick. The tube wall temperature is 200°C, and the fin is exposed to a fluid at 20°C with a convection hheat-transfer coetficient of 80 Wim? » °C, Calculate the heat loss from the fin. A 1.0-cm-diameter steel rod (k = 20 Wim? - "C) is 20 cm long. It has one end maintained at S0°C and the other at 100°C. It is exposed to a convection envi- Fonmept at 20°C with k= 85 Wim? - °C. Calculate the temperature at the center of the rod, AA straight rectangular fin of steel (1% C) is 2.6 em thick and 17 em long. It is placed on the outside of a wall which is maintained at 230°C, The surrounding290 291 202 299 204 298 296 ‘Stesdysate conduction one cimension air temperature is 25°C.and the convection heat-transfer coefficient is 23 Wim? «°C. Calculate the heat lost from the fin per unit depth and the fin efficiency. AA straight fin having a triangular profile has a length of S cm and a thickness of 4 mm and is constructed of a material having k = 23 Wim - °C. The fin is exposed to surroundings with a convection coefficient of 20 Wim? - °C and a temperature ‘of 40°C. The base of the fin is maintained at 200°C. Calculate the heat lost per unit depth of fin ‘A circumferential aluminum fin is installed on a l-in-diameter tube. The length of the fin is 0.5 in and the thickness is 1.0 mm. It is exposed to a convection environment at 30°C with a convection coefficient of $6 Wim? -°C. The base temperature is 125°C. Calculate the heat lost by the fin, A circumferential fin of rectangular profile is constructed of stainless steel (18% Cr, 8% Ni). The thickness of the fin is 2.0 mm, the inside radius is 2.0 em, and the length is 8.0 cm. The base temperature is maintained at 135°C and'the fin is ‘exposed to a convection environment at 15°C with h = 20 Wim? «°C. Calculate ‘the heat lost by the fin. AA rectangular fin has a length of 2.5 cm and thickness of 1.1 mm. The thermal ‘conductivity is $5 Wim « °C. The fin is exposed to a convection environment at 20°C and h = $00 Wim? - °C. Calculate the heat loss for a base temperature of 128°. ‘A L0-mm-thick aluminum fin surrounds a 2.5-cm-diameter tube, The length of the fin is 1.25 em. The fin is exposed to a convection environment at 30°C with he = 75 Wim? + °C. The tube surface is maintained at 100°C. Calculate the heat lost by the fin A glass rod having a diameter of | cm and length of S cm is exposed toa convection ‘environment at a temperature of 20°C. One end of the rod is maintained at a temperature of 180°C. Calculate the heat lost by the rod if the convection heat transfer coefficient is 15 Wim? » °C. ‘A stainless steel rod has a square cross-section measuring 1 by 1 em. The rod length is 8 cm, and k = 18 Wim? - °C. The base temperature of the rod is 300°C. ‘The rod is exposed to a convection environment at SO°C with = 45 Wim? - °C. Calculate the heat lost by the rod and the fin efficiency. Copper fins with a thickness of 1.0 mm are installed on a 2.5-cm-diameter tube. ‘The length of each fin is 12 mm. The tube temperature is 250°C and the fins are ‘exposed to air at 30°C with a convection heat-transfer coefficient of 120 Wim: + °C. Calculate the heat lost by each fin, A straight fin of rectangular profile is constructed of stainless steel (18% Cr, 89% Ni) and has a length of 5 cm and a thickness of 2.5 cm. The base temperature is ‘maintained at 100°C and the fin is exposed to a convection environment at 20°C with k= 47 Wim? «°C. Calculate the heat lost by the fin per meter of depth, and the fin efficiency. A circumferential fin of rectangular profile is constructed of Duralumin and sur- rounds a 3-cm-diameter tube. The fin is 3 cm long and I mm thick. The tube wall temperature is 200°C, and the fin is exposed to a fuid at 20°C with a convection heat-transfer coefficient of 80 Wim? - “C. Calculate the heat loss from the fin.Reterences 08 2.97 A circular fin of rectangular profile is attached to a 3.0-cm-diameter tube main- tained at 100°C. The outside diameter of the fin is 9.0 cm and the fin thickness is 1.0 mm. The environment has a convection coefficient of 50 Wim? - C and a temperature of 30°C. Calculate the thermal conductivity of the material for a fin efficiency of 60 percent. 2.98 A circumferential fin of rectangular profile having a thickness of 1.0 mm and a length of 2.0 em is placed on a 2.0-cm-diameter tube. The tube temperature is 150°C, the environment temperature is 20°C, and h = 200 Wim? - °C. The fin is aluminum, Calculate the heat lost by the fin. 299 Two I-in-diameter bars of stainless steel {k = 17 Wim -*C} are brought into end-to-end contact so that only 0.1 percent of the cross-sectional area is in ‘contact at the joint. The bars are 7.5 cm long and subjected to an axial temper- ature difference of 300°C. The roughness depth in each bar (L,/2) is estimated to be 1.3 jum, The surrounding fluid is air, whose thermal conductivity may be taken as 0.035 Wim - °C for this problem. Estimate the value of the contact resistance and the axial heat flow. What would the heat flow be for a continuous 15-cm stainless-steel bar? 2-100 When the joint pressure for two surfaces in contact is increased, the high spots of the surfaces are deformed so that the contact area A, is increased and the roughness depth L, is decreased. Discuss this effect in the light of the presen- tation of Sec. 2-11. (Experimental work shows that joint conductance varies ‘almost directly with pressure.) 24101 Two aluminum plates $ mm thick with a ground roughness of 100 uin are bolted together with a contact pressure of 20 atm. The overall temperature difference ‘across the plates is 80°C. Calculate the temperature drop across the contact joint. ‘24402 Fins are frequently installed on tubes by a press-fit process. Consider a circum- ferential aluminum fin having a thickness of 1.0 mm to be installed on a 2.5-em- diameter aluminum tube. The fin length is 1.25 cm, and the contact conductance may be taken from Table 2-2 for a 100-uin ground surface. The convection environment is at 20°C, and h = 125 Wim? «°C. Calculate the heat transfer for ‘each fin for a tube wall temperature of 200°C. What percentage reduction in heat transfer is caused by the contact conductance? 2403 An aluminum fin is attached to a transistor which generates heat at the rate of ‘300 mW. The fin has a total surface area of 9.0 em: and is exposed to surrounding. air at 27°C. The contact conductance between transistor and fin is 0.9 x 10-* m °C/W, and the contact area is 0.5 em?. Estimate the temperature of the transistor, assuming the fi is uniform in temperature. lS REFERENCES 4) Schneider, P. J.: “Conduction Heat Transfer,”” Addison-Wesley Publishing Com- pany, Inc., Reading, Mass., 1955, 2 Harper, W. B., and D. R. Brown: Mathematical Equations for Heat Conduction in the Fins of Air-cooled Engines, NACA Rep. 158, 1922. 3. Gardner, K. A.: Efficiency of Extended Surfaces, Trans. ASME, vol. 67, pp. 621-631, 1945,10 ‘Steagy sate consuction--one dimension Moore, C. J.; Heat Transfer across Surfaces in Contact: Studies of Transients in ‘One-dimensional Composite Systems, Southern Methodist Univ., ThermaliFluid Sci. Ctr, Res. Rep. 67-2, Dallas, Tex., March 1967 Ybarrondo, L. J., and J. E. Sunderland: Heat Transfer from Extended Surfaces. Bull. Mech. Eng. Educ., vol. 5, pp. 229-234, 1966. Moore, C. J., Jr, H. A. Blum, and H. Atkins: Subject Classification Bibliography for Thermal Contact Resistance Studies, ASME Pap. 68-WAJHT-18, December 1968. Clausing, A. M.: Transfer at the Interface of Dissimilar Metals: The Influence of ‘Thermal Strain, Int, J. Heat Mass Transfer, vol. 9. p. 791, 1966. Kem, D. Q., and A. D. Kraus: “Extended Surface Heat Transfer," McGraw-Hill Book Company, New York, 1972. Siege!, R., and J. R. Howell: “Thermal Radiation Heat Transfer,” 2d ed.. MeGraw- Hill Book Company, New York, 1980. Fried E.: Thermal Conduction Contribution to Heat Transfer at Contacts, “Thermal (R. P, Tye, ed.) vol. 2, Academic Press, Inc., New York, 1969.STEADY-STATE CONDUCTION— MULTIPLE DIMENSIONS 3-1. INTRODUCTION In Chap. 2 steady-state heat transfer was calculated in systems in which the temperature gradient and area could be expressed in terms of one space co- ordinate. We now wish to analyze the more general case of two-dimensional heat flow. For steady state, the Laplace equation applies, er, er ene Go) assuming constant thermal conductivity. The solution to this equation may be obtained by analytical, numerical, or graphical techniques. The objective of any heat-transfer analysis is usually to predict heat flow or the temperature which results from a certain heat flow. The solution to Eq. (3 1) will give the temperature in a two-dimensional body as a function of the two independent space coordinates x and y. Then the heat flow in the x and y directions may be calculated from the Fourier equations ar od 62) ar iy G3) ‘These heat-flow quantites are directed either in the x direction or in the y direction. The total heat flow at any point in the material is the resultant of the 4. and q, at that point. Thus the total heat-flow vector is directed so that it is Perpendicular to the lines of constant temperature in the material, as shown in Fig. 3-1. So if the temperature distribution in the material is known, we may easily establish the heat flow.172 Steaoy sale conductonulioe amensions Fig. 341 Sketch showing the heat flow in wo mensions 3.2. MATHEMATICAL ANALYSIS OF TWO-DIMENSIONAL HEAT CONDUCTION We first consider an analytical approach to a 4wo-dimensional problem and then indicate the numerical and graphical methods which may be used to ad- vantage in many other problems. It is worthwhile to mention here that analytical solutions are not always possible to obtain; indeed, in many instances they are very cumbersome and difficult to use. In these cases numerical techniques are frequently used to advantage. For a more extensive treatment of the analytical methods used in conduction problems, the reader may consult Refs. 1, 2, 12, and 13. Consider the rectangular plate shown in Fig. 3-2. Three sides of the plate are maintained at the constant temperature 7,, and the upper side has some temperature distribution impressed upon it. This distribution could be simply a constant temperature or something more complex. such as a sine-wave dis- tribution. We shall consider both cases, To solve Eq, (3-1), the separation-of-variables method is used. The essential point of this method is that the solution to the differential equation is assumed to take a product form n ae Fig.3-2 Igotnerms and heat tow »——— lines in a rectangular plateMatpeatical aneysis of two-dimensional haat conduction 72 T=XY where X= Xx) G4) Y= YO) ‘The boundary conditions are then applied to determine the form of the functions X and Y. The basic assumption as given by Eq. (3-4) can be justified only if it is possible to find a solution of this form which satisfies the boundary conditions. First consider the boundary conditions with a sine-wave temperature di tribution impressed on the upper edge of the plate. Thus T=Taty T=Tatx=0 65) T= Tar=W n(™ . rersin(Z) en aye where T,, is the amplitude of the sine function, Substituting Eq. (3-4) in (3-1) gives Lax _1eY "¥ae "Yar G6) Observe that each side of Eq. (3-6) is independent of the other because x and y are independent variables. This requires that each side be equal to some constant. We may thus obtain two ordinary differential equations in terms of this constant, £2. ex =o en ay foe we -MY=0 G8) where A? is called the separation constant. Its value must be determined from the boundary conditions. Note that the form of the solution to Eqs. (3-7) and 3-8) will depend on the sign of 4°; a different form would also result if A? were zero. The only way that the correct form can be determined is through an application of the boundary conditions of the problem. So we shall first write down all possible solutions and then see which one fits the problem under consideration, For 2 = X=C + Cx Y=Q+Cy GH T= (C+ CoG + Cy) ‘This function cannot fit the sine-function boundary condition, so that the A? = 0 solution may be excluded.“74 Sandy ate conducton—mutiple dmarsons For’ <0: X= Cye™ + Coe ¥ = Cycos Ay + Cy sin Ay G-10) T = (Cye™™ + Cee*V(C; cos Ay + Cy sin Ay) Again, the sine-function boundary condition cannot be satisfied, so this solution is excluded also. For? >0: X= Cycos Ax + Cyo sin Ax Y= Ce + Cue” Gt) T= (Coe Cro sin Ax\Cue“” + Cue”) Now, it is possible to satisfy the sjne-function boundary condition; so we shall attempt to satisfy the other conditions. The algebra is somewhat easier to handle ‘when the substitution o=T-7, is made. The differential equation and the solution then retain the same form in the new variable 0, and we need only transform the boundary conditions. Thus e=0 aty=0 =0 atx =0 B-12) 0 atx = W ™ =TysinZt ty =H Applying these conditions, we have 0 = (Cy cos Ax + Cio sin AKXCy + Cia) (a) 0 = C&Cye-” + Cpe () 0 = (Cy cos AW + Cw sin AWKCne™ + Cue”) (0) Tm sin TE = (Cy 08 Ax + Cw sin ANCye + Cue) — (A) ‘Accordingly, Cu = -Cn 0 co and from (c), 2 sin AW(e"” — e-™) This requires that sin AW = 0 6-13)atnematecal analyse of two-dimensional heal conducton 78 Recall that A was an undetermined separation constant. Several values will satisfy Eq. (3-13), and these may be written a-% 14) where n is an integer. The solution to the differential equation may thus be written as a sum of the solutions for each value of n. This is an infinite sum, so that the final solution is the infinite series o=T-T, Pa 1 sin 2 sinh G5) ‘where the constants have been combined and the exponential terms converted to the hyperbolic function. The final boundary condition may now be applied: nal Tp sin Wy Bo sin sinh — which requires that C, = 0 for n > 1, The final solution is therefore sinh (xy/W) (z sinh OW) sin (2) + 7, 3-16 a sinh (H/W) W, ay The temperature field for this problem is shown in Fig. 3-2. Note that the heat- flow lines are perpendicular to the isotherms. ‘We now consider the set of boundary conditions 7, aty=0 T atx=0 T=T atx=W T=7, aty=H Using the first three boundary conditions, we obtain the solution in the form of Eq. (3-15) amy Ti sinh 22 317 T 5 C. sin = sin SO G17) ‘Applying the fourth boundary condition gives na nm 1. Ty = Cy sin sinh “Ht (3-18) ‘This is a Fourier sine series, and the values of the C, may be determined by expanding the constant temperature difference T; ~ T; ina Fourier series over the interval 0 » Fig. 9. Skaicn snowing element used for curvilinear-square anal ysis of two-dienensionai Neat ow.‘he conducton shape tector 77 sketched to aid in this calculation. The isotherms and heat-flow lines form groupings of curvilinear figures like that shown in Fig. 3-3b. The heat flow across this curvilinear section is given by Fourier's law, assuming unit depth of material: ar a= kan G21) This heat flow will be the same through each section within this heat-flow lane, and the total heat flow will be the sum of the heat flows through all the lanes. If the sketch is drawn so that Ax = Ay, the heat flow is proportional to the AT across the element and, since this heat flow is constant, the AT across each element must be the same within the same heat-flow lane. Thus the AT across an element is given by ATroverat ar = Ss where N is the number of temperature increments between the inner and outer surfaces. Furthermore, the heat flow through each lane is the same since it is independent of the dimensions Ax and Ay when they are constructed equal. ‘Thus we write for the total heat transfer M M ad = Ti 22) = Movers = 5 MT2 ~ 1) 3-22) where M is the number of heat-flow lanes. So, to calculate the heat transfer, we need only construct these curvilinear-square plots and count the number of temperature increments and heat-flow lanes. Care must be taken to construct the plot so that Ax ~ Ay and the lines are perpendicular. The accuracy of this method is dependent entirely on the skill of the person sketching the curvilinear squares. Even a rude sketch, however, can frequently help to give fairly good estimates of the temperatures that will occur in a body; and these estimates may then be refined with numerical techniques discussed in Sec. 3-5. An electrical analogy may be employed to sketch the curvilinear squares, as discussed in Sec. 3-9. The graphical method presented here is mainly of historical interest to show the relation of heat-flow lanes and isotherms. It may not be expected to be used for the solution of many practical problems. 3-4 THE CONDUCTION SHAPE FACTOR, In a two-dimensional system where only two temperature limits are involved, we may define a conduction shape factor S such that g = KS BT overat 3-23) The values of $ have been worked out for several geometries and are sum- marized in Table 3-1. A very comprehensive summary of shape factors for aee (qt) 447 wid — c'd — Ses sees coon eas ar ae — ieee “ eon oe < MONOH — oy 1004 1 uoysuaunp-249 ¥ res aueig 2oepns wo. (@) a LG] 90" reuuoyrosr survey ‘wnpow ayuyurusas ‘ur pauing padidayayresed LRU RS 7e[n80E1291[OULIINOST enous ‘ wispawy yuyu err 41 suas ut panel spe oe Tet TEUTOuIOST Jo s9purAo feunaynosy nnee (Tan) a (My) 41 a =< A ae ao. nga aa 7 od oe aman at Soci peremeaiens cot eso ss —— cass ag — | a — —- 2 ore 47, =0 4m, = 0 4m. = 0 47, = 0 Fig. Ex. 39‘Accuracy consioratons 148 For node 7 we can use a resistance formulation and obtain UR = k _ Hal + Axi) Rave = SR Fg URs-1 = 2 ‘and we find 1000 + T, + 2Te - ST; = 0 ‘Similar resistors are obtained for node 10. VRies = k UR io = 2 = UR so that 2, + Ty + 2M ~ ST = 0 For node 1, Aay/6 + Ay) VR aad 2k is = MHS + BH ayy ay UR = 2k and the nodal equation becomes 31 + Te + Ty Th = 0 For node 11, Mayl6 + Ay) UR i107 = UR a Rin =k Max WR yy = SO = AB ‘and the nodal equation becomes 600 + 6Tn + 3% + Ta ~ 167 = 0 Similarly, the equation for node 12 is 30 + Oy + OT, + For node 1. day ana URyvay = WR = WS UR 3k = URis0116 Steady state concicton—mutiole dimensions and we obtain 1000 + 974 + Ty ~ 207, = 0 Similarly for node 14, 100 + 97, + 97, + Ty ~ 2074 Finally, from resistances already found, the nodal equation for node 3 is 200 + 97), + 27, - 137, = 0 We choose to solve the set of equations by the Gauss-Seidel iteration technique and thus write them in the form T, = 17). The solution was set up-on a computer with all initial values for the T,'s taken as zero. The results of the computations are showa in the following table umber of iterations Node 2 0 20 E 1 39,3066 232.6668 247.1479 247.7605 247.7875, 2 59,30662, 232.6668 247.1479 247.7605 247.7875 3 50.11073 139.5081 147.2352 147.5629 147.773, 4 50.11073 139.5081 147.2352 147.5629, 147.5773, 5 206.25 288.358 293.7838 294.0129 294.023 6 48.75 359.025 366.9878 367.3243 367.3391 7 29145 390.989 398.7243 399.0813 399.0657 8 102.9297 200.5608 208.4068 208.7384 208.753 9 121.2334 268.2423 275.7592 76.2462 276.2677 10 164.5493 302.3108 313.5007 313.974 313.9948 " 70.95459 156.9976 164.3947 164.7076 164.7215 2 73.8905 203.6437 214.5039 214.9634 214.9836 B 0.18905 115.2635 119.2079 119.3752 119.3826 ry 62.8298 129.8294 135.6246 135.8703, 135.8811 Note that these solutions for 7 = T; = 247.79°C and T, = T, = 147.58°C are some- What below the values of 250°C and 150°C obtained when only four nodes were em- ployed, but only modestly so. 2.9 ELECTRICAL ANALOGY FOR TWO-DIMENSIONAL CONDUCTION ‘Steady-state electric conduction in a homogeneous material of constant resis- tivity is analogous to steady-state heat conduction in body of simlarsvometnc shape. For two-dimensional electric conduction the Laplace equation applies: aE, ve ax? * ay where E is the electric potential. A very simple way of solving a two-dimensional heat-conduction problem is to construct an electrical analog and experimentallySummary 197 determine the geometric shape factors for use in Eq. (3-23). One way to ac- ‘complish this is to use a commercially available paper which is coated with a thin conductive film. This paper may be cut to an exact geometric model of the two-dimensional heat-conduction system. At the appropriate edges of the paper, good electrical conductors are attached to simulate the temperature boundary conditions on the problem. An electric-potential difference is then impressed on the model. It may be noted that the paper has a very high resis- tance in comparison with the conductors attached to the edges, so that a con- stant-potential condition can be maintained at the region of contact. ‘Once the electric potential is impressed on the paper. an ordinary voltmeter may be used to plot lines of constant electric potential. With these constant- potential lines available. the flux lines may be easily constructed since they are ‘orthogonal to the potential lines. These equipotential and flux lines have pre- cisely the same arrangement as the isotherms and heat-flux lines in the cor- responding heat-conduction problem. The shape factor is calculated immedi- ately using the method which was applied to the curvilinear squares. It may be noted that the conducting-sheet analogy is not applicable to prob- lems where heat generation is present; however, by addition of appropriate resistances, convection boundary conditions may be handled with little trouble. ‘Schneider (2] and Ozisik [12] discuss the conducting-sheet method, as well as other analogies for treating conduction heat-transfer problems, and Kayan [4. 5] gives a detailed discussion of the conducting-sheet method. 3-10 SUMMARY There is a myriad of analytical solutions for steady-state conduction heat- transfer problems available in the literature. In this day of computers most of these solutions are of small utility, despite their exercise in mathematical fa- cilities. This is not to say that we cannot use the results of past experience to anticipate answers to new problems. But, most of the time, the problem a person wants to solve can be attacked directly by numerical techniques, except when there is an easier way to do the job. As a summary, the following sug: gestions are offered: 1, When tackling a two- or three-dimensional heat-transfer problem first try to reduce it to a one-dimensional problem. An example is a cylinder with length much larger than its diameter. 2. If possible, select a simple shape factor model which may either exactly or approximately represent the physical situation. See comments under items 4and 5. 3. Seek some simple analytical solutions but, if solutions are too complicated, g0 directly to the numerical techniques. 4. In practical problems recognize that convection and radiation boundary con-198 Steady-state conduction mestine deensions ditions are subject to large uncertainties. This means that, in most practical situations, undue concern over accuracy of solution to numerical nodal equa- tions is unjustified. . In general, approach the solution in the direction of simple to complex, and make use of checkpoints along the way. REVIEW QUESTIONS ‘What is the main assumption in the separation-of-variables method for solving Laplace's equation? Define the conduction shape factor. ‘What is the basic procedure in setting up a numerical solution to a two-dimensional conduction problem? Once finite-dfference equations are obtained for a conduction problem, what meth- ‘ods are available to effect a solution? What are the advantages and disadvantages of each method, and when would each technique be applied? Investigate the computer routines that are available at your computer center for solution of conduction heat-transfer problems. a Beginning with the separation-of-variables solutions for 4° = 0 and A? < 0 (Eqs. G-9) and (3-10)], show that it is not possible to satisfy the boundary conditions for the constant temperature at y = H with either of these two forms of solution. That is, show that, in order to satisfy the boundary conditions T=T ay=0 T=T ar=0 T= atxeW Tet aty=H either a trivial or physically unreasonable solution results when either Eq. (3-9) or G-10) is used. 32 Write out the first four nonzero terms of the series solutions given in Eq. (3-20). What percentage error results from using only these four terms at y = Hf and x= WR? 33 A6.0-cm-diameter pipe whose surface temperature is maintained at 210°C passes through the center of a concrete slab 45 cm thick. The outer surface temperatures Of the slab are maintained at 15°C. Using the flux plot, estimate the heat loss from the pipe per unit length. 34 A heavy-wall tube of Monel, 2.5-cm ID and S-cm OD, is covered with a 2.5-cm layer of glass wool. The inside tube temperature is 300°C, and the temperature at the outside of the insulation is 40°C. How much heat is lost per foot of length? Take k = 11 Btwh- ft «°F for Monel.35 A symmetrical furnace wall has the dimensions shown, Using the fux plot, obtain the shape factor for this wall Fig. P3-5 38 A furnace of | by 2 by 3 ft inside dimensions is constructed of a material having fa thermal conductivity of 0.5 Btu ft - °F. The wall thickness is 6 in. The inner ‘and outer surface temperatures are 1000 and 200°F, respectively. Calculate the hheat loss through the furnace wall. 3-7 A cube 35 em on each external side is constructed of fireclay brick. The wall thickness is 5.0 cm. The inner surface temperature is SOO*C, and the outer surface temperature-is 80°C. Compute the heat flow in watts. 38 Two long cylinders 8.0 and 3.0 cm in diameter are completely surrounded by a medium with k = 1.4 Wim -°C. The distance between centers is 10 cm, and the cylinders are maintained at 200 and 35°C. Calculate the heat-transfer rate per unit length. 38 A L-mdiameter sphere maintained at 30°C is buried in the earth at a place where k= 1.7 Wim -°C. The depth to the centerline is 2.4 m, and the earth surface temperature is 0°C. Calculate the heat lost by the sphere. 340 A 20-cm-diameter sphere is totally enclosed by a large mass of glass wool. A heater inside the sphere maintains its outer surface temperature at 170°C while the temperature at the outer edge of the glass wool is 20°C. How much power ‘must be supplied to the heater to maintain equilibrium conditions? 311 A large spherical storage tank, 2 m in diameter. is buried in the earth ata location ‘where the thermal conductivity is 1.5 Wim - °C. The tank is used for the storage of an ice mixture at O°C, and the ambient temperature of the earth is 20°C. Caleulate the heat loss from the tank. 342. The solid shown in the figure has the upper surface, including the half-cylinder cutout, maintained at 100°C. Ata large depth in the solid the temperature is 300 K;4 = 1 Wim + °C. What is the heat transfer at the surface for the region where L = Wemand D = Wem?1120 Steady-state conduction-mutipe dimensions a3 a8, 316 a7 318 a9 321 In certain locales, power transmission is made by means of underground cables. In one example an 8.0-cm-diameter cable is buried at a depth of 1.3 m, and the resistance of the cable is 1.1 x 10-* f/m. The surface temperature of the ground is 25°C, and k = 1.2 Wim - °C for earth. Calculate the maximum allowable current if the outside temperature of the cable cannot exceed 110°C. ‘A copper sphere 4.0 cm in diameter is maintained at 70°C and submerged in a large earth region where k = 1.3 Wim » °C. The temperature at a large distance from the sphere is 12°C. Calculate the heat lost by the sphere, ‘Two long, eccentric cylinders having diameters of 15 and 4 cm respectively are maintained at 100 and 20°C and separated by a material with k = 3.0 Wim = °C. ‘The distance between centers is 4.5 cm. Calculate the heat transfer per unit length between the cylinders. Two pipes are buried in the earth and maintained at temperatures of 300 and 125°C. The diameters are 8 and 16 cm, and the distance between centers is 40 ‘em. Calculate the heat-transfer rate per unit length if the thermal conductivity of ‘earth at this location is 0.7 Wim «°C. ‘A hot sphere having a diameter of 1.5 m is maintained at 300°C and buried in a ‘material with k = 1.2 Wim -°C and outside surface temperature of 30°C. The depth of the centerline of the sphere is 3.75 m. Calculate the heat loss. ‘A scheme is devised to measure the thermal conductivity of soil by immersing @ long electrically heated rod in the ground in a vertical position. For design pur- poses, the rod is taken as 2.5 cm in diameter with a length of 1 m. To avoid improper alteration of the soil, the maximum surface temperature of the rod is 55°C while the soil temperature is 10°C. Assuming a soil conductivity of 1.7 ‘Wim - °C, what are the power requirements of the electric heater in watts? ‘Two pipes are buried in an insulating material having k = 0.8 Wim «°C. One pipe is 10 cm in diameter and carries a hot fuid at 300°C while the other pipe is 2.8 ‘cmin diameter and carries a cool fuid at 15°C. The pipes are parallel and separated by a distance of 12 em on centers. Calculate the heat-transfer rate between the sper meter of length. ‘At a certain location the thermal conductivity of the earth is 1.5 Wim °C. At this location an isothermal sphere having a temperature of S°C and a diameter of, 2.0 m is buried at a centerline depth of 5.0 m. The earth temperature is 25°C. Calculate the heat lost from the sphere. People are sometimes careless at universities and bury steam pipes in the earth without insulation. Consider a 4-in pipe carrying steam at 300°F buried at a depth of 9 in to centerline. The buried length is 100 yards. Assuming that the earth thermal conductivity is 1.2 Wim? - °C and the surface temperature is 60°F, esti= imate the heat lost from the pipe. ‘Two parallel pipes 5 cm and 10.cm in diameter are totally surrounded by loosely packed asbestos. The distance between centers for the pipes is 20 cm. One pipe carries steam at 110°C while the other carries chilled water at 3°C. Calculate the hheat lost by the hot pipe per unit length.a2 s § é s g ‘A long cylinder has its surface maintained at 135°C and is buried in a material ‘having a thermal conductivity of 15.5 Wim - °C. The diameter of the cylinder is 3.cm and the depth to its centerline is 5 em. The surface temperature of the material is 46°C. Calculate the heat lost by the cylinder per meter of length. ‘A 3-m-diameter sphere contains a mixture of ice and water at 0°C and is buried ina semi-infinite medium having a thermal conductivity of 0.2 W/m - °C. The top surface of the medium is isothermal at 30°C and the sphere centerline is ata depth ‘of 8.5 m. Calculate the heat lost by the sphere. ‘An electric heater in the form of a S0-by-100-cm plate is laid on top of a semi infinite insulating material having a thermal conductivity of 0.74 Wim - °C. The heater plate is maintained at a constant temperature of 120°C over all its surface, and the temperature of the insulating material a large distance from the heater is 1S. Calculate the heat conducted into the insulating material. ‘A small furnace has inside dimensions of 60 by 70 by 80 cm with a wall thickness ‘of S cm. Calculate the shape factor for this geometry A 15-cm-diameter steam pipe at 150°C is buried in the earth near a S-cm pipe carrying chilled water at 5°C. The distance between centers is 15 cm and the thermal conductivity of the earth at this location may be taken as 0.7 Wim «°C. Calculate the heat lost by the steam pipe per unit length. Derive an equation equivalent to Eq. (3-24) for an interior node in a three- dimensional heat-flow problem. Derive an equation equivalent to Eq. (3-24) for an interior node in a one-dimen- sional heat-low problem. Derive an equation equivalent to Eq. (3-25) for @ one-dimensional convection boundary condition. Considering the one-dimensional fin problems of Chap. 2, show that a nodal equation for nodes along the fin in the accompanying figure may be expressed as AP(Ax)? APA? 1. [2882 2] Mane = Tas + Tad 20 fray mt [mst Fig. P31 Show that the nodal equation corresponding to an insulated wall shown in the accompanying figure is Tats + Tan + 2ay-in ~ Tr,4122, Stoady-state conducion-matiple mensions aaa ge a7 a8 Fig. P3.32 For the insulated corner section shown, derive an expression for the nodal equa- tion of node (m,n) under steady-state conditions. Insulte, Fig. P33 Derive the equation in Table 3-2 Derive an expression for the equation of a boundary node subjected to a constant heat flux from the environment. Use the nomenclature of Fig. 3-7 Set up the nodal equations for a modification of Example 3-5 in which the left half of the wire is insulated and the right half is exposed to a convection envi- ronment with h = 200 Wim? «°C and T = 20°C. In a proposed solar-energy application, the solar flux is concentrated on a S-cm- OD stainless-steel tube [k = 16 Wim - °C] 2 m long. The energy flux on the tube surface is 20,000 Wim?, and the tube wall thickness is 2 mm. Boiling water flows inside the tube with a convection coefficient of 5000 Wim? « °C and a temperature ‘of 250°C. Both ends of the tube are mounted in an appropriate supporting bracket, ‘which maintains them at 100°C. For thermal-stress considerations the temperature gradient near the supports is important. Assuming a one-dimensional system, set ‘up a numerical solution to obtain the temperature gradient near the supports. ‘An aluminum rod 2.5 cm in diameter and 15 cm long protrudes from a wall ‘maintained at 300°C. The environment temperature is 38°C. The heat-transfer ‘coefficient is 17 Wim? « °C. Using a numerical technique in accordance with the result of Prob. 3-31, obtain values for the temperature along the rod. Subsequently obtain the heat flow from the wall at x = 0. Hint: The boundary condition at the end of the rod may be expressed by hax , APs 7. [SEs kA +i]- hax | APtasY [A+ Ser] - reo339 Problems 123 where m denotes the node at the tip of the fin. The heat flow at the base is Gene = GE Tus = Tod where T,, ment, isthe base temperature and T,,. is the temperature of the first incre- Repeat Prob. 3-38, using a linear variation of heat-transfer coefficient between base temperature and the tip of the fin. Assume h = 28 Wim? «°C at the base and h = 1 Wim? «°C at the tip. For the wall in Prob. 3-5 a material with k = 1.4 Wim - °C is used. The inner and outer wall temperatures are 650 and 150°C, respectively. Using a numerical technique, calculate the heat flow through the wall. Repeat Prob. 3-40, assuming that the outer wall is exposed to an environment at 38°C and that the convection heat-transfer coefficient is 17 Wim? - °C. Assume that the inner surface temperature is maintained at 650°C. Repeat Prob. 3-3, using the numerical technique. In the section illustrated, the surface 1-4-7 is insulated. The convection heat- transfer coefficient at surface 1-2-3 is 28 Wim? «°C. The thermal conductivity of the solid material is 5.2 Wim - °C. Using the numerical technique, compute the ‘temperatures at nodes 1, 2, 4, and 5. Insulated Fig. P43, ‘A glass plate 3 by 12 by 12 in [k = 0.7 Wim - °C} is oriented with the 12 by 12 face in a vertical position. One face loses heat by convection to the surroundings at 70°F. The other vertical face is placed in contact with a constant-temperature block at 400°F. The other four faces are insulated. The convection heat-transfer ‘coefficient varies approximately as hy = 0.22T, ~ T.)!e-™ — Brulh- f° where Ts and T. are in degrees Fahrenheit, Ty is the local surface temperature, and x is the vertical distance from the bottom of the plate, measured in feet. Determine the convection heat loss from the plate, using an appropriate numerical analysis.124 Steady-state condcton—mtiple drensons 3-45. Calculate the temperatures at points 1, 2, 3, and 4 using the numerical method. sore socre oore sore Fig. P45, 248 For the block shown, calculate the steady-state temperature distribution at ap- propriate nodal locations using the nimercal method. Employ a digital computer for the solution if possible, and take advaniage of library subroutines avalable ’ in the computer center, k= 2.0 Btuh ft. { sok Insulated oF | L2in Fig. P3486 347 Rework Prob. 3-43, using the Gauss-Seidel iteration method. vr 348° The composite strip in the accompanying figure is exposed to the convection environment at 300°C and h = 40 Wim - °C. The material properties are ky = 20 Wim «°C, ke = 1.2 Wim »"C, and ke = 0.5 Wim - °C. The strip is mounted on plate maintained at the constant temperature of SO*C. Calculate the heat transfer from the strip to plate per unit length of strip. Assume two-dimensional heat flow. ‘349 The fin shown has a base maintained at 300°C and is exposed to the convection ‘environment indicated. Calculate the steady-state temperatures of the nodes shown and the heat loss if k = 1.0 W/m -*C.3-80 Calculate the steady-state temperatures for nodes | to 16 in the figure. {om h=30wintrc Ta=10C Insulated Treated Tem Tam wimec Fig. P3-50 351 Calculate the steady-state temperatures for nodes | to 9 in the figure. 5 Wim-*C qi Tr 100% 3] r= io0re 3 Wimerc Fig. F3514198 Steady tte conducion-—mutiple mensions \ 352 Calculate the steady-state temperatures for nodes | to 6 in the figure. he 12Wim Ta81sC rese¢| oe iy resee x= dy =25em eS Wine Fig. P3-82 359 Calculate the temperatures for the nodes indicated in the accompanying figure. ‘The entire outer surface is exposed to the convection environment and the entire inner surface is at a constant temperature of 300°C. Properties for materials A and B are given in the figure. “c Fig. P35,384 355 3-56 387 Problems 187 ‘A rod having a diameter of 2 cm and a length of 10 em has one end maintained ‘at 200°C and is exposed toa convection environment at 25°C with h = 40 Wim? + °C. ‘The rod generates heat internally at the rate of 50 MWim* and the thermal con- ductivity is 35 Wim - °C. Calculate the temperatures of the nodes shown in the figure assuming one-dimensional heat flow T= 200°C Fig. P3-54 Calculate the steady-state temperatures of the nodes in the accompanying figure. The entire outer surface is exposed to the convection environment at 20°C and the entire inner surface is constant at SO0°C. Assume k = 20 Wim - °C. Fig. P3-65 A liner of stainless steel (k = 20 Wim - “C), having a thickness of 3 mm, is placed ‘on the inside surface of the solid in Problem 3-55. Assuming now that the inside surface of the stainless steel is at $00°C, calculate new values for the nodal temperatures in the low-conductivity material. Set up your nodes in the stainless steel as necessary Calculate the steady-state temperatures for the nodes indicated in the accom panying figure.128 Steady-state conducion—mutiple dmeneione hers wine moe af E05 on] 10°C 5 e roore’ Inaultes #40 Wim" Fig, P3-87 ‘The two-dimensional solid shown in the accompanying figure generates heat in- ternally atthe rate of 90 MWim®. Using the numerical method calculate the steady state nodal temperatures for k = 20 Wim - °C. = 100 Wim?" T= onc Inssated Tralee = 20Wime Fig. P3-58 359 The half-eylinder has k = 20 Wim - °C and is exposed to the convection envi ronment at 20°C. The lower surface is maintained at 300°C, Compute the tem- peratures for the nodes shown and the heat loss for steady state 106m he sowie rae Fig. P3-59Relernces 129 3-60 Atube has diameters of 4 mm and 5 mm and a thermal conductivity 20 W/m? «°C. Heat is generated uniformly in the tube at a rate of 500 MW/m? and the outside surface temperature is maintained at 100°C. The inside surface may be assumed to be insulated. Divide the tube wal into four nodes and calculate the temperature at each using the numerical method. Check with an analytical solution. 2461 Repeat Prob. 3-60 with the inside of the tube exposed to a convection condition with h = 40 Wim? - °C, Check with an analytical calculation. 3-62 Rework Prob. 3-50 with the surface absorbing a constant heat flux of 300 W/m? instead of the convection boundary condition. The bottom surface still remains at 200°C, Rework Prob, 3-53 with the inner surface absorbing a constant heat ux of 300 ‘Wim instead of being maintained at 2 constant temperature of 300°C. é é Rework Prob, 3-58 with the surface marked at a constant 100°C now absorbing ‘a constant heat flux of 500 Wim. Add nodes as necessary. REFERENCES 4 Carslaw, H. S., and J. C. Jaeger: “Conduction of Heat in Solids," 2d ed., Oxford University Press, Fair Lawn, N.J., 1959. 2. Schneider, P. J.: “Conduction Heat Transfer.” Addison-Wesley Publishing Com- pany, Inc., Reading, Mass., 1955. 3. Dusinberre, G. M.: “Heat Transfer Calculations by Finite Differences,” Interna tional Textbook Company, Scranton, Pa., 1961. 4. Kayan, C. F.: An Electrical Geometrical Analogue for Complex Heat Flow, Trans. ASME, vol. 67, p. 713, 1945. 5 Kayan, C. F.: “Heat Transfer Temperature Patterns of a Multicomponent Structure ‘by Comparative Methods,” Trans. ASME, vol. 71. p. 9, 1949. 6 Rudenberg, R.: Die Ausbreitung der Luft- und Erdfelder und Hochspannungslei- tungen, besonders bei Erd- und Kurzschlussen, Eleckirotech. Z., vol. 46. p. 1342, 1925. 7 Andrews, R. V.: Solving Conductive Heat Transfer Problems with Electrical- analogue Shape Factors, Chem. Eng. Prog.. vol. 51, n0. 2, p. 67, 1955. 8 Sunderland, J. E., and K. R. Johnson: Shape Factors for Heat Conduction through Bodies with Isothermal or Convective Boundary Conditions, Trans. ASHAE, vol 10, pp. 237-241, 1964 9 Richtmeyer, R. D.: “Difference Methods for Initial Value Problems,” Interscience Publishers, Inc., New York, 1957 40 Cranck, J., and P. Nicolson: A Practical Method for Numerical Evaluation of Solutions of P.D.E. of Heat Conduction Type, Proc. Camb, Phil. Soc., vol. 43, p. 50, 1947, 44 Barakat, H. Z.,and J, A. Clark: On the Solution of Diffusion Equation by Numerical Methods. J. Heat Transfer, p. 421, November 1966120 Steacyciat conducton—mutiple mensions 2 a “ 6 7 * ” Orisik, M. N.: “Boundary Value Problems of Heat Conduction," International Textbook Company, Scranton, Pa., 1968, Arpaci, V. S.: Conduction Heat Transfer,"” Addison-Wesley Publishing Company, Inc., Reading, Mass., 1966. ‘Ames, W. F.: "Nonlinear Partial Differential Equations in Engineering,” Academic Press, Inc., New York, 1965. Myers, R. F.: “Conduction Heat Transfer," McGraw-Hill Book Company, New York, 1972. ‘Adams, J. A., and D. F. Rogers: “Computer Aided Analysis in Heat Transfer.” McGraw-Hill Book Company, New York, 1973. Bowers, J. C., and S. R. Sedore: “SCEPTRE: A Computer Program for Circuit and Systems Analysis,” Prentice-Hall, Inc., Englewood Cliffs, N.J., 1971. Rohsenow, W. M.,and J. P. Hartnett (eds.): “Handbook of Heat Transfer,” McGraw- Hill Book Company, New York, 1973. Kem, D. Q., and A. D. Kraus: “Extended Surface Heat Transfer Book Company, New York, 1972. ‘Strong, P. F., and A. G. Emslie: “The Method of Zones for the Calculation of, ‘Temperature Distribution," NASA CR-56800, July 1963 Sepetoski, W. K.,C. H. Sox, and P. F. Strong: “Description of a Transient Thermal ‘Analysis Program for Use with the Method of Zones,” NASA CR-S6722, August 1963, ‘Schultz, H. D.: “Thermal Analyzer Computer Program for the Solution of General ‘Heat Transfer Problems,” NASA CR-65581, July 1965 Hahne, E., and U. Grigull: Formfaktor und Formweiderstand der stationaren mehr- ddimensionalen Warmeleitung, Int. J. Heat Mass Transfer, vol. 18, p. 751, 1975. * McGraw-HillUNSTEADY-STATE CONDUCTION @ 4.1 INTRODUCTION If a solid body is suddenly subjected to a change in environment, some time must elapse before an equilibrium temperature condition will prevail in the body. We refer to the equilibrium condition as the steady state and calculate the temperature distribution and heat transfer by methods described in Chaps. 2 and 3. In the transient heating or cooling process which takes place in the interim period before equilibrium is established, the analysis must be modified to take into account the change in internal energy of the body with time, and the boundary conditions must be adjusted to match the physical situation which is apparent in the unsteady-state heat-transfer problem. Unsteady-state heat- transfer analysis is obviously of significant practical interest because of the large number of heating and cooling processes which must be calculated in industrial applications To analyze a transient heat-transfer problem, we could proceed by solving the general heat-conduction equation by the separation-of-variables method, similar to the analytical treatment used for the two-dimensional steady-state problem discussed in Sec. 3-2. We give one illustration of this method of solution for a case of simple geometry and then refer the reader to the references for analysis of more complicated cases. Consider the infinite plate of thickness 2L shown in Fig. 4-1. Initially the plate is at a uniform temperature T,, and at time zero the surfaces are suddenly lowered to T = T,, The differential equation is er _ lar ee ale “Dp ‘The equation may be arranged in a more convenient form by introduction of the variable @ = T = T,. Then (42) 138122 nsaaay sae conduction Fig. 441 infinte plate subjected to sudden cooing of sur ae eee oe with the initial and boundary conditions O=G=T.-T atr=00sx52% (a) a=0 atx =0,7>0 (by o=0 atx = 2L,7>0 wo Assuming a product solution Ox, 1) = X(x)%(z) produces the two ordinary differential equations @X Ga thx=0 n= qt ant = 0 where A? is the separation constant. In order to satisfy the boundary conditions it is necessary that A? > 0 so that the form of the solution becomes @ = (Cycos Ax + Cy sin Axje“"or From boundary condition (6), C, = 0 for + > 0. Because C; cannot also be zero, we find from boundary condition (c) that sin 2LA = 0, or oT a n= 1,23. ‘The final series form of the solution is therefore on nents gin ME = Cyne sin or This equation may be recognized as a Fourier sine expansion with the constants C, determined from the initial condition (a) and the following equation: 1 pt, ame 4 Cap fl osin d= n= 1,35...Lumpedneat-capacty system 133 ‘The final series solution is therefore @_T-h% 451 " 4S Li twatier gin "™ p= aT en 2 3,5 (43) In Sec. 4-4, this solution will be presented in graphical form for calculation purposes. For now, our purpose has been to show how the unsteady-heat- conduction equation can be solved, for at least one case, with the separation- of-variables method. Further information on analytical methods in unsteady- state problems is given in the references 4-2 LUMPED-HEAT-CAPACITY SYSTEM We continue our discussion of transient heat conduction by analyzing systems which may be considered uniform in temperature. This type of analysis is called the lumped-heat-capacity method. Such systems are obviously idealized be- cause a temperature gradient must exist in a material if heat is to be conducted into or out of the material. In general, the smaller the physical size of the body, the more realistic the assumption of a uniform temperature throughout; in the limit a differential volume could be employed as in the derivation of the general heat-conduction equation. a hot steel ball were immersed in a cool pan of water, the lumped-heat- capacity method of analysis might be used if we could justify an assumption ‘of uniform ball temperature during the cooling process. Clearly, the temperature distribution in the ball would depend on the thermal conductivity of the ball material and the heat-transfer conditions from the surface of the ball to the surrounding fluid, i.e., the surface-convection heat-transfer coefficient. We should obtain a reasonably uniform temperature distribution in the ball if the resistance to heat transfer by conduction were small compared with the con- vection resistance at the surface, so that the major temperature gradient would ‘occur through the fluid layer at the surface. The lumped-heat-capacity analysis, then, is one which assumes that the internal resistance of the body is negligible in comparison with the external resistance. ‘The convection heat loss from the body is evidenced as a decrease in the internal energy of the body, as shown in Fig. 4-2. Thus ar ar where A is the surface area for convection and V is the volume. The initial condition is written q= nA(T ~ To) = -cp¥ 4) T=T atr=0 0 that the solution to Eq. (4-4) is T-T. = enter 3)4124, Unsteaay state conduction =n Tayo ana -Tay=-eov Fig. 42 Nomenciature for sin le-ump heat-capacty analysis © ‘The thermal network for the single-capacity system is shown in Fig. 4-2b. In this network we notice that the thermal capacity of the system is “charged” initially at the potential To by closing the switch S. Then, when the switch is ‘opened, the energy stored in the thermal capacitance is dissipated through the resistance I/hA. The analogy between this thermal system and an electric sys- tem is apparent, and we could easily construct an electric system which would behave exactly like the thermal system as long as we made the rat nA eV Roca equal to 1/R,C,, where R- and C, are the electric resistance and capacitance, respectively. In the thermal system we store energy, while in the electric system we store electric charge. The flow of energy in the thermal system is called heat, and the flow of charge is called electric current. The quantity cpVihA is called the time constant of the system because it has the dimensions of time. When Co = pe ev A it is noted that the temperature difference T ~ T. has a value of 36.8 percent of the initial difference Ty ~ T.. 0 Applicability of Lumped-Capacity Analysis We have already noted that the lumped-capacity type of analysis assumes a uniform temperature distribution throughout the solid body and that the as- sumption is equivalent to saying that the surface-convection resistance is large compared with the internal-conduction resistance. Such an analysis may be expected to yield reasonable estimates when the following condition is met: HVIA) k where k is the thermal conductivity of the solid. In sections which follow we examine those situations for which this condition does not apply. We shall see that the lumped-capacity analysis has a direct relationship to the numerical * 10 r RT 150 ~ 100 4350 — 100 1 = 5819s = 162h 14.3 TRANSIENT HEAT FLOW IN A SEMLINFINITE SOLID Consider the semi-infinite solid shown in Fig. 4-3 maintained at some initial temperature T,. The surface temperature is suddenly lowered and maintained ata temperature Ty, and we seek an expression for the temperature distribution in the solid as a function of time. This temperature distribution may subse quently be used to calculate heat flow at any x position in the solid as a function of time, For constant properties, the differential equation for the temperature distribution Tix, 2) is #T lat at 7 gar Ca ‘The boundary and initial conditions are Tex, 0) = 7, 10,7) =, forr>0 This is a problem which may be solved by the Laplace-transform technique The solution is given in Ref. 1 as (4-8) Fig. 4-9 Nomenclature for wansient neat tow inva. Semeininte solidTransit hal fom in a sembfinie sold 137 where the Gauss error function is defined as pee Rhea “s) It will be noted that in this definition » is a dummy variable and the integral is a function of its upper limit. When the definition of the error function is inserted in Eq. (4-8), the expression for the temperature distribution becomes Te) = To _ 2 (Vg Me LO tan (410) ‘The heat flow at any x position may be obtained from (411) (4-12) The surface heat flux is determined by evaluating the temperature gradient at x = 0 from Eq, (4-11). A plot of the temperature distribution for the semi- infinite solid is given in Fig. 4-4. Values of the error function are tabulated in Ref. 3, and an abbreviated tabulation is given in Appendix A. ° Fig. 4-4 Temperature distrbuton in the sem infinite sobs#30 Unsiacy state conaucton Constant Heat Flux on Semi-Infinite Solid For the same uniform initial temperature distribution, we could suddenly expose the surface to a constant surface heat flux qy/A. The initial and boundary conditions on Eq. (4-7) would then become 7.0) = T, we. 42 — A a leco for 0 The solution for this case is ayer in r-1,- 4 43) prawn 42 Alara block of stel A = 45 Wim °C, © 14. 10-‘ sis itly ata uniform temperature of 35°C. The surface is exposed to a heat flux (a) by suddenly raising the surface temperature to 250°C and (b) through a constant surface heat flux of 3.2 x 10° Wim. Calculate the temperature at a depth of 2.5 em after a time of 0.5 min for both these cases. Solution We can make use of the solutions for the semi-infinite solid given as Eqs. (4-8) and (4-13). For case a, 0.025 War” Qa x 10 9G0F ‘The error function is determined from Appendix A as erf 0.61 = 0.61164 We have T, = 35°C and Ty = 250°C, so the temperature at x = 2.5 em is determined from Eq, (48) as 1) = To + (I, ~ To) orf Tx.) , ~ Tete 250 + (35 — 250K0.61164) = 118.5°C For the constant-heat-fux case b, we make use of Eq. (4-13). Since qd/A is given as 3.2 x 10° Wim?, we can insert the numerical values to Q)B.2 «1011.4 & 10-9G0Vn)"? a5 + OG Nts = tomer Tx.) — 0.025)3.2 > 10% , 2 1 = 0.61164) IC x= 2Sem,7= 30sCenvecton boundary conditens 438 For the constant-heat-fux case the surface temperature after 30 $ would be evaluated with x = 0 in Eq, (4-13). Thus 8.2 x1 x 10-")30Va)"" Te = 0) =35 + % ONL x 10 VBOV AI gy soe 4 EXAMPLE 43 ‘A large slab of aluminum at a uniform temperature of 200°C suddenly has its surface temperature lowered to 70°C. What is the total heat removed from the slab per unit surface area when the temperature at a depth 4.0 cm has dropped to 120°C? Solution We first find the time required to attain the 120°C temperature and then integrate Eq (4-12) to find the total heat removed during this time interval. For aluminum, @ = 84x WO tmis k= 21S Wim °C (124 Buh f= °F] We also have 7. = 200°C Using Eg, (4-8) gives WC Tx) = 1200 120-70 x a7. og te ao = 70 ~ Taq = OM? From Fig. 4-4 or Appendix A. te = 03553 Ver (0.04) and a OBB 109 ‘The total heat removed at the surface is obtained by integrating Eq. (4-12) 8 [ef dr = 2kTo = Td y * 21.13 x 104m? (1861 Bruit?) yn yt Fea 109 44 CONVECTION BOUNDARY CONDITIONS In most practical situations the transient heat-conduction problem is connected with a convection boundary condition at the surface of the solid. Naturally. the boundary conditions for the differential equation must be modified to take into account this convection heat transfer at the surface, For the semi-infinite- solid problem above. this would be expressed by: Heat convected into surface = heat conducted into surface or AAT. ~ Theo wt] (aH140 Unsteady state conduction ‘The solution for this problem is rather involved, and is worked out in detail by Schneider [1]. The result is T1. he | ar ae 1a etx- [exw (E+ “s)| x [r-er(x+ BE)] cos daVar) ial temperature of solid vironment temperature where X 1. This solution is presented in graphical form in Fig. 4-5, Solutions have been worked out for other geometries. The most important spar Fig, 45 Temperature dstebuton in the somi-intnte sold with ‘convection boundary condtionConvection boundary condtens 141 cases are those dealing with (1) plates whose thickness is small in relation to the other dimensions, (2) cylinders where the diameter is small compared to the length, and (3) spheres. Results of analyses for these geometries have been presented in graphical form by Heisler [2], and nomenclature for the three cases is illustrated in Fig. 4-6. In all cases the convection environment temperature is designated as T. and the center temperature for x = 0 or r = Ois Tp. At time zero, each solid is assumed to have a uniform initial temperature 7, ‘Temperatures in the solids are given in Figs. 4-7 to 4-13 as functions of time and spatial position, In these charts we note the definitions eel O=Tea-T. or Tr ~ Te =~ T. = To - Ts Q Fig. 4-6 Nomenclature for ne-cimensional solids su enly subjected to convection envionment at. = see temperate ‘inte piate of ickness 21: (6) inne eyinder ot , fe (e) sphere of radius04 009 bv 2. 94 won Te SSonION 3 DIEM HU UE 20 aose oo one 081061 2 124 wos “7 snipes jo opus oyu Ue 19 BIBIOdWIO SHY BP BLS sono e or ¢ oo r00 100 1% woo v0 132 4 wos ®) sripes jo a10ydS # 40} meLaduse 212 6+ “Bld ose et ot so ore ort ont zo eo soFig. 4410 Temperature. atunction of center temperature nan mtinte plate of mickness 21. tom Ret. 2 os oa 07 tos oa 03 02 on ° Dor 002 0080102 05 10 23 5 10 % 50 100 Fig. 411 Temperature as a function of axis temperature in an infinite cinder of radius ro, fom Ret. 2.Fig.4-12 Temperature asa tunctionot center temperature ora sphere ote sis 025 Fig. 4-13 Center temperature for plates, cylinders, and spheres, for small valves of h, ac cording o Ret. 2. (s = L for plate, and s = re lor cyinder and sphere ) 140 on 07 06 os os oH 1 4 001002005 v1.02 05 19 235 19 20 ig Bi fads %, Wom Ret. 2 09 ox 07 06 os 03 o1 0020406081012 161.82023 o10 oo 006 20 003 00: oo 0.00 0.006 2.005 0.004 003 0.002 000 = 8 4 3Convection boundary conations $47 If a centerline temperature is desired, only one chart is required to obtain a value for 8 and then To. To determine an off-center temperature two charts are required to calculate the product For example, Figs. 4-7 and 4-10 would be employed to calculate an off-center temperature for an infinite plate, ‘The heat losses for the infinite plate, infinite cylinder, and sphere are given in Figs. 4-14 to 4-16, where Qy represents the initial internal energy content of the body in reference to the environment temperature Q, = eV UT, = T.) = pevV9, facie} In these figures Q is the actual heat lost by the body in time 7. Obviously, there are many other practical heating and cooling problems of interest, The solutions for a large number of cases are presented in graphical form by Schneider [7], and readers interested in such calculations will find this reference to be of great utility. Figure 4-13 gives the center temperatures of the three types of solids for small values of ft, or for conditions where the solids behave as a lumped capacity. In this figure the characteristic dimensions s is L for the plate and ry for the cylinder and sphere. Simplified charts are available in Ref. (17) wh ot Fig. 414 Dimensionless heat loss 0/Qp of an infinte plane of thickness 2L with tim, from Ret 6148 Unsieaoy state concucton bel ft ie 4 AY ni a Fig. 415 Dimensionless heat oss /Q, of an infinte cylinder of racus fo wih me, trom Ret 6 ‘The Blot and Fourier Numbers A quick inspection of Figs. 4-5 to 4-16 indicates that the dimensionless tem- erature profiles and heat flows may all be expressed in terms of two dimen- sionless parameters called the Biot and Fourier numbers: hs Biot number = Bi = ar br Fourier number = Fo = “7 = “7 ° pes os on o7 06 2 a 8 oa oa 0 01 0: oe Pe Pet coat Fig. 416 Dimensionless neat 108s 0/0, of a sphere of radius fo with time, Hom Rel. 6,CCorvacton boundary conditions 449 In these parameters s designates some characteristic dimensions of the body; for the plate it is the half-thickness, whereas for the cylinders and sphere it is the radius. The Biot number compares the relative magnitudes of surface- convection and internal-conduction resistances to heat transfer. The Fourier modulus compares a characteristic body dimension with an approximate tem- perature-wave penetration depth for a given time 7. ‘A very low value of the Biot modulus means that internal-conduction resis- tance is negligible in comparison with surface-convection resistance. This in turn implies that the temperature will be nearly uniform throughout the solid, and its behavior may be approximated by the lumped-capacity method of anal- ysis. It is interesting to note that the exponent of Eq. (4-5) may be expressed in terms of the Biot and Fourier numbers if one takes the ratio V/A as the characteristic dimension s. Then. hs kr = Dag 7 BFR Applicability of the Heisler Charts ‘The calculations for the Heisler charts were performed by truncating the infinite series solutions for the problems into a few terms. This restricts the applicability of the charts to values of the Fourier number greater than 0.2. Fo = 7 >02 For smaller values of this parameter the reader should consult the solutions and charts given in the references at the end of the chapter. Calculations using the truncated series solutions directly are discussed in Appendix C EXAMPLE 4-4 The slab of Example 4-3 is suddenly exposed to a convection-surface environment of J0°C with a heat-transfer coefficient of 525 Wim? - °C. Calculate the time required for the temperature to reach 120°C at the depth of 4.0 cm for this circumstance. ‘Soliton We may use either Eq. (415) or Fig, 45 for solution of this problem, but Fig, 4-5 is easier to apply because the time appears in two terms. Even when the figure is used, fan iterative procedure is required because the time appears in both of the variables Jy Var and xia. We seek the value of 7 such that ean T,~ 7 ~ 200 ois a)180 Unseady-stte conducten We therefore try values of r and obtain readings of the temperature ratio from Fig. 4-5 until agreement with Eq. (a) is reached. The iterations are listed below. Values of and a are obtained from Example 4-3, War 7% - rom Fig. 4 ns bar 77m ie 5 i000 0.708 0.069 Omi 3012280 ot 4000146 0.035 0.68 Consequently, the time required is approximately 3000 s. EXAMPLE 45 A large plate of aluminum 5.0 cm thick and initially at 200°C is suddenly exposed to the convection environment of Example 4-4. Calculate the temperature at a depth of 1.25 cm from one of the faces 1 min after the plate has been exposed to the environment, How much energy has been removed per unit area from the plate in this time? ler charts of Figs. 4-7 and 4-10 may be used for solution of this problem. We first calculate the center temperature of the plate, using Fig. 4-7, and then use Fig. 4-10 to calculate the temperature at the specified x position, From the conditions of the problem we have 0, = T, - Ts = 200 ~ 10 =.130 a = 8.4 x 10-$ m/s [3.26 fh) = SOom L=2Sem += 1 min = 6s k= 215 Wim -*C_ [124 Bluth ft °F] fa = S25 Wie? °C. (92.5 Beuth - £2 °F] 2525-125 = 1.258em ar _ (84 x 10-160) kas ene ae -1 The vc (0.025)? aoe AL (5254 0.025) a x _ as £~ 1B sos From Fig. 47 % = Ty ~ Te = (0.619030) = 793Convection boundary condtions From Fig. 4-10 at z/L = 0.5, o £ = 038 % @ = T~ Te = (0.98\79.3) T2717 +70 = 4770 We compute the energy lost by the slab by using Fig. 4-14. For this calculation we require the following properties of aluminum: p= 2700 kg/m’ 0.9 kuikg For Fig. 4-14 we need Barr _ (S25)(8.4 x 10-5460) Ad _ (52530.025) B any Ce ee From Fig. 414 Q 2-041 For unit area Oe wees Oe AYE pe(2tse (2700,900)0.05)(130) 15.8 x 108 Yim? so that the heat removed per unit surface area is 2 = as x v0.44 A 6.48 x 10° Jim? (571 Bruit) EXAMPLE 46 ‘A long aluminum eylinder 5.0 em in diameter and initially at 200°C is suddenly exposed toa convection environment at 70°C and h = 525 Wim? - °C. Calculate the temperature at a radius of 1.25 em and the heat lost per unit length I min after the cylinder is exposed to the environment. Solution ‘This problem is like Example 4-5 except that Figs. 48 and 4-11 are employed for the solution, We have Te = 20-10 = 130 @ = 8.4 « 10S ms ro = 250m min = 60s k= US Win "Ck p= 700 kgim’ c= 09 KKE *C Ss Wim! °C r= 1.28 em182 Unctoasy-sate conduction t We compute ar _ (84 x 10-5160) k 2is S064 iim” Ganon ~ 1638 & & ~ 0.38 and from Fig. 4-11 at ri = 0.5 a qo a 4 sothat $= %2 ~ @38y0.98) = 0.372 that Bn oa, 7 ONO) = 03 and = 1 To = O21) = 44 T= 10 +484 = 1188C To compute the heat lost, we determine ’ (525n0.025), War _ (S2548.4 x 10 *)60) 0.03 0.061 a 2 «on Q, Forni ee and the actual heat lost per unit length is (6.203 x 10°"0.681 = 4.032 x 10° Sm. (116.5 Bruit) 1 4.5 MULTIDIMENSIONAL SYSTEMS, ‘The Heisler charts discussed above may be used to obtain the temperature distribution in the infinite plate of thickness 2L., in the long cylinder. or in the sphere. When a wall whose height and depth dimensions are not large compared i with the thickness or a cylinder whose length is not large compared with its diameter is encountered. additional space coordinates are necessary to specify the temperature. the above charts no longer apply, and we are forced to seek Fl another method of solution. Fortunately. it is possible to combine the solutions \Mutomensonal systems 189 for the one-dimensional systems in a very straightforward way to obtain so- lutions for the multidimensional problems. Iti clear that the infinite rectangular bar in Fig. 4-17 can be formed from ‘two infinite plates of thickness 21, and 2L, respectively. The differential equa- tion governing this situation would be eT eT _ lat at oe ut aa ay and to use the separation-of-variables method to effect a solution. we should assume a product solution of the form Thx, 2. 7) = XOOZ@VOW It can be shown that the dimensionless temperature distribution may be ex- pressed as a product of the solutions for two plate problems of thickness 22, and 223, respectively FB) Baca Bare where 7, is the initial temperature of the bar and T. is the environment tem- perature. For two infinite plates the respective differential equations would be e (4.19) and the product solutions assumed would be T=Ma) h= he (4-20) ‘We shall now show that the product solution to Eq. (4-17) can be formed from a simple product of the functions (Ty, 7), that is, Tx, 2.7) = Te, OTe. 21) Fig. 417 Infinte rectangular bar184 Unsteady sate concucton The appropriate derivatives for substitution in Eq, (4-23) are obtained from Eq. (4-21) as ar Using Eqs. (4-19), we have ar a ‘or the assumed product solution of Eq. (4-21) does indeed satisfy the original differential equation (4-17). This means that the dimensionless temperature distribution for the infinite rectangular bar may be expressed as a product of the solutions for two plate problems of thickness 2, and 2L:. respectively, as indicated by Eq. (4-18). In a manner similar to that described above, the solution for a three-dimen- sional block may be expressed as a product of three infinite-plate solutions for plates having the thickness of the three sides of the block. Similarly, a solution for a cylinder of finite length could be expressed as a product of solutions of the infinite cylinder and an infinite plate having a thickness equal to the length of the cylinder. Combinations could also be made with the infinite-cylinder and infinite-plate solutions to obtain temperature distributions in semi-infinite bars and cylinders. Some of the combinations are summarized in Fig. 4-18, where C(®) = solution for infinite cylinder P(X) = solution for infinite plate ‘S(X) = solution for semi-infinite solid The general idea is then eas” Gn (8) nape (8) ge Heat Transfer in Multidimensional Systems Langston {16] has shown that it is possible to superimpose the heat-loss so- lutions for one-dimensional bodies, as shown in Figs. 4-14, 4-15, and 4-16 toMutamensional systems 488 Pus Para, qT a \ Product solutions for temperatures ‘in mukigimensional systems: (a) som-intite plate; (0) infinte rectangular bar, (¢) semint- nite rectangular bar; (d) rectangular paral lepiped (e) semrinfinte cylinder: (9 short cy © ” inden obtain the heat for a multidimensional body. The results of this analysis for intersection of two bodies is ou @). @.D-@.] « where the subscripts refer to the two intersecting bodies. For a multi sional body formed by intersection of three one-dimensional systems, the heat loss is given by (Gan (8),* GL @.] “@b-@lb-@) >186. Urceady state conduction Ifthe heat loss is desired after a given time, the calculation is straightforward. On the other hand. if the time to achieve a certain heat loss is the desired quantity, a trial-and-error or iterative procedure must be employed. The fol- lowing examples illustrate the use of the various charts for calculating tem- peratures and heat flows in multidimensional systems. EXAMPLE 47 A semi-infinite aluminum cylinder 5 em in diameter is initially at a uniform temperature of 200°C. It is suddenly subjected to a convection boundary condition at 70°C with /h = 525 Wim? - °C. Calculate the temperatures at the axis and surface of the cylinder 10 em from the end | min after exposure to the environment. ‘Solution ‘This problem requires a combination of solutions for the infinite cylinder and semi- infinite slab in accordance with Fig. 4-18, For the slab we have = Wem a= 84% 10% mis k= 215 Wim-°C so thatthe parameters for use with Fig, 4-5 are Var _ (5258.4 x 10-5)60))7 k 215 ae A ou Var ~ Bites x 10H = 97% From Fig. 45 (2) oe 1 006 = 0964 = su For the infinite cylinder we seek both the axis- and surface-temperature ratios. The parameters for use with Fig. 48 are sem A = 16.38 ne z = 8064 This the axistemperature ratio. To find the surface-temperature ratio, we enter Fig. 411, using : ° Lato fn0 ro ho be ” Thus @ : ar=0 ao (3)... f re Combining the solutions forthe semi-infinite slab and infinite cylinder, we have 8 (3) at an ” SOSH = (0.38)0.964) = 0366 at = 0 = (0360¥0.964) = 0386 atr=Maismensions systema 187 ‘The corresponding temperatures are T = 70 + (0.366)(200 - 70) T = 10 + (0.356)(200 ~ 70) = 176 atr=0 163 atr= ry EXAMPLE 48 ‘A short aluminum cylinder 5.0 cm in diameter and 10.0 cm long is initially at a uniform ‘temperature of 200°C. It is suddenly subjected to a convection environment at 70°C, and h = 525 Wim? : °C. Calculate the temperature at a radial position of 1.25 em and a distance of 0.625 em from one end of the cylinder 1 min after exposure to the envi- ronment. Solution To solve this problem we combine the solutions from the Heisler charts for an infinite cylinder and an infinite plate in accordance with the combination shown in Fig. 4-18f. For the infnite-plate problem L= Sem ‘The x position is measured from the center of the plate so that x= 50.625 4375 om For aluminum = 84K 10 mls k= 21S Wim °C ks ar _ BA x 10-7460) a AL ~ G25)0.05) ~ Gu ve 0.05)? oi From Figs. 47 and 410, respectively, 2075 2-095 a 6 sot (2) = e099 = ons 1 ee For the eylinder ry = 2.5 em 1as kas "E575 ig” Gayo. 7 8 ar _ (4 x 10-7460) _ re omg 8.084 and from Figs. 48 and 4-11, respectively, f Foe @ so that (§),, = ©381098 = 3724188 Unsieacysiate conduction Combining the solutions for the plate and cylinder gives (2) ou," 0712103729 = 025 Thus T= Te + (0.265, ~ T.) = 10 + (0.265)(200 - 70) = 104.5°C = EXAMPLE 49 Calculate the heat loss for the short cylinder in Example 4-8. Soluton We first calculate the dimensionless heat-loss ratio for the i cylinder which make up the multidimensional body. For the plate we have em = 0.05 m, Using the properties of aluminum from Example 4-8, we calculate AL _ (325X0.05) _ nite plate and infinite =5 Me POD - oun War _ (525) B.4 x 10-1@) eo eusy From Fig. 4-14, for the plate, we read @) hry _ (525)/0.025) eS @)-» ‘The two heat ratios may be inserted in Eq. (4-22) to give 22 = 0.061 (8), 022+ @sox - 929 = 068 ‘The specific heat of aluminum is 0.896 ki/kg - °C and the density is 2707 kgm’, so we calculate Qs as 21 = pee, = (2707)(0.896)11(0.025)10.1)7200 ~ 70) 61.910 The actual heat loss inthe I-min time is thus Q* (61.9 kI}(0.649) = 40.2 KI 46 TRANSIENT NUMERICAL METHOD The charts described above are very useful for calculating temperatures in + certain regular-shaped solids under transient peat-flow conditions. Unfortu-Transent camera method: $88 nately, many geometric shapes of practical interest do not fall into these cat- egories; in addition, one is frequently faced with problems in which the bound- ary conditions vary with time. These transient boundary conditions as well as the geometric shape of the body can be such that a mathematical solution is not possible. In these cases, the problems are best handled by a numerical technique with computers. It is the setup for such calculations which we now describe. For ease in discussion we limit the analysis to two-dimensional sys- tems, An extension to three dimensions can then be made very easily. Consider a two-dimensional body divided into increments as shown in Fig. 4-19. The subscript m denotes the x position. and the subscript n denotes the y position. Within the solid body the differential equation which governs the heat flow is eT a (+ SF) = (Gare) = assuming constant properties. We recall from Chap. 3 that the second partial derivatives may be approximated by (4-24) ert FT stn + Tn tn ~ (428) ae” OF , #1 FE He mast + Tma-t ~ Tn (4-26 a @s : ” The time derivative in Eq. (4-24) is approximated by aT _ hid ~ The a ar (4.27) — Fig. 419 Nomenclatue for numerical solu Paral ton of two-dimensional unsteady-state con suction probier100, Unsioadystatecondvcton In this relation the superscripts designate the time. increment. Combining the relations above gives the difference equation equivalent to Eq. (4-24) 2Than , Thanos + Th 27, 3+ Th. 17 wy a Ar (4-28) Thus, if the temperatures of the various nodes are known at any particular time, the temperatures after a time increment Ar may be calculated by writing ‘an equation like Eq. (4-28) for each node and obtaining the values of T%!. The procedure may be repeated to obtain the distribution after any desired number of time increments. If the increments of space coordinates are chosen such that Ax = Ay the resulting equation for 7%,,' becomes wear : _ da dr Test = FOE hein + Theta + Thanet + That) + [1 = this equation becomes identical with Eq. (3-32), the formulation we employed for the iterative solution in Chap. 3 The stability requirement in the explicit formulation may be examined by solving Eq. (4-41) for 72"! wes(eeBREs(-EEg) ‘The value of g, can influence the stability, but we can choose a safe limit by ‘observing the behavior of the equation for g, = 0. Using the same type of thermodynamic argument as with Eq. (4-31), we find that the coefficient of T? cannot be negative. Our minimum stability requirement is therefore ar 1 c = 20 (4-45) Suppose we have a complicated numerical problem to solve with a variety of boundary conditions, perhaps nonuniform values of the space increments, ete. Once we have all the nodal resistanzes and capacities formulated, we then have the task of choosing the time increment Ato use for the calculation, To ensure stability we must Keep Az equal to or less than a value obtained from the most restrictive nodal relation like Eq. (4-45). Solving for Ar gives o_ i ars [sam ne for stability (4-46) While Eq. (4-44) is very useful in establishing the maximum allowable time increment, it may involve problems of round-off errors in computer solutions when small thermal resistances are employed. The difficulty may be alleviated by expressing 7?*? in the following form for calculation purposes: Ar 1-1 124 {a+ 3B i"]i rn 48 Rae [« z Fe | TF (4-47)$70 Unsieagysateconaucton We should remark that the resistance-capacity formulation is easily adapted to take into account thermal-property variations with temperature. One need only calculate the proper values of p, c, and k for inclusion in the C, and R,, Depending on the nature of the problem and accuracy required. it may be necessary to calculate new values of C, and R,, for each iteration. Example 4-16 illustrates the effects of variable conductivity EXAMPLE 4-10 Sudden cooling of a 103 AA steel rod {k = 50 Wim + °C] 3 mm in diameter and 10 cm long is initially at uniform temperature of 200°C. At time zero it is suddenly immersed in a fluid having h = 50 Wim! «°C and T. = 40°C while one end is maintained at 200°C. Determine the tem: perature distribution in the rod after 100 s. The properties of steel are p = 7800 k/m and c = 0.47 kilkg »°C. ‘Solution ‘The selection of increments on the rod is as shown in the figure, The cross-sectional area of the rod is A = 1.5)" = 7.069 mm. The volume element for nodes 1, 2. and Bis av=Aar (7.069425) = 176.725 mm” Node 4 has @ AV of half this value, or 88.36 mm’, We can now tabulate the various resistances and capacities for use in an explicit formulation. For nodes 1, 2. and 3 we have 0.025 Be arenes 7 BTW 1 1 © hed Bx) ~ (S0ym3 x 10-*40.025) C= pe AV = (7800)470\1,7673 x 10-7) = 0.6479 IFC and Re = 84.880 For node 4 we have 1 ax ay = & = wnrer Rae FCW Ra. = EE = TO.TSNCIW av . 2 7 c= PM = ommoure R= i = arc Fig. Ex. 4-10To determine the stability requirement we form the following table Node SUR.) ' 0.04006 2 0.04006 3 4 Thus node 4 is the most restrictive, and we mus! select A> 15.9. Since we wish: to find the temperature distribution at 100 s. Tet us use Ar = 10-5 and make the calculation for 10 time increments using Eq. (4-47) for the computation. We note St course. that g, = 0 because there is no heat generation. The calculations are shown in the following table. Time — nee 7 increment th qT qr o 200 200 200 ' 17087 17087 170.87 169.19 2 153.40 147.08 146.68 145.05 3 141S4 12886 126.98 125.58 4 133.06 11S.08 1.24 109.70 5 126.79 1448 98-76 96.96 6 122.10 96.36 88.92 86.78 7 11853 9009 81.78.71 8 11880 85.23 75.08 72.34 9 sas 7031 67.31 0 TS 66.57 63.37 We can calculate the heatransfer rate at the end of 100 s by summing the convection heat losses on the surface of the rod. Thus eat a= B25 and 200 = 40. , 112.08 + 78.51 + 66.57 = G40) 4 ~ (64-883) 84.883 feet « (Gaia * ig) 27 ~ 40 = 20172 Unseaay-stae consueton ’ EXAMPLE 4-11 We can iustrate the calculation scheme for the implicit formulation by reworking Example 4-10 using only two time increments, that is, A7 = 50's For this problem we employ the formulation indicated by Eq. (4-43), with & 50s. The following quantities are needed, Ne GG ar Rtas ; — { 1 0.01296 0.05302 2 oor oassa2 3 oa136 = os3a2 4 oasis o.02686 ‘We have already determined the Ry in Example 4-10 and thus can insert them into Eq. (4-43) to write the nodal equations for the end of the frst time increment, taking all Tr = 200°C. We use the prime to designate temperatures at the end of the time | increment. For node 1, assoary = 2H + EE + 2 + ons6xm) ' For node 2, 4 7 qT T co + @01 oasoer; = Ei + a + gM + eosizsoxa00 For nodes and 4, ; so, , 0 0.083027; = Fo7m * Tara + Geass + (001296200) Ts 40. 40 ‘ ooneser, = Ee + M+ onsnva09 ‘These equations can then be reduced to ; { 0.053027; ~ 0.014147; = 5.8911 he ~0014147; + 0.083007; - 0.014147; = 3.0635 = 00141473 + 0.083027; — 0.014147; = 3.0635, ~ 0.014147; + 0.026867; = 1.5457 ‘which have the solution Ty, = 145.81" 130.12 Ty" 12943C 7; = 123.56°C y ‘We can now apply the backward-difference formulation a second time using the double prime to designate the temperatures at the end of the second time increment:Tarai esstance and capaci frmeaton 173 20 | | 0.053007; = sa; + soar * aags + (0.012901045.81) oossorry = Py SB 4 4 + 0.01296K130.12) * = jo.731 * 70.7 * #4855 qt 0.083027 Foam + gees + (0:012960125.43) oonsser; = Bo + ee + (0.00648)(123.56) ao.731 * 2a29 * 169.7 and this equation set has the solution T= 13sec T= 972°C Ty = 8832C Ty = 85.59°C We find this calculation in substantial disagreement with the results of Example 4-10. With a larger number of time increments better agreement would be achieved. In a problem involving a large number of nodes, the implicit formulation might involve less Computer time than the explicit method, and the purpose of this example has been to show how the calculation is performed. EXAMPLE 4-12 Cooling of a coramic Alby 2cm ceramic strip {k = 3.0 Wim + °C]is embedded ina high-thermal-conductivity ‘material, as shown, so that the sides are maintained at a constant temperature of 300°C. The bottom surface of the ceramic is insulated, and the top surface is exposed to a convection environment with h = 200 Wim? °C and T. = SOC, At time zero the ceramic is uniform in temperature at 300°C. Calculate the temperatures at nodes 1 to 9 after a time of 12 s. For the ceramic p = 1600 kpim' and c= 0.8 ki/kg “°C. Also calculate the total heat loss in this time. 2m—+ ATe2src | Inalated Fig. Ex. 4-12 Solution We treat this as a two-dimensional problem with Ax = Ay = 0.5 em. From symmetry T= Ty Te Ta, $0 we have six unknown nodal temperatures. We now tabulate the various nodal resistances and capacities. For nodes 4 and S ax 00s kA ~ G0x0.005) Rs ® Ru = Ruy = Rye = 3u3176 yrmeaay sate conduction For nodes 1 and 2 ae ~ kA =k, R, = 0.3337C1W , L ~ hee ~ (200)00.005) ~ R, Lorcw ' For nodes 7 and 8 Ra © Ru = O.6667°C/W Re, = 0.3333°C/W For nodes 1, 2, 7. and 8 the capacities are 1600(80010.005)? , = 169°C J For nodes 4 and 5 The stability requirement for an explicit solution is now determined by tabulating the following quantities: 1 com =e, Bair, 1 7 16 2.286 2 7 16 2.286 4 2 2 2.667 5 2 32 2.667 7 6 16 2.667 i 8 6 16 2.667 . ‘Thus the two convection nodes control the stability requirement, and we must choose ‘br = 2.286 s. Let us choose Ar = 2.0.s and make the calculations for six time increments with Eq. (4-47). We note once again the symmetry considerations when calculating the temperatures of nodes 2, 5, and 8, that is, 7, = Ts, etc. The calcu- lations are shown in the following table. Node temperature h r, t m tr 300 300 300 300 300 ' 268.75 300 300 300 300 253.13 294.14 294.14 300 300 24531 289.75 287.55 297.80 297.80 284.73 239.48 «285.81 282.38 295.19 293.96 4 3 246.67 235.35 8263-27779 29234 290.08 6 243.32 231.97 279.87 273.95 289.71 286.32‘Thermal resstance ana capacayfomulaton $78 The total heat loss during the 12-s time interval is calculated by summing the heat loss of each node relative to the intial temperature of 300°C. Tis 4 = 26.00 - 1) ‘where q isthe heat Joss. For this summation, since the constant-temperature boundary nodes experience no change in temperature, they can be left out. Recalling that T, = Ts, T. = To and T; = To, we have 3C,300 ~ T) = nodes (1, 2,3, 7, 8,9) + nodes (4, 5. 6) 16{(64300) — (2243.32) ~ 231.97 — 2289.71) 86.32] + 32{(39300) — (2279.87) ~ 273.95), $572.3 Jim length of strip The average rate of heat loss for the 12-s time interval is = 464.4 W [1585 Buu’) EXAMPLE 4-13 Cooling ofa ste! rod, nonuniform fh A nickel-stee! ros having a diameter of 2.0 cm is 10 em long and initially at a uniform temperature of 200°C. It is suddenly exposed to atmospheric air at 30°C while one end Fig. Ex. 413178 Unsteady-site conducton of the rod is maintained at 200°C. The convection heat-transfer coefficient can be com- puted from fh = 9.0. AT Wim? - °C where AT is the temperature difference between the rod and air surroundings. The properties of nickel steel may be taken as k = 12 Wim -°C, c = 0.48 ki/kg «°C, and 1p = 7800 kg/m. Using the numerical method, (a) determine the temperature distribution in the rod after 250, $00, 750, 1000, 1250 s, and for steady state; (b) determine the steady-state temperature distribution for a constant h = 22.11 Wim? - °C and compare with an analytical solution, Solution Five nodes are chosen as shown in the accompanying figure with Ax ‘capacitances are then 0.em. The Com Cy = Cy = Cm PONEB IO. 02/7002) 23.524 IPC Ge 11.762 °C Rae Ry- Be ~~ (40.02) FE = HP Ax» 0.01m0.02)0.0247 ~ 30” For node 5 LBL x 10-2 — 309915 0.188496 = = 9920 er gpm = any nt — 30" Re (5.655 x 10-7 — 309°" where T. = 30°C for all nodes. We can compute the following table for worst-case conditions of T = 200°C throughout the rod. The stability requirement so established will then work for all other temperatures. 5 G Node SUR saie* It 0.4048 58.11 2 0.4048 58.11 3 0.4058 58.11 4 0.4048 58.11 3 0.2093 36.197‘arma esstance and capacty formulation 477 ‘Thus, time steps below 56 s will ensure stability. The computational procedure is complicated by the fact that the convection-resistance elements must be recal culated for each time step. Selecting Ar = 50's, we have Av, It 2.1255 2 2.1285 3 2.1285 4 2.1255 5 4251 We then use the explicit formulation of Eq. (4-47) with no heat generation. The com- putational algorithm is thus: 1. Compute R. values for the intial condition, 2. Compute temperatures at next time increment using Eq. (4-47) 3. Recalculate Rx values based on new temperatures. 4. Repeat temperature calculations and continue until the temperature distributions are obtained at the desired times. Results of these calculations are shown in the accompanying figure To determine the steady-state distribution we could carry the unsteady method for- ward a large number of time increments or use the steady-state method and an iterative approach, The iterative approach is required because the equations are nonlinear as a result of the variations in the convection coefficient, We still use a resistance formulation, which is now given as Eq. (3-31): T- = =o Re ‘The computational procedure is 1. Caleulate R. values forall nodes assuming all 7; = 200°C. 2, Formulate nodal equations for the 7's 3. Solve the equations by an appropriate method 4. Recalculate R. values based on 7; values obtained in step 3. 5, Repeat the procedure until there are only small changes in T's178 Unsteaay state concen ‘The results of this iteration are shown in the following table: eration Bcc Mises scciagl Tatchell accleaern cc) 1 148.462 114.381 92.726 80.310 75.302 2 ISL 381 119.557 99.409 87.853 8B. 1RS 3 151.105 119.038 98.702 87.028 82.306 4 151132 119.090 98.774 87.109 82.396 ‘This steady-state temperature distribution is also plotted with the transient profiles The value of h for T, = 200°C is 22.11 Wim? - °C, so the results of the first iteration correspond to a solution for a constant h of this value, The exact analytical solution is given in Eq. (2-34) as Fi @ T= Te _ cosh mil. ~ x) + [hikm) a "ht ‘cosh mL. + (hikm} sinh mL . The required quantities are 2 Fez ne(0.02 m= (i) = [Sedo | * 9-64 mL = (19.1968)0.1) = 1.91964 nn bikm z Gs (12)(19. 1964) ee ‘The temperatures at the nodal points can then be calculated and compared with the ‘numerical results in the following table. As can be seen, the agreement is excellent. ' Percent ! Node (810.)oun deviation ' 0.02 0.6968 027 2 0.08 0.4968 0.59 3 0.06 0.3650 09 4 0.08 0.2989 16 s on 0.2665 13 ‘We may also check the heat loss with that predicted by the analytical relation in Eq. (2:34). When numerical values are inserted we obtain i gent = MATA W A ‘The heat loss for the numerical model is computed by summing the convection loss ' from the six nodes (including base node at 200°C). Using the temperatures for the first iteration corresponding to h = 22.11 Wim®*C, 4g = (22.11) (0.02)(0.02){(200 ~ 30,4) + (148.462 ~ 30) + (114.381 ~ 30) + (92,726 ~ 30) + (80.31 ~ 30) + (75.302 ~ 30,0] + (22.11)(0.01)5(75.302 = 30) 12.082 WTremaleetarce ane capactyfrmulaion 178 We may make a further check by calculating the energy conducted in the base. This must be the energy conducted to node | plus the convection lost by the base node or 2)x¢0,01)2 22 48.46 O02 + (22.11)m(0,0240.01)200 ~ 30) 12.076 W This agrees very well with the convection calculation and both are within 1.8 percent of the analytical value. The results of this example illustrate the power of the numerical method in solving problems which could not be solved in any other way. Furthermore, only a modest ‘number of nodes, and thus modest computation facilities, may be required to obtain a sufficiently accurate solution. For example. the accuracy with which h will be known is typically + 10 to 15 percent. This would overshadow any inaccuracies introduced by using relatively large nodes, as was done here. EXAMPLE 4-14 Radiation heating and cooling ‘The ceramic wall shown is initially uniform in temperature at 20°C and has a thickness, ‘of 3.0 em. It is suddenly exposed to a radiation source on the right side at 1000°C. The left side is exposed to room air at 20°C with a radiation surrounding temperature of 20°C. Properties of the ceramic are k = 3.0 Wim °C, p = 1600 kgim’, and ¢ = 0.8 dike - °C. Radiation heat transfer with the surroundings at T, may be calculated from 4. = eA TOW w where @ = 5.669 x 10-%,¢ = 0.8, and Tis in degrees K. The convection heat-transfer coefficient from the left side of the plate is given by f= 192AT™ Wim! °C w Convection on the right sie is negligible. Determine the temperature distribution in the plate ater 1S, 30, 45, 60, 90, 120, and 150s. Also determine the steady-state temperature distribution. Calculate the total heat gained by the plate for these times Solution We divide the wall into five nodes as shown and must express temperatures in degrees Kelvin because of the radiation boundary condition, For node I the transient energy equation is e298 Te) ncry — zaps + ~ near ~ 2998 + Lars Similarly. for node S are «res Lam = =Thermal resistance and capactyformulaton 194 Equations (c) and (d) may be subsequently written r= B[ouaae + mmya9s + my239) — Lowry —amya9y + A a] 21 cease og + {1-8 [occas + re a38 + my — 19207 ~ aonie + El}n © Ie nm [az + mun + maa + Ln] ry ar ; k + {1-8 focaar + m2 + m+ s]}a wo Where C, = Cs = pe Ax/2. For the other three nodes the expressions are much simpler: atthe, _ thar nen Game ms (1- SE) m a) vet kame m+ (i - 2M mee Game m+ (i- BE) m a mar ; (- 3a)" ® where C: = C, = Cs = pe Ax. So, to determine the transient response, we simply choose a suitable value of Ar and march through the calculations. The stability criterion is such that the coefficients of the last term in each equation cannot be negative. For Eqs. (g) (A), and (i) the maximum allowable time increment is Cx Ax _ (1600(800%0.0075)? _ Aro = 3) For Eq. (fh, the worst case is at the start when 78 = 20°C = 293 K, We have 6 = OCOD, syyg so that Ce ee Toe = 5.665 x 10 OBKIATS + 29FHIIT + 293) 4 3.010.0095, = 9.435 For node 1 {Ea. (e)] the most restrictive eondition occurs when Tt = 293. We have C= C= 4800 so that Srna = 0 (5.669 x 10-HO.BAIT" + 293}Q93 + 293) — 0 + 3.010.0075 = 11.86 5 So, from these calculations we see that node 5 is most restrictive and we must choose Ar < 9.43, The calculations were performed with Ar = 3.0 s, and the results are shown in the02 Unstesaysate conduction accompanying figures. To compute the heat added at any instant of time We perform the sum: Oe) = ECM, ‘and plot the results in Fig. Ex. 4-14c. 293) w EXAMPLE 4-15 Transient conduction with heat generation ‘The plane wall shown has internal heat generation of SO MW/m and thermal properties of k = 19 Wim -°C, p = 7800 kg/m?, and ¢ = 460 J/kg °C. Iti intially at a uniform temperature of 100°C and is suddenly subjected tothe heat generation and the convective boundary conditions indicated in the figure. Calculate the temperature distribution after several time increments. Ing = 400 Wim *C To 0c y= S00Wim?“C T= 20°C are 10mm Fig. Ex. 4-15(a) Solution We use this resistance and capacity formulation and write, for unit area, UR. = KAI = (1991)0.001 = 19,000 WFC All the conduction resistances have this value. Also, UR,, = HA = (400)(1) = 400 WPC Rig = HA = (S00K1) = S00 WP ‘The capacities are C= C= platy e (7800)(0.001/2\460) = 1794 IPC phAx)e = 3588.I°C ¢ Suk) SR) 1940017940. 38000 3588 0.084 36.000 3888 0.094 38.000 3588 (0.094 38,000 3588 (0.094 1794 0.092‘Thormalressiance and capactyforrulaten. 183 Any time increment Sr less than 0.09 s will be satisfactory. The nodal equations are ‘now written in the form of Eq. (4-47) and the calculation marched forward on a computer. ‘The heat generation terms are a Gay, so that 41 = ge = (50 % 1090)0.00172) 25,000 W Gs = as = Ge = Qs = (50 X 109(1}(0.001) = 50,000 W ‘The computer results for several time increments of 0.09 s are shown in the following. table. Because the solid stays nearly uniform in temperature at any instant of time it behaves almost like a lumped capacity. The temperature of node 3 is plotted versus time to illustrate this behavior. Number of time increments (Sx = 0.09 5) Node 5 0 100 200 ! 106.8826 123.0504 190.0725 246.3855 2 106.478 122.8867 190.9618 248.1988, 3 106.1888 122.1404 190.7033, 248.3325 4 105.372 120.9763, 189.3072 246.7933, 5 104.4622 119.2217 186.7698 243.5786 6 102.4416 117.0056 183.0735, 238.673, Node 500 300 1200 3000 1 320.5766 340.1745 346.0174 347.2085 2 323.6071 343.5267 349.4654 350.676 3 324.2577 344.3137 350.2931 351.512 4 322.5298, 342.536 348.5006 349.7165 5 318.4229 338.1934 344.0877 345.2893 6 311.9341 331.2853 337.0545 338.2306 i i ime, Fig. Ex. 4-1540)‘104 Unsieacystteconcucton EXAMPLE 4-18 Numerical solution for variable conductivity ‘A.4,0-cm-thick slab of stainless steel (18% Cr, 8% Ni) i initially t a uniform temperature ‘of O°C with the left face perfectly insulated as shown in the accompanying figure. The right face is suddenly raised to a constant 1000°C by an intense radiation source. Cal- ‘culate the temperature distribution after (a) 25 s, (6) 50, (c) 100s, (d) an interval long ‘enough for the slab to reach a steady state, taking into account variation in thermal , ‘conductivity. Approximate the conductivity data in Appendix A with a linear relation. Repeat the calculation for the left face maintained at 0°C. breton Fig. Ex. 416%) ‘Solution From Table A-2 we have k = 16.3 Wim» °C at °C and k = 31 Wim «°C at 1000°C. A linear relation for k is assumed so that ke kdl + BD where T is in degrees Celsius. Inserting the data gives k = 16.301 + 9.02 x 10°47) Wim «°C We also have p = 7817 kg/m? and c = 460 J/kg “°C, and use the thermal resistance ’ capacitance formula assuming that the resistances are evaluated at the arithmetic mean aw ‘of their connecting nodal temperatures; i.e., Rs-is evaluated at (7, +T02. First, the thermal capacities are evaluated for unit area: Cy = phasic = (181790.01/2X460) = 17980 Jim? °C p= 6 = Co= plane = (1817H0.01)460) = 35960 Jim? = °C For the resistances we have the form, for u UR = Wax = kil + BTA Evaluating at the mean temperatures between nodes gives UR. = (6.31 + 4.51 x 10-47, + TOI = VR. ; UR, = (163) + 451 x 10-4; + TVOOL = UR UR. = (63) + 4.51 x 10-47, + TIVOOI = URas Rea) = (V6.3) + 451 X 10-AT + Tiawd/O.01 = URiow4‘haa’ eeatance and capaci lrmulaton 188 ‘The stability requirement is most severe on node 1 because it has the lowest capacity. ‘To be on the safe side we can choose a large & of about 31 Wim - *C and calculate 17,9800.01) 58s 3 ous ‘The nodal equations are now written in the form of Eq. (4-47); viz., the equation for node 2 would be {1630U + 4.51 x 10-47 + PONTE ~ 7H) + 1630{1 + 4.51 10-4 (TS + TOUTS - THE + Te ‘A computer solution has been performed with Ar = 5 s and the results are shown in the tables, The steady-state solution for the insulated left face is, of course, a constant 1000°C. The steady-state distribution for the left face at 0°C corresponds to Ea. (2-2) of Chap. 2. Note that, because of the nonconstant thermal conductivity, the steady- state temperature profile is not a straight line ‘Temperatures for Loft Face at Constant 0°C, Ar = 5 8 Steady Node 28s 0s Os sate 1 0 0 0 0 2 94.5788 736.9619 308.2699 317.3339 3 318.7637 «486.5802 565.7786 575.9138 4 653.5105 748.1389 793.7976 799.7735 ‘Temperatures for Left Face Insulated, Ar = 5.8 ‘Steady Node 25s Sos 1005 state 1 3055758 232.8187 $87,021 1000 a 96.67601 310.1737 623.5018, 1000 3 318,7637 505.7613. 721.5908, 1000 4 653.5105 752.3268 855.6965, 1000 ‘These temperatures are plotted in the accompanying figure. ‘The purpose of this example has been to show how the resistance-capacity formu lation can be used to take into account property variations in a rather straightforward ‘way. These variations may or may not be important when one considers uncertainties in boundary conditions.Fig, Ex. 416(0) 2 4.8 GRAPHICAL ANALYSIS_THE SCHMIDT PLOT A graphical technique of historical note may be employed to establish transient temperature distributions when the problem is one-dimensional. The method is based on the choice of the parameter (ax? aie (4-48) so that the temperature at any node after the time increment Aris the arithmetic average of the temperatures of the adjacent nodes at the beginning of the time increment. Such an arithmetic average is very easy to construct graphically, as shown in Fig. 4-22. The value of T2,°' is obtained by drawing a line between Tp. and Te, «1. Thus, to find the temperature distribution in a solid after some specified time, the solid is divided into increments of Ax. Then, by using Eq. (4-48), the value of Aris obtained. This value of Az, when divided into the total time, gives the number of time increments necessary to establish the desired temperature distribution. The graphical construction is repeated until the final temperature distribution is obtained. A nonintegral number of time increments is usually required, and it will probably be necessary to interpolate between the last two increments to obtain the final temperature distribution.Graorca!anavas—tne Serena okt 187 fF Fig. 4-22. Graphical constuction for one-cimen. be arelearl sonal uneteady-siae conduction problems, ‘An example of the method is shown in Fig. 4-23, where an initial temperature distribution is given and the construction is carried out for four time increments. The boundary temperatures are maintained at constant values throughout the cooling process indicated in this example, Notice that the construction ap- proaches the steady-state straight-line temperatire distribution with increasing time. Fig. 423. Schmct-plot constuction lor four time crerents‘198 Unstongy state cancion When a convection boundary condition is involved, the construction at the boundary must be modified. Rewriting Eq. (4-39), we have 7 I, -T. 7 a (4-49) and the temperature gradient at the surface is approximated by the construction shown in Fig. 4-24. A line is drawn between the temperature T,,,, and the environment temperature 7... The intersection of this line with the surface determines the surface temperature at that particular time. This type of con- struction is used at each time increment to establish the surface temperature. Once this temperature is established, the construction to determine the internal temperatures in the solid proceeds as described above. An example of the ‘construction for the convection boundary-condition problem with four time increments is shown in Fig. 4-25. In this example the temperature of the right face and the environment temperature 7. are maintained constant. If the en- vironment temperature changes with time, according to some known variation, this could easily be incorporated in the construction by moving the T. point up or down as required. In a similar fashion a variable heat-transfer coefficient could be considered by changing the value of k/h, according to some specified variation, and moving the environment point in or out a corresponding distance. Refinements in the Schmidt graphical method are discussed by Jakob [5], Particularly the techniques for improving accuracy at the boundary for either convection or other boundary conditions. The accuracy of the method is im- Proved when smaller Ax increments are taken, but this requires a larger number of time increments to obtain a temperature distribution after a given time. Graphical techniques are seldom sed anymore because of the ready avail- Fig. 424 Graphical technique of representing con- ‘ection boundary condition wit the Schmidt pitsummary 189 Fig. 4.25 Schmit plot forfour ume ncremen mn baunaary condiion Hg con ability of computing power. We ~ention them here mainly to indicate the resourcefulness of heat-transfer workers before the time of computers. 49 SUMMARY In progressing through this chapter the reader will have noted analysis tech- niques of varying complexity, ranging from simple lumped-capacity systems to numerical computer solutions. At this point some suggestions are offered for a general approach to follow in the solution of transient heat-transfer prob- lems, First, determine if a lumped capacity analysis can apply. If so, you may be led to a much easier calculation. 2. Check to see if an analytical solution is available with such aids as the Heisler charts and approximations. 3. If analytical solutions are very complicated, even when already available, move directly to numerical techniques. This is particularly true where re- petitive calculations must be performed. 4. When approaching a numerical solution recognize the large uncertainties present in convection and radiation boundary conditions. Do not insist upon,1180. Unsiesaysiate conaucton a large number of nodes and computer time (and chances for error) which cannot possibly improve upon the basic uncertainty in the boundary con- ditions. 5. Finally, recognize that it is a rare occurrence when one has a “‘pure’’ con- duction problem; there is almost always a coupling with convection and radiation. The reader should keep this in mind as we progress through sub- sequent chapters which treat heat convection and radiation in detail. @ REVIEW QUESTIONS 1 ‘What is meant by a lumped capacity? What are the physical assumptions necessary for a lumped-capacity unsteady-state analysis to apply” 2. What is meant by a semi-infinite solid? ‘What initial conditions are imposed on the transient solutions presented in graphical form in this chapter? ‘What boundary conditions are applied to problems in this chapter? Define the error function. Define the Biot and Fourier numbers. Describe how one-dimensional transient solutions may be used for solution of two- ‘and three-dimensional problems. ‘What are the advantages and disadvantages of the forward- and backward-difference formulations in the unsteady-state numerical method” Under what conditions would ‘you choose one method over the other? = PROBLEMS 441 A.copper sphere initially ata uniform temperature To is immersed in aluid. Electric heaters are placed in the fluid and controlled so that the temperature of the fluid follows a periodic variation given by where T,, = time-average mean fluid temperature ‘A = amplitude of temperature wave @ = frequency Derive an expression for the temperature of the sphere as a function of time and the heat-transfer coefficient from the fluid to the sphere. Assume that the tem- peratures of the sphere and fluid are uniform at any instant so that the lumped- ‘capacity method of analysis may be used. 42 Am infinite plate having a thickness of 2.5 cm is initially at a temperature of 150°C, and the surface temperature is suddenly lowered to 30°C. The thermal diffusivity Of the material is 1.8 x 10-* m/s. Calculate the center-plate temperature after 1 min by summing the first four nonzero terms of Eq. (43). Check the answer using the Heisler chartsas ar 49 410 an “2 Proven 194 What error would result from using the first four terms of Eq. (4-3) to compute the temperature at r= Oand x = L? |A solid body at some initial temperature Ty is suddenly placed in a room where the air temperature is T. and the walls of the room are very large. The heat- transfer coefficient for the convection heat loss is h, and the surface of the solid nay be assumed black. Assuming that the temperature in the solid is uniform at ‘Ty instant, write the differential equation for the variation in temperature with time, considering both radiation and convection. ‘A.W by 30 cm slab of copper 5 cm thick at a uniform temperature of 260°C caddenly has its surface temperature lowered to 35°C. Using the concepts of thermal resistance and capacitance and the lumped-capacity analysis, find the time at which the center temperature becomes 97°C; p = 8900 kg/m’, cp = 0.38 klikg «SC, and & = 370 Wim » °C. 'A piece of aluminum weighing 5.5 kg and initially at « temperature of 290°C is fuddenly immersed in a fluid at 15°C. The convection heat-transfer coefficient is $8 Wis "C. Taking the aluminum as a sphere having the same weight as that given, estimate the time required to cool the aluminum to 90°C, using the lumped ‘capacity method of analysis, “Two identical 7.S-cm cubes of copper at 425 and 90°C are brought into contact ‘heouming that the blocks exchange heat only with each other and that there is fo resistance to heat flow as a result of the contact of the blocks, plot the tem- erature of each block as a function of time, using the lumped-capacity method Prranalysis, That is, assume the resistance to heat transfer is the conduction esistance of the two blocks. Assume that all surfaces are insulated except those in contact. Repeat Prob. 47 fora 7.S-cm copper cube at 425°C in contact witha 7.5-em steel cube at 90°C, Sketch the thermal circuit ‘Am infinite plate of thickness 22 is suddenly exposed to a constant-temperature fadiation heat source oF sink of temperature T,. The plate has a uniform initial Temperature of T,. The radiation heat loss from each side of the plate is glven by se yea(T®— 7-1), where and € are constants and is the surface area, 4 cumming that the plate behaves as a lumped capacity, that is. k—> », derive an “expression for the temperature of the plate as a function of time. ‘A stainless-steel rod (18% Cr, 8% Ni) 6.4 mm in diameter is initially at a uniform ‘emperature of 50°C and is suddenly immersed ina liquid at 200°C with A = 120 Wins C. Using the lumped-capacity method of analysis, calculate the time necessary for the rod temperature to reach 120°C. |A S-emediameter copper sphere is initially at a uniform temperature of 250°C * suddenly exposed to an environment at 30°C having 2 heat transfer coefficient Fn-28 Winn? *C. Using the lumped-capacity method of analysis. calculate the time necessary for the sphere temperature to reach °C. [A stack of common building brick I m high, 3 m long, and 0.5 m thick leaves an aie where it has been heated {0 a uniform temperature of 300°C, The stack is Slowed to cool in a room at 35°C with an air-convection coefficient of 15 Wim? °C.on ou as cory ow 8 on ‘The bottom surface of the brick is on an insulated stand. How much heat will hhave been lost when the bricks cool to room temperature? How long wil it take to lose half this amount, and what will the temperature at the geometric center Of the stack be at this time? A.copper sphere having a diameter of 3.0 cm is intially at a uniform temperature of 50°C. It is suddenly exposed to an airstream of 10°C with h = 15 Wim? °C. How long does it take the sphere temperature to drop to 25°C? ‘An aluminum sphere, 5.0 cm in diameter, is initially at a uniform temperature of ‘50°C. It is suddenly exposed to an outer-space radiation environment at 0 K (no convection). Assuming the surface of aluminum is blackened and lumped-capacity analysis applies, calculate the time required for the temperature of the sphere to drop to 110°C. ‘An aluminum can having a volume of about 350 cm? contains beer at 1°C. Using 4 lumped-capacity analysis, estimate the time required for the contents to warm to 15°C when the can is placed in a room at 22°C with a convection coefficient of 15 Wim? - °C. Assume beer has the same properties as water, A 12-mm-diameter aluminum sphere is heated to a uniform temperature of 400°C and then suddenly subjected to room air at 20°C with a convection heat-transfer coefficient of 10 Wim? - °C. Calculate the time for the center temperature of the sphere to reach 200°C. ‘A 4-cm-diameter copper sphere is intially at a uniform temperature of 200°C. It is suddenly exposed to a convection environment at 30°C with h = 20 Wine «°C. Caleulate the time necessary for the center of the sphere to reach a temperature of 80°C. When a sine-wave temperature distribution is impressed on the surface ofa semi- infinite solid, the temperature distribution is given by tae= Ta = Aer (-= sn (200-2) where T,, = temperature at depth x and time rafter start of temperature wave at surface T,, = mean surface temperature in = frequency of wave, cycles per unit time A = amplitude of temperature wave at surface Ifa sine-wave temperature distribution is impressed on the surface ofa large slab of concrete such that the temperature varies from 35 10 90°C and a complete cycle is accomplished in 15 min, find the heat flow through a plane Sm from the surface 2h after the start of the initial wave. Using the temperature distribution of Prob. 4-18, show that the time lag between ‘maximum points in the temperature wave at the surface and at a depth xis given by420 aa 422 a2 44 426 427 428 429 420 432 Probleme 193) AA thick concrete wall having a uniform temperature of 54°C is suddenly subjected to an airstream at 10°C. The heat-transfer coefficient is 2.6 Wim? «°C. Calculate the temperature in the concrete slab at a depts. of 7 em after 30 min. |A very large slab of copper is initially at a temperature of 300°C, The surface temperature is suddenly lowered to 38°C. What is the temperature at a depth of 7.S em 4 min after the surface temperature is changed” On a hot summer day a conerete driveway may reach a temperature of SO°C. ‘Suppose that a stream of water is directed on the driveway so that the surface temperature is suddenly lowered to 10°C. How long will take to cool the concrete to 25°C at a depth of 5 em from the surface? A semi-infinite slab of copper is exposed to a constant heat flux at the surface of. 0.32 MWim?, Assume that the slab is in a vacuum, so that there is no convection atthe surface. What isthe surface temperature after S min ifthe initial temperature fof the slab is 30°C? What is the temperature at a distance of 15 cm from the surface after 5 min? A large slab of copper is initially at a uniform temperature of 90°C. Its surface temperature is suddenly lowered to 30°C. Caleulate the heat-transfer rate through ‘a plane 7.5 cm from the surface 5 after the surface temperature is lowered. A large slab of aluminum at a uniform temperature of 30°C is suddenly exposed to a constant surface heat fux of 15 kWim?. What is the temperature at a depth of 2.5 em after 2 min? For the slab in Prob. 4-25, how long would it take for the temperature to reach 130°C at the depth of 2.5 m2. A piece of ceramic material (k = 0.8 Wim *C.p © 2700kgim!.¢ = 0.8 kitkg - °C] is quite thick and initially at a uniform temperature of 30°C. The surface of the material is suddenly exposed to a constant heat flux of 650 Wim? - °C. Plot the temperature at a depth of 1 cm as a function of time. AA large thick layer of ice is initially at a uniform temperature of ~ 20°C. If the Surface temperature is suddenly raised to ~ 1°C, calculate the time required for the temperature at a depth of 1.5 cm to reach -11I°C. The properties of ice are p= 57 Ibe/f0, cy = 0.46 Btuilb, k= 1.28 Btuhh ft °F, a = 0.088 ft) ‘A large slab of concrete (stone 1-2-4 mix) is suddenly exposed to a constant radiant heat flux of 900 Wim?, The slab is initially uniform in temperature at 20°C. Calculate the temperature at a depth of 10 cm in the slab after a time of 9h. A very thick plate of stainless steel (18% Cr, 8% Ni) at a uniform temperature of 300°C has its surface temperature suddenly lowered to 100°C. Calculate the time required for the temperature at a depth of 3 cm to attain a value of 200°C. ‘A large slab has properties of common building brick and is heated to a uniform temperature of 40°C. The surface is suddenly exposed to a convection environ- ment at 2°C with A = 25 Wim? °C. Calculate the time for the temperature to reach 20°C at a depth of & em. ‘A large block having the properties of chrome brick at 200°C is at a uniform temperature of 30°C when itis suddenly exposed to a surface heat flux of 3 x 10*435, 436 a7 438 439 Unstendy-stae conduction Wim. Calculate the temperature at a depth of 3 em after a time of 10 min. What is the surface temperature at this time? A slab of copper having a thickness of 3.0 cm is initially at 300°C. It is suddenly exposed to a convection environment on the top surface at 80°C while the bottom surface is insulated. In 6 min the surface temperature drops to 140°C. Calculate the value of the convection heat-ransfer coefficient. A large slab of aluminum has a thickness-of 10 cm and is initially unifofm in temperature at 400°C. Suddenly itis exposed to a convection environment at 90°C with h = 1400 Wim? - °C. How long does it take the centerline temperature to drop to 180°C? A horizontal copper plate 10 cm thick is initially uniform in temperature at 250°C. The bottom surface of the plate is insulated, The top surface is suddénly exposed to a fluid stream at 100°C. After 6 min the surface temperature has dropped to 150°C. Calculate the convection heat-transfer coefficient which causes this drop. AA large slab of aluminum has @ thickness of 10 cm and is initially uniform in temperature at 400°C. It is then suddenly exposed to a convection environment at 90°C with h = 1400 Wim? - °C. How long does it take the center to cool to 180°C? A plate of stainless steel (18% Cr. 8% Ni) has a thickness of 3.0 cm and is initially uniform in temperature at 500°C. The plate is suddenly exposed to a convection environment on both sides at 40°C with k = 150 Wirt » °C. Calculate the times for the center and face temperatures to reach 100°C (A steel cylinder 10 cm in diameter and 10 em long is initially at 300°C. It is suddenly immersed in an oil bath which is maintained at 40°C. with h ~ 280 Wim: - °C. Find (a) the temperature at the center of the solid after 2 min and (b) the temperature at the center of one of the circular faces after 2 min, ‘An aluminum bar has a diameter of 11 em and is initially uniform in temperature at 300°C. If it is suddenly exposed to a convection environment at $0°C with hh = 1200 Wim? - °C, how long does it take the center temperature to cool to 80°C? Also calculate the heat loss per unit length, A‘S.1b roast initially at 70°F is placed in an oven at 380°F. Assuming that the heat-transfer coefficient is 2.$ Biu/h - ft °F and that the thermal properties of the roast may be approximated by those of water, estimate the time required for the center of the roast to attain a temperature of 20°F. A fused-quartz sphere has a thermal diffusivity of 9.5 x 10-7 ms, a diameter of 2.5 em, and a thermal conductivity of 1.52 Wim - °C. The sphere is intially at 8 uniform temperature of 25°C and is suddenly subjected to @ convection envi ronment at 200°C. The convection heat-transfer coefficient is 110 Wim? - °C. Calculate the temperatures at the center and at a radius of 6.4 mm after a time of 4 min, Lead shot may be manufactured by dropping molten-lead droplets into water. ‘Assuming that the droplets have the properties of solid lead at 300°C, calculate the time for the center temperature to reach 120°C when the water is at 100°C with h = $000 Wim? -°C, d = 1.5 mm483 449 ast 482 Problems 108 A steel sphere 10 cm in diameter is suddenly immersed in a tank of oil at 10°C. ‘The initial temperature of the sphere is 220°C: h = 5000 Wim? - °C. How long will it take the center of the sphere to cool to 120°C? [A boy decides to place his glass marbles in an oven at 200°C. The diameter of the marbles is 15 mm. After a while he takes them from the oven and places them in room air at 20°C to cool. The convection heat-transfer coefficient is approxi- mately 14 Wim» °C. Calculate the time the boy must wait until the center tem- perature of the marbles reaches 35°C. A lead sphere with d = 1.5 mm and initial temperature of 200°C is suddenly exposed to.a convection environment at 100°C and hk = 5000 Wim? «°C. Calculate the time for the center temperature to reach 120°C. ‘A long steel bar 5 by 10 cm is initially maintained at a uniform temperature of 250°C. Itis suddenly subjected to a change such that the environment temperature is lowered to 35°C. Assuming a heat-transfer coefficient of 23 Wim? «°C, use a ‘numerical method to estimate the time required for the center temperature to reach 90°C. Check this result with a calculation, using the Heisler charts. AA steel bar 2.5 em square and 7.5 cm long is initially at a temperature of 250°C. Itis immersed in a tank of oil maintained at 30°C. The heat-transfer coefficient is $70 Wim? - °C. Calculate the temperature in the center of the bar after 2 min. ‘A.cube of aluminum 10 cm on each side is initially at a temperature of 300°C and is immersed in a fluid at 100°C. The heat-transfer coefficient is 900 Win? «°C. Calculate the temperature at the center of one face after 1 min. |A short concrete cylinder 1S cm in diameter and 30 cm long is initially at 25°C It is allowed to cool in an atmospheric environment in which the temperature is heat-transfer coeffi ‘A 4.0-0m cube of aluminum is initially at 450°C and is suddenly exposed to a convection environment at 100°C with h = 120 Wim? ~ °C. How long does it take the cube to cool to 200°C? ‘A cube of aluminum 11 em on each side is initially at a temperature of 400°C. It js suddenly immersed in a tank of oil maintained at 85°C. The convection coef- ficient is 1100 Wim? - °C. Calculate the temperature at the center of one face after atime of I min. ‘An aluminum cube S em on a side is initially at a uniform temperature of 100°C ‘and is suddenly exposed to room air at 25°C. The convection heat-transfer coef- ficient is 20 Wim - °C, Calculate the time required for the geometric center tem- perature to reach S0°C. A stainless stee! cylinder (18% Cr, 8% Ni) is heated to a uniform temperature of 200°C and then allowed to cool in an environment where the air temperature is ‘maintained constant at 30°C. The convection heat-transfer coefficient may be taken as 200 Wim? » C, The cylinder has a diameter of 10 cm and a length of 15 em. Calculate the temperature of the geometric center of the cylinder after a time of 10 min. Also calculate the heat loss.196 457 as i Unsteady state conduction A cylinder having a diameter of 15 cm and a length of 30 em is intially uniform in temperature at 300°C. It is suddenly exposed to a convection environment at 20°C with h = 35 Wim? - °C. Properties of the solid are k = 2.3 Wim -°C,p = 3000 kgim’, and c = 840 J/kg «°C. Calculate the time for (a) the center and (6) the center of one face to reach a temperature of 150°C. Also calculate the heat loss for each case, ‘A rectangular solid is 15 by 10 by 20 cm and has the properties of fireclay brick. 1 is intially uniform in temperature at 300°C and then suddenly exposed to a convection environment at 80°C and f = 110 Wim? «°C. Calculate the time for (a) the geometric center and (6) the center of each face to reach a temperature of 200°C. Also calculate the heat loss for each of these times, Caleulate the heat loss for both cases in Prob. 438. Calculate the heat loss for the bar in Prob. 4-47 per unit length, Calculate the heat loss for the cube in Prob. 4-48. Develop a backward-difference formulation for a boundary node subjected to a convection environment. Check with Table 4:3. The stainless-steel plate is surrounded by an insulating block as shown and is initially at a uniform temperature of S0°C with a convection environment at S0°C. The plate is suddenly exposed to a radiant heat ux of 20 kW/m?. Calculate the temperatures at the indicated nodes after 19 s, 1 min, and 10 min, Take the properties of stainless steel as k = 16 Wim °C. p = 7800 kp/m’, and c = 0.46 Kuikg «°C, A = 30 Wim? «°C. Assume all the radiation is absorbed, Fig. P4-60 The composite plate shown has one face insulated and is initially at a uniform temperature of 100°C. At time zero the face is suddenly exposed to a convection environment at 10°C and hk = 70 Wim? « C. Determine the temperatures at the indicated nodes after 1's, 10 s, 1 min, and 10 min, . b, 6, Material Wim °C__kgim? uke = °C A 20 7800 0.46 B 12 1600 os c os 2500 08Prooloms 197 Osom—a} bode peas 1Sem 462 The 4.0-mm-diameter stainless-steel wire shown is initially at 20°C and is exposed tw a convection environment at 20°C where h may be taken as 200 Wim? °C. ‘An electric current is applied to the wire such that there is a uniform internal heat generation of 500 MW/m?. The left side of the wire is insulated as shown. Set up the nodal equations and stability requirement for calculating the temer- ‘ature in the wire as a function of time, using increments of Ar = 0.5 mm and Ag = m/4. Take the properties of stainless steel as k = 16 Wim -*C, p = 7800 kg/m’, and c = 0.46 kilkg -°C. Fig. Paez 463. The comer shown is inially uniform at 200°C and then suddenly exposed to convection around the edge with h = SO Wim? °C and T= 30°C. Assume the solid has the proparties of fireclay brick. Examine nodes 1, 2, 3, and 4 and determine the maximum time increment which may be used for a transient nu- ‘merical calculation bo | is Fig. Pasa 4464 Write the rodal equation for node 3 in the figure for use in a transient analysis. Determine the stability criterion for this node.198 Unctenoy state conducton Fig. Pé-6e 485 Write a nodal equation for analysis of node (m,n) in the figure to be used in a transient analysis of the solid Ccamsection, hos Fig. Paes 458 Write the nodal equation and establish the stability criteria for node | in the figure (ransient analysis). Materials A and B have the properties given in Prob, 4-61 T.=30°C A240 Wim? “c Fig. P4668 467 Write a computer program which will solve Example 4-15 for different input properties. For nomenclature take T(N) = temperature of node N at beginning bf time increment, TP(N) = temperature of node at end of time increment X = number of nodes, W = width of plate, TA = temperature of left fluid, HA ~ convection coefficient of left fuid, TB = temperature of right fluid, HB+60 69 = convection coetficient of right fuid, DT = time increment, C = specific heat, D = density. K = thermal conductivity, Q = heat-generation rate per unit vol ume, TH = total time. Write the program so that the user can easily rerun the program for new times and print out the results for each CCaleulate the maximum time inerement that can be used for node $ in the ac- ‘companying figure fora iransient qumerical analysis. Also write the nodal equation for this node = atm Fig ‘The corner shown is initially uniform at 300°C and then suddenly exposed to a convection environment at 50°C with h = 60 Wim? «°C, Assume the solid has the properties of fireclay brick. Examine nodes 1. 2, 3. 4, and 5 and determine the maximum time increment which may be used for a transient numerical cal- culation Fig. Pa-69 Write a steady-state nodal equation for node 3 in the figure assuming unit depth perpendicutit to the page and using the node spacing shown. The thermal con200 Unsteady sate conduction ductivity of the solid is 15 W/m - °C and the convection heat-transfer coefficient ‘on the side surface is 25 Wim? - °C. qo so T Fie * pg. Paro 471 For the section shown, calculate the maximum time increment allowed for node 2 in a transient numerical analysis. Also write the entire nodal equation for this node. f= s0Wint-"c Mates e 0 2 Wimsc > 6500 2000 kgim® ¢ 03 07 Whe"C pig. part ‘472 A transient numerical analysis is to be performed on the composite material section shown. Calculate the maximum time increment that can be used for node 5 to ensure convergence. Loc 4 Toem : A gypsum plaster B= Als (darsumin) C2 Tas ce ae Fig. Pé-72on on as “78 on an an For the section shown, calculate the maximum time increment allowed for node 4 in a transient numerical environment. Also write the complete nodal equation for node 4 20 2 Wine > 700 1600 Agia? © 0S 8 he"C lg. Pare ‘A node like that shown in Table 3-2d has both x and y increments equal to 1.0 ‘em. The convection boundary condition is at 50°C and h = 60 Wim? «°C. The solid material is stainless stee! (18% Cr, 8% Ni). Using the thermal resistance ‘and capacitance formulation for @ transient analysis write the nodal equation for this node and determine the maximum allowable time increment. ‘The solid in Prob. 3-43 is intially uniform in temperature at 10°C. At time zero the right face is suddenly changed to 38°C and the left face exposed to the con- vection environment. Nodes 3 and 6 remain at 10°C. Select an appropriate value for 4 and calculate the temperatures of nodes 1, 2, 4, and 5 after 10 time incre- ‘ments. Carry the calculation forward to verify the steady-state distribution. Take (p= 3000 kg/m? and c = 840 Jikg -°C. The solid in Prob. 3-45 has k = 11 Wim - °C andis initially uniform in temperature at 1000°C. At time zero the four surfaces are changed to the values shown. Select aan appropriate Ar and calculate the temperatures of nodes 1, 2, 3, and 4 after 10 time increments. Also obtain the limiting steady-state temperatures. Take p = 2800 kg/m? and c = 940 Jikg - °C. ‘The fin in Prob, 3-49: initially uniform in temperature at 300°C and then suddenly exposed to the convection environment. Select an appropriate Ar and calculate the nodal temperatures after 10 time increments. Take p = 2200 kg/m! and c = 820 kg °C, ‘The fin in Prob. 3-50 is initially uniform in temperature at 200°C and then is suddenly exposed to the convection environment shown while maintaining the bottom face at 200°C. Select an appropriate Ar and calculate the nodal temper- atures after 10 time increments. Repeat for 100 Ar. Take p = 7800 kg/m? and = 460 Jikg «°C. The solid in Prob. 3-51 is initially uniform in temperature at 100°C and then suddenly exposed to the convection condition while the right and bottom faces are held constant at 100°C. Select a value for Ar and calculate the nodal tem- peratures after 10 time increments. Take p = 3000 kg/m? and c = 800 J/kg - °C.02 Unateedystate conduction +00 é & ‘The solid in Prob. 3-52 is initially uniform in temperature at SO°C and suddenly is exposed to the convection condition. Select a value for Ar and calculate the nodal temperatures after 10 time increments. Take p = 2500 kg/m? and c = 900 ig -*C. ‘The solids in Prob. 3-53 are initially uniform in temperature at 300°C and suddenly are exposed to the convection boundary, while the inner temperature is kept constant at 307°C. Select a value for Av and calculate the nodal temperatures after 1O time increments. Take py = 2900 kgm’, cy, = 810,J0kg **C, pp = 7800 kim’, and cp = 470 Stkg °C. ‘The fin in Prob. 3-54 is initially uniform in temperature at 200°C, and then suddenly exposed to the convection boundary and heat generation. Select a value for Ar ‘and calculate the nodal temperatures for 10 time increments. Take p = 7600 kg/m? and c = 450 J/kg °C. The base stays constant at 200°C. ‘The solid in Prob. 3-55 is intially uniform in temperature at SOO°C and suddenly exposed to the convection boundary while the inner surface is kept constant at 500°C. Select a value for Ar and calculate the nodal temperatures after 10 time increments. Take p = 500 kg/m? and c = 810 Jikg «°C. Repeat Prob. 4-83 for the steel liner of Prob. 3-56. Take p = 7800 kg/m? and ¢ = 460 Jikg «°C for the steel. ‘The plate in Prob. 3-57 is intially uniform in temperature at 100°C and suddenly exposed to the convection boundary. Select a value for A and calculate the nodal temperatures after 10 time increments. Take p = 7500 kg/m? and c = 440 Jikg - °C. The solid shown in Prob. 3-8 is initially uniform in temperature at 100°C and ‘suddenly exposed to the convection boundary and heat generation while the right face is kept at 100°C. Select a value for A and calculate the nodal temperatures after 10 time increments. Take p = 7600 kg/m? and c = 460 J/kg - °C. AA steel rod 12.5 mm in diameter and 20 cm long has one end attached to a heat reservoir at 250°C. The bar is initially maintained at this temperature throughout. It is then subjected to an airstream at 30°C such that the convection heat-transfer coefficient is 35 Wm - °C. Estimate the time required for the temperature midway along the length of the rod to attain a value of 19C. A concrete slab 15 cm thick has a thermal conductivity of 0.87 Wim « °C and has ‘one face insulated and the other face exposed to an environment. The slab is initially uniform in temperature at 300°C, and the environment temperature is suddenly lowered to 90°C. The heat-transfer coefficient is proportional to the fourth root of the temperature difference between the surface and environment and has a value of 11 Wm? -°C at time zero. The environment temperature increases linearly with time and has a value of 200°C after 20 min. Using the ‘numerical method, obtain the temperature distribution in the slab after 5, 10, 15, ‘and 20 mit perature as shown and calculated in Table 3-3. At time zero the 500°C face is suddenly lowered to 30°C. Taking Ax = Ay = 1Semand a = 1.29 x 10°* mis, calculate the temperatures at nodes 1, 2, 3, and 4 after 30 min. Perform theoo Problems 209 calculation using both a forward- and backward-difference method. For the back- ward-difference method use only two time increments. Take k = 45 Wim » °C. ‘The strip of material shown has a thermal conductivity of 20 Wim °C and is placed firmly on the isothermal surface maintained at SO°C. At time zero the strip is suddenly exposed to an airstream with T. = 300°C and h = 40 Wim? » °C. Using @ numerical technique, calculate the temperatures at nodes | to 8 after 1 5, 10's, 1 min, and steady state; p = 7000 kg/m? and c = 0.5 ki/kg -°C. Fig. Pé-50 Réwork Probs. 4-7 and 4-8 using the numerical technique. Rework Prob. 4-87 using the numerical technique. A blackened ceramic sphere of 10 cm diameter is initially uniform in temperature at 1000°K and is suddenly placed in outer space where it loses heat by radiati (a0 convection) according to AT* Tin degrees Kelvin 5.669 x 10-* Wim?» Kt Calculate the temperatures of the nodes shown for several increments of time and the corresponding heat losses. Use the values of k, p, and c from Prob. 4-60. in Fig. Pasa Anhollow concrete sphere (k = 1.3 Wim -°C, a = 7 x 10°? m’/s] has inside and ‘outside diameters of 0.5 and 1.0m and is initially uniform in temperature at 200°C. ‘The outside surface is suddenly lowered to 20°C. Calculate the nodal temperatures1204 Unsteay sate conduction shown for several increments of time. Assume the inside surface acts as though it were insulated. Fig. Pass 495. The stainless-steel plate shown is initially at a uniform temperature of 150°C and is suddenly exposed to a convection environment at 30°C with k = 17 Wim? -°C. Using numerical techniques, calculate the time necessary for the temperature at a depth of 6.4 mm to reach 65° bs0em—a} Fig. P45 4496 Repeat Prob. 4-50 with the top surface also losing heat by radiation according to = oAdT* - 7) Tin degrees Kelvin = 5.669 x 10° Wim? K* «=07 i 497 A fireproof safe is constructed of loosely packed asbestos contained between thin sheets of stainless steel. The safe is built in the form of a cube with inside , and outside dimensions of 0.5 and 1.0 m. If the safe is initially uniform in 4 ‘temperature at 30°C and the outside is suddenly exposed to a convection envi- ronment at 600°C, h = 100 Wim? «°C, calculate the time required for the inside temperature to reach 150°C. Assume the inside surface is insulated, and neglect the resistance and capacitance of the stainless steel. Take the properties of asbestos as k = 0.16 Wim -°C, a = 3.5 x 10-7 ms, 498 The half-cylinder in Prob. 3-59 is intially uniform in temperature at 300°C and then suddenly exposed to the convection boundary while the bottom side is maintained at 300°C, Calculate the nodal temperatures for several time incre- ‘ments, and compute the heat loss in each period. Take a = 0.5 x 10-% m/s. 499 A large slab of brick [k = 1.07 Wim -°C, a = 5.4 x 10-7 m/s] is initially at a uniform temperature of 20°C. One surface is suddenly exposed to a uniform heat , ffux of 4500 Wim?. Calculate and plot the surface temperature as a function of 4 time. Also calculate the heat lux through the plane 2.0cm deep when the surface ' temperature reaches 150°CReterences 208 4100 Repeat Prob. 4-95 with the top surface also losing heat by radiation according to ne = GAA T* = T.1) Tin degrees Kelvin 5.669 x 10-* Wim? = K* «207 Repeat the calculation for 10 and 20 min, 4101 Work Prob. 4-22 using the numerical method, 4102 Work Prob. 4-20 using the numerical method, 4103 Work Prob. 4-22 using the Schmidt plot. 4108 Work Prob. 4-20 using the Schmidt plot. 4105 A ceramic plate having a thickness of 2.0 em is heated to a uniform temperature ‘of 1000°K and suddenly exposed to radiation on both sides at 300°K. The prop- erties of the solid are k = 1.2 Wim °C, p = 2500 kg/m’, ¢ = 0.9 klikg : °C, and € = 0.85. Divide the plate into eight segments (Ax ='0.25 cm) and, using ‘2 numerical technique, obtain information to plot the center and surface tem peratures as a function of time 4106 Suppose the ceramic of Prob. 4-105 is in the form of a long cylinder having a diameter of 2.0 cm. Divide the cylinder into four increments (Ar = 0.25 em) and obtain information to plot the center and surface temperatures as a function of REFERENCES 1. Schneider. P. J.: “Conduction Heat Transfer."" Addison-Wesley Publishing Com- pany. Inc_, Reading, Mass... 1955 2 Heisler. M. P.: Temperature Charts for Induction and Constant Temperature Heat ing. Trans. ASME, vol. 69, pp. 227-236, 1947. 3. Abramowitz, M.. and I. Stegun (eds.): Handbook of Mathematical Functions, NBS [AMS 5S. U.S. Government Printing Office. 1964 4 Dusinberre, G. M.: “Heat Transfer Calculations by Finite Differences,” Interna- tional Textbook Company. Scranton, Pa... 1961 Jakob. M.: “Heat Transfer,” vol. 1, John Wiley & Sons, Ine, New York, 1949. Graber, H., S. Erk, and U, Grigull: “Fundamentals of Heat Transfer,” MeGraw- Hill Book Company. New York, 1961 7 Schneider, P. J.: “Temperature Response Charts,” John Wiley & Sons, Ine., New York, 1963. 8 Schenck. H.: “Fortran Methods in Heat Flow,” The Ronald Press Company, New York. 1963. 9 Richardson. P. D.. and Y. M. Shum: Use of Finite-Element Methodls in Solution oof Transient Heat Conduction Problems, ASME Pap. 69-WA/HT-3.1208 Unsteady-state conduction 10 8 “ 6 6 7 Emery, A. F., and W. W. Carson: Evaluation of Use of the Finite Element Method in Computation of Temperature, ASME Pap. 69-WAJHT-38. Wilson, E. L., and R. E. Nickell: Application of the Finite Element Method to Heat Conduction Analysis, Nucl. Eng. Des., vol. 4, pp. 276-286, 1966. Zienkiewicz, O. C.: “The Finite Element Method in Structural and Continuum Mechanics,”* McGraw-Hill Book Company, New York, 1967. Myers, G. E.: “Conduction Heat Transfer,” McGraw-Hill Book Company, New York, 1972. Arpaci, V. S.: “Conduction Heat Transfer,"’ Addison-Wesley Publishing Company, Ine., Reading, Mass., 1966. Orisik, M. N.: “Boundary Value Problems of Heat Conduction,”" International Textbook Company, Scranton, Pa., 1968. Langston, L. S.: Heat Transfer from Multidimensional Objects Using One-Dimen- sional Solutions for Heat Loss, Int. J. Heat Mass Transfer, vol. 25, p. 149, 1982. Colakyan M., R. Turton, and O. Levenspiel: Unsteady State Heat Transfer to ‘Variously Shaped Objects, Heat Transfer Engr., vol. 5, p. 82, 1984.PRINCIPLES OF CONVECTION 151 INTRODUCTION The preceding chapters have considered the mechanism and calculation of conduction heat transfer. Convection was considered only insofar as it related to the boundary conditions imposed on a conduction problem. We now wish to ‘examine the methods of calculating convection heat transfer and, in particular, the ways of predicting the value of the convection heat-transfer coefficient h. The subject of convection heat transfer requires an energy balance along with an analysis of the fluid dynamics of the problems concerned. Our discussion in this chapter will first consider some of the simple relations of fluid dynamics and boundary-layer analysis which are important for a basic understanding of convection heat transfer. Next, we shall impose an energy balance on the flow system and determine the influence of the flow on the temperature gradients in the fluid. Finally, having obtained a knowledge of the temperature distri- bution, the heat-transfer rate from a heated surface to a fluid which is forced over it may be determined. Our development in this chapter is primarily analytical in character and is concerned only with forced-convection flow systems, Subsequent chapters will present empirical relations for calculating forced-convection heat transfer and will also (reat the subjects of natural convection and boiling and condensation heat transfer. = 5.2 viscous FLOW Consider the flow over a flat plate as shown in Figs, 5-1 and 5-2. Beginning at the leading edge of the plate, a region develops where the influence of viscous forces is felt. These viscous forces are described in terms of a shear stress + between the fluid layers. If this stress is assumed to be proportional to the normal velocity gradient, we have the defining equation for the viscosity, 207208 Principles of convection Lamina eon “Turbulent > Laminar sablayer + Fig. 5-1 Sketch showing diferent boundary-layer flow regimes on a ft plate du Teese S-) ‘The constant of proportionality jis called the dynamic viscosity. A typical set of units is newton-seconds per square meter; however, many sets of units are used for the viscosity, and care must be taken to select the proper group which will be consistent with the formulation at hand. ‘The region of flow which develops from the leading edge of the plate in which the effects of viscosity are observed is called the boundary layer. Some arbitrary point is used to designate the y position where the boundary layer ends; this point is usually chosen as the y coordinate where the velocity becomes 99 percent of the free-stream value. Initially, the boundary-layer development is laminar, but at some critical distance from the leading edge, depending on the flow field and fluid properties. small disturbances in the flow begin to become amplified, and a transition process takes place until the flow becomes turbulent. The turbulent-flow region may be pictured as a random churning action with chunks of fluid moving to and fro in all directions. The transition from laminar to turbulent flow occurs when Fig. 2. Laminar velocty profile ona Ht)Viscous tow 208 where w. = free-stream velocity x = distance from leading edge » = plo = kinematic viscosity ‘This particular grouping of terms is called the Reynolds number, and is di mensionless if a consistent set of units is used for all the properties: it (52) v Re, Although the critical Reynolds number for transition on a flat plate is usually taken as 5 x 10° for most analytical purposes, the critical value in a practical situation is strongly dependent on the surface-roughness conditions and the “turbulence level” of the free stream. The normal range for the beginning of transition is between $x 10° and i0*, With very large disturbances present in the flow, transition may begin with Reynolds numbers as low as 10°, and for flows which are very free from fluctuations, it may not start until Re = 2 x 10° or more. In reality, the transition process is one which covers a range of Reynolds numbers, with transition being complete and with developed turbulent flow usually observed at Reynolds numbers twice the value at which transition began. The relative shapes for the velocity profiles in laminar and turbulent flow are indicated in Fig. 5-1. The laminar profile is approximately parabolic, while the turbulent profile has a portion near the wall which is very nearly linear. ‘This linear portion is said to be due to a laminar sublayer which hugs the surface very closely. Outside this sublayer the velocity profile is relatively flat in com- parison with the laminar profile ‘The physical mechanism of viscosity is one of momentum exchange. Con- sider the laminar-flow situation. Molecules may move from one lamina to an- other, carrying with them a momentum corresponding to the velocity of the flow. There is a net momentum transport from regions of high velocity to regions of low velocity, thus creating a force in the direction of the flow. This force is the viscous-shear stress which is calculated with Eq. (5-1) The rate at which the momentum transfer takes place is dependent on the rate at which the molecules move across the fluid layers. In a gas, the molecules would move about with some average speed proportional to the square root of the absolute temperature since. in the kinetic theory of gases, we identify temperature with the mean kinetic energy of a molecule. The faster the mol- ecules move, the more momentum they will transport. Hence we should expect the viscosity of a gas to be approximately proportional to the square root of temperature, and this expectation is corroborated fairly well by experiment. The viscosities of some typical fluids are given in Appendix A. In the turbulent-flow region distinct fluid layers are no longer observed, and we are forced to seek a somewhat different concept for viscous action. A qualitative picture of the turbulent-flow process may be obtained by imagining macroscopic chunks of fluid transporting energy and momentum instead of240 Principles of convection microscopic transport on the basis of individual molecules. Naturally, we should expect the larger mass of the macroscopic elements of fluid to transport more energy and momentum than the individual molecules, and we should also expect a larger viscous-shear force in turbulent flow than in laminar flow (and a larger thermal conductivity as well). This expectation is verified by experiment, and it is this larger viscous action in turbulent flow which causes the flat velocity profile indicated in Fig. 5-1 Consider the flow in a tube as shown in Fig. 5-3. A boundary layer develops at the entrance, as shown. Eventually the boundary layer fills the entire tube, and the flow is said to be fully developed. If the flow is laminar, a parabolic velocity profile is experienced, as shown in Fig. 5-3a. When the flow is tur- bulent, a somewhat blunter profile is observed, as in Fig. 5-36. In a tube, the Reynolds number is again used as a criterion for laminar and turbulent flow. For d Reg > 2300 53) v the flow is usually observed to be turbulent. Again, a range of Reynolds numbers for transition may be observed, de- pending on the pipe roughness and smoothness of the flow. The generally accepted range for transition is 2000 < Rey < 4000 Uniform Starting ents Laminar siblayee— Fig. 6:3 Velocty proiefor (a laminar flow in tube and (0) turbulent tube How.Invscid om 294 although laminar flow has been maintained up to Reynolds numbers of 25,000 in carefully controlled laboratory conditions. ‘The continuity relation for one-dimensional flow in a tube is Fit = Ply (5-4) ras i fo mean velsty We dfn the mas elo as \ Mass velosty = G = 3 = pn os) so that the Reynolds number may also be written Gd Rey = (5-6) # Equation (5-6) is sometimes more convenient to use than Eq. (5-3). 5-3 INVISCID FLOW ‘Although no real fluid is inviscid, in some instances the fluid may be treated ‘as such, and it is worthwhile to present some of the equations which apply in these circumstances. For example, in the flat-plate problem discussed above, the flow at a sufficiently large distance from the plate will behave as a nonvis- cous flow system. The reason for this behavior is that the velocity gradients normal to the flow direction are very small. and hence the viscous-shear forces are small If a balance of forces is made on an element of incompressible fluid and these forces are set equal to the change in momentum of the fluid element, the Bernoulli equation for flow along a streamline results: ei ple = const (5-7a) or, in differential form, (5-76) where p = fluid density _p = pressure at particular point in flow V = velocity of flow at that point ‘The Bernoulli equation is sometimes considered an energy equation because the V22g. term represents kinetic energy and the pressure represents potential energy; however, it must be remembered that these terms are derived on the242 Prnciles of convection basis of a dynamic analysis, so that the equation is fundamentally a dynamic equation. In fact, the concept of kinetic energy is based on a dynamic analysis, When the fluid is compressible, an energy equation must be written which will take into account changes in internal thermal energy of the system and the corresponding changes in temperature. For a one-dimensional flow system this equation is the steady-flow energy equation for a control volume, aa ita V+ O= bt oe Ve + Wk 68) where jis the enthalpy defined by e+ py (59) and where ¢ = internal energy Q = heat added to control volume Wk = net external work done in the process v = specific volume of fluid (The symbol i is used to denote the enthalpy instead of the customary h to avoid confusion with the heat-transfer coefficient.) The subscripts 1 and 2 refer to entrance and exit conditions to the control volume. To calculate pressure drop in compressible flow, itis necessary to specify the equation of state of the fluid, viz., for an ideal gas, pa pRT Be=c AT Ai= cy AT ‘The gas constant for a particular gas is given in terms of the universal gas constant 3 as a ne M where M is the molecular weight and 3 ~ 8314.5 J/kg - mol - K. For air, the appropriate ideal-gas properties are Ray = TIRE K pair = 1.005 Kk Ce, ay = 0.718 kik «°C To solve a particular problem, we must also specify the process. For example, reversible adiabatic flow through a nozzle yields the following familiar expres: sions relating the properties at some point in the flow to the Mach number and the stagnation properties, i.e, the properties where the velocity is zero:Invsed tow 213 where To. Po. po = stagnation properties 'y = ratio of specific heats cy/c, M = Mach number a=V (5-10) for an ideal gas.t For air behaving as an ideal gas this equation reduces to a = 20.045 VT mis (S11) where T is in degrees Kelvin. EXAMPLE 5-1 Water at 20°C flows at & kg/s through the diffuser arrangement shown in the accom- panying figure. The diameter at section | is 3.0 cm, and the diameter at section 2 is 7.0 em. Determine the increase in static pressure between sections ! and 2. Assume frictionless flow. Flow, solution 1069 x 10+ m* = 3.848 < 107? m* “The density of water at 20°C is 1000 kg/m, and so we may calculate the velocities from the mass-continuity relation “A 1 The isentropic Now formul are derived in Ref. 7, p. 629.244, Prncpes of conection 80 Tip007.069 10-5 8 TOONS. 848 x10) ‘The pressure difference is obtained from the Bernoulli equation (5-74): 11.32 mis (37.1 fs) = 2.079 m/s (6.82 tvs} BaP ys yy oR 28 an.» — eon = 61.91 kPa. [8.98 Ibvin? abs} EXAMPLE 5-2 Air at 300°C and 0.7 MPa pressure is expanded isentropically from a tank until the velocity is 300 m/s. Determine the static temperature, pressure, and Mach number of the air at the high-velocity condition. y= 1.4 for air. Solution We may write the steady-flow energy equation as ate because the initial velocity is small and the process is adiabatic. In terms of temperature, AT) = = G00) oxo 255.2°C = 282K [491.4°F] (1005300 ~ 73) h ‘We may calculate the pressure from the isentropic relation ee ‘$28.2)°% 7 p= 07 (82) = osnemrs63 mins ‘The velocity of sound at condition 2 is a; = (20,045)(528.2)"® = 460.7 mis (1511 fs} so that the Mach number is 300. _ a; ~ 4607 = 068!Carns coundary layer on a at pate 245 m1 5-4 LAMINAR BOUNDARY LAYER ON A FLAT PLATE. Consider the elemental control volume shown in Fig. 5-8, We derive the equa- tion of motion for the boundary layer by making a force-and-momentum balance on this element. To simplify the analysis we assume: 1. The fluid is incompressible and the flow is steady. 2. There are no pressure variations in the direction perpendicular to the plate. 3. The viscosity is constant, 4. Viscous-shear forces in the y direction are negligible. We apply Newton's second law of motion, dim), r= = a ‘The above form of Newton's second law of motion applies to a system of constant mass. In fluid dynamics it is not usually convenient to work with elements of mass: rather, we deal with elemental control volumes such as that shown in Fig, 5-4, where mass may flow in or out of the different sides of the Fig. 54. Elemental conto! volume for fore balance on laminar boundary layer246. Princpies of connection volume, which is fixed in space. For this system the force balance is then written & F, = increase in momentum flux in x direction ‘The momentum flux in the x direction‘is the product of the mass flow through 4 particular side of the control volume and the x component of velocity at that point. ‘The mass entering the left face of the element per unit time is pu dy if we assume unit depth in the z direction. Thus the momentum entering the left face per unit time is pu dy u = pu dy The mass flow leaving the right face is : wea oles ar)o au\? e (« +S)o ‘The mass flow entering the bottom face is pv de and the mass flow leaving the top face is ° (- + =a) dx A mass balance on the element yields pty + pw ds = 9 (u + Bde) dy + p(y + May) de au, ar * ay This is the mass continuity equation for the boundary layer. Returning to the momentum-and-force analysis, the momentum in the x direction which enters the bottom face is or 0 (5-12) pvu dx and the momentum in the x direction which leaves the top face is av ou o(v tzu) (« +a) aeaminar tounsuy layer on att piste 247 We are interested only in the momentum in the x direction because the forces considered in the analysis are those in the x direction. These forces are those due to viscous shear and the pressure forces on the element. The pressure force on the left face is p dy, and that on the right is (p+ (p/ax) dx] dy, so that the net pressure force in the direction of motion is ap x dx dy ‘The viscous-shear force on the bottom face is ~ hax and the shear force on the top is au, 8 (au nde [s ae ‘The net viscous-shear force in the direction of motion is the sum of the above: #u Net viscous-shear force = aS dx dy a Equating the sum of the viscous-shear and pressure forces to the net momentum twansfer in the x direction, we have au » Z au) : wt gcay ~ Baca =o (u-+ Man) dy pea co (o+ Bay) (w+ Hap) ae ~ pra Clearing terms, making use of the continuity relation (5-12) and neglecting second-order differentials, gives iu aw Fu ap p (ume —-2 5-13) , ( ax *) Mae” ax Ca) ‘This is the momentum equation of the laminar boundary layer with constant properties. The equation may be solved exactly for many boundary conditions. and the reader is referred to the treatise by Schlichting {1] for detaits of the various methods employed in the solutions. In Appendix B we have included the classical method for obtaining an exact solution to Eq, (5-13) for laminar flow over a flat plate. For the development in this chapter we shall be satisfied ‘with an approximate analysis which furnishes an easier solution without a loss in physical understanding of the processes involved. The approximate method is due to von Karman [2] Consider the boundary-layer flow system shown in Fig. 5-5. The free-stream velocity outside the boundary layer is w., and the boundary-layer thickness is248. Prncipes of convection Fig. 55. Elementalcontolvol ume for integral mornentum an: a aiysis of laminar boundary r 2 layer 3. We wish to make a momentum-and-force balance on the control volume bounded by the planes 1, 2, A-A, and the solid wall. The velocity components. normal to the wall are neglected, and only those in the x direction are consid- ered, We assume that the control volume is sufficiently high that it always encloses the boundary layer; that is, H > 6. ‘The mass flow through plane 1 is J" om ay «@ and the momentum flow through plane | is i oe ay o ‘The momentum flow through plane 2 is dy a UL ont fi we ty + 2 (fmt ay) de io) and the mass flow through plane 2 is "i apy J oe ay + (at) @ Considering the conservation of mass and the fact that no mass can enter the control volume through the solid wall, the additionat mass flow ia expression, (d) over that in (a) must enter through plane A-A. This mass flow carries with it a momentum in the x direction equal to E(t)Laminar boundary layer on a fa pate 299 ‘The net momentum flow out of the control volume is therefore al A (fl ne ay) ds — ag (foe) a This expression may be put in a somewhat more useful form by recalling the product formula from the differential calculus: did) = ndé + odn or nd = dnd) ~ ddn In the momentum expression given above, the integral [Rem ay is the @ function and w,. is the » function. Thus wb (Lime) ee fm) (fms) (fone) 1-3 (fn) ‘The us may be placed inside the integral since it is not a function of y and thus may be treated as a constant insofar as an integrabwith respect to y is concerned. ‘Returning to the analysis, the force on plane 1 is the pressure force pH and that on plane 2 is [p + (dp/dx)dx]H. The shear force at the wall is (S14) au ated = —p ae] ‘There is no shear force at plane A-A since the velocity gradient is zero outside the boundary layer. Setting the forces on the element equal to the net increase in momentum and collecting terms gives ® 4 phy du. Bye ok flu ana He fds 68 ‘This is the integral momentum equation of the boundary layer. If the pressure is constant throughout the flow, dp ax ince the pressure and free-stream velocity are related by the Bernoulli equa- tion. For the constant-pressure condition the integral boundary-layer equation becomes (5-17) od (Pu. ~ andy220 Principio of convecton The upper limit on the integral has been changed to & because the integrand is zero for y > 8 since u = us for y > 6. Ifthe velocity profile were known, the appropriate function could be inserted in Eq. (5-17) to obtain an expression for the boundary-layer thickness. For our approximate analysis we first write down some conditions which the velocity function must satis u=0 aty=0 (a) usu aty=6 ) au ro ays © For a constant-pressure condition Eq. (5-13) yields Feng myno @ ay" since the velocities w and v are zero at y = 0. We assume that the velocity Profiles at various x positions are similar; i.e., they have the same functional dependence on the y coordinate, There are four conditions to satisfy, The simplest function which we can choose to satisfy these conditions is a poly. ‘nomial with four arbitrary constants. Thus = Ct Cyt Cy + Cy? (5-18) Applying the four conditions (a) to (d), T/y\* a-}# 3(3) (5-19) Inserting the expression for the velocity into Eq. (5-17) gives 4 fue {* [321 (x) ][, - =a ale 018-30) ] ~ 6S. 3 le Carrying out the integration leads to Since p and u. are;constants, the variables may be separated to give Woy) | Woe Bde TS pe 8 i F140 ox and eetLaminar bowery layer on aa pile 4 Atx = 0,5 = 0, so that 6 = 4.64 i (3-20) cs ‘This may be written in terms of the Reynolds number as b_ Ass x Rey? 6-21) where Re, = y ‘The exact solution of the boundary-layer equations as given in Appendix B yields a (5-210) EXAMPLE 5:3 Air at 27°C and 1 atm flows over a lat plate ata speed of 2 m/s. Calculate the boundary- ayer thickness at distances of 20 and 40 cm from the leading edge of the plate. Calculate the mass flow which enters the boundary layer between x = 20 em and x = 40 cm. ‘The viseosity of airat 27°C is 1,85 10. “kg/m - s, Assume unit depth inthe z direction. Solution The density of air is calculated from p_ 1.0132 « 10° ?* RT (28700) 1.177 kg/m’ (0.073 tb, ft) ‘The Reynolds number is calculated as LATH2.000.21 ty = 206m = 208 Ary = 20 185 x 10> “0 .17792.010.4) x = 40 cm e = HATIAWO) 4 55, atx = 40 Re = ETE 55.160 The boundary-layer thickness is calculated from Eq. (5-21) tx em 8 = SPIES. 9.00559 m in| A Pegs = 0.00889 m_ [0.24 i] (4.64N0.4) Atx=qocm 6 = SOOR_ m (04 in 0 ST © 0.0079 m (04 in] Toc msn whet he oun ete sabe [mo‘222 Principe of comecton \ By dt @)] 7 eee oe fm [B5-40)]o fo eee Am = 4 puslba ~ 630) (he = ((1.177)2.0(0.0079 — 0.00559) 5-5 ENERGY EQUATION OF THE BOUNDARY LAYER ‘The foregoing analysis considered the fluid dynamucs of a laminar-boundary- layer flow system. We shall now develop the energy equation for this system and then proceed to an integral method of solution. Consider the elemental control volume shown in Fig. 5-6. To simplify the analysis we assume t nail at Fg. 66 Elemental volume for energy analy of laminar boundary layer ‘Energy equabon ofthe bouncary ayer 223 1. Incompressible steady flow 2, Constant viscosity, thermal conductivity, and specific heat 3. Negligible heat conduction in the direction of flow (x direction) ‘Then, for the element shown, the energy balance may be written Energy convected in left face + energy convected in bottom face "> heat conducted in bottom face + net viscous work done on element = energy convected out right face + energy convected out top face + heat conducted out top face ‘The convective and conduction energy quantities are indicated in Fig, 5- and the energy term for the viscous work may be derived as follows. The viscous work may be computed as a product of the net viscous-shear force ‘and the distance this force moves in unit time, The viscous-shear force is the product of the shear-stress and the area dx, and the distance through which it moves per unit time in respect to the elemental control volume dx dy is, au Say an so that the net viscous energy delivered to the element is »(e) a0 ‘Writing the energy balance corresponding to the quantities shown in Fig. 5-6, assuming unit depth in the z direction, and neglecting second-order differentials yields aT aT po [ote viEer(t Using the continuity relation and dividing by pc, gives # wy +#(% (5-22) Pe, & ‘This is the energy equation of the laminar boundary layer. The left side rep- resents the net transport of energy into the control volume, and the right side224 Prncipies of convection Fepresents the sum of the net heat conducted out of the control volume and the net viscous work done on the element. The viscous-work term is of im- portance only at high velocities since its magnitude will be small compared with the other terms when low-velocity flow is studied. This may be shown with an order-of-magnitude analysis of the two terms on the right side of Eq. (5-22). For this order-of-magnitude analysis we might consider the velocity as having the order of the free-stream velocity u. and the y dimension of the order of 8. Thus sothat “te # (x) ~~ pep \ay) ~ pe If the ratio of these quantities is small, that is, ey (5-23) then the viscous dissipation is small in comparison with the conduction term. Let us rearrange Eq. (5-23) by introducing where Pris called the Prandtl number, which we shall discuss later. Equation (5-23) becomes prver (5-24) As an example, consider the flow of air at 4 = m/s T= 20°C = 293K p= iam For these conditions c, = 1005 J/kg - °C and Pr = 0.7 so that a? 0.2170)? eet (1005293) indicating that the viscous dissipation is small, Thus, for low-velocity incom- Pressible flow, we have = 0.012 < 1.0 (5-25) In reality, our derivation of the energy equation has been a simplified one, and several terms have been left out of the analysis because they are small in comparison with others. In this way we arrive at the boundary-layer approx-‘he nermal boundary ayer 228 imation immediately, without resorting to a cumbersome elimination process to obtain the final simplified relation. The general derivation of the boundary- layer energy equation is very involved and quite beyond the scope of our discussion. The interested reader should consult the books by Schlichting [1] and White [5] for more information. ‘There is a striking similarity between Eq. (5-25) and the momentum equation for constant pressure (5-26) The solution to the two equations will have exactly the same form when a = ¥. Thus we should expect that the relative magnitudes of the thermal diffusivity and kinematic viscosity would have an important influence on convection heat transfer since these magnitudes relate the velocity distribution to the temper- ature distribution. This is exactly the case, and we shall see the role which these parameters play in the subsequent discussion. 5-6 THE THERMAL BOUNDARY LAYER Just as the hydrodynamic boundary layer was defined as that region of the flow where viscous forces are felt, a thermal boundary layer may be defined as that region where temperature gradients are present in the low. These temperature gradients would result from a heat-exchange process between the fluid and the wall. Consider the system shown in Fig. 5-7. The temperature of the wall is T.. the temperature of the fluid outside the thermal boundary layer is T., and the thickness of the thermal boundary layer is designated as 8,. At the wall, the velocity is zero. and the heat transfer into the fluid takes place by conduction. Thus the local heat flux per unit area, q”. is ($2 Fig. &7 Temperature profile in the thermal boundary layer238 Prnciles of conection From Newton's law of cooling (Ea. (1-8)), q = WT. - 7) (5-28) where h is the convection heat-transfer coefficient. Combining these equations, we have =KTIay aan : T. — Te (5.29) so that we need only find the temperature gradient at the wall in order to evaluate the heat-transfer coefficient. This means that we must obtain an expression for the temperature distribution. To do this, an approach similar to that used in the momentum analysis of the boundary layer is followed. The conditions which the temperature distribution must satisfy are T=T, aty=0 @ ar a at ) > y T=T. ay=5, © and by writing Eq. (5-25) at y = 0 with no viscous heating we find Seo ay=0 @ oF since the velocities must be zero at the wall Conditions (a) to (d) may be fitted to a cubic polynomial as in the case of the velocity profile, so that @_T-T. _3y $(2) T= T. 28, 2\8, where @ = T — T... There now remains the problem of finding an expression for 6, the thermal-boundary-layer thickness. This may be obtained by an i tegral analysis of the energy equation for the boundary layer. Consider the control volume bounded by the planes 1, 2, A-A, and the wall as shown in Fig. 5-8. It is assumed that the thermal boundary layer is thinner than the hydrodynamic boundary layer, as shown. The wall temperature is T,, the free-stream temperature is T., and the heat given up to the fluid over the length dx is dq. We wish to make the energy balance 5-30) Energy convected in + viscous work within element + heat transfer at wall = energy convected out (5-31) ‘The energy convected in through plane 1 is [ure“Te heal boundary ayer 227 ig. 58 Control volume for ic9ra energy analysis laminar boundary tow ‘and the energy convected out through plane 2 is wo (fara) + (oy fat ds) The mass flow through plane A-A is d (meds) a and this carries with it an energy equal to ot Ef mrt) The net viscous work done within the element is (du el L(G) afer and the heat transfer at the wall is ty = hale aI Combining these energy quantities according to Eq. (5-81) and collecting terms EUG) a] eR] oe This is the integral energy equation of the boundary layer for constant properties and constant free-stream temperature 7. To calculate the heat transfer at the wall, we need to derive an expression for the thermal-boundary-layer thickness which may be used in conjunction ra ae228 recipies of convection with Eqs. (5-29) and (5-30) to determine the heat-transfer coefficient, For now, we neglect the viscous-dissipation term; this term is very small unless the velocity of the flow field becomes very large, and the calculation of high-velocity heat transfer will be considered later. The plate under consideration need not be heated over its entire length. The situation which we shall analyze is shown in Fig. 5-9, where the hydrodynamic boundary layer develops from the leading edge of the plate, while heating does hot begin until x = Xo Inserting the temperature distribution Eq. (5-30) and the velocity distribution Eq. (5-19) into Eq, (5-32) and neglecting the viscous-dissipation term, gives afyn afen al r. - ua} 4If @.- oud] = oun Af ("fy 3% 1 (x) P32 102? = tan $f f [! 25,7 3(3) 26 3(3) ty a) _ 3b. dy }y-0 28, Let us assume that the thermal boundary layer is thinner than the hydrodynamic boundary layer. Then we only need to carry out the integration to y = 6, since the integrand is zero for y > 5. Performing the necessary algebraic manipu- lation, carrying out the integration, and making the substitution { = 8/6 yields ab. 4[;(3p-3;0 tae [6 Ge soe) | = 3 Because 8, <8, <1, and the term involving {* is small compared with the term, we neglect the ¢* term and write (5-33) 0. 3, Ley 3 3p Oe GO) = 5 (534) Performing the differentiation gives — a | Fig. $8 Hyarosynamc and 4 tmermat bouncary layers on a 5 fat plate Heating stars atx =The thermal boundary layar 229 But and so that we have &_ Ba pra d- 2s (535) Noting that cea Pa tae we see that Eq, (5-35) is a linear differential equation of the first order in °, and the solution is, When the boundary condition 6, =0 atx=%y is applied, the final solution becomes & 1pm, (2)"]" dl [: (*) ] (5:36) where pr=2 (37) has been introduced. The ratio v/a is called the Prandtl number after Ludwig Prandtl, the German scientist who introduced the concepts of boundary-layer theory When the plate is heated over the entire length, x» ~ 0, and eee 38> Tom? (5-38) In the foregoing analysis the assumption was made that <1. This as- sumption is satisfactory for fluids having Prandtl numbers greater than about230 Principles of convection 0.7, Fortunately, most gases and liquids fall within this category. Liquid metals are a notable exception, however, since they have Prandtl numbers of the order of 0.01. ‘The Prandtl number via has been found to be the parameter which relates the relative thicknesses of the hydrodynamtic and thermal boundary layers. The kinematic viscosity of a fluid conveys information about the rate at which momentum may diffuse through the fluid because of molecular motion. The thermal diffusivity tells us the same thing in regard to the diffusion of heat in the fluid. Thus the ratio of these two quantities should express the relative magnitudes of diffusion of momentum and heat in the fluid. But these diffusion rates are precisely the quantities that determine how thick the boundary layers will be for a given external flow field; large diffusivities mean that the viscous or temperature influence is felt farther out in the flow field. The Prandtl number is thus the connecting link between the velocity field and the temperature field The Prandtl number is dimensionless when a consistent set of units is used. y _ nlp. 2 Set a kipc,k (5-39) In the SI system a typical set of units for the parameters would be «in kilograms per second per meter, cp in kilojoules per kilogram per Celsius degree, and k in kilowatts per meter per Celsius degree. In the English system one would typically employ in pound mass per hour per foot. c, in Btu per pound mass per Fahrenheit degree, and k in Btu per hour per foot per Fahrenheit degree Returning now to the analysis, we have ~keTiay), T, ~ Te h (5-40) Substituting for the hydrodynamic-boundary-layer thickness from Eq. (5-21) and using Eq. (5-36) gives vee (2) [-("]" om The equation may be nondimensionalized by multiplying both sides by x/k, producing the dimensionless group on the left side, hex Nu, = (5-42) called the Nusselt number after Wilhelm Nusselt, who made significant con- tributions to the theory of convection heat transfer. Finally, Nu, = 0.332 Pr? Rel? [) - (2) ] (5-43) he or, for the plate heated over its entire length, x» = 0 and“he thermal boundary layer 234 Nu, = 0.332 Pr"? Re!? (5-44) Equations (5-41), (5-43), and (5-44) express the local values of the heat-transfer coefficient in terms of the distance from the leading edge of the plate and the fluid properties. For the case where xo = 0 the average heat-transfer coefficient and Nusselt number may be obtained by integrating over the length of the plate: (5-45) Nu, (5-46a) wy ik 2 or Nu, e 0.664 Re,"? Pr'® (5-466) where Re, = Me rs The reader should carry out the integrations to verify these results. ‘The foregoing analysis was based on the assumption that the fluid properties were constant throughout the flow. When there is an appreciable variation between wall and free-stream conditions, itis recommended that the properties be evaluated at the so-called film temperature T,, defined as the arithmetic ‘mean between the wall and free-stream temperature, Ty + To 1 6-47) Anexact solution to the energy equation is given in Appendix B. The results of the exact analysis are the same as those of the approximate analysis given above. Constant Hi The above analysis has considered the laminar heat transfer from an isothermal surface. In many practical problems the surface heat flux is essentially constant, and the objective is to find the distribution of the plate-surface temperature for given fluid-flow conditions. For the constant-heat-flux case it can be shown that the local Nusselt number is given by hx. Nu, = 7 = 0.453 Ret? Pri® (5-48) which may be expressed in terms of the wall heat flux and temperature differ- ence as it Flux1232 Principles of convection Nu, (5-49) ST, = To) The average temperature difference along the plate, for the constant heat flux condition, may be obtained by performing the integration rT ape al ff a thm - meer Ae aubik * 0.6795 Rel? Pri? (5-50) In these equations q.. is the heat flux per unit area and will have the units of watts per square meter (W/m?) in SI units or British thermal units per hour per square foot (Btwh - f°) in the English system. Other Relations Equation (5-44) is applicable to fluids having Prandtl numbers between about 0.6 and 50. It would not apply to fluids with very low Prandtl numbers like liquid metals or to high-Prandtl-number fluids like heavy oils or silicones. For fa very wide range of Prandtl numbers, Churchill and Ozoe {9] have correlated a large amount of data to give the following relation for laminar flow on an isothermal flat plate: _0.3387 Rel? Pri” 0.0468)" 1+ Gre) | For the constant-heat-flux case, 0.3387 is chamged to 0.4637 and 0.0868 is changed to 0.0207. Properties are still evaluated at the film temperature. Nu, for Re, Pr > 100 (51) EXAMPLE 5-4 For the low system in Example 5-3 assume that the plate is heated over its entre length toa temperature of 60°C, Calculate the heat transferred in (a) the first 20 cm of the plate and (h) the first 0). em of the plate Solution The toa het transfer over cera length ofthe plat is desired: wo we wih 10 cree ee RSucwtntercociente For ths purpose we use Eqs. 88) and {Ba eating the papers tthe lm copra = 228 ease nesk Wose From Appendit A the properties are 17.36 x 10° m/s [1.87 x 10°¢ ft/s} f= 0.0219 Wn °C. (0.0199 Blu f°)‘The ermal soundary ayer 233 07 1.006 Ki/kg + °C [0.24 Buu, -*F] Atx = 20m we @X0.2) 7 1736 x 196 = OHH het BES 0.392 Rey? Pr? ne = 0.332 Rel = (0.3329(23,041)"" (0.7)! = 44.74 a () (44.74y(0.02749) : 02 = 6.15 Wim? -*C (1.083 Bau ft °F “The average value of the heat-transfer coefficient is twice this value, or B= Q\6.1S) = 12.3 Wim? -*C (2.17 Bush - f+ °F) ‘The heat flow is a= hat, - 7) If we assume unit depth in the z direction, q = (12.340.2460 ~ 27 81.18 W (277 Buh} ux v0.4) v ~ 736 x 10% (0.332)(46,082)"* (0.7)"° = 63.28 (63.28)0.02749) _ ae AO SETPD 4.349 Wim? °C (2)64,349) = 8.698 Wit «°C (1.53 Beulh - £2 -*F] 4 = (8.698)0.4)(60 — 27) = 114.8 W (392 Btu/h) = 46,082 ss EXAMPLE 5-5 ‘A 1.0-KW heater is constructed of a glass plate with an electrically conducting film ‘which produces a constant heat flux. The plate is 60 by 60 em and placed in an airstream at 27°C, 1 atm with we = 5 mvs, Calculate the average temperature difference along the plate and the temperature difference at the trailing edge. Solution Properties should be evaluated at the film temperature, but we do not know the plate Temperature so for an initial calculation we take the properties at the free-stream con- ditions of, tr v= 15.69 x 10-* mis234 Principles of convection 0.6K5) 15.69 x or ~ 191 * 1° Re From Eq. (5-50) the average temperature difference is = —11000/0.6)2K0.60.02624 ‘O16795(1.91 x 10577(0.708)" ~ Now, we go back and evaluate properties at 240 +27 2 Te 240°C T, 133.5°C = 406.5 K 6.65 x 10" mis Pr = 0.687 k 0.45 Hees x I —_10010.6)210.6/0 0384 Tar9SE.13 x 100.687 ‘At the end of the plate (x = L = 0.6m) the temperature difference is obtained from Eqs. (5-48) and (5-50) with the constant 0.453 to give (240.390.6795) Oy ~ Tahoe = SST = 30S ‘An alternate solution would be to base the Nusselt number on Eq. (5-51). 0344 Wim °C LAB 10 240,3°C EXAMPLE 56 Engine oil at 20°C is forced over a 20-cm-square plate at a velocity of 1.2 mls. The plate is heated to a uniform temperature of 60°C. Calculate the heat lost by the plate Solution We first evaluate the film temperature: + Z 2 arc ‘The properties of engine oil are p= 816 kp/m? = 0.00024 mis k= 0.146 Wim °C Pr = 2870 The Reynolds number is we, _ (1.20002) v ~ 0,00024 1000 Because the Prandtl number is so large we will employ Eq, (5-51) for the solution, We see that A varies with x in the same fashion as in Eq. (5-44), ie., hy © x7", so that we get the same solution as in Eq. (5-45) for the average heat-iransfer coefficient. Evaluating Eq. (551) at x = 0.2 givesio, = $0:3387H1000)287097 _ 455 5 +(e 2 2870 wl (152.200.148) 399 6 Wren? 02 The average value of the convection coefficient is = (2X1096) = 219.2 Wim? °C so that the (otal heat transfer is = HALT, ~ To) = (219.240.2360 ~ 20) = 350.6 W 5-7 THE RELATION BETWEEN FLUID FRICTION AND HEAT TRANSFER We have already seen that the temperature and flow fields are related. Now we seek an expression whereby the frictional resistance may be directly related to heat transfer. The shear stress at the wall may be expressed in terms of a friction coefficient, & n= OS (5-52) Equation (5-52) is the defining equation for the friction coefficient. The shear stress may also be calculated from the relation aw mwa Bae Using the velocity distribution given by Eq. (5-19), we have = 3 Hue ™ S28 and making use of the relation for the boundary-layer thickness gives 3 Mite (de)? 5. w= 38s (4) 639 Combining Eqs. (5-52) and (5-53) leads to a 2 3 pte (ue\\"® 1 te - a 54) a (2) pad 7 0323 Rex (5-54) Equation (5-44) may be rewritten in the following form: Die a Re, Pr pepita = 0.332 Pr-®* Rey? The group on the left is called the Stanton number,298 Procines of convecton hy St. = Pep so that St, Pr2? = 0.332 Re>!® (5-55) Upon comparing Eqs. (5-54) and (5-55), we note that the right sides are alike except for a difference of about 3 percent in the constant, which is the result of the approximate nature of the integral boundary-layer analysis. We recog this approximation and write st, pes = Ge (5-56) Equation (5-56), called the Reynolds-Colburn analogy, expresses the relation between fluid friction and heat transfer for laminar flow on a flat plate. The heat-transfer coefficient thus could be determined by making measurements of the frictional drag on a plate under conditions in which no heat transfer is involved It turns out that Eq. (5-56) can also be applied to turbulent flow over a flat plate and in a modified way to turbulent flow in a tube. It does not apply to laminar tube flow. In general, a more rigorous treatment of the governing ‘equations is necessary when embarking on new applications of the heat-trans- fer-fluid-friction analogy, and the results do not always take the simple form of Eq. (5-56). The interested reader may consult the references at the end of the chapter for more information on this important subject. At this point, the simple analogy developed above has served to amplify ouf understanding of the physical processes in convection and to reinforce the notion that heat- transfer and viscous-transport processes are related at both the microscopic and macroscopic levels. EXAMPLE 57 For the flow system in Example 5-4 compute the drag force exerted on the first 40.em. Of the plate using the analogy between fluid friction and heat transfer Solution We use Eg. (5-56) t0 compute the friction coefficient and then calculate the drag force. ‘An average friction coefficient is desired, so Sipe w The density at 316.5 K is 1.0132 x 10! AON X10 1115 kgm (287)(316.5) as For the 40-em length St =< a = 3.88 x 10°" Fras ~ OATS O06Trout toundary ayer test vans 237 Then from Eq. (a) = (3.88 10-5H0.70 = 3.06 > 10°? The average shear stress at the wall is computed from Eq. (5-52); 13.06% 10 LEI 1.0136 Nim? The drag force is the product of this shear stress and the area, D (0.013604) = $.44 mN [1.23 5 10-71) 5-8 TURBULENT-BOUNDARY-LAYER MEAT TRANSFER Consider a portion of a turbulent boundary layer as shown in Fig. $-10. A very thin region near the plate surface has a laminar character. and the viscous action and heat transfer (ake place under circumstances like those in laminar fiow. Farther out, at larger y distances from the plate, some turbulent action is experienced, but the molecular viscous action and heat conduction are still important. This region is called the buffer laver. Still farther out, the flow is fully turbulent, and the main momentum- and heat-exchange mechanism is one involving macroscopic lumps of fluid moving about in the flow. In this fully turbulent region we speak of eddy viscosity and eddy thermal conductivity. These eddy properties may be 10 times as large as the molecular values. ‘The physical mechanism of heat transfer in turbulent flow is quite similar to that in laminar flow; the primary difference is that one must deal with the eddy properties instead of the ordinary thermal conductivity and viscosity. The main difficulty in an analytical treatment is that these eddy properties vary across the boundary laver, and the specific variation can be determined only from experimental data, This is an important point, All analyses of turbulent Turbuent Baler fIovet | sinae Fig. 510. Velocity protien turbulent oundary layer on 8 Hat plateflow must eventually rely on experimental data because there is no completely adequate theory to predict turbulent-flow behavior. If one observes the instantaneous macroscopic velocity in a turbulent-flow system, as measured with a laser anemometer or other sensitive device, sig- nificant fluctuations about the mean flow velocity are observed as indicated in Fig. 5-11, where w is designated as the mean velocity and u’ is the fluctuation: from the mean. The instantaneous velocity is therefore usdew (557) ‘The mean value of the fluctuation u’ must be zero over an extended period for steady flow conditions. There are also fluctuations in the y component of ve- locity, so we would write + (5-58) The fluctuations give rise to a turbulent-shear stress which may be analyzed by referring to Fig. 5-12 For a unit area of the plane P-P, the instantaneous turbulent mass-transport rate across the plane is pv’. Associated with this mass transport is a change in the x component of velocity u’. The net momentum flux per unit area, in the x direction, represents the turbulent-shear stress at the plane P-1 a ‘When a turbulent lump moves upward (v' > 0). it enters a region of higher and is therefore likely to effect a slowing-down fluctuation in u’, that is, u’ <0. ‘A similar argument can be made for v’ < 0, so that the average turbulent-shear stress will be given as n= pw (5-59) W = 0, the average of the fluctuation ‘We must note that even though 7” product wv" is not zero. | a Fig. 5-11. Turbulent fuctuations win ome. Tine,urulentooundary layer hea! wansler 238 Fig. 5-12 Turtulentshearsvess land mixing length Eddy Viscosity and the Mixing Length Let us define an eddy viscosity or eddy diffusivity for momentum ey such that _, du 1 BW = peu (5-60) We have already likened the macroscopic transport of heat and momentum in turbulent flow to their molecular counterparts in laminar flow, so the definition in Eq, (5-60) is a natural consequence of this analogy. To analyze molecular- transport problems (see. for example, Ref. 7, p. 369) one normally introduces the concept of meun free path, or the average distance a particle travels between collisions. Prandtl introduced a similar concept for describing turbulent-flow phenomena. The Prandtl mixing length is the distance traveled, on the average, by the turbulent lumps of fluid in a direction normal to the mean flow. Let us imagine a turbulent lump which is located a distance ¢ above or below the plane P-P. as shown in Fig. 5-12. These lumps of fluid move back and forth across the plane and give rise to the eddy o turbulent-shear-stress effect. At ¥ + Cthe velocity would be approximately au uy + O= ayy + ee av) + OS while at y ~ € au my ~ O~ uy ~ OF Prandi! postulated that the turbulent fluctuation u/ of the above (wo quantities, or proportional to the mean wae (5-61)The distance ¢ is called the Prandtl mixing length. Prandtl also postulated that v' would be of the same order of magnitude as u’ so that the turbulent-shear stress of Eq. (5-60) could be written (5-62) (3) (5.63) We have already noted that the eddy properties, and hence the mixing length, vary markedly through the boundary layer. Many analysis techniques have been applied over the years to take this variation into account, Prandti’s hy- pothesis was that the mixing length is proportional to distance from the wall. or 6 = ky (5-64) where K is the proportionality constant. The additional assumption was made that in the near-wall region the shear stress is approximately constant so that 1, ~ Tox When this assumption is used along with Eq. (5-64). Eq. (5-62) yields - (2) ‘Taking the square root and integrating with respect to y gives KY» where C is the constant of integration. Equation (5-65) matches very well with experimental data except in the region very close to the wall. where the faminar sublageris present. In the sublayer the velocity distribution is essentially linear. Let us now quantify our earlier qualitative description of a turbulent bound: ary layer by expressing the shear stress as the sum of a molecular and turbulent part: Iny + 5-65) au = Ota (5-66) a The so-called universal velocity profile is obtained by introducing two nondi- mensional coordinates ut (5-67) Vreb Ve y = ee 5-68)cen. poundaryayer heat vars 244 Using these parameters and assuming + = constant, we can rewrite Eq, (5:66) dy du (5-69) T+ ewe In terms of our previous qualitative discussion. the laminar sublayer is the region where ey ~ 0, the buffer layer has ey ~ ¥, and the turbulent kayer has eu, > ». Therefore, taking éy = 0 in Eq, (5-69) and integrating yields At the wall. w° = Ofory™ = so that ¢ = O and y (5-70) is the velocity relation (a linear one) for the laminar sublayer. In the fully turbulent region ey” > 1. From Eq. (5-65) aut feel a KVpy Substituting this relation along with Eq. (5-64) into Eq. (5-63) gives wee wf or ky" 71) ‘Substituting this relation in Eq. (5-69) for ey/v ® | and integrating gives l ui =ainyt te (5.7) This same form of equation will also be obtained for the buffer region. The limits of each region are obtained by comparing the above equations with experimental velocity measurements, with the following generally accepted constants: Laminar sublayer: 0 (5-75) ‘Turbulent Heat Transfer Based on Fluid-Friction Analogy Various analyses, similar to the one for the universal velocity profile above, have been performed to predict turbulent-boundary-layer heat transfer. The analyses have met with good success, but for our purposes the Colburn analogy between fluid friction and heat transfer is easier to apply and yields results, which are in agreement with experiment and of simpler form. In the turbulent-flow region, where ey, > v and € > a, we define the tur bulent Prandtl number as : (5-76) If we can expect that the eddy momentum and energy transport will both be increased in the same proportion compared with their molecular values, we might anticipate that heat-transfer coefficients can be calculated by Eq. (5-56) with the ordinary molecular Prandtl number used in the computation. It turns ‘out that the assumption that Pr, = Pr is a good one because heat-transfer cal- culations based on the fluid-friction analogy match experimental data very well. For this calculation we need experimental values of C, for turbulent flow. Schlichting [1] has surveyed experimental measurements of friction coeffi- cients for turbulent flow on flat plates. We present the results of that survey so that they may be employed in the calculation of turbulent heat transfer with the fluid-friction-heat-transfer analogy. The local skin-friction coefficient is iven by Cr for Reynolds numbers between 5 x 10' and 10°. At higher Reynolds numbers from 10” to 10? the formula of Schultz-Grunow (8] is recommended: Ch = 0.370 (log Re,)-2 (5-78) 0.0592 Re; "s (S-77) The average-friction coefficient for a flat plate with a laminar boundary layer up to Rey and turbulent thereafter can be calculated from 0455 AL flog Rey? ~ Re, G Recu < Rex < 10° 65-79)‘urbulet ocundary yer hoa rarer 24 where the constant A depends on Ren, in accordance with Table 5-1. A some- what simpler formula can be obtained for lower Reynolds numbers as oor A a = Rec = Rey < 10" (5-80) C= Rel® ~ Rex Equations (5-79) and (5-80) are in agreement within their common range of applicability. and the one to be used in practice will depend on computational ‘Applying the fluid-friction analogy St Pr = C,/2, we obtain the local tur: bulent heat transfer as: St, Pre = 0.0296 Rev! Sx 10" < Re, < 107 or St, Pre? = O.18S(log Rey 7% 107 = Re, < 10" The average heat transfer over the entire laminar-turbulent boundary layer is Sipe = S83) For Regu = 5 * 10 and Re, < 107. £4, (5-80) can be used to obtain Pr} = 0,037 Re; "> — 850 Re, ' (5-84) Recalling that Si = NUiiRe, Pr). we can rewrite Eq. (S-84) as Nu, = ME = pet o.0y7 Ret* ~ 850) (5-85) For Reynolds numbers above 10°, Eq. (5-83) can be used in combination with C, obtained from Eq. (5-79). The reader Should note that if transition Reynolds number different from 500,000 is chosen. then Eqs. (5-84) and (5-85) must be Changed accordingly. An alternative equation is suggested by Whitaker [10] which may give better results with some liquids because of the viscosity-ratio term: Wi, = 0.036 Pr (Re?* ~ 9200) (4) (5-86) me) The viscosity-ratio term is dropped when the equation is used for gases. All properties except yt. and ys, are evaluated at the film temperature. Table 5-1 4 1050 100 3300 S700244 Procines of convection Constant Heat Flux For constant wall heat flux in turbulent flow it is shown in Ref. 12 that the local Nusselt number is only about 4 percent higher than for the isothermal surface: i.e., Nu, = 1.04 wo, fre = come (587) Some more comprehensive methods of correlating turbulent-boundary-layer heat transfer are given by Churchill [11] EXAMPLE 5.6 Air at 20°C and 1 atm flows over a flat plate at 35 m/s. The plate is 75 cm long and is maintained at 60°C. Assuming unit depth in the : direction, calculate the heat transfer from the plate. Solution We evaluate properties atthe film temperature: 20+ T= BES _ we = 313K p_ 10132 x10 = RE “aang ~ 1128 keine = 1.906 x 10-5 kal -s Pr=0.7 k= 0.0273Wim-C cy The Reynolds number is 007 kiikg + °C 1.553 x 10° and the boundary layer is turbulent because the Reynolds number is greater than 5 x 10° Therefore, we use. Eq. (5-85) to calculate the average heat transfer over the plate: = Pr'* (0.037 Ret* ~ 850) (0.70"" (0.037101,553 «10 219340,02723) a 7 oo UTS = 79.6 Wim? °C [14.02 Burk fe °F] (79.60.7560 ~ 20) = 2388 W- [8150 Btwh) 293 850) ae Nu, > 4 = FAT, ~ 7) 5-0 TURBULENT-BOUNDARY-LAYER THICKNESS A number of experimental investigations have shown that the velocity profile in a turbulent boundary layer. outside the laminar sublayer. can be described by @ one-seventh-power relation a ss = ( 3) (5-88)Tutbulen-ooundaryiayer thickness 248 where 6 is the boundary-layer thickness as before. For purposes of an integral analysis the momentum integral can be evaluated with Eq. (5-88) because the laminar sublayer is so thin. However, the wall shear stress cannot be calculated from Eq. (5-88) because it yields an infinite value at y = 0. To determine the turbulent-boundary-layer thickness we empioy Eq. (5-17) for the integral momentum relation and evaluate the wall shear stress from the empirical relations for skin friction presented previously. According to Eq (5-52), and so for Re, < 10° we obtain from Eq. (5-77) T. = 0.0296 (| — ta? (5-89) (Sy 589) Now, using the integral momentum equation for zero pressure gradient (Eq. (5-17)] along with the velocity profile and wall shear stress, we obtain ELE G)"] G9 = 0065)" Integrating and clearing terms gives 2.2 coarse) (2)""x "s (5.90) We shall integrate this equation for two physical situations 1, The boundary layer is fully turbulent from the leading edge of the plate. 2. The boundary layer follows a laminar growth pattern up to Regi, = 5 10° and a turbulent growth thereafter. For the first case we integrate Eq. (5-89) with the condition that 8 = 0 at x = 0, to obtain For case 2 we have the condition B= Bam al ayy = 5% 108 Now, Bim is calculated from the exact relation of Eq. (5-214) Baum = S.0xey(5. 108)" (5-93)1248, Precpias of convection Integrating Eq. (5-89) gives y 8 ~ Bam = F co0rs6 ( ) Sas = Kent?) (5-94) Combining the various relations above gives 8 = 0.381 Re," — 10,256 Re, (5-95) This relation applies only for the region 5 x 10° < Re, < 10”, EXAMPLE 5-9 Calculate the turbulent-boundary-layer thickness at the end of the plate for Example 5-7, assuming that it develops (a) from the leading edge of the plate and (b) from the transition point at Reany = 5x 10% Solution ‘Since we have already calculated the Reynolds number as Re, = 1.553 x 10*, itis a simple matter to insert this value in Eqs. (5-91) and (5-95) along with x = L = 0.75 m to give (a 5 (0.75}0,381)1,583 x 109)-*2 = 0.0165 m = 16.5 mm [0.65 in} (0.75)(0.381\1.553 x 108-9 ~ 10,256(1.553 x 10%)-1] 0.0099 m = 9.9 mm (0.39 in} ‘The two values differ by 40 percent 5-10 HEAT TRANSFER IN LAMINAR TUBE FLOW Consider the tube-flow system in Fig. 5-13. We wish to calculate the heat transfer under developed flow conditions when the flow remains laminar. The wall temperature is 7. the radius of the tube is r,, and the velocity at the center of the tube is uo, It is assumed that the pressure is uniform at any cross section. The velocity distribution may be derived by considering the fluid ele- ment shown in Fig. 5-14. The pressure forces are balanced by the viscous- ss ac ae ne n Fig. 5:19 Control volume for energy analysis in tube towHeal enter in aminar tube tow 247 oer) paw Deanne?» Le shear forces so that Fig. 5-14 Force balance on fd element in tube flow. or du = and Bp + const (5-96) With the boundary condition 1Be_ py uaeo®! the velocity at the center of the tube is given by ra dp a 6-97) so that the velocity distribution may be written Hat 698) uw re which is the familiar parabolic distribution for laminar tube flow. Now consider the heat-transfer process for such a flow system. To simplify the analysis, we assume that there is a constant heat flux at the tube wall; that is, cre a? The heat flow conducted into the annular element is and the heat conducted out is denen “tone + dyer» a240, Punciles of comvecton ‘The net heat convected out of the element is ar dar dr popu dx The energy balance is, Net energy convected out = net heat conducted in or, neglecting second-order differentials, (5-99) wa") We assume that the heat flux at the wall is constant, so that the average fluid temperature must increase linearly with «, or oF const con This means that the temperature profiles willbe similar at various x distances along the tube. The boundary conditions on Eq, 5-99 are 7 ee ar To obtain the solution to Eq. (5-99), the velocity distribution given by Eq. (5-99) must be inserted. It is assumed that the temperature and velocity fields are independent; i.e. a temperature gradient does not affect the calculation of the velocity profile. This is equivalent to specifying that the properties remain ition of the velocity profile, Eq. (5-99) “Ea (i-8): aT _ lar é Integration yields and a second integration gives r= 1 a ( eo a joa) tCrlar +Applying the first boundary condition, we find that cso The second boundary condition has been satisfied by noting that the axial temperature gradient 87/ax is constant. The temperature distribution may finally be written in terms of the temperature at the center of the tube: T=T. atr=0 sothat C;= 7, (QQ) om T-T, ‘The Bulk Temperature, In tube flow the convection heat-transfer coefficient is usually defined by Local heat flux = g” = A(T. ~ T,) (-101) where T. is the wall temperature and 7, is the so-called bulk temperature, oF cenergy-average fluid temperature across the tube, which may be calculated from [from deur Bete os. ff rordrn, ‘The reason for using the bulk temperature in the definition of heat-transfer coefficients for tube flow may be explained as follows. In a tube flow there is 1no easily discernible free-stream condition as is present in the flow over a flat plate. Even the centerline temperature T, is not easily expressed in terms of the inlet flow variables and the heat transfer. For most tube- or channel-flow heat-transfer problems the topic of central interest isthe total energy transferred to the fluid in either an elemental length of the tube or over the entire length of the channel. At any x position, the temperature that is indicative of the total energy of the flow is an integrated mass-energy average temperature over the entire flow area. The numerator of Eq. (5-102) represents the total energy flow through the tube, and the denominator represents the product of mass flow and specific heat integrated over the flow area, The bulk temperature is thus rep- resentative of the total energy of the flow at the particular location. For this reason, the bulk temperature is sometimes referred to as the “mixing cup" temperature, since it is the temperature the fluid would assume if placed in a mixing chamber and allowed to come to equilibrium. For the temperature distribution given in Eq. (5-100), the bulk temperature is a linear function of x because the heat flux at the tube wall is constant. Catculating the bulk tem- perature from Eq. (5-102), we have 2 urd aT, eee 96 ores (5-103)280 Props of cowecton and for the wall temperature = 7.4 Mee oT T= 1. + wed (5-104) ‘The heat-transfer coefficient is calculated from 4 sat, - 1) = 44 (3) ar, (5-105) Marion, T. - To ‘The temperature gradient is given by ar MoT (r_ oP More AT . 7 @ ax G @)... “da ax Ci) Substituting Eqs. (5-103), (5-104), and (5-106) in Eq. (5-105) gives Mk _ 48 k "Tin ind, Expressed in terms of the Nusselt number, the result is Nuy = M2 = 4.366 ($-107) k which is in agreement with an exact calculation by Sellars, Tribus, and Klein G3], which considers the temperature profile as it develops. Some empirical relations for calculating heat transfer in laminar tube flow will be presented in Chap. 6. We may remark at this time that when the statement is made that a fluid centers a tube at a certain temperature, it is the bulk temperature to which we refer. The bulk temperature is used for overall energy balances on systems, 5-11 TURBULENT FLOW IN A TUBE ‘The developed velocity profile for turbulent flow in a tube will appear as shown in Fig. 5-15. A laminar sublayer, or “film,” occupies the space near the surface, while the central core of the flow is turbulent. To determine the heat transfer analytically for this situation, we require, as usual, a knowledge of the tem- perature distribution in the flow. To obtain this temperature distribution, the Fig. 5:15 Velocity profile in tr ‘bulant tube flow.Turtulant town a ibe 284 analysis must take into consideration the effect of the turbulent eddies in the transfer of heat and momentum. We shall use an approximate analysis which relates the conduction and transport of heat to the transport of momentum in the flow, i.e., viscous effects. “The heat flow across a fluid element in laminar flow may be expressed by a__,4a 4. 42 Ae ae Dividing both sides of the equation by pc, q__,4 pep dy It will be recalled that @ is the molecular diffusivity of heat. In turbulent flow fone might assume that the heat transport could be represented by 4 pA where éy is an eddy diffusivity of heat. Equation (5-108) expresses the total heat conduction as a sum of the mo- lecular conduction and the macroscopic eddy conduction. In a similar fashion, the shear stress in turbulent flow could be written ar = eta (5-108) ri (e du du ta (Eten)Z-wtr ae Gra erer es where ej is the eddy diffusivity for momentum. We now assume that the heat ‘and momentum are transported at the same rate; that is, ew = ¢y and ¥ = a, or Pr Dividing Eq. (5-108) by Eq. (5-108) gives Edu = -a7 (5-109) ‘An additional assumption is that the ratio of the heat transfer per unit area to the shear stress is constant across the flow field. This is consistent with the ‘assumption that heat and momentum are transported at the same rate, Thus ae) aoe : Z = const = (S-110) ‘Then, integrating Eq. (5-109) between wall conditions and mean bulk conditions gives aig S ee fl fall. Aureey (-l)1282 Principe of convection But the heat transfer at the wall may be expressed by Gu = by (Te ~ Ts) and the shear stress may be calculated from Sp (ed. 4, ddl ~ 4 L ‘The pressure drop may be expressed in terms of a friction factor f by 12) so that 5-113) L pg? te =F pt Substituting the expressions for 1. and q. in Eq. (5-111) gives st=- 4. Nu f 5-114) Pepin ~ Reg Pr~ 8 Equation (5-114) is called the Reynolds analogy for tube flow. It relates the heat-transfer rate to the frictional loss in tube flow and is in fair agreement with experiments when used with gases whose Prandtl numbers are close to unity. (Recall that Pr = 1 was one of the assumptions in the analysis.) ‘An empirical formula for the turbulent-friction factor up to Reynolds num- bers of about 2 x 10° for the flow in smooth tubes is 0.316 foe 115) Inserting this expression in Eq. (5-113) gives Nug a Res Pr 7 0.0395 Rez or Nug = 0.0395 Rey* 116) since we assumed the Prandtl number to be unity. This derivation of the relation for turbulent heat transfer in smooth tubes is highly restrictive because of the Pr ~ 1.0 assumption. The heat-transfer-fluid-friction analogy of Sec. 5-7 in- dicated a Prandtl-number dependence of Pr°* for the flat-plate problem and, as it tums out, this dependence works fairly well for turbulent tube flow. Equations (5-114) and (5-116) may be modified by this factor to yield spre = £ (S-114a) Nu, = 0.0395 Re} Pr’? (S-116a)Heat wanserinrigh-speed tow 283 As we shall see in Chap. 6, Eq. ( that are somewhat higher than those observed in experiments. The purpose of the discussion at this point has been to show that one may arrive at a relation for turbulent heat transfer in a fairly simple analytical fashion. As we have indicated earlier, a rigorous development of the Reynolds analogy between heat transfer and fluid friction involves considerations beyond the scope of our discussion and the simple path of reasoning chosen here is offered for the purpose of indicating the general nature of the physical processes, For calculation purposes, a more correct relation to use for turbulent flow in a smooth tube is Eq. (64), which we list here for comparison. Nuy = 0.023 Re Pr’ (6-4) All properties in Eq. (6-4) are evaluated at the bulk temperature. 512 HEAT TRANSFER IN HIGH-SPEED FLOW Our previous analysis of boundary-layer heat transfer (Sec. 5-6) neglected the effects of viscous dissipation within the boundary layer. When the free-stream velocity is very high, as in high-speed aircraft, these dissipation effects must be considered. We begin our analysis by considering the adiabatic case, i.e... perfectly insulated wall. In this case the wall temperature may be considerably higher than the free-stream temperature even though no heat transfer takes place. This high temperature results from two situations: (1) the incréase in temperature of the fluid as it is brought to rest at the plate surface while the kinetic energy of the flow is converted to internal thermal energy and (2) the heating effect due to viscous dissipation. Consider the first situation. The kinetic energy of the gas is converted to thermal energy as the gas is brought to rest, and this process is described by the steady-flow energy equation for an adiabatic process: thu? (5-117) 2K where ig is the stagnation enthalpy of the gas. This equation may be written in terms of temperature as Une fp To ~ To) = 5g where Tp is the stagnation temperature and T. is the static free-stream tem- perature. Expressed in terms of the free-stream Mach number, it is To yol ete 2 5-118) zo! 7M. (5-118) where M. is the Mach number, defined as M. = ua/a, and a is the acoustic284 Principles of convection velocity, which for a perfect gas may be calculated with a= VygRT (S119) where R is the gas constant. In the actual case of a boundary-layer flow problem, the fluid is not brought to rest reversibly because the viscous action is basically an irreversible process in a thermodynamic sense. In addition, not all the free-stream kinetic energy is converted to thermal energy—part is lost as heat, and part is dissipated in the form of viscous work. To take into account the irreversibilities in the boundary-layer flow system, a recovery factor is defined by Tow Ty ~ To r (3-120) where Tow is the actual adiabatic wall temperature and T~ is the static tem- perature of the free stream. The recovery factor may be determined experi- mentally, or, for some flow systems, analytical calculations may be made. ‘The boundary-layer energy equation * @) Pep \ay, has been solved for the high-speed-flow situation, taking into account the ‘cous-heating term. Although the complete solution is somewhat tedious, the final results are remarkably simple. For our purposes we ‘Present only the results ‘and indicate how they may be applied. The reader is referred to Appendix B for an exact solution to Eq. (5-22). An excellent synopsis of the high-speed +heat-transfer problem is given in a report by Eckert [4]. Some ‘typical boundary- layer temperature profiles for an adiabatic wall in high-speed flow are given in Fig. B-3. ‘The essential result of the high-speed heat-transfer analysis is that heat transfer rates may generally be calculated with the same relations used for low- ‘speed incompressible flow when the average heat-transfer coefficient is rede- fined with the relation = FAT. — Tow) (5-121) Notice that the difference between the adiabatic wall temperature and the actual wall temperature is used in the definition so that the expression will yield a Value of zero heat flow when the wall is at the adiabatic wall temperature. For gases with Prandtl numbers near unity the following relations for the recovery factor have been derived: Laminar flow: Pra Turbulent flow: r= Pri (5-122) (6-123)Heat taster high-speed tow 288 ‘These recovery factors may be used the adiabatic wall temperature. In high-velocity boundary layers substantial temperature gradients may oc- cur, and there will be correspondingly large property variations across the boundary layer. The constant-property heat-transfer equations may still be used. if the properties are introduced at a reference temperature T* as recommended by Eckert conjunction with Eq. {5-119} to obiain Tt = Te + 0.50(Ty ~ To) + 0.22(Tow — T-) (5-124) ‘The analogy between heat transfer and fluid friction (Eq. (5-56)] may also be used when the friction coefficient is known. Summarizing the relations used for high-speed heat-transfer calculations: Laminar boundary layer (Re, <5 x 10°): St? Pre? = 0,332 Ref! (5-125) Turbulent boundary layer (5 x 10° < Re, < 10"): St? Pr*2 = 0.0296 Ret-"* (5-126) Turbulent boundary layer (10" < Re, < 10°) Stf Pr? = 0.185 (log Ref)? (5-127) ‘The superscript * in the above equations indicates that the properties are eval- uated at the reference temperature given by Eq. (5-124) To obtain an average heat-transfer coefficient, the above expre_.ions must be integrated over the length of the plate. If the Reynolds number falls in a range such that Eq. (5-127) must be used, the integration cannot be expressed in closed form, and a numerical integration must be used. Care must be taken in performing the integration for the high-speed heat-transfer problem since the reference temperature is different for the laminar and turbulent portions of the boundary layer. This results from the different value of the recovery factor used for laminar and turbulent flow as given by Eqs. (5-122) and (5-123). When very high flow velocities are encountered, the adiabatic wall temper- ature may become so high that dissociation of the gas will take place and there will be a very wide variation of the properties in the boundary layer. Eckert [4] recommends that these problems be treated on the basis of a heat-transfer coefficient defined in terms of enthalpy difference: 4 = HA lie ~ ted (5-128) ‘The enthalpy recovery factor is then defined as " (5-129) fy = be where ia, is the enthalpy at the adiabatic wall conditions. The same relations£256 Principles of convection 4s before are used to calculate the recovery factor and heat transfer except that all properties are evaluated at a reference enthalpy i* given by i = i + OSs — ba) + 022i ge ~ (8-130) The Stanton number is redefined as sya (5-131) Pun This Stanton number is then used in Eq. (5-125), (5-126), or (5-127) to calculate the heat-wansfer coefficiegt. When calculating the enthalpies for use in the above relations, the ‘otal enthalpy must be used; i.e., chemical energy of dissociation as well as internal thermal energy must be included. The reference- enthalpy method has proved successful for calculating high-speed heat transfer with an accuracy of better than 10 percent. EXAMPLE 5-10 A flat plate 70 cm long and 1.0 m wide is placed in a wind tunnel where the flow conditions are M = 3, p = sh atm, and T = ~40°C. How much cooling must be used ‘to maintain the plate temperature at 35°C? Solution ‘We must consider the laminar and turbulent portions of the boundary layer separately because the recovery factors, and hence the adiabatic wall temperatures, used to es- tablish the heat flow wil be different for each flow regime, It turns out that the difference is rather small in this problem, but we shall follow a procedure which would be used if the difference were appreciable, so that the general method of solution may be in- dicated. The free-stream acoustic velocity is calculated from = Vyg.RT. = ((1.4)1.09287(233)]"" = 306 er/s [1003 fs) so that the free-stream velocity is 3)(306) = 918 m/s [3012 fs} ‘The maximum Reynolds number is estimated by making a computation based on prop erties evaluated at free-stream conditions: (1.0132 x 106%) (287)033) be = 1.434 x 10° kgm + § (0.0347 Ib.Jh - fe) (0.0758\918X0.70) _ x TTA x 10-3895 10 0.0758 kem* (4.73 x 10° tbg/f?) Rein ‘Thus we conclude that both laminar and turbulent boundary-layer heat transfer must bbe considered. We first determine the reference temperatures for the two regimes and then evaluate properties at these temperatures.wea! vanster in highspeed tow 287 Laminar Portion teeta (0+ 252 we) = amon 29079 = 682K Assuming a Prandtl number of about 0.7, we have Pr? = (0.77 = 0.837 Tx —Ts Tn — 2383 Ty Te” 652 — 233 11°C (592°F}. Then the reference temperature from Eq. (5-123) is 233 + (O.5N3S + 40) + (0.22584 — 233) = 447.8 K Checking the Prandtl number at this temperature, we have Prt = 0.697 so that the calculation is valid. If there were an appreciable difference between the value of Pr* and the value used to determine the recovery factor, the calculation would have to be repeated until agreement was reached. The other properties to be used in the laminar heat-transfer analysis are (1.0132 x 105)11720) (@87)387.8) 2.07 10-7 kgm: 5 ke = 0.03 Wim °C (0.0173 Buh - ft» *F 1.008 akg «°C = 0.0508 kel Turbulent Porton Assuming Pr = 0.7 gives Te Tn = 23 Ty~ Te ~ 652 = 233 To = 605 K = 332°C Tr = 233 + (0.5135 + 40) + (0.224605 - 233) = 3523 K Prt = 0.695 eel 0.888 ‘The agreement between Pr* and the assumed value is sufficiently close. The other properties to be used in the turbulent heat-transfer analysis are fo .0132. x 101/20) oe 0852.3) = 2.09 x 10° kgim 0.0302 Wim -*C_ cf = 1.009 ki/ke = 0.0501 kgim* a258 Principe of convection Laminar Host Transtor We assume Rely = 5x 10° = (5 x 109207 * 10°) _ 9 995 7 “o.0sosyors) = 972? 0.664 (Ret)"* Pr* (0.66445 x 1051%(0,697)" (416,310.03) _ 0.222 4163 $6.25 Wim?» °C (9.91 Btwh - f2 - °F] This is the average heat-transfer coefficient for the laminar portion of the boundary layer, and the heat transfer is calculated from 49 = AC. ~ T..) (656.26)0.222\35 - 311) 3445 W [~11,750 Bw/h} so that 3445 W of cooling is required in the laminar region of the plate per meter of depth in the z direction. Turbulent Heat Transtor To determine the turbulent heat transfer we must obtain an expression for the local heat-transfer coefficient from ‘Std Prt = 0.0296 Re?-" and then integrate from x = 0.222 m tox 0.7 mto determine the total heat transfer ene mrncsoois (Bt) ” Insite us ales rte ropes ges jw Minhide Fm de = 16 Wim? «°C [19.6 Buh - fe» "F] Using this value we may calculate the heat transfer in the turbulent region of the flat plate 9 = FAC, ~ To) = (111.46y0.7 ~ 0.222935 — 332 = = 15,823 W_ [54,006 Bru/h}Fewew quostons 260 ‘The total amount of cooling required is the sum of the heat transfers for the laminar and turbulent portions: Total cooling = 445 + 15,823 = 19,268 W (65,761 Brulh] ‘These calculations assume unit depth of 1 m in the z direction 513 SUMMARY Most of this chapter has been concerned with flow over flat plates and the associated heat transfer. For convenience to the reader we have summarized the equations in Table 5-2 along with restrictions which apply. The general procedure then is to: 1. Evaluate the fluid properties: this will usually be at the film temperature, 2, Establish the flow regime. 3. Select the appropriate equation, taking into account the flow regime and any fluid property restrictions which may apply. 4, Calculate the value(s) of the convection heat-transfer coefficient and/or heat transfer. REVIEW QUESTIONS 1) What is meant by a hydrodynamic boundary level? Define the Reynolds number. Why is it important? 2 3. What is the physical mechanism of viscous action? 4 Distinguish between laminar and turbulent flow in a physical sense. 5 ‘What is the momentum equation for the laminar boundary layer on a flat plate? What assumptions are involved in the derivation of this equation?” How is the boundary-layer thickness defined? 7 What is the energy equation for the laminar boundary layer on a flat plate? What assumptions are involved in the derivation of this equation? ‘8 What is meant by a thermal boundary layer? Define the Prandtl number. Why is it important? 10 Describe the physical mechanism of convection. How is the convection heat- transfer coefficient related to this mechanism? 11. Describe the relation between fluid friction and heat transfer. 12 Define the bulk temperature. How is it used? 13. How is the heat-transfer coefficient defined for high-speed heat-transfer calcula tions?es) (058 — 679 L600 rtd = "ON lx s = oy ors) 1799 058 ~ r629M L600 = ccd IS ‘ol > aa 7 uss) "ON HL = "ON OL > "2H > Ol (es) rece ('2Y 8008810 = cel 1S OT > "UD ass) £0191 96200 = celd SOL > "94 > 01 Feo} “wwainqiny, 0s > 4 > 90 EN ealeoeeree “oi xs> our oro LAG) ~ otertco «na sayy nener (steiauprmby) | 44 SOL x S>"2y 18009 ="Z a0] “seu ors) Ox § > Tey afesaKe seuRE OLX § > "2M ISU09 =~ e00) “seu 0s > 4d > 90 rs) end of E5¥'0 = "ON SOL x 5 > "RE suo =" yeoySseUMET 4) ] . +1 eliGsoo) +] 7 ass) cold en PE LEED SOL xs > "yO = "y——yes0] seu] 05 >44>90 brs) ld en9 780 = “ON OLX §> "2H "SU = "Z—yeso| “eure efiaaas aH Sequmu wowonby wopomby Tuonauiney as org otou osmuowo ssorun 2/1 + “1) = 1 1 poyenene souedoid ‘Sorel eLJ J0n0 mot4 20} suoNenb3 0 ARCUALNS 2-3 ONgeL,FS aged woay ves f724 SON) . 4 =! Soy > Soy afezone w2]nqin (64-5) ssv0 2 Cl 4s dines (8L-9) Pay Bopie'0 = > Al > "2A > OL (ues) a 76500 = "D OL > "2A > Ol x (pss) “yp = %D OLX S> 3M eo wT Saye = es ~ ea ecoe sol xs = oy (56-5) sear “OL > 2M > OL XS quamnqany seco =* o-rwo=8 Sonam S| 01> OU qwaqnquny, ons caueune rx s> "mH rooney 7 ‘ssauyotys sadny Gropunog _ Epos st T= patenexs CL "pvy => soruadoud yum moy poods-no 0} se sues “shoo = "Moy pods “rie ” oe 11 ateione 7) (0026 - 60!2¥) wo 9 “uaynqara-seurure282. Princes of convection a gt oT ‘A certain nozzle is designed to expand aur from stagnation conditions of 1.38 MPa and 200°C to 0.138 MPa. The mass rate of flow is designed to be 4.5 kg/s. Suppose this nozzle is used in conjunction with a blowdown wind-tunnel facility so that the nozzle is suddenly allowed to discharge into a perfectly evacuated tank. What will {the temperature of the air inthe tank be when the pressure in the tank equals 0.138 MPa? Assume that the tank is perfectly insulated and that air behaves as a perfect ‘gas. Assume that the expansion in the nozzle is isentropic, Using a linear velocity profile {for a flow over a flat plate, obtain an expression for the boundary-level thickness as a function of Using the continuity elation uy a ty along with the velocity distribution ude tty a7 35-2(3) and the expression fo the boundary ayer thickness 5. 464 derive an expression for the y component of velocity ¥ as a function of x and y. Calculate the value of v at the outer edge of the boundary layer at distances of 6 and 12 in from the leading edge for the conditions of Example 5:3. Repeat Prob. 5-3 for the linear velocity profile of Prob. 5-2 Using the linear-velocity profile in Prob. 5-2 and a cubic-parabola temperature distribution (Ea, (5-30)], obtain an expression for heat-transfer coefficient as a function of the Reynolds number for a laminar boundary layer on a fat plate, Air at 20 kPa and 5°C enters a 2.5-cm-diameter tube at a velocity of 1.5 m/s. Using a flat-plate analysis, estimate the distance from the entrance at which the flow becomes fully developed. A fluid flows between two large parallel plates. Develop an expression for the velocity distribution as a function of distance from the centerline between the two plates under developed flow conditions. Using the energy equation given by Eq. (5-32), determine an expression for heat transfer coefficient under the conditions T-T, _y ror 8 te = const where 6, is the thermal-boundary-layer thickness.5 510 on sa 513 o4 515 516 57 s18 519 520 Derive an expression for the heat transfer in a laminar boundary layer on a flat plate under the condition u = u. = constant. Assume that the temperature dis- tribution is given by the cubic-parabola relation in Eq. (5-30). This solution ap- proximates the condition observed in the flow of a liquid metal over a fiat plate. Show that @ulay? = 0 at y = 0 for an incompressible laminar boundary layer on a flat plate with zero-pressure gradient. [Review the analytical developments of this chapter and list the restrictions which apply to the following equations: (5-25), (5-26), (5-44), (5-46), (5-85), and (5-107). Calculate the ratio of thermal-boundary-layer thickness to hydrodynamic-bound- ary-layer thickness for the following fluids: air at 1 atm and 20°C, water at 20°C, helium at 1 atm and 20°C, liquid ammonia at 20°C, glycerine at 20°C. For water flowing over a flat plate at 15°C and 3 mvs, calculate the mass flow through the boundary layer at a distance of 5 cm from the leading edge of the plate Air at 90°C and 1 atm flows over a flat plate at a velocity of 30 mis. How thick is the boundary layer at a distance of 2.5 em from the leading edge of the plate? [Air flows over a flat plate at a constant velocity of 20 m/s and ambient conditions ‘of 20 kPa and 20°C. The plate is heated to a constant temperature of 75°C, starting at a distance of 7.5 cm from the leading edge. What is the total heat transfer from the leading edge to a point 35 cm from the leading edge? Water at 15°C flows between two large parallel plates at a velocity of 1.5 mis. The plates are separated by a distance of 1S mm. Estimate the distance from the leading edge where the flow becomes fully developed, ‘Air at standard conditions of 1 atm and 30°C flows over a flat plate at 20 mis. ‘The plate is 60 em square and is maintained at 90°C. Calculate the heat transfer from the plate. ‘Air at 7 kPa and 35°C flows across a 30-cm-square flat plate at 7.5 m/s. The plate is maintained at 65°C. Estimate the heat lost from the plate. Air at 90°C and atmospheric pressure flows over a horizontal flat plate at 60 m/s. ‘The plate is 60 cm square and is maintained at a uniform temperature of 10°C. What is the total heat transfer? Plot the heat-transfer coefficient versus length for flow over a I-m-long flat plate under the following conditions: (a) helium at 1 Ibfin’ abs, 80°F, us mls}; (b) hydrogen at 1 Ibvin® abs, 80°F, u- = 10 fs; (c) ar at 1 Ib/in? abs, 10 fUs; (d) water at 80°F, u. = 10 fUs; (e) helium at 20 Ibvin* abs, 80°F, 10 fus. Calculate the heat transfer from a 30-cm-square plate over which air flows at 35°C ‘and 14 kPa, The plate temperature is 250°C, and the free-stream velocity is 6 m/s. Air at 20 kPa and 20°C flows across a flat plate 60 em long. The free-stream Velocity is 30 m/s, and the plate is heated over its total length to a temperature of SSC. For x = 30cm calculate the value of y for which w will equal 22.5 m/s. For the flow system in Prob. 5-22 calculate the value of the friction coefficient at a distance of 15 cm from the leading edge.528 sar é é Air at $°C and 70 kPa flows over a flat plate at 6 m/s. A heater strip 2.5 em long. is placed on the plate ata distance of 15 cm from the leading edge. Calculate the heat lost from the strip per unit depth of plate for a heater surface temperature of 65°C. Airat | atm and 27°C blows across a large concrete surface 15 m wide maintained at SS°C. The flow velocity is 4.5 m/s. Calculate the convection heat loss from the surface. ‘Air at 300 K and 75 kPa flows over a I-m-square plate at a velocity of 45 mi. ‘The plate is maintained at a constant temperature of 400 K, Calculate the heat lost by the plate ‘A horizontal flat plate is maintained at 50°C and has dimensions of 50 by 50 em. Air at 50 kPa and 10°C is blown across the plate at 20 m/s. Calculate the heat Tost from the plate Air flows across a 20-cm-square plate with a velocity of $ mis. Free-stream conditions are 10°C and 0.2 atm. A heater in the plate surface furnishes a constant hheat-flux condition at the wall so that the average wall temperature is 100°C. Calculate the surface heat flux and the value of h at an x position of 10 cm: Calculate the flow velocity necessary to produce a Reynolds number of 10° for flow across a I-m-square plate with the following fluids: (a) water at 20°C. (b) air at 1 atm and 20°C, (c) Freon 12 at 20°C, (d) ammonia at 20°C, and (e) helium at 2. Calculate the average heat-transfer coefficient for each of the cases in Prob. 5-29 assuming all properties are evaluated at 20°C. Calculate the boundary-layer thickness at the end of the platé for each case in Prob. 5-29 AA blackened plate is exposed to the sun so that a constant heat lux of 800 Wim? is absorbed. The back side of the plate is insulated so that all the energy absorbed is dissipated to an airstream which blows across the plate at conditions of 25°C, 1 atm, and 3 m/s. The plate is 25 em square. Estimate the average temperature of the plate. What is the plate temperature at the trailing edge? Air at 1 atm and 300 K blows across a S0-cm-square flat plate at a velocity such that the Reynolds number at the downstream edge of the plate is 1.1 < 10° Heating does not begin until halfway along the plate and then the surface tem- perature is 400 K. Calculate the heat transfer from the plate. Air at 20°C and 14 kPa flows at a velocity of 150 m/s past a flat plate 1 m long which is maintained at a constant temperature of 150°C. What is the average heat- transfer rate per unit area of plate? Derive equations equivalent to Eq. (5-85) for critical Reynolds numbers of 3 x 10°, 10", and 3 x 10% ‘Assuming that the local heat-transfer coefficient for flow on a flat plate can be represented by Eq. (5-81) and that the boundary layer starts at the leading edge of the plate, determine an expression for the average heat-transfer coefficient ‘A 10 by 10-cm-square plate has an electric heater installed which produces538 g i 546 constant heat flux. Water at 10°C flows across the plate at a velocity of 3 mis. ‘What is the total heat which can be dissipated if the plate temperature is not to exceed 80°C? Repeat Prob, 5-37 for air at 1 atm and 300 K. Helium at 1 atm and 300 K is used to cool a I-m-square plate maintained at 500 K. The flow velocity is $0 m/s. Calculate the total heat loss from the plate. What is the boundary-layer thickness as the flow leaves the plate? For the flow system in Prob. 5-40 calculate the y position in the boundary layer at the trailing edge where u = 25 mis. A low-speed wind tunnel isto be designed to study boundary layers upto Re, = 107 with air at 1 atm and 25°C. The maximum flow velocity which can be expected from an existing fan system is 30 m/s. How long must the flat-plate test-section be to produce the required Reynolds numbers? What will the maximum boundary- layer thickness be under these conditions? What would the maximum boundary- layer thicknesses be for low velocities at 7 and 12 m/s? [Aight breeze at 10 mi/h blows across a metal building in the summer. The height ff the building wall is 12 ft, and the width is 20 ft. A net energy flux of 110 Btu/h - fe from the sun is absorbed in the wall and subsequently dissipated 10 the surrounding air by convection, Assuming that the air is 80°F and 1 atm and blows across the wall as on a flat plate, estimate the average temperature the wall, will attain for equilibrium conditions, ‘The bottom of a corn-chip fryer is 10 ft long by 3 ft wide and is maintained at 2 temperature of 420°F. Cooking oil flows across this surface at a velocity of 1 ts and has a free-stream temperature of 400°F. Calculate the heat transfer to the oil ‘and estimate the maximum boundary-layer thickness. Properties of the oil may be taken as ¥ = 2 x 10° mils, k= 0.12 Wim -°C, and Pr = 40, ‘Air at 27°C and | atm blows over a 4.0-m-square flat plate at a velocity of 40 mis. ‘The plate temperature is 77°C. Calculate the total heat transfer. ‘The roof of a building is 30 m by 60 m, and because of heat loading by the sun it attains a temperature of 300 K when the ambient air temperature is 0°C. Cal: culate the heat loss from the roof for a mild breeze blowing at S mi/h across the roof (L = 30m). Air at 1 atm and 30°C flows over a 1S-cm-square plate at a velocity of 10 mis. Calculate the maximum boundary layer thickness. ‘Air at 0.2 MPa and 25°C flows over a flat plate at a velocity of 60 mis. The plate js 0.5 on a side and is maintained at a constant temperature of 166°C. Calculate the heat lost from the plate Helium at a pressure of 150 kPa and a temperature of 20°C flows across a I-m-square plate at a velocity of 50 m/s. The plate is maintained at a constant temperature of 100°C. Calculate the heat lost by the plate Air at 50 kPa and 250 K flows across a 2-m-square plate at a velocity of 20 m/s. ‘The plate is maintained at a constant temperature of 350 K. Calculate the heat lost by the plate551 sz 559 Princes of convection Using Eqs. (5-55), (5-81), and (5-82) for the local heat transfer in their respective ranges, obtain an expression for the average heat transfer coefficient, or Nusselt ‘umber, over the range 5 x 10° < Re, < 10? with Recm = 5 x 10, Use a nu- ‘merical technique to perform the necessary integration and a curve fit to simplify the results ‘An experiment is to be designed to demonstrate measurement of heat loss for ‘water flowing over a flat plate. The plate is 30 cm square and it will be maintained ‘early constant in temperature at 50°C while the water temperature will be about 10°C. (a) Calculate the flow velocities necessary to study a range of Reynolds ‘numbers from 10* to 10’. (6) Estimate the heat-transfer coefficients and heat- transfer rates for several points in the specified range. [Nitrogen at 50 kPa and 300 K flows over a flat plate at a veRity of 100 mis. The length of the plate is 1.2 m and the plate is maintained at a constant temperature ‘of 400 K. Calculate the heat lost by the plat. Hydrogen at 2 atm and 15°C flows across a I-m-square fat plate at a velocity of 6 mis, The plate is maintained at a constant temperature of 139°C. Calculate the heat lost by the plate. Liquid ammonia at 10°C is forced across a square plate 45 cm on a side at a velocity of $ m/s. The plate is maintained at SO°C. Calculate the heat lost by the plate. Helium flows across a 1.0-m-square plate at a velocity of 50 m/s. The helium is at a pressure of 45 kPa and a temperature of 50°C. The plate is maintained at a constant temperature of 136°C. Calculate the heat lost by the plate. Air at 0.1 atm flows over a flat plate ata velocity of 300 mis. The plate temperature is maintained constant at 100°C and the free-stream air temperature is 10°C. Calculate the heat transfer for a plate which is 80 cm square. Water at 70°F flows across a I-ft-square flat plate at a velocity of 20 fs. The plate is maintained at a constant temperature of 130°F. Calculate the heat lost by the plate. Plot h, versus x for air at 1 atm and 300 K flowing at a velocity of 30 m/s across a flat plate. Take Rew = 5 % 10" and use semilog plotting paper. Extend the plot to an x value equivalent to Re = 10°. Also plot the average heat-transfer coefficient over this same range. Air flows over a flat plate at 1 atm and 350 K with a velocity of 30 m/s. Calculate the mass flow through the boundary layer at x locations where Re, = 10* and 0 Air flows with a velocity of 6 m/s across a 20 by 20-cm-square plate at 50 kPa ‘and 300 K. An electrical heater is installed in the plate such that it produces a constant heat flux. What is the total heat which can be dissipated if the plate temperature cannot exceed 600 K? “Slug” flow in a tube may be described as that flow in which the velocity is constant across the entire low area of the tube. Obtain an expression for the heat-transfer coefficient in this type of flow with a constant-heat-iux conditioné : 570 maintained at the wall. Compare the results with those of Sec. 5-10. Explain the reason for the difference in answers on a physical basis. ‘Assume that the velocity distribution in the turbulent core for tube flow may be represented by ‘where u, is the velocity at the center of the tube and r, is the tube radius. The velocity in the laminar sublayer may be assumed to vary linearly with the radius. Using the friction factor given by Eq. (5-115), derive an equation for the thickness of the laminar sublayer. For this problem the average flow velocity may be cal- culated using only the turbulent velocity distribution, Using the velocity profile in Prob. 5-62, obtain an expression for the eddy dif fusivity of momentum as.a function of radius. In heat-exchanger applications, it is frequently important to match heat-transfer requirements with pressure-drop limitations. Assuming a fixed total heat-transfer requirement and a fixed temperature difference between wall and bulk conditions as well as a fixed pressure drop through the tube, derive expressions for the length and diameter of the tube, assuming turbulent flow ofa gas with the Prandtl number near unity. Water flows in a 2.5-cm-diameter pipe so that the Reynolds number based on diameter is 1500 (laminar flow is assumed). The average bulk temperature is 35°C. Calculate the maximum water velocity inthe tube, (Recall that u,, = 0.5t9.) What would the heat-transfer coefficient be for such a system if the tube wall was subjected to a constant heat flux and the velocity and temperature profiles were completely developed? Evaluate properties at bulk temperature. ‘A slug flow is encountered in an annular-flow system which is subjected to a ‘constant heat flux at both the inner and outer surfaces. The temperature is the same at both inner and outer surfaces at identical x locations. Derive an expression for the temperature distribution in such a flow system, assuming constant prop- ‘erties and laminar flow. Air at Mach 4 and 3 lbvin? abs, 0°F, flows past a flat plate. The plate is to be ‘maintained at a constant temperature of 200°F. If the plate is 18 in long, how ‘much cooling will be required to maintain this temperature? Air flows over an isothermal flat plate maintained at a constant temperature of 65°C. The velocity of the air is 600 mis at static properties of 15°C and 7 kPa Calculate the average heat-transfer coefficient for a plate 1 m long, Ai at 7 kPa and —407C flows over a flat plate at Mach 4. The plate temperature is 35°C, and the plate length is 60 cm. Calculate the adiabatic wall temperature for the laminar portion of the boundary layer. ‘A wind tunnel is to be constructed to produce flow conditions of Mach 2.8 at T. = ~40°C and p = 0.05 atm. What is the stagnation temperature for these conditions? What would be the adiabatic wall temperature for the laminar and turbulent portions of a boundary layer on a flat plate? Ifa flat plate were installedon on 673 574 575 578 a7 Princito of convacton in the tunnel such that Re, = 107, what would the heat transfer be for a constant. wall temperature of 0°C? Compute the drag force exerted on the plate by each of the systems in Prob. 5.20. Glycerin at 30°C flows past a 30-cm-square flat plate ata velocity of 1.5 m/s. The drag force is measured as 8.9 N (both shies of the plate). Calculate the heat. transfer coefficient for such.a flow system. Calculate the drag (viscous-friction) force on the plate in Prob. 5-21 under the conditions of no heat transfer. Do not use the analogy between fluid friction and heat transfer for this calculation; ie., calculate the drag directly by evaluating the viscous-shear stress at the wal. ‘Nitrogen at | atm and 20°C is blown across a 130-cm-square flat plate at a velocity ‘f 3.0 mvs. The plate is maintained at a constant temperature of 100°C. Calculate the average-friction coefficient and the heat transfer from the plate Using the velocity distribution for developed laminar flow expression for the friction factor as defined by Eq. 5-112. Engine oil at 10°C flows across a 15-cm-square plate upon which is imposed a constant heat flux of 10 kW/m?, Determine (a) the average temperature difference, (®) the temperature difference atthe tailing edge, and (c) the average heat-transfer ‘coefficient. Use the Churchill relation (Eq. (S-Si)]. wa = 0.5 m/s. ‘Work Prob: 5-76 for a constant plate surface temperature equal to that at the trailing edge, and determine the total heat transfer. ‘a tube, derive an REFERENCES 1 3 Schlichting, H.: “Boundary Layer Theory,""7th ed., McGraw-Hill Book Company, New York, 1979, von Kérmén, T.: Uber laminaire und turbulente Reibung, Angew. Math. Mech., vol. 1, pp. 233-282, 1921; also NACA Tech, Mem, 1092, 1946. Sellars, J. R., M. Tribus, and J. S. Klein: Heat Transfer to Laminar Flows in a Round Tube or Flat Conduit: The Graetz Problem Extended, Trans. ASME, vol 78, p. 441, 1956. Eckert, E. R. G.: Survey of Boundary Layer Heat Transfer at High Velocities and High Temperatures, WADC Tech. Rep., pp. 59-624, April 1960. White, F. M.: "Viscous Fluid Flow,” McGraw-Hill Book Company, New York, 1974, Knudsen, J. D., and D. L. Katz: “Fluid Dynamics and Heat Transfer," McGraw- Hill Book Company, New York, 1958. Holman, J. P.: “Thermodynamics.” 34 ed., McGraw-Hill Book Company, New York, 1980, ‘Schultz-Grunow. F.: Neues Widerstandsgesetz flr glatte Platten, Luffahrifor- schung, vol. 17, p. 239, 1940; also NACA Tech. Mem, 986, 194110 " eteences 200 Churchill, $. W., and H. Ozoe: Correlations for Laminar Forced Convection in Flow over an Isothermal Flat Plate and in Developing and Fully Developed Flow in an Isothermal Tube, J. Heat Transfer, vol. 95, p. 46, 1973, Whitaker, S.: Forced Convection Heat Transfer Correlation for Flow in Pipes, Past Flat Plates, Single Cylinders, Single Spheres, and for Flow in Packed Beds and Tube Bundies, AICKE J., vol. 18, p. 361, 1972. Churchill, 8. W.: A. Comprehensive Correlating Equation for Forced Convection from Flat Plates, AICAE J., vol. 22, p. 264, 1976.EMPIRICAL AND PRACTICAL RELATIONS FOR FORCED- CONVECTION HEAT TRANSFER 1 INTRODUCTION The discussion and analyses of Chap. 5 have shown how forced-convection heat transfer may be calculated for several cases of practical interest; the problems considered, however, were those which could be solved in an ana- lytical fashion. In this way, the principles of the convection process and their relation to fluid dynamics were demonstrated, with primary emphasis being devoted to a clear understanding of physical mechanism. Regrettably. it is not always possible to obtain analytical solutions to convection problems, and the individual is forced to resort to experimental methods to obtain design infor- mation, as well as to secure the more elusive data which increase the physical understanding of the heat-transfer processes. Results of experimental data are usually expressed in the form of either empirical formulas or graphical charts so that they may be utilized with a maximum of generality. It is in the process of trying to generalize the results ‘of one’s experiments, in the form of some empirical correlation, that difficulty is encountered. If an analytical solution is available for a similar problem, the correlation of data is much easier, since one may guess at the functional form of the results, and hence use the experimental data to obtain values of constants or exponents for certain significant parameters such as the Reynolds or Prandtl numbers. If an analytical solution for a similar problem is not available, the individual must resort to intuition based on physical understanding of the prob- lem, or shrewd inferences which one may be able to draw from the differential equations of the flow processes based upon dimensional or order-of-magnitude{272 Empécal anc practic! relations for lrced-convecton heat ans estimates, In any event, there is no substitute for physical insight and under- standing. To show how one might proceed to analyze a new problem to obtain an important functional relationship from the differential equations, consider the problem of determining the hydrodynamic-boundary-layer thickness for flow over a flat plate. This problem was solved in Chap. 5, but we now wish to make an order-of-magnitude analysis of the differential equations to obtain the functional form of the solution. The momentum equation a a must be solved in conjunction with the continuity equation u,% Wg ax * ay Within the boundary layer we may say that the velocity u is of the order of the free-stream velocity u.. Similarly, the y dimension is of the order of the boundary-layer thickness 8. Thus yn8 and we might write the continuity equation in an approximate form as au ity Then, by using this order of magnitude for v, the analysis of the momentum equation would yield or Dividing by x to express the result dimensionless form gives iEmotica!relatons for pipe and tube tow 27, 6 fel x Vix VRe. This functional variation of the boundary-layer thickness with the Reynolds number and x position is precisely that which was obtained in Sec. 5-4. Although this analysis i rather straightforward and does indeed yield correct results, the order-of-magnitude analysis may not always be so fortunate when applied to more complex problems, particularly those involving turbulent- or separated- flow regions. Nevertheless, one may often obtain valuable information and physical insight by examining the order of magnitude of various terms in a governing differential equation for the particular problem at hand. ‘A conventional technique used in correlation of experimental data is that of dimensional analysis, in which appropriate dimensionless groups such as the Reynolds and Prandtl numbers are derived from purely dimensional and func tional considerations. There is, of course, the assumption of flow-field and temperature-profile similarity for geometrically similar heating surfaces. Gen- erally speaking, the application of dimensional analysis to any new problem is extremely difficult when a previous analytical solution of some sort is not available, It is usually best to attempt an orger-of-magnitude analysis such as, the one above if the governing differential equations are known. In this way it may be possible to determine the significant dimensionless variables for cor- relating experimental data. In some complex flow and heat-transfer problems a clear physical model of the processes may not be available, and the engineer ‘must first try to establish this model before the experimental data can be cor- related. Schlichting {6}, Giedt (7], and Kline (28] discuss similarity considerations and their use in boundary-layer and heat-transfer problems. ‘The purpose of the foregoing discussion has not been to emphasize or even to imply any new method for solving problems, but rather to indicate the necessity of applying intuitive physical reasoning to a difficult problem and to point out the obvious advantage of using any and all information which may be available, When the problem of correlation of experimental data for a pre- viously unsolved situation is encountered, one must frequently adopt devious methods to accomplish the task. 6-2. EMPIRICAL RELATIONS FOR PIPE AND TUBE FLOW The analysis of Sec. 5-10 has shown how one might analytically attack the problem of heat transfer in fully developed laminar tube flow. The cases of undeveloped laminar flow, flow systems where the fluid properties vary widely with temperature, and turbulent-low systems are considerably more compli- cated but are of very important practical interest in the design of heat exchang- ers and associated heat-transfer equipment. These more complicated problems may sometimes be solved analytically, but the solutions, when possible, arevery tedious. For design and engineering purposes, empirical correlations are usually of greatest practical utility. In this section we present some of the more important and useful empirical relations and point out their limitations, ‘The Bulk Temperature First let us give some further consideration to the bulk-temperature concept which is important in all heat-transfer problems involving flow inside closed channels. In Chap. 5 we noted that the bulk temperature represents energy average or ‘mixing cup" conditions. Thus. for the tube flow depicted in Fig. 6-1 the total energy added can be expressed in terms of a bulk-temperature difference by q = mie lTa: ~ To) 1) provided c, is reasonably constant over the length. In some differential length dx the heat added dq can be expressed either in terms of a bulk-temperature difference or in terms of the heat-transfer coefficient dq = rheydT, = har) dx (Ty — To) (62) where 7, and T, are the wall and bulk temperatures at the particular x location, ‘The total heat transfer can also be expressed as q = AAT. = Trav 63) where A is the total surface area for heat transfer. Because both T. and 7, can vary along the length of the tube, a suitable averaging process must be adopted for use with Eq. (6-3). In this chapter most of our attention will be focused on methods for determining h, the convection heat-transfer coefficient. Chapter 10 will discuss different methods for taking proper account of temperature variations in heat exchangers. For fully developed turbulent flow in smooth tubes the following relation is recommended by Dittus and Boeiter (1) Nuy = 0.023 Reg Pre (4) ‘The properties in this equation are evaluated at the fluid bulk temperature, and, the exponent 1 has the following values: Fig. 61 Total heat transferin terms of buktemperature diferenceEmpica lao erpipe ano tbe Row TH n= {08 — for beating {03 for cooting Equation (6-4) is valid for fully developed turbulent flow in smooth tubes {or fluids with Prandtl numbers ranging from about 0.6 to 100 and with moderate temperature differences between wall and fluid conditions ‘Ae cay ask the reason for the functional form of Eq. (6-4). Physical rea soning, based on the experience gained with the analyses of Chap. 5, would certainly indicate a dependence of the heat-transfer process on the flow field, ‘and hence on the Reynolds number. The relative rates of diffusion of heat and momentum are related by the Prandt! number, so that the Prandtl number is expected to be a significant parameter in the final solution. We can be rather confident of the dependence of the heat transfer on the Reynolds and Prandtl numbers. But the question arises as to the correct functional form of the re- lation: .¢., would one necessarily expect a product of two exponential functions of the Reynolds and Prandtl numbers? The answer is that one might expect this functional form since it appears in the flat-plate analytical solutions of Chap. 5, as well as the Reynolds analogy for turbulent flow. In addition, this type of functional relation is convenient to use in correlating experimental data, as described below. ‘Suppose a number of experiments are conducted with measurements taken of heattransfer rates of various fluids in turbulent flow inside smooth tubes under different temperature conditions, Different-diameter tubes may be used to vary the range of the Reynolds number in addition to variations in the mass- flow rate. We wish to generalize the results of these experiments by arriving ‘at one empirical equation which represents all the data. As described above, We may anticipate that the heat-transfer data will be dependent on the Reynolds and Prandtl numbers. An exponential function for each of these parameters is perhaps the simplest type of relation to use, so we assume Nuy = C Rey” Prt where C, m, and n are constants to be determined from the experimental data. ‘A log-log plot of Nuy versus Rey is first made for one fuid to estimate the dependence of the heat transfer on the Reynolds number, i-e., to find an ap- proximate value of the exponent m. This plot is made for one fluid at a constant temperature, so that the influence of the Prandtl number will be small, since the Prandtl number will be approximately constant for the one fluid. By using this first estimate for the exponent m, the data for all fluids are plotted as log (NuyRe/”) versus log Pr, and a value for the exponent 1 is determined. Then, by using this value of m, all the data are plotted again as log (Nu/Pr") versus log Rey, and @ final value of the exponent m is determined as well as @ value for the constant C. An example of this final type of data plot is shown in Fig. 6-2. The final correlation equation usually represents the data within £25 percent. If wide temperature differences are present in the flow, there may be an4276 Empirical and practical reatons lr forced convection heat tansler Fig. 62. Typical data corlation for focced carwection in smooth tubes, tu bulent tow. appreciable change in the fluid properties between the wall of the tube and the central flow. These property variations may be evidenced by a change in the velocity profile as indicated in Fig. 6-3. The deviations from the velocity profile for isothermal flow as shown in this figure are a result of the fact that the iscosity of gases increases with an increase in temperature, while the viscos- ies of liquids decrease with an increase in temperature. To take into account the property variations, Sieder and Tate [2] recommend the following relation: Nuy = 0.027 Re9! Pri? (£) 65) Fig. 63 Influence of heating on veloety prof in larinar tube Row.monica ration lor oe and tube tow 277 All properties are evaluated at bulk-temperature conditions, except i, which is evaluated at the wail temperature. Equations (6-4) and (6-5) apply to fully developed turbulent flow in tubes. In the entrance region the flow is not developed, and Nusselt (3] recommended the following equation: a\onss L rays aomenarren (2) orweb ewe — 0 where L is the length of the tube and d is the tube diameter. The properties in Eq. (6-6) are evaluated at the mean bulk temperature. Hartnett (24) has given experimental data on the thermal entrance region for water and oils. Definitive studies of turbulent transfer with water in smooth tubes and at uniform heat flux have been presented by Allen and Eckert [25]. ‘The above equations offer simplicity in computation, but errors on the order of +25 percent are not uncommon. Petukhov [42] has developed a more ac- curate, although more complicated, expression for fully developed turbulent flow in smooth tubes: Nuy © 07 + 12.7(f18)"(Pr? — 1) \p.., where n = 0.11 for T,, > T,, 2 = 0.25 for T. < T,, and n = 0 for constant heat flux or for gases. All properties are evaluated at T; = (T.. + Ts)/2 except for jp and y2.. The friction factor may be obtained either from Fig. 6-4 or from the following for smooth tubes: Sf = (1.82 logiy Reg — 1.64)-? (6-8) Equation (6-7) is applicable for the following ranges: 0.5 10 ‘The product of the Reynold and Prandtl numbers which occurs in the laminar- flow correlations is called the Peclet number. tee Pe= Rey Pr 1) ‘The calculation of laminar heat-transfer coefficients is frequently complicated by the presence of natural-convection effects which are superimposed on the forced-convection effects. The treatment of combined forced- and free- convection problems is discussed in Chap. 7. ‘The empirical correlations presented above, with the exception of Eq (6-7), apply to smooth tubes. Correlations are, in general, rather sparse where rough tubes are concerned, and it is sometimes appropriate that the Reynolds analogy between fluid friction and heat transfer be used to effect a solution under these circumstances. Expressed in terms of the Stanton number, sunp =f on The friction coefficient f is defined by Lt? ap =f a? 2. (6-13) where um is the mean flow velocity Values of the friction coefficient for different roughness conditions are shown in Fig. 6-4 Note that the relation in Bq. (6-12) is the same as Eq. (5-114), except that the Stanton number has been multiplied by Pr? to take into account the vari- ation of the thermal properties of different fluids. This correction follows the recommendation of Colburn [15], and is based on the reasoning that fluid friction ‘and heat transfer in tube flow are related to the Prandtl number in the same way as they are related in flat-plate flow (Eq. (5-56)]. In Eq. (6-12) the Stanton,1280 Empirical and practeal eaten lr forced convection host ranser ‘number is based on bulk temperature, while the Prandtl number and friction factor are based on properties evaluated at the film temperature. Further in- formation on the effects of tube roughness on heat transfer is given in Refs. 27, 29, 30, and 31. If the channel through which the fluid flows is not of circular cross section, it is recommended that the heat-transfer correlations be based on the hydraulic diameter Dy, defined by 4A n= 14) where A is the cross-sectional area of the flow and P is the wetted perimeter. This particular grouping of terms is used because it yields the value of the physical diameter when applied to a circular cross section. The hydraulic di- ameter should be used in calculating the Nusselt and Reynolds numbers, and in establishing the friction coefficient for use with the Reynolds analogy. Although the hydraulic-diameter concept frequently yields satisfactory re- lations for fluid friction and heat transfer in many practical problems, there are ‘some notable exceptions where the method does not work. Some of the prob- Jems involved in heat transfer in noncircular channels have been summarized by Irvine (20] and Knudsen and Katz (9}. The interested reader should consult, these discussions for addi transfer and fluid-friction information for fully developed laminar flow in ducts with a variety of flow cross sections as shown in Table 6-1. In this table the following nomenclature applies: ‘Nuy, = average Nusselt number for uniform heat flux in flow direction and uniform wall temperature at particular flow ‘cross section Nuya = average Nusselt number for uniform heat flux both in flow direction and around periphery Nur = average Nusselt number for uniform wall temperature Ff Re = product of friction factor and Reynolds number Kays (36] and Sellars, Tribus, and Klein (Ref. 3, Chap. 5) have calculated the local and average Nusselt numbers for laminar entrance regions of circular tubes for the case of a fully developed velocity profile. Results of these analyses are shown in Fig. 6-5 in terms of the inverse Graetz number, where Graetz number = Gz = Re Pr 4 15) Entrance effects for turbulent flow in tubes are more complicated than for laminar flow and cannot be expressed in terms of a simple function of the Graetz number. Kays [36] has computed the influence for several values of Re and Pr with the results summarized in Fig. 6-6. The ordinate is the ratio of theEnplical relations for ipe a8 Woetow 284 ‘Table 6-1 Heat Transfer and Flid Friction for Fully Dbveloped Laminar Flow in Ducts of Various Cross Sectors. Geometry 5 (LD, > 100) Nur sRe 1992-247 713.333 OMB soar asses 4353.66 18.700 09 + a 53m 2.930 4.439 18.233 8235 8.235 7.541 24.000 0 s3a5 4861 24.0002 i 2 ° 3x10" 10 = (rid Reg Pr. Inverse Grate number Fig. 6-5 Local and average Nusselt numbers for citcular ube thermal entrance regions in fully developed laminar tow.Emplical relation for pipe and tube fow 289 local Nusselt number to that along distance from the inlet, or for fully developed thermal conditions. In general, the higher the Prandtl number, the shorter, the eniry length. We can see that the thermal entry lengths are much shorter for turbulent flow than for the laminar counterpart EXAMPLE 6-1. Turbulent low in a tube Air at 2 atm and 200°C is heated as it flows through a tube with a diameter of 1 in (2.54 cm) at a velocity of 10 mvs, Calculate the heat transfer per unit length of tube if @ constant-heat-flux condition is maintained at the wall and the wall temperature is 20°C fbove the air temperature, all along the length of the tube. How much would the bulk temperature increase over a 3-m length of the tube? Solybon We first calculate the Reynolds number to determine if the flow is laminar or turbulent. and then select the appropriate empirical correlation to calculate the heat transfer. The properties of air at a bulk temperature of 200°C are p_ _ 2K1.0132 x 10) 2 RE ~~ QBINETS) 1.493 kg/m? (0.0932 Ib,./f) so that the flow is turbulent, We therefore use Eq. (6-4) to calculate the heat-transfer coefficient we = M002 Rest es OUEHLATSOFNO GRIN = 247 k 0.038642.67) hs > a fn = Eg, = CORRES — 4.85 Wim? °C {11.42 Biuth °F] "The heat flow per unit tength is then 4 = nd (1. ~ Ta) = (64.85)m(0.025420) = 103.5 Wim [107.7 Bruit) We can now make an energy balance to calculate the increase in bulk temperature in 3.0m length of tube: = nic, Te = (2 a= me, at, = (2) We also have Pa (0.025992 1 = pn 2E = (1495 )00)7 OE = 7.565 «10° kg/s (0.0167 Ib,/s}‘284 Empics and practical relations fr frced-convcton heat rarer : so that we insert the numerical values in the energy balance to obtain (7.565 x 10-91025) AT, = (3.0)(103.5) and AT, = 40.08C [104.07 F] EXAMPLE 62 Laminar tubo fow 4 ‘Water at 60°C enters a tube of 1-in (2.54-cm) diameter at a mean flow velocity of 2 cm/s. Calculate the exit water temperature ifthe tube is 3.0 m long and the wall temperature is constant at 80°C. : ‘Sotuton " We first evaluate the Reynolds number atthe inlet bulk temperature to determine the flow regime. The properties of water at 60°C are p= 9S kgm? —c, = 4.18 kJ/kg - °C \ w= 471 x 10+ kgim-s [1.139 Ib./h + ft) k= 0651 Win-"C Pr = 3.02 : Rey = Bad _ (985)0.0240.0254) _ a AIX 10 £0 the flow is laminar. Calculating the additional parameter, we have , Rested = LOBED 7155 1g ! 80 Ea, (6-10) is applicable. We do not yet know the rean bulk temperature to evaluate ' properties so we first make the calculation on the byfis of 60°C, determine an exit bulk {emperature, and then make’a second iteration to @btain a more precise value. When i inlet and outlet conditions are designated with the subscripts 1 and 2, respeofively, the energy balance becomes 1082 q= heal (" S Se = mies(Tn ~ To) @ ‘At the wall temperature of 80°C we have fi Me = 3.55 x 10-* kg/m -s It From Eq. (6-10) = 5816 Nas = 1.86 [ (4062)(3.02)(0.0254)]" 4.71 3 3.55, = 69.1 Wim? °C (26.26 Btwh - fe ‘The mass flow rate is 2 tm pg, = SOIMOSINOY 5 10-rtye Inserting the value for h into Eq. (a) along with mn and T,, = 60°C and T. = 80°C gives \Emoircal retains for pioe and ube tow 288 (1.)900.025010.0 (0 - 8 (9.982 x 10->KA180NT., - 60) 6) ‘This equation can be solved to give ie ngec Thus, we shoud g ue and eva pope Tan = BEE oe We obtain p= 982 kgm’ — cy = 4185 Sky °C = 4.36 x 10-¢ kg/m 5 ke oes winec pra 278 ayer Re, = LADET ye were - UMTEIOON _ y p= es a (4:36)" wa. = .soaran (236) = 378 cosens.29 . = OSHE an win We insert this value of h back into Eq. (a) to obtain Ty = 1188 (161.4°F) ‘The iteration makes very litle difference in this problem. If a large bulk-temperature difference had been encountered, the change in properties could have had a larger effect. = EXAMPLE 6-3 Air at | atm and 27°C enters a 5.0-mm-diameter smooth tube with a velocity of 3.0 mvs, ‘The length of the tube is 10 cm. A constant heat flux is imposed on the tube wall Calculate the heat transfer if the exit bulk temperature is 77°C. Also calculate the exit wall temperature and the value of h at exit. ‘soution We first must evaluate the flow regime and do so by taking properties at the average bulk temperature 7- BAT. xe = 25k Y= 1822 10-* mis Pr = 0.703 k = 0.02814 Wim = °C Re, = Mf = 80.05) _ 923 - (@) v 1822 x 10 so that the flow is laminar. The tube length is rather short, so we expect a thermal1206 Empirical and practca rlatons fo erced-convection heat taneloe ‘entrance effect and shall consult Fig. 6-5. The inverse Graetz number is computed as er on Re, Pr d ~ (823X0.703X0.005) constant, we obtain the Nusselt number at exit from Fig. 6-5 as Ge 0.0346 Therefore, for a a = tet Nu P47 = ee o “The total heat tranafer is obtained in terms of the overall energy balance: = mey(To, ~ Tod At entrance p = 1.1774 kg/m’, so the mass flow is ri = (1.1774) x40.0025)°3.0) 94 x 10"? kes and 4 = (6.94 x 10-9\1006\77 — 27) = 3.49. W ‘Thus we may find the heat transfer without actually determining wall temperatures or values of h. However, to determine T,, we must compute q. for insertion in Eq. (b) We have fal 9 = qd = 3.49 W ge = 2222 Wim? Now, from Eq. (b) (2222,0.005) (> Thee = Gy 02818 sec The wall temperature at exit is thus Teleon = 44+ 77 = 161°C and the heatansfer coetiient is nn ty CoH Tes "eg = 78.45 Went °C EXAMPLE 6-4 Repeat Example 6-3 for the case of constant wall temperature. Soluton ‘We evaluate properties as before and now enter Fig. 6-5 to determine Nu, for T. = constant. For Gz~* = 0.0346 we read Nuy = 5.15 We thus calculate the average henttransfer coefficient as = 29.98 Wim? °CEmpirical reations for pe and tube tow 287 We base the heat transfer on mean bulk temperature of 52°C, so that q = Fd (7, ~ T) = 3.49 and T. = 1667 + $2 = 12867°C EXAMPLE 6-5. Heat transtor in rough tube ‘A20-cmediameter tube having a relative roughness of 0.001 is maintained at constant ‘wall temperature of 90°C, Water enters the tube at 40°C and leaves at 60°C. If the femtering velocity is 3 ms, calculate the length of tube necessary to accomplish the heafing. ‘Solution We first calculate the heat transfer from (989)(3,0)m(0.01)5(4174)(60 ~ 40) = 77,812 W For the rough-tube condition, we may employ the Petukhov relation, Eq. (6-7). The ‘mean film temperature is 4 = me, MT, 020 we and the fluid properties are TR kgim? — x = 4.0.x 10-* kgim = s A = 0.668 Wim °C Pr'= 2.84 Also. dae = 585. 10 “kam +s gm. = 281 x 10-* Reims “The Reynolds number is thus (978)(3)0.02) ax 10+ Consulting Fig. 6-4. we find the fiction factor as f= 00218 f= 0.002725 = Because T, > Tr we take n= 0.11 and obtain (0,002725)(146,700)(2.54) 5.35)°" ToT + (12.7H0.002725)702.54° — 1) (i) 666.8 (656.810.6648) _ 092 ‘The tube length is then obtained from the energy balance Tew dLAT, ~ T,) = 77812 W 1.40 me Nuy = 22138 Wim? «°C hs 1m208 Empirical and practical latons fr force. convecton het ster €3 FLOW ACROSS CYLINDERS AND SPHERES While the engineer may frequently be interested in the heat-transfer charac- teristics of flow systems inside tubes or over flat plates, equal importance must be placed on the heat transfer which may be achieved by a cylinder in cross flow, as shown in Fig. 6-7. As would be expected, the boundary-layer devel- ‘opment on the cylinder determines the heat-transfer characteristics. As long ‘as the boundary layer remains laminar and well behaved, it is possible to ‘compute the heat transfer by a method similar to the boundary-layer analysis of Chap. 5. It is necessary, however, to include the pressure gradient in the analysis because this influences the boundary-layer velocity profile’ to an ap- preciable extent. In fact, it is this pressure gradient which causes a separated- flow region to develop on the back side of the cylinder when the free-stream velocity is sufficiently large. ‘The phenomenon of boundary-layer separation is indicated in Fig. 6-8. The physical reasoning which explains the phenomenon in a qualitative way is as follows. Consistent with boundary-layer theory, the pressure through the boundary layer is essentially constant at any x position on the body. In the case of the cylinder, one might measure x distance from the front stagnation point of the flinder. Thus the pressure in the boundary layer should follow that of the free stream for potential flow around a cylinder, provided this behavior would not contradict some basic principle which must apply in the boundary layer. As the flow progresses along the front side of the cylinder, the pressure would deciease and then increase along the back side of the cylinder, resulting in an increase in free-stream velocity on the front side of the cylinder and a decrease on the beck side. The transverse velocity (that velocity parallel to the surface) would decrease from a value of uw. at the outer edge of the boundary layer to 270 at the surface. As the flow proceeds to the back side of the cylinder, the Pressure increase causes a reduction in velocity in the free stream and through- ‘out the boundary layer. The pressure increase and reduction in velocity are related through the Bernoulli equation written along a streamline: dp ¢) aoa (es ° 2a. Since the pressure is assumed constant throughout the boundary layer, we note that reverse flow may begin in the boundary layer near the surface; iue., the momentum of the fluid layers near the surface is not sufficiently high to over- —— “ge, ae Flow bt‘Flow across cyinders and epherce 298 FIg.6-48 Velocity distributions Indicating tow ‘separation on ‘ cylinder in cross flow. come the increase in pressure. When the velocity gradient at the surface be- comes zero, the flow is said to have reached a separation poi . rain ptt) This separation point is indicated in Fig. 6-8. As the flow proceeds past the separation point, reverse-flow phenomena may occur, as also shown in Fig. 68. Eventually, the separated-flow region on the back side of the cylinder becomes turbulent and random in motion, The drag coefficient for bluff bodies is defined by Pla? Drag force = Fo = CoA 616) 28. where Cp is the drag coefficient and A is the frontal area of the body exposed to the flow, which, for a cylinder, is the product of diameter and length. The values of the drag coefficient for cylinders and spheres are given as a function of the Reynolds number in Figs. 6-9 and 6-10. ‘The drag force on the cylinder is a result of a combination of frictional resistance and so-called form, or pressure drag, resulting from a low-pressure region on the rear of the cylinder created by the flow-separation process. At low Reynolds numbers of the order of unity, there is no flow separation, and all the drag results from viscous friction. At Reynolds numbers of the order of 10, the friction and form drag are of the same order, while the form drag resulting from the turbulent separated-flow region predominates at Reynolds numbers greater than 1000. At Reynolds numbers of approximately 10°, based on di- ameter, the boundary-layer flow may become turbulent, resulting in a steeper velocity profile and extremely late flow separation. Consequently, the form drag is reduced, and this is represented by the break in the drag-coefficient curve at about Re = 3 X 10%. The same reasoning applies to the sphere as to the circular cylinder. Similar behavior is observed with other bluff bodies, such as elliptic cylinders and airfoils. The flow processes discussed above obviously influence the heat transfer from a heated cylinder to a fluid stream. The detailed behavior of the heat transfer from a heated cylinder to air has been investigated by Giedt (7), andi ve wot? 48 49° 2 48 ag 1? Ree undie SE ig 2 48 ge 1 vot Fig. 69 Drag costicient for circular cyinders asa funtion of the Reynolds number. fom Rot 6 the results are summarized in Fig. 6-11. At the lower Reynolds numbers (70,800 and 101,300) a minimum point in the heat-transfer coefficient occurs at ap- proximately the point of separation. There is a subsequent increase in the heat. transfer coefficient on the rear side of the cylinder, resulting from the turbulent \ eddy motion in the separated flow. At the higher Reynolds numbers two min. : imum points are observed. The first occurs at the point of transition from laminar to turbulent boundary layer, and the second minimum occurs when the turbulent boundary layer separates. There is a rapid increase in heat transfer 96 B 7 zl WO $6 Q? 46 QE 46 Gq? 46 4912 46 9? 461g 2 48 196 Fig. 610 Drag cootcient for spheres as a function ofthe Reynolds number, tom Ret. 6.Frow acros cylinders and spheres 204 8007-4 700 ‘0 Rex, son teach. soo ya aU TTA Fig. 619 Local Nusselt number for heat rans fer tom a cylinder i cross flow. trom Ret. 7 when the boundary layer becomes turbulent and another when the increased eddy motion at separation is encountered Because of the complicated nature of the flow-separation processes. itis not possible to calculate analytically the average heat-transfer coefficients in cross flow; however, correlations of the experimental data of Hilpert [8] for gases and Knudsen and Katz (9} for liquids indicate that the average heat-transfer coefficients may be calculated with ta ko where the constants C and n are tabulated in Table 6-2. The heat-transfer data for air ate plotted in Fiy, 6-12. Properties for use with Eq. (6-17) ary evaluated atthe film temperature as indicated by the subseript Figure 6-13 shows the temperature field around heated cylinders placed in transverse airstream, The dark lines are lines of constant temperature, made Visible through the use of an interferometer. Note the separated-flow region Which develops on the back side of the cylinder atthe higher Reynolds numbers andthe turbuleot field which is present in that region. We may note that the original correlation for gases omitted the Prandtt number term in £4. (6-17) with little error because most diatomic gases have 1)£282 Empinca and practical relation for otced convection hea taster Table 62 ConstantstorUsewithEa.(6-17) Based on Rets.8 and 9, eine c ” O44 0.989 0.330 440 0.911 0.385 40-4000 0.683 0.466, 4000-40,000 0.193 0.618 40,000-400,000 0.0266 0.805 .7. The introduction of the Pr'? factor follows from the previBus rea- soning in Chap. 5. Fand [21] has shown that the heat-transfer coefficients from liquids to cyl- inders in cross flow may be better represented by the relation Nu = (0.35 + 0.56 Ref) Pry? (18) This relation is valid for 10-' < Re, < 10° provided excessive free-stream tur- bulence is not encountered. In some instances, particularly those invol~ing calculations on a computer, it may be more convenient to utilize a more complicated expression than Eq, (6-17) ifit can be applied over a wider range of Reynolds numbers. Eckert and Drake [34] recommend the following relations for heat transfer from tubes in ross flow, based on the extensive study of Refs. 33 and 39: Pry Na = (045 + 00 Rer9 Pe ( ) forl 0.2. (6-21)Fow across cyiedors and spheres, 283 1b Yet wow ‘siapuIAD aBuIs 0 ewH0U BuImoy se Jo Oujood pu BuNeaY 10; ee ZEB Bld age 1 alf RE=23 0S"DIA. RE=120 10"DIA. RE=597 1S"DIA. RE=85 0S"DIA. RE*218 — 1.0°DIA RE= 1600 1.S"DIA Fig. 6-13. Interferometer photograph showing isotherms around heatec horizontal cyinders placed ina trangverse airstream. (Photograph courtesy E. Soehngen) This relation underpredicts the data somewhat in the midrange of Reynolds ‘numbers between 20,000 and 400,000, and itis suggested that the following be employed for this range: Nuy = 0.3 + 2:82 Rea! 7 [r+ ( a.) ] [aT ee ‘The heat-transfer data which were used to arrive at Eqs. (6-21) and (6-22) include fluids of air, water, and liquid sodium. Still another correlation equation is given by Whitaker [35] as for 20,000 < Re, < 400,000; Peg > 0.2 (6-22) nu = 4. ares 0.06 Rem) pee ( ) (623) % ra for 40 < Re < 10°, 0.65 < Pr < 300, and 0.25 < ju/jty < 5.2. All properties are evaluated at the free-stream temperature except that is at the wall temperature. Below Pey = 0.2, Nakai and Okazaki [38] present the following relation: Nug = [0.8237 — In(Pey®)]-! for Pey < 0.2 (24) Properties in Eqs. (6-21), (6-22), and (6-24) are evaluated at the film temperature.From across oyndets anc spheres 208. Choice of Equation for Cross Flow Over Cylinders ‘The choice of equation to use for cross flow over cylinders is subject to some conjecture. Clearly, Eq, (6-17) is easiest to use from a computational standpoint, and Eq. (6-21) is the most comprehensive. The more comprehensive relations are preferable for computer setups because of the wide range of fluids and Reynolds numbers covered. For example. Eq. (6-21) has been successful in correlating data for fluids ranging from air to liquid sodium. Equation (6-17) could not be used for liquid metals. If one were making calculations for air either relation would be satisfac Noncircular Cylinders Jakob [22] has summarized the results of experiments with heat transfer from noncirculat cylinders. Equation (6-17) is employed in order to obtain an em- pirical correlation for gases, and the constants for use with this equation are summarized in Table 6-3 Spt McAdams [10] recommends the following relation for heat transfer from spheres toa flowing gas: meee ky ‘Achenbach [43] has obtained relations applicable over a still wider range of Reynolds numbers for air with Pr = 0.71 Nu =2+ (0.25 +3 x 10 4Re!)!? for 100< Re <3 x 10° (6-26) Nu = 430 + aRe + bRe? + cRe' for 3 x 10°< Re <5 x 10° (6-27) ) for 17 < Rey < 70,000 (6-25) with a@=5xM% 6 =025xK 10% c= -31K 107 For flow of liquids past spheres, the data of Kramers [1] may be used to obtain the correlation Iya 057 008 (SY rt Regn a9 Viiet and Leppe:t 19] recommend the followmg expression for heat transfer from spheres to oil and water over a more extended range of Reynolds numbers from 1 10 200,000: Nu pres (He) = 12 + 0.5 Rey 629) a‘296 Empirical and practical eatons lr frcee-convecton neat vanser ‘Table -3 Constants for Heat Transer trom Noncicular Cyinders According to ' Fat. 22, for Uso with Ea. (6-17). Rey c 5x10 - 10° 0.286 0.588 = ge: 5x 1010: oto os 5x10 195% 10 010638 19s 1010 ouases 0.982 ' = a sx 10 0.153 0.638 4x10 ~ 1.5 x 108 03% O71 where all properties are evaluated at free-stream conditions, except ja. Which Fi is evaluated at the surface temperature of the sphere. Equation (6-26) represents \ the data of Ref. 11, as well as the more recent data of Ref. 19. ! All the above data have been brought together by Whitaker [35] to develop a single equation for gases and liquids flowing past spheres Nu = 2 + (0.4 Re}? + 0.06 Re3) Pr°* (e/a)! (6-30) which is valid for the range 3.5 < Rey <8 x 10* and 0.7 < Pr < 380. Prop- erties in Eq. (6-27) are evaluated at the free-stream temperature. EXAMPLE 6-6 Air at 1 atm and 35°C flows across a 5.0-cm-diameter cylinder at a velocity of $0 mis, , The cylinder surface is maintained at a temperature of 150°C. Calculate the heat loss \ per unit length of the eylinder. ‘Solution We first determine the Reynolds number and then find the applicable constants from Table 6-2 for use with Eq. (6-17). The properties of air are evaluated at the film tem= perature:ow across cylinders and spheres 297 045 grse o us5K . P| M0132 x 108 5 ; or = Be = “anyaessy — 0-96 kei? (0.0603 1b fy = 214 10°? kg/m + 5 (0.0486 lbh « fe} ky = 0.0312 Wim °C [0.018 Btu/h - ft - “F] Pr, = 0.695 ad _ (0.966)50)0.5) _ ET Rabo toe ~ M29 * 10 Rey From Table 6-2 C= 00265 n= 0.805 so, that from Eq. (6-17) = (0.0266)(1.129 x 105°*(0.695)" = 275.1 AS00912) 9,7 Wim VC G02 Baw =F ‘The heat transfer per unit length is therefore d(T, ~ To) = (180,5)x(0.05)(150 ~ 35) 3100 Wim (3226 Bru/t) EXAMPLE 67 ‘A fine wire having a diameter of 0.001 in (2.54 x 10~* m)is placed in a L-atm airstream at 25°C having a flow velocity of 50 m/s perpendicular to the wire. An electric current is passed through the wire, raising its surface temperature to 50°C. Calculate the heat loss per unit length. Solution We first obtain the properties at the film temperature: T, = QS + S02 = 37C = NOK y= 16.7 10-% mils k = 0.02704 Wim °C Pr, = 0.706 ‘The Reynolds number is ad _ (503.94 % 10° ty 61x ng Rey298 Empirical and practical elaons lor forced.conacton heat anser ‘The Peclet number is Pe = Re Pr = 83.3, and we find that Eqs. (6-17), (6-21), or 46-19) apply. Let us make the calculation with both the simplest expression, (6-17), and ‘the most complex, (6-21), and compare results Using Eq. (6-17) with C = 0.683 and n = 0.466, we have Nav = 48LNF 0.705" = 54 andi vale of te heaves coetient i Tne nae (8) = sens QR, = 205 wie he tet afer ut eth shen dil = nak ~ 73 = w0.98 x 10-9084 ~ 24 = 1131 Win Using £9 (21, we eau Se Nut umber as 0.62K118)"7(0.708)"9 Nive = 03 + Ty ar 70579) [1 + (118/282,000%¢ 4377 and. (4.3770.02704) - 1 Ie © 3004 Wim? °C and GIL = (3004)7(3.98 x 10°(50 ~ 25) = 9.30 Wim Here, we find the two correlations differing by 22 percent ifthe value from Eq. (6-21) is taken as corect. or 11 percent from the mean value. Examining Fig. 6 12.\we sce that data scatter of £15 percent isnot unusual EXAMPLE 6-8 Air at {atm and 27°C blows across a 12-mm-diameter sphere ata fre-stream velocity of 4 v/s. A small heater inside the sphere maintains the surface temperature at 7 Caleutate the heat lost by the sphere ! Soliton Consulting Eq. (6-30) we find that the Reynolds number is evaluated at the free-stream temperature. We therefore need the following properties: at T. = 27°C = 300 K, Y= 15.69 x 10-* m/s k= 0.02624 Wim? «°C, Pr = 0.708 He = 1.8462 5 10-5 kgim +s 350K, = 2075 x 10-5 ‘The Reynolds number is thus (40.012) Rey = = 3059 15.69 1From Eq. (6-30), Ra = 2 + (0.410599 + onyaessyyo.709%4{ LS)” 2075 = 3140 and w(t) « Shannon . 0 ea The heat transfer is then 4g = FACT, ~ To) = (68.66)(4)(0.006)°07 ~ 27) = 1.553 W For comparison purposes let us also calculate the heat-transfer coefficient using Eq. (6-25). The film temperature is T; = (350 + 300)/2 = 325 K so that Vy 1823 x We mls ky © 0.02814 Wim «°C ‘and the Reynolds number is 0.012) Res = ig33 x From Eq. (6-25) Nuy = (0.374633) = 41.73 and F is caleulated as i u(t) = EHO 9 Wigs oc d) 0.012 for about 42 percent higher than the value calculated before 4 FLOW ACROSS TUBE BANKS Because many heat-exchanger arrangements involve multiple rows of tubes, the heat-transfer characteristics for tube banks are of important practical in- terest. The heat-transfer characteristics of staggered and in-line tube banks ‘were studied by Grimson [12]. and on the basis of a correlation of the results of various investigators, he was able to represent data in the form of Eq. (6-17). The values of the constant C and the exponent m are given in Table 6- 4 in terms of the geometric parameters used to describe the tube-bundle ar- rangement. The Reynolds number is based on the maximum velocity occurring, in the tube bank, i.e., the velocity through the minimum-low area. This area will depend on the geometric tube arrangement. The nomenclature for use with Table 6-4 is shown in Fig. 6-14. The data of Table 6-4 pertain to tube banks having 10 or more rows of tubes in the direction of flow. For fewer rows the ratio of h for N rows deep to that for 10 rows is given in Table 6-5,200. Empincal and practical relaton for ores: convection reat waster ‘able 6-4 Correlation of Grimson for Heat Transfor in Tube Banks of 10 Rows or More, From ol. 12, for Use with Eq. (6-17) 20 30 1.25 0386 0.592 0.305 0.608 0.111 0.704 0.0703. 0.752 1S 0.407 0.586 0.278 0.620 0.112 0.702 0.0753. 0.744 20 0.464 0.570 0.332 0.602 0.254 0632-020 0.648 3.00322 0.601 0.415 0.581 0317 __ 0.608 06 one) 09 — = = = 049s os 0.445 0.581 1.0 0552 05 = = 1125 — = 031 0.565 0.575 0.560 125 0.554 0.576 0.556 0379 0.562. 1s 0.562 0.502 0.568 0582 0.568 20 0568 0.535 0.556 0.498 0.370 3.0 0.580 0.488 0.562 0.467 0.574 bs, 1 Fig. 6-14 Nomenciatute for use wih Table 6-4 (a) inti tube “ ‘ows; (b) staggered tube rons,Fiow across tube barks 304 Table 6-5 Ratio of h for N Rows Deep to that for 10 Rows Deep. From Ref. 17. N 12 3 4 $ 6 7 8 9 Ratio for staggered 068 0.75 0.83 0.89 0.92 0.95 0.97 098 0.99 1.0 tubes Ratio for in-line 0.64 0.80 0.87 0.90 0.92 0.94 0.9% 098 0.99 1.0 tubes Pressure drop for flow of gases over a bank of tubes may be calculated with Eq. (6-31), expressed in pascals: Ap = LGiasN (2) (631) pw where Grax = mass velocity at minimum flow area, ke/m? - s ’D = density evaluated at free-stream conditions, kg/m? N = number of transverse rows by = average free-stream viscosity ‘The empirical friction factor J” is given by Jakob [18] as 25 + OB __} Re ate fuss + qth a} ec oxy for staggered tube arrangements, and 0.088 /d 1G, = did f {ooss + aman} Regn (633) for in-line arrangements. Zukauskas {39] has presented additional information for tube bundles which takes into account wide ranges of Reynolds numbers and property variations, The correlating equation takes the form MC Retna POM (®) 634) Pre where all properties except Pr, are evaluated at T- and the values of the constants are given in Table 6-6 for greater than 20 rows of tubes. This equation is applicable for 0.7 < Pr < S00 and 10 < Revimay < 10°. For gases the Prandtl number ratio has little influence and is dropped. Once again, note that the Reynolds number is based on the maximum velocity in the tube bundle. For tess than 20 rows in the direction of flow the correction factor in Table 6-7 should be applied. It is essentially the same as for the Grimson correlation ‘Additional information is given by Morgan [44]. Further information on pressure drop is given in Ref. 39,302 Empict and practical relations for forced-convecon heat ansier ‘Table 6-6 Constant for Zukauskes Correlation (Eq, (6-34) for Heat Transtor in Tube Banks of 20 Rows or More. From Ret. 99, onary Rema é ; inioe 1-0 08 0 (350i tes wae on os tie om om Sugeest stn os 0 (3S hae ini ne 19 - 2x 10° Sa)? 0 Se 0.60 038 (8 ae <2 W241 guygeSoe 08 sx own ont ‘The reader should keep in mind that these relations correlate experimental data with an accuracy of about +25 percent. EXAMPLE 6-9 Air at 1 atm and 10°C flows across a bank of tubes 15 rows high and $ rows deep at a velocity of 7 m/s measured at a point in the flow before the air enters the tube bank. ‘The surfaces of the tubes are maintained at 65°C. The diameter of the tubes is 1 in (2.54 cml; they are arranged in an in-line manner so that the spacing in both the normal and parallel directions to the flow is 1.5 in [3.81 cm]. Calculate the total heat transfer per ‘unit length for the tube bank and the exit air temperature. ‘Solution ‘The constants for use with Eq. (6-17) may be obtained from Table 6-4, using. Sal S381, d 234 d 2560 sothat C= 0a n= 0620 ‘Table 6-7 Ratio of for N Rows Deep to that for 20 Rows Deep According to Ret. 99 and for Use with Eq, (6-34). Ny 2 3 4 § 6 8 WW wb Staggered 0.77 0.84 0.89 092 094 097 0.98 0.99 10 Inline 0.70 0.80_0.90_0.92_0.94 0.97 (0.98 _(0.99_1.0The properties of air are evaluated at the film temperature, which at entrance to the tube bank is T.+T. 65+ an) | 10132 x 10° RT G8nGI0.5 by = 1.894 10? kim = ky = 0.027 Wim -°C (0.0156 Btu/h - R= °F] p= 1007 Jikg + °C [0.24 Buus, - °F] 0.706 = 37SC = MOSK [558.9°R] A Then L137 kei To calculate the maximum velocity, we must determine the minimum fowarea. From Fig. 614 we find thatthe ratio ofthe minimum Row atea (othe total fromal area (Se ~ dS. The maximum velocity is thus Sa 18.81) SORE = 2 mls (68.9 105) @ where us is the incoming velocity before entrance to the tube bank. The Reynolds ‘number is computed by using the maximum velocity. pial _ (1137)2190,0254) LRM x IDS Re 32,000 o The heat-transfer coefficient is then calculated with Eq. (6-17) #4 - zres2.0n0re0.306" = 1838 6 43 sy0.027 . - 15.00.02 og, win C288 Ua 2007 ig, win HC R88 BUNA? OF) This is the heat-transfer coefficient which would he obtained if there were 10 sows of tubes in the direction of the flow. Because there are only 5 rows, this value must be ‘multiplied by the factor 0,92, as determined from Table 6-5 The total surface area for heat transfer, considering unit length of tubes. is AN ml) = (ASNS)er(0.0254) » 5.985 me Where Nis the total number of tubes, Before calculating the heat transfer. we must recognize that the ait temperature increases as the air lows through the tube bank. Therefore. this must be taken into ‘account when wsing 4 = AT. = To © ‘As a good approximation. we can use an arithmetic average value of T. and write for the energy balance BE) iedTs = Tod o wa (n= Eat204 Empirical and practos relations lo torcad-convecton heat ranaler where now the subscripts 1 and 2 designate entrance and exit to the tube bank. The ‘mass flow at entrance to the 15 tubes is. mh = paulI3)S, _p _ 10132 < 10° RT. ~~ 287,283) 1.247)(7K15}10.0381) = 4.99 kg/s [11.0 Iba} so that Eq. () becomes 1.246 kg/m? @ cornensses (5-258) « asmneeers~ 1 which may be solved to give 19.08°C ‘The heat transfer is then obtained from the right side of Eq, (f) q = (4.99)1005)(19.08 ~ 10) = 45.6 kWim ‘This answer could be improved somewhat by recalculating the air properties based on ‘a mean value of T., but the improvement would be small and well within the accuracy of the empirical heat-transfer correlation of Eq. (6-17) EXAMPLE 6-10 ‘Compare the heat-transfer coefficient calculated with Eq. (6-34) with the value obtained in Example 6-9. Solution Properties for use in Eq, (6-34) are evaluated at free-stream conditions of 10°C, so we have Y= 1486 x 10 Pr= O71 k= 0.049 The Reynolds number is (210.0254) Resme = 7486 x 10 35,895 so that the constants for Eq. (6-34) are C = 0.27 and m = 0.63, Inserting values, we obtain . omusassesonps « 168 and (176.890.0249) wer 173.3 Wim? °C for a value about 9 percent higher than in Example 6-9. Both values are within the accuracies of the correlations.Louis metal neat transfer 908 96.5 LIQUID-METAL HEAT TRANSFER In recent years concentrated interest has been placed on liquid-metal heat transfer because of the high heat-transfer rates which may be achieved with these media. These high heat-transfer rates result from the high thermal con- ductivities of liquid metals as compared with other fluids; as a consequence they are particularly applicable to situations where large energy quantities must be removed from a relatively small space, as in a nuclear reactor. In addition, the liquid metals remain in the liquid state at higher temperatures than con- ventional fluids like water and various organic coolants. This also makes more compact heat-exchanger design possible. Liquid metals are difficult to handle because of their corrosive nature and the violent action whigh may result when they come into contact with water or air; even so, their advantages in certain heat-transfer applications have overshadowed their shortcomings, and suitable techniques for handling them have been developed. Leet us first consider the simple flat plate with a liquid metal flowing across it, The Prandtl number for liquid metals is very low, of the order of 0.01. so that the thermal-boundary-layer thickness should be substantially larger than the hydrodynamic-boundary-layer-thickness. The situation results from the high values of thermal conductivity for liquid metals and is depicted in Fig. 6-15. Since the ratio of 6/5, is small, the velocity profile has a very blunt shape over ‘most of the thermal boundary layer. AS a first approximation, then, we might assume a slug-flow model for calculation of the heat transfer; i-c.. we take wea (635) throughout the thermal boundary layer for purposes of computing the energy transport term in the integral energy equation (Sec. 5-6); £ [ffir ray] 9], (639 Fig. 615 Boundarylayer regimes for analysis! touic-metal Hest transter1308 Empincal and practcal ators fr forced convection hea anser The conditions on the temperature profile are the same as those in Sec. 5-6, 0 that we use the cubic parabola as before: Oe @.° T.- 7. 28, Inserting Eqs. (6-31) and (6-33) in (6-32) gives. aun {f° [i 5 3k + 5 (%)] ay} (6-38) which may be integrated to give 37 28, d8, = dx (639) ‘The solution to this differential equation is (6-40) for a plate heated over its entire length. ‘The heat-transfer coefficient may be expressed by =KaTayre _ 3k _ V3 ear 76 Vas ow This relationship may be put in dimensionless form as ‘Nu, = te = 0,530 (Re, Pr)’? = 0.530 Pe!? (6-42) Using Eq. (5-21) for the hydrodynamic-boundary-layer thickness, 3 468 x” Rel? (6-43) we may compute the ratio 4/5,: uy 4 VP i = 1.64 VPr (6-44) Using Pr ~ 0.01, we obtain which is in reasonable agreement with our slug-flow model. The flow model discussed above illustrates the general nature of liquid-metal heat transfer, and it is important to note that the heat transfer is dependent onLousimeta neat raster 207 the Peclet number. Empirical correlations are usually expressed in terms of this parameter, four of which we present below. Extensive data on liquid metals are given in Ref. 13, and the heat-transfer characteristics are summarized in Ref. 23. Lubarsky and Kaufman [14] rec- ‘ommended the following relation for calculation of heat-transfer cvefficients in fully developed turbulent flow of liquid metals in smooth tubes with uniform heat flux at the wall: hd Nuy = 0,625 (Rey Pry'* (6-45) [All properties for use in Eq. (6-45) are evaluated at the bulk temperature. Equation (6-45) is valid for 102 < Pe < 10" and for Lid > 60. Seban and Shi mazaki [16] propose the following relation for calculation of heat transfer t0 liquid metals in tubes with constant wall temperature: Nug = 5.0 + 0.025 (Rey Pr)?* (6-46) where all properties are evaluated at the bulk temperature. Equation (6-42) is valid for Pe > 10? and Lid > 60. ‘More recent data by Skupinshi, Tortel, and Vautrey (26) with sodium- potassium mixtures indicate that the following relation may be preferable to that of Eq. (6-45) for constant-heat-flux conditions: Nu = 4.82 + 0.0185 Pe°®” 647) This relation is valid for 3.6 x 10° < Re < 9.05 x 10% and 10? < Pe < 10. ‘Witte [32] has measured the heat transfer from a sphere to liquid sodium during forced convection, with the data being correlated by Nu = 2 + 0.386 (Re Pr (6-48) for the Reynolds number range 3.56 x 10* < Re < 1.525 x 10%. Kalish and Dwyer [41] have presented information on liquid-metal heat trans- fer in tube bundles. In general, there are many open questions concerning liquid-metal heat trans fer, and the reader is referred to Refs. 13 and 23 for more information. EXAMPLE 6-11 Liquid bismuth flows ata rate of 4.5 kg/s through a 5.0-cm-diameter stainless-steel tube. ‘The bismuth enters at 415°C and is heated to 440°C as it passes through the tube. If constant heat flux is maintained along the tube and the tube wall is at a temperature 20°C higher than the bismuth bulk temperature, calculate the length of tube required to effect the heat transfer. Suton ‘Because a constant heat flux is maintained, we may use Eq. (6-47) t6 calculate the heat- transfer coefficient. The properties of bismuth are evaluated at the average bulk tem-are (ors) (9) ws 9) 9) 9) saquina woqwonbey soy 1uainqum padojanap Aung 1 ot < Jagr pare “moy seurure) podo}24ap syn, Jeune “1 o> To “Ol xs > "4 > or “0007 > 4d > $0 “moy 1uaingimy padorasap Atm P oor > > 01 ‘moy wainqun, ‘Moy wajnqim padoranap Ain O01 > 4d > 9.0 S8uyoo> soy ¢'9 = 4 Aurreay 10} 59 = 4 ‘mous wamngim padoranop Aung 8 ybaw Bo (aba Fe hd ag Uwoneja1 aoymiag, T yoy =" sel) otto se = = CG) oatoron sav =m el eo? ZOO = PON, a ag s2qm ysnoy mou 2qnu, mou qn oy aan, oraz ouenua ROU QL mou 9qn. moy qn. Gaawe2 (oerene Avadoxs 0) 18) 095) suoneje4 uoWoanLED-peDI0, JO AMUAUNG @-9 O1FEL01 Ue9) oe urs (0£-9) 01 (57-9) (79) 0 (07-9) 0 (81-9) az ws os) x91 295 sol > 94 > OL os > 44 > 0 1x91 995 x21 225 1x91 295 cosed >a > Ol 00'00r > "2H > FO mou aqm 40 Uonenba sejnanied se ures 9 ge, 903 ¥ ome 9 ead “dt 2 = AN £9 91981 995, = PAN Be 0 "ON 7-9 aIqe. wou} w pue > wld £2 = 70 dw won ‘PUP auimeapcy jo siseq uo payenjeas soquinu splousay, sreiow pinbri ssqueg 2am Ss019e MOL syueg 2qm ssou2e mold sarayds 80398 MOLL si9puyao jnasiauou $8019" OI su9punyd> ssou> OL srapuyso 50198 MOL srapus 80190 OL sionp seynou}9UON340 Empirical and practia!relanons fr lorced-canvecon heal transl erature of (415 + 440VE = 427.5°C 134 107 kgim $13,282 Ibn =] p= 0.149 ki/kg - °C (0.0356 Btulib,. - °F] 15.6 Wim -°C (9.014 Btu/h ft °F] 0.013 ‘The total transfer is calculated from ® " y 4 = ric, ATs = (4,5)(149)440 ~ 415) = 16.76 KW [57,186 Btwh] (a) We calculate the Reynolds and Peclet numbers as 054s) (m(0.05)7/4](1.34 x 10) ae o Pe = Re Pr = (552000019 = Hit ‘The heat-transfer coefficient is then calculated from Eq. 6-47 Nu, = 4.82 + (0.0185)1111)°"” = 10.93 Co) = MORNE sa What C0 Be) ‘The total required surface area of the tube may now be computed from q= AAT. ~ 7) @ ‘where we use the temperature difference of 20°C; 0.246 m* [2.65 £0] This area in turn can be expressed in terms of the tube length 0.246, 710.05) A mdb and Ls7m (5.15 8) 66 SUMMARY REMARKS In contrast to Chap. 5, which was mainly analytical in character, this chapter hhas dealt almost entirely with empirical correlations which may be used to calculate convection heat transfer. The general calculation procedure is as follows: 1, Establish the geometry of the situation. 2. Make a preliminary determination of appropriate fluid properties. 3. Establish the flow regime by calculating the Reynolds or Peclet number.Problems a4 4, Select an equation which fits the geometry and flow regime and reevaluate properties, if necessary, in accordance with stipulations and the equation. 5, Proceed to calculate the value of h and/or the heat-transfer rate, We should note that the data upon which the empirical equations are based ‘ste most often taken under laboratory conditions where it is possible to exert careful control over temperature and flow variables. In a practical application such careful control may not be present and there may be deviations from heat- transfer rates calculated from the equations given here. Our purpose is not to incourage the reader by this remark, but rather to indicate that sometimes it wil be quite satisfactory to use a simple correlation over a more elaborate expression even if the simple relation has a larger scatter in its data represen- tation, Our purpose has been to present a variety of expressions (where avail- able) so that some choices can be made. Finally. the most important relations of this chapter are listed in Table 6-8 for quick reference purposes. REVIEW QUESTIONS 1 What is the Dittus-Boelter equation? When does it apply? 2 How may heat-transfer coefficients be calculated for low in rough pipes? 3. What is the hydraulic diameter? When is it used? 4. What is the form of equation used to calculate heat transfer for flow over cylinders and bluff bodies? 5 Why does a slug-flow model yield reasonable results when applied to liquid-metal heat transfer? What is the Peclet number? What is the Graetz number? PROBLEMS 61 Engine oil enters a 5.0-mm-diameter tube at 120°C. The tube wall is maintained at S0°C, and the inlet Reynolds number is 1000. Calculate the heat transfer, average heat-transfer coefficient, and exit oil temperature for tube lengths of 10, 20, and 50 em, 6:2 Water at an average bulk temperature of 80°F flows inside a horizontal smooth tube with wall temperature maintained at 180°F. The tube length is 6.0 ft, and diameter is 0.125 in. The flow velocity is 0.125 fs. Calculate the heat-transfer rate. 63 Liquid ammonia flows in a duct having a cross section of an equilateral triangle 0'cm on @ side. The average bulk temperature is 20°C, and the duct wall tem-2 or e10 et on on ow Empirea and practea relations lor forced con perature is $0°C. Fully developed laminar flow is experienced with a Reynolds ‘number of 1000. Calculate the heat transfer per unit length of duct. Water flows in a duct having a cross section $ x 10 mm with a mean bulk temperature of 20°C. If the duct wall temperature is constant at 60°C and fully ‘developed laminar flow is experienced, calculate the heat transfer per unit length Water at the rate of 3 kp/s is heated from 5 to 15°C by passing it through a S-cm- ID copper tube. The tube wall temperature is maintained at 90°C, What is the length of the tube? Water at the rate of 0.8 kp/s is heated from 35 to 40°C in a 2.S-cm-diameter tube whose surface is at 90°C. How long must the tube be to accomplish this heating? Water flows through a 2.5-cm-ID pipe 1.5 m long ata rate of 1.0 ke/s. The pressure drop is 7 kPa through the 1.5-m length. The pipe wall temperature is maintained at a constant temperature of 50°C by a condensing vapor, and the inlet water temperature is 20°C, Estimate the exit water temperature. Water at the rate of 1 kgis is forced through a tube with a 2.5-cm ID. The inlet Water temperature is 15°C, and the outlet water temperature is S0°C. The tube wall temperature is 14°C higher than the water temperature all along the length Of the tube. What is the length of the tube? Engine oil enters a 1,25-cm-diameter tube 3 m long at a temperature of 38°C. The tube wall temperature is maintained at 65°C, and the flow velocity is 30 em/s. Estimate the total heat transfer to the oil and the exit temperature of the oil Air at 1 atm and 15°C flows through a long rectangular duct 7.5 by 15.em. A 1.8- 1m section of the duct is maintained at 120°C, and the average air temperature at exit from this section is 65°C. Calculate the air flow rate and the total heat transfer. A heat exchanger is constructed so that hot flue gases at 700 K flow inside a 2.5- em-ID copper tube with 1.6-mm wall thickness. A 5.0-cm tube is placed around the 2.5-cm-diameter tube, and high-pressure water at 150°C flows in the annular space between the tubes. If the flow rate of water is 1.5 kg/s and the total heat transfer is 17.5 kW, estimate the length of the heat exchanger for a gas mass flow of 0.8 ke/s. Assume that the properties of the flue gas are the same as those of air at atmospheric pressure and 700 K. Water at the rate of 0.5 kg/s is forced through a smooth 2.5-em-ID tube 15 m long. The inlet water temperature is 10°C, and the tube wall temperature is 15°C higher than the water temperature all along the length of the tube. What is the ‘exit water temperature? Water at an average temperature of 300 K flows at 0.7 ke/s in a 2.S-cm-diameter tube 6 m long. The pressure drop is measured as 2 kPa. A constant heat flux is imposed, and the average wall temperature is 55°C. Estimate the exit temperature of the water, Anoil with Pr = 1960,p = 860kg/m?, V = 1.6 x 10~+ms, andk = 0.14Wim «°C ‘enters a 2.5-mm-diameter tube 60 cm long. The oil entrance temperature is 20°C,615 616 or e190 620 on on Problems 349 the mean flow velocity is 30 cmvs, and the tube wall temperature is 120°C. Calculate the heat-transfer rate. Liquid ammonia flows through a 2.5-cm-diameter smooth tube 2.5 m long at a rate of I Iba/s. The ammonia enters at 10°C and leaves at 38°C, and a constant heat flux is imposed on the tube wall. Calculate the average wall temperature necessary to effect the indicated heat transfer. Liquid Freon 12 (CC1F,) flows inside a 1.25-cm-diameter tube at a velocity of 3 ims, Calculate the heat-transfer coefficient for a bulk temperature of 10°C, How ddoes this compare with water at the same conditions? Water at an average temperature of 10°C flows in a 2.5-cm-diameter tube 6m Jong at a rate of 0.4 kg/s. The pressure drop is measured as 3 kPa. A constant heat flux is imposed, and the average wall temperature is 50°C, Estimate the exit temperature of the water Water at the rate of 0.5 kp/s is to be cooled from 71 to 32°C. Which would result in less pressure drop—to run the water through a 12.S-mm-diameter pipe at a constant temperature of 4°C or through a constant-temperature 25-mm-diameter pipe at 20°C? Air at 1400 kPa enters a duct 7.5 em in diameter and 6 m long at a rate of 0.5 kes. The duct wall is maintained at an average temperature of 500 K. The average air temperature in the duct is 550 K. Estimate the decrease in temperature of the air as it passes through the duct ‘An annulus consists of the region between two concentric tubes having diameters ‘of 4 cm and 5 em. Ethylene glycol flows in this space at a velocity of 6.9 mis. Entrance temperature is 20°C and the exit temperature is 40°C. Only the inner tube is a heating surface and itis maintained constant at 80°C. Calculate the length of annulus necessary to effect the heat transfer. ‘An air-conditioning duct has a cross section of 45 by 90 cm. Air flows in the duct at a velocity of 7.5 m/s at conditions of 1 atm and 300 K. Calculate the heat- transfer coefficient for this system and the pressure drop per unit length. Water flows in a 3.0-cm-diameter tube having a relative roughness of 0.002 with ‘a constant wall temperature of 80°C. If the water enters at 20°C, estimate the convection coefiicient for a Reynolds number of 10° Liquid Freon 12 (CCLF,) enters a 3.5-mm-diameter tube at 0°C and with a flow rate such that the Reynolds number is 700 at entrance conditions. Calculate the length of tube necessary to raise the fluid temperature to 20°C if the tube wall temperature is constant at 40°C. Air enters a small duct having a cross section of an equilateral triangle, 3.0 mm fon a side. The entering temperature is 27°C and the exit temperature is 77°C. If the flow rate is 5 x 10-* kg/s and the tube length is 30 cm, calculate the tube wall temperature necessary to effect the heat transfer. Also calculate the pressure drop. The pressure is 1 atm. 74 oa 628 620 62 633 oo4 635 636 oa7 Epica and practice latons to ood convecton heat aster Air at 90 kPa and 27°C enters a 4.0-mm-diameter tube with a mass flow rate of 7 x 10-2 kg/s. A constant heat flux is imposed at the tube surface so that the tube wall temperature is 70°C above the fluid bulk temperature. Calculate the exit air temperature for a tube length of 12 em. ‘Air at 110 kPa and 40°C enters a 6.0-mm-diameter tube with a mass flow rate of 8 x 10-5 kg/s, The tube wall temperature is maintained constant at 140°C. Cal- ‘culate the exit air temperature for a tube length of 14 cm. Engine oil at 40°C enters a I-cm-diameter tube ata flow rate such that the Reynolds number at entrance is $0. Calculate the exit oil temperature for a tube length of 8 cm and a constant tube wall temperature of 80°C. Water flows in a 2-cm-diameter tube at an average flow velocity of 8 mvs. If the ‘water enters at 20°C and leaves at 30°C and the tube length is 10 m, estimate the average wall temperature necessary to effect the required heat transfer. Engine oil at 20°C enters a 2.0-mm-diameter tube at a velocity of 1.2 m/s. The tube wall temperature is constant at 60°C and the tube is 1.0 m long. Calculate the exit oil temperature. Water flows inside a smooth tube at a mean flow velocity of 10 ft/s. The tube diameter is 1.0 inch and a constant heat flux condition is maintained at the tube wall such that the tube temperature is always 20°C above the water temperature. ‘The water enters the tube at 30°C and leaves at SOFC. Calculate the tube length necessary to accomplish the indicated heating. Water enters a 3-mm-diameter tube at 21°C and leaves at 32°C. The flow rate is such that the Reynolds number is 600. The tube length is 10 cm and is maintained at a constant temperature of 60°C. Calculate the water flow rate. Water enters a 3.0-cm-diameter tube at 60°F and leaves at 100°F. The flow rate is 1.0 kg/s and the tube wall temperature is 140°F. Calculate the length of the tube. Glycerin flows in a S-mm-diameter tube at such a rate that the Reynolds number is 10. The glycerine enters at 10°C and leaves at 30°C. The tube wall is maintained ‘constant at 40°C. Calculate the length of the tube. A S-cm-diameter cylinder maintained at 100°C is placed in a nitrogen flow stream ‘at 2 atm pressure and 10°C. The nitrogen flows across the cylinder with a velocity ‘of $ mis. Calculate the heat lost by the cylinder per meter of length Airat | atm and 0°C blows across a4-cm-diameter cylinder maintained at a surface temperature of S4°C. The air velocity is 25 m/s. Calculate the heat loss from the cylinder per unit length, Air at 200 kPa blows across a 20-cm-diameter cylinder at a velocity of 25 m/s and temperature of 10°C. The cylinder is maintained at a constant temperature of 80°C. Calculate the heat transfer and drag force per unit length, Water at 43°C enters a S.cm-ID pipe having a relative roughness of 0.002 at a rate of 6 ke/s. If the pipe is 9 m long and is maintained at 71°C, calculate the exit water temperature and the total heat transfer.ose 639 640 ear Proves 998 ‘A short tube is 6.4 mm in diameter and 15 em long. Water enters the tube at 1.5 ims and 38°C, and a constant-heat-lux condition is maintained such that the tube ‘wall temperature remains 28°C above the water bulk temperature. Calculate the hheat-transfer rate and exit water temperature. Ethylene glyco! is to be cooled from 60 to 40°C in a 3.0-em-diameter tube. The tube wall temperature is maintained constant at 20°C. The glycol enters the tube with a velocity of 10 m/s. Calculate the length of tube necessary to accomplish this cooling Using the values of the local Nusselt number given in Fig. 6-11, obtain values for the average Nusselt number as a function of the Reynolds number. Plot the results as log Nu versus log Re, and obtain an equation which represents all the data. Compare this correlation with that given by Eq. (6-17) and Table 6-2. Air at 70 kPa and 20°C flows across a S-cm-diameter cylinder at a velocity of 20 mis. Compute the drag force exerted on the cylinder. A heated cylinder at 450 K and 2.5 cm in diameter is placed in an atmospheric airstream at 1 atm and 325 K. The air velocity is 30 mis. Calculate the heat loss per meter of length for the cylinder. Assuming that a human can be approximated by a cylinder 1 ft in diameter and 6 ft high with a surface temperature of 75°F, calculate the heat the person would lose while standing in a 30-mi/h wind whose temperature is 30°F. Assume that one-half the heat transfer from a cylinder in cross flow occurs on the front half ofthe cylinder. On this assumption, compare the heat transfer from ‘a cylinder in cross flow with the heat transfer from a flat plate having a length ‘equal to the distance from the stagnation point on the cylinder. Discuss this ‘comparison. ‘A.0.13-mm-diameter wire is exposed to an airstream at ~30°C and 54 kPa. The low velocity is 230 m/s, The wire is electrically heated and is 12.5 mm long, Calculate the electric power necessary to maintain the wire surface temperature at 175°C. Air at 90°C and 1 atm flows past a heated sh-in-diameter wire at a velocity of 6 rms. The wire is heated toa temperature of 150°C. Calculate the heat transfer per unit length of wire. A fine wire 0.025 mm in diameter and 15 cm long is to be used to sense flow velocity by measuring the electrical heat which can be dissipated from the wire when placed in an airflow stream, The resistivity of the wire is 70 nS? - em. The temperature of the wire is determined by measuring its electric resistance relative to some reference temperature Ty so that R= Rll + aT ~ Td For this particular wire the value of the temperature coefficient a is 0.006°C.* ‘The resistance can be determined from measurements of the current and voltage impressed on the wire, andae oss os7 ess 650 660 Empirical and practical relations fr lrced-convecton heal tansy ‘Suppose a measurement is made for air at 20°C with a flow velocity of 10 mis ‘and the wire temperature is 40°C. What values of voltage and current would be ‘measured for these conditions if Re is evaluated at Ty = 20°C? What values of voltage and current would be measured for the same wire temperature but flow velocities of 15 and 20 m/s? ‘Helium at 1 atm and 325 K flows across a 3-mm-diameter cylinder which is heated to 425 K. The flow velocity is 9 mvs. Calculate the heat transfer per unit length of wire. How does this compare with the heat transfer for air under the same conditions? Calculate the heat-transfer rate per unit length for flow over a 0.025-mm-diameter cylinder maintained at 65°C. Perform the calculation for (a) air at 20°C and 1 atm and (6) water at 20°C; ue = 6 mls. Compare the heat-transfer results of Eqs. (6-17) and (6-18) for water at Reynolds ‘numbers of 10%, 10, and 10 and a film temperature of 90°C. Compare Eqs. (6-19), (6-20), and (6-21) with Eq. (6-17) for a gas with Pr = 0.7 at the following Reynolds numbers: (a) 500, (4) 1000, (c) 2000, (d) 10,000, (e) 100,000. A pipeline in the Arctic carries hot oil at $0°C. A strong arctic wind blows across the 50-cm-diameter pipe at a velocity of 13 m/s and a temperature of —35°C. Estimate the heat loss per meter of pipe length. ‘Two tubes are available, a 4.0-cm-diameter tube and a 4.0-cm-square tube. Air at 1 atm and 27°C is blown across the tubes with a velocity of 20 mis. Calculate the heat transfer in each case if the tube wall temperature is maintained at 50°C. ‘A3.0-cm-diameter cylinder is subjected to cross flow of carbon dioxide at 200°C and a pressure of 1 atm. The cylinder is maintained at a constant temperature of. 50°C and the carbon dioxide velocity is 40 m/s. Calculate the heat transfer to the cylinder per meter of length. Helium at 150 kPa and 20°C is forced at 50 m/s across a horizontal cylinder having a diameter of 30 cm and a length of 6 m. Calculate the heat lost by the cylinder if its surface temperature is maintained constant at 100°C. ‘A.0.25-inch-diameter cylinder is maintained at a constant temperature of 300°C and placed in a cross flow of CO; at p = 100 kPa and T = 30°C. Calculate the heat loss for a 4.5-m length of the cylinder if the CO, velocity is 35 ms. A 20-cm-diameter cylinder is placed in a cross low CO: stream at 1 atm and 300 K. The cylinder is maintained at a constant temperature of 400 K and the CO, velocity is 50 m/s, Calculate the heat lost by the cylinder per meter of length Air lows across a 4-cm-square cylinder at a velocity of 10 mis. The surface temperature is maintained at 85°C. Free-stream air conditions are 20°C and 0.6 atm. Calculate the heat loss from the cylinder per meter of length. Water flows over a 3-mm-diameter sphere at 6 m/s. The free-stream temperature is 38°C, and the sphere is maintained at 93°C, Calculate the heat-transfer rate. A spherical water droplet having a diameter of 1.3 mm is allowed to fall fromost en Problems 347 rest in atmospheric air at 1 atm and 20°C. Estimate the velocities the droplet will attain after a drop of 30, 60, and 300 m. AA spherical tank having a diameter of 4.0 m is maintained at a surface temperature of 40°C. Air at | atm and 20°C blows across the tank at 6 m/s, Calculate the heat loss. Anheated sphere having a diameter of 3 cm is maintained at a constant temperature cof 90°C and placed in a water flow stream at 20°C. The water flow velocity is 3.5 rms, Calculate the heat lost by the sphere. A small sphere having a diameter of 6 mm has an electric heating coil inside which maintains the outer surface temperature at 220°C. The sphere is exposed to an airstream at 1 atm and 20°C with a velocity of 20 mvs. Calculate the heating rate which must be supplied to the sphere. Air at a pressure of 3 atm blows over a flat plate at a velocity of 100 m/s. The plate is maintained at 200°C and the free-stream temperature is 30°C. Calculate the heat loss for a plate which is 1 m square Air at 3.5 MPa and 38°C flows across a tube bank consisting of 400 tubes of 1.25- ‘em OD arranged in a staggered manner 20 rows high; S, = 3.75 cmand S, = 2.5 ‘em. The incoming-flow velocity is 9 m/s, and the tube wall temperatures are ‘maintained constant at 200°C by a condensing vapor on the inside of the tubes. The length of the tubes is 1.5 m. Estimate the exit air temperature as it leaves the tube bank, ‘A tube bank uses an in-line arrangement with $, = §, = 1.9 em and 6.33-mm- diameter tubes. Six rows of tubes are employed with a stack 50 tubes high. The surface temperature of the tubes is constant at 90°C, and atmospheric air at 20°C is forced across them at an inlet velocity of 4.5 mis before the flow enters the tube bank. Calculate the total heat transfer per unit length for the tube bank. Estimate the pressure drop for this arrangement. Repeat Prob. 6-66 for a staggered-tube arrangement with the same values of S, and 5, ‘A more compact version ofthe tube bank in Prob. 6-66 can be achieved by reducing. the S, and 5, dimensions while stil retaining the same number of tubes. Investigate the effect of reducing S, and S, in half, that is, $, = S, = 0.95 cm. Calculate the heat transfer and pressure drop for this new arrangement. Condensing steam at 150°C is used on the inside of a bank of tubes to heat a ‘cross-flow stream of CO; which enters at 3 atm, 35°C, and 5 m/s. The tube bank consists of 100 tubes of 1.25-cm OD in a square in-line array with $, = S, = 1.875 ‘em. The tubes are 60 cm long. Assuming the outside tube wall temperature is constant at 150°C, calculate the overall heat transfer to the CO; and its exit ‘temperature ‘An in-line tube bank is constructed of 2,5-cm-diameter tubes with 15 rows high and 7 rows deep. The tubes are maintained at 90°C, and atmospheric air is blown across them at 20°C and.u. = 12 mvs. The arrangement has $, = 3.75andS, = 5.0 em. Calculate the heat transfer from the tube bank per meter of length. Also calculate the pressure drop.210 en on 673 om 75 676 on 67 679 Empical anc practical relton for lrcea-convecton neat vansor ‘Air at 300 K and 1 atm enters an in-line tube bank consisting of five rows of 10 tubes each. The tube diameter is 2.5 em and 5, = S, = 5.0 cm. The incoming velocity is 10 mvs and the tube wall temperatures are constant at 350 K. Calculate the exit air temperature. Atmospheric air at 20°C flows across a S-cm-square rod at a velocity of 15 mi, ‘The velocity is normal to one of the faces of the rod. Calculate the heat transfer per unit length for a surface temperature of 50°C. A certain home electric heater uses thin metal strips to dissipate heat. The strips are 6 mm wide and are oriented normal to the airstream, which is produced by ‘a small fan. The air velocity is 2 mis, and seven 35-cm strips are employed. If the strips are heated to 870°C, estimate the total convection heat transfer to the room air at 20°C. (Note that in such a heater, much of the ‘otal transfer will be by thermal radiation.) ‘A square duct, 30 by 30 cm, is maintained at a constant temperature of 30°C and ‘an airstream of 50°C and | atm is forced across it with a velocity of 6 mvs, Calculate the heat gained by the duct. How much would the heat flow be reduced if the flow velocity were reduced in half? ‘The drag coefficient for a sphere at Reynolds numbers less than 100 may be approximated by C, = b Re ', where bis a constant, Assuming that the Colburn analogy between heat transfer and fluid friction applies, derive an expression for the heat loss from a sphere of diameter d and temperature T,, released from rest and allowed to fallin a uid of temperature T.. (Obtain an expression for the heat lost between the time the sphere is released and the time it reaches some velocity , Assume that the Reynolds number is less than 100 during this time and that the sphere remains at a constant temperature.) Using the slug-flow model, show that the boundary-layer energy equation reduces to the same form as the transient-conduction equation for the semi-infinite solid lof Sec. 4-3. Solve this equation and compare the solution with the integral analysis of Sec. 65, Liquid bismuth enters a 2.5-cm-diameter stainless-steel pipe at 400°C at a rate of 1 kg/s. The tube wall temperature is maintained constant at 450°C. Calculate the bismuth exit temperature ifthe tube is 60 cm long. Liquid sodium is to be heated from 120 to 149°C at a rate of 2.3 kgls. A 2.5-cm- diameter electrically heated tube is available (constant heat flux). Ifthe tube wall temperature is not to exceed 200°C, calculate the minimum length required ‘Determine an expression for the average Nusselt number for liquid metals flowing ‘over a flat plate. Use Eq. (6-42) as a starting point. Water at the rate of 0.8 kg/s at 93°C is forced through a S.cm-ID copper tube at a suitable velocity. The wall thickness is 0.8 mm. Air at 15°C and atmospheric pressure is forced over the outside ofthe tube ata velocity of 15 mvs in a direction normal to the axis of the tube. What is the heat loss per meter of length of the tube?Feterences at lm REFERENCES 10 1" 2 a “ 6 6 v7 w ® Dittus, F. W., and L. M. K. Boelter: Univ. Calif. (Berkeley) Pub, Eng., vol. 2, p. 443, 1930. Sieder. E. N., and C, E, Tate: Heat Transfer and Pressure Drop of Liquids in Tubes, Ind. Eng. Chem., vol. 28, p. 1429, 1936. Nusselt, W.: Der Warmeaustausch zwischen Wand und Wasser im Rohr, Forsch. Geb. Ingenicurwes., vol. 2, . 309, 1931 Hausen, H.; Darstellung des Warmeuberganges in Rohren durch verallgemeinerte Potenzbeziehungen, VDI Z., no. 4, p. 91, 1943, Moody, F. F.: Friction Factors for Pipe Flow, Trans. ASME, vol. 66, p. 671, 1944. Schlichting, H.: “Boundary Layer Theory,”*7th ed., McGraw-Hill Book Company, New York, 1979, Giedt, W. H.: Investigation of Variation of Point Unit-Heat-Transfer Coefficient around a Cylinder Normal to an Air Stream, Trans. ASME, vol. 71, pp. 375-381, 1949. Hilpert, R.; Warmeabgabe von geheizen Drahten und Rohren. Forsch. Geb. In senicurwes., vol. 4, p. 220, 1933 Knudsen, J. D., and D. L, Katz: “Fluid Dynamics and Heat Transfer,”” McGraw. Hill Book Company, New York, 1958. McAdams, W. H.: “Heat Transmission,” 3d ed., McGraw-Hill Book Company, New York, 1954, Kramers, H.: Heat Transfer from Spheres to Flowing Media, Physica, vol. 12, p. 61, 1946, Grimson, E. D.: Correlation and Utilization of New Data on Flow Resistance and Heat Transfer for Cross Flow of Gases over Tube Banks, Trans. ASME, vol. 59. pp. $83-594, 1937, Lyon, R. D. (ed.): “Liquid Metals Handbook,” 3d ed., Atomic Energy Commission and U.S. Navy Department, Washington, D.C.. 1952 Lubarsky, B., and S. 1, Kaufman: Review of Experimental Investigations of Liquid- Metal Heat Transfer, NACA Tech. Note 3336, 1955, Colburn, A. P.: A Method of Correlating Forced Convection Heat Transfer Data ‘and a Comparison with Fluid Friction, Trans. AIChE, vol. 29, p-174, 1933. Seban, R. A., and T, T. Shimazaki: Heat Transfer to a Fluid Flowing Turbulently in a Smooth Pipe with Walls at Constant Temperature, Trans. ASME, vol. 73, p. 803, 1951, Kays, W. M., and R. K. Lo: Basic Heat Transfer and Flow Friction Data for Gas Flow Normal to Banks of Staggered Tubes: Use of a Transient Technique. Stanford Univ. Tech, Rep. 15, Navy Contract N6-ONR251 T.0. 6, 1952. Jakob, M.: Heat Transfer and Flow Resistance in Cross Flow of Gases over Tube Banks, Trans, ASME, vol. 60, p. 384, 1938, Vliet. G. C., and G. Leppert: Forced Convection Heat Transfer from an Isothermal Sphere to Water, J. Heat Transfer, serv. C. vol. 83, p. 163, 1961‘320 Emplical and practical elatons fr frced-convecton heat ransor BBR Irvine, T. R.: Noncircular Duct Convective Heat Transfer, in W. Ibele (ed.), “Mod cern Developments in Heat Transfer,” Academic Press, Inc., New York, 1963. Fand, R. M.: Heat Transfer by Forced Convection from a Cylinder to Water in Crossflow, Int. J. Heat Mass Transfer, vol. 8, p. 995, 1965. Jakob, M.: “Heat Transfer,” vol. 1, John Wiley & Sons, Inc., New York, 1949. Stein, R.: Liquid Metal Heat Transfer, Adv. Heat Transfer, vol. 3, 1966. Hartnett, J. P.: Experimental Determination of the Thermal Entrance Length for the Flow of Water and of Oil in Circular Pipes, Trans. ASME, vol. 77, p. 1211, 1955, Allen, R. W., and E.R. G. Eckert: Friction and Heat Transfer Measurements to ‘Turbulent Pipe Flow of Water (Pr = 7 and 8) at Uniform Wall Heat Flux, J. Heat Transfer, ser. C, vol. 86, p. 301, 1964 Skupinshi, E., J. Tortel, and L. Vautrey: Détermination des coefficients de con- vection a'un alliage sodium-potassium dans un tube circulaire, Int. J. Heat Mass Transfer, vol. 8, p. 937, 1965. Dipprey, D. F., and R. H. Sabersky: Heat and Momentum Transfer in Smooth and Rough Tubes at Various Prandtl Numbers, Int. J. Heat Mass Transfer, Vol. 6, P. 329, 1963. Kline, S. J.: “Similitude and Approximation Theory, pany, New York, 1965 McGraw-Hill Book Com- 29 Townes, H. W., and R. H. Sabersky: Expetiments on the Flow over a Rough 4 7 Surface, Int. J. Heat Mass Transfer, vol. 9, p. 729, 1966. Gowen, R. A., and J. W. Smith: Turbulent Heat Transfer from Smooth and Rough Surfaces, Int. J. Heat Mass Transfer, vol. 11, p. 1657, 1968. ‘Sheriff, N., and P. Gumley: Heat Transfer and Friction Properties of Surfaces with Discrete Roughness, Int. J. Heat Mass Transfer, vol. 9, p. 1297, 1966. Witte, L. C.: An Experimental Study of Forced-Convection Heat Transfer from a Sphere to Liquid Sodium, J. Hear Transfer, vol. 90, p. 9, 1968. Zukauskas, A. A., V. Makarevicius, and A. Schlanciauskas: “Heat Transfer in Banks of Tubes in Crossflow of Fluid,”” Mints, Vilnius, Lithuania, 1968. Eckert, E. R. G.,.and R. M. Drake: “Analysis of Heat and Mass Transfer,” McGraw- Hill Book Company, New York, 1972. Whitaker, S.: "Forced Convection Heat-Transfer Correlations for Flow in Pipes, ast Flat Plates, Single Cylinders, Single Spheres, and Flow in Packed Bids and Tube Bundles,"" AICHE J., vol. 18. p. 361, 1972. Kays, W. M.: “Convective Heat and Mass Transfer,"" pp. 187-190, McGraw-Hill Book Company, New York, 1966. Churchill, S. W., and M. Bernstein: A Correlating Equation for Forced Convection from Gases and Liquids to a Circular Cylinder in Crossflow, J. Heat Transfer, vol. 99, pp. 300-306, 1977, Nakai, S., and T. Okazaki: Heat Transfer from a Horizontal Circular Wire at Small Reynolds and Grashof Numbers—1 Pure Convection, Int. J. Heat Mass Transfer, vol. 18, p. 387, 1975.“ a Relerences 924 Zukauskas, A.: Heat Transfer from Tubes in Cross Flow, Adv, Heat Transfer, vol 8, pp. 93-160, 1972, Shah, R. K., and A. L. London: ® ‘Academic Press, New York, 1978, Kalish, S., and O. E. Dwyer: Heat Transfer to NaK Flowing through Unbafled Rod Bundles, Int. J. Heat Mass Transfer, vol. 10, p. 1533, 1967, Petukhov, B. S.: Heat Transfer and Friction in Turbulent Pipe Flow with Variable Physical Properties, in J. P. Hartnett and T. F. Irvine, (eds.), “Advances in Heat Transfer” Academic Press, Inc., New York, pp. 504-564, 1970, Achenbach, E.: Heat Transfer from Spheres up to Re = 6 x 10#, Proc. Sixth In. Heat Trans. Conf., vol. 5, Hemisphere Pub. Co., Washington, D.C., pp. 341-346, 1978. ‘Morgan, V. T.: The Overall Convective Heat Transfer from Smooth Cireular Cyl- inders, in T. F. Irvine and J. P. Hartnett, (eds.). “Advances in Heat Transfer" vol. 11, Academic Press,.tnc., 1975, aminar Flow: Forced Convection in Ducts,”NATURAL- CONVECTION SYSTEMS 7-4, INTRODUCTION Our previous discussions of convection heat transfer have considered only the calculation of forced-convection systems where the fluid is forced by or through the heat-transfer surface. Natural, or free, convection is observed as a result fof the motion of the fluid due to density changes arising from the heating process. A hot radiator used for heating a room is one example of a practical device which transfers heat by free convection, The movement of the fluid in free convection, whether itis a gas or a liquid, results from the buoyancy forces imposed on the fluid when its density in the proximity of the heat-transfer surface is decreased as a result of the heating process. The buoyancy forces ‘would not be present if the fluid were not acted upon by some external force field such as gravity, although gravity is not the only type of force field which ‘can produce the free-convection currents; a fluid enclosed in a rotating machine is acted upon by a centrifugal force field, and thus could experience free- convection currents if one ar more of the surfaces in contact with the fluid ‘were heated. The buoyancy forces which give rise to the free-convection cur- rents are called body forces. 1H 7-2 FREE.CONVECTION HEAT TRANSFER ON A VERTICAL FLAT PLATE Consider the vertical flat plate shown in Fig. 7-1. When the plate is heated, a free-convection boundary layer is formed, as shown. The velocity profile in this boundary layer is quite unlike the velocity profile in a forced-convection boundary layer. At the wall the velocity is zero because of the no-slip condition; it increases to some maximum value and then decreases to zero at the edge of the boundary layer since the “free-stream” conditions are at rest in the free- convection system. The initial boundary-layer development is laminar; but at324 Nawaconvecton systems & some distance from the leading edge, depending on the fluid properties and the temperature difference between wall and environment, turbulent eddies are formed, and transition to a turbulent boundary layer begins. Farther up the plate the boundary layer may become fully turbulent, To analyze the heat-transfer problem, we must first obtain the differential equation of motion for the boundary layer. For this purpose we choose the x coordinate along the plate and the y coordinate perpendicular to the plate as in the analyses of Chap. 5. The only new force which must be considered in the derivation is the weight of the element of fluid. As before, we equate the sum of the external forces in the x direction to the change in momentum flux through the control volume dx dy. There results au, au) ap 2u a MH) = 2 — pg 2H ot o(uit +S) = Bowes ws > where the term —pg represents the weight force exerted on the element. The pressure gradient in the x direction results from the change in elevation up the plate. Thus Pe pg 72)Free-conecton hes Vanler on a vertical plate 328 In other words, the change in pressure over a height dx is equal to the weight per unit area of the fluid element. Substituting Eq. (7-2) into Eq. (7-1) gives au) au! eu 0 (Wits 8) = mo Ze 73) The density difference p. — p may be expressed in terms of the volume coefficient of expansion B, defined by pa h(M) 1 Vv ene “viar), ViT-T. pT -To au au’ # sotto (wart) eon ra eZ om This is the equation of motion for the free-convection boundary layer. Notice that the solution for the velocity profile demands a knowledge of the temper- ature distribution. The energy equation for the free-convection system is the same as that for a forced-convection system at low velocity: ar an aT (02 4) os) The volume coefficient of expansion @ may be determined from tables of properties for the specific fluid. For ideal gases it may be calculated from (see Prob. 7-3) 1 ay where T is the absolute temperature of the gas. Even though the fluid motion is the result of density variations, these vari- ations are quite small, and a satisfactory solution to the problem may be ob: tained by assuming incompressible flow, that is, p = constant. To effect a solution of the equation of motion, we use the integral method of analysis similar to that used in the forced-convection problem of Chap. 5. Detailed boundary-layer analyses have been presented in Refs. 13, 27, and 32. For the free-convection system, the integral momentum equation becomes a(fma) = + peer - To dy au “wh. = + ff espur - 1.) dy 7 and we observe that the functional form of both the velocity and the temperature distributions must be known in order to arrive at the solution. To obtain these functions. we proceed in much the same way as in Chap. 5. The following conditions apply for the temperature distribution:{388 Natal convection systems T-T ay= T=T. aty= ar ate ae so that we obtain for the temperature distribution: Ee - (1-3) on ‘Three conditions for the velocity profile are u=0 aty=0 u=0 aty=5 Meo aty x An additional condition may be obtained from. Eq. (7-4) by noting that a ~ 9p = a aty = As in the integral analysis for forced-convection problems, we assume that the velocity profiles have geometrically similar shapes at various x distances along the plate. For the free-convection problem, we now assume that the velocity may be represented as a polynomial function of y multiplied by some arbitrary function of x. Thus, HZ asatbytott ay a Ny + oy? + ay where u, is a fictitious velocity which is a function of x, The cubic-polynomial form is chosen because there are four conditions to satisfy. and this is the ‘simplest type of function which may be used. Applying the four conditions to the velocity profile listed above, we have ¥ _ BB eT. = T)¥(, _ x) iu quy BMS ‘The term involving the temperature difference, 8*, and u, may be incorporated into the function u, so that the final relation to be assumed for the velocity profile is 3) 08) A plot of Eq. (7-8) is given in Fig. 7-2. Substituting Eqs. (7-7) and (7-8) into Eq. (7-6) and carrying out the integrations and differentiations yieldsFree-convecion hea ranler ona vertical fal pate 327 Fig.7-2 Free-convection velooty prctie given byEa (78) a9) 1 (u2 8) = 3 eB. ~ To) 6 ~ 105 dx ‘The integral form of the energy equation for the free-convection system is, afr ar if wr ra] 2] . 7-10) and when the assumed velocity and temperature distributions are inserted into this equation and the operations are performed, there results ha. oan noe on Inserting this type of relation in Eq. (7-9) yields the result that b~ xt (7-13) We therefore assume the following exponential functional variations for 1, and 6: Cx? (7-14) 8 = Cx" (7-15) Introducing these relations into Eqs. (7-9) and (7-11) gives 3 Cac = = ey Se — By ap CPE = BAT. - TH a” (7-16) 1 and BOcw ™ 7. woes ay These two equations may be solved for the constants C, and C3 to give328 Natura-convecton systems 20 v\ Pept. — T.))'* c= 5.7 (R +4) [: | 018) 20, v\ 0 gore c.=3n(2+2)"[ By" om ‘The resultant expression for the boundary-layer thickness is y where the Prandtl number Pr = via has been introduced along with a new dimensionless group called the Grashof number Gr. = eB. ~ Tae Gr, - 21) ‘The heat-transfer coefficient may be evaluated from ar sat 7, — Te tA, |, = ATs ~ Td) Using the temperature distribution of Eq. (7-7), one obtains * bx x =F oo Panu a25 so that the dimensionless equation for the heat-transfer coefficient becomes Nu, = 0,508 Pr (0.952 + Pr)" Gri (7-22) Equation (7-22) gives the variation of the local heat-transfer coefficient along, the vertical plate. The average heat-transfer coefficient may then be obtained by performing the integration hed 0-3) t To For the variation given by Eq. (7-22), the average coefficient is Fi= thee (7-24) ‘The Grashof number may be interpreted physically as a dimensionless group representing the ratio of the buoyancy forces to the viscous forces in the free- convection flow system. It has a role similar to that played by the Reynolds number in forced-convection systems and is the primary variable used as a criterion for transition from laminar to turbulent boundary-layer flow. For air in free convection on a vertical flat plate, the critical Grashof number has been observed by Eckert and Soehngen [1] to be approximately 4 x 10%. Values ranging between 10% and 10° may be observed for different fluids and environ- ment “turbulence levels.”convacton neat vanearon a veicl fat plate 228 A very complete survey of the stability and transition of free-convection houndary layers has been given by Gebhart et al. [13-15] ‘The foregoing analysis of free-convection heat transfer on a vertical flat plate is the simplest case that may be treated mathematically, and it has served to introduce the new dimensionless variable, the Grashof number,t which is im- portant in all free-convection problems. But as in some forced-convection prob- lems, experimental measurements must be relied upon to obtain relations for heat transfer in other circumstances. These circumstances are usually those in which it is difficult to predict temperature and velocity profiles analytically ‘Turbulent free convection is an important example, just as is turbulent forced convection, of a problem area in which experimental data are necessary; how- ever, the problem is more acute with free-convection flow systems than with forced-convection systems because the velocities are usually so small that they are very difficult to measure. Despite the experimental difficulties, velocity measurements have been performed using hydrogen-bubble techniques (26]. hot-wire anemometry [28]. and quartz-fiber anemometers. Temperature field measurements have been obtained through the use of the Zehnder-Mach in- terferometer. The laser anemometer (29] is particularly useful for free-convec- tion measurements because it does not disturb the flow field. ‘An interferometer indicates lines of constant density in a fluid flow field. For 4 gas in free convection at low pressure these lines of constant density are equivalent to lines of constant temperature. Once the temperature field is ob- tained, the heat transfer from a surface in free convection may be calculated by using the temperature gradient at the surface and the thermal conductivity of the gas. Several interferometric studies of free convection have been made [1-3]. and some typica! photographs of the flow fields are shown in Figs. 7-3 to 7-6. Figure 7-3 shows the lines of constant temperature around a heated vertical flat plate, Notice that the lines are closest together near the plate surface, indicating a higher temperature gradient in that region. Figure 7-4 shows the lines of constant temperature around a heated horizontal cylinder in free convection, and Fig. 7-5 shows the boundary-layer interaction between € group of four horizontal cylinders. A similar phenomenon would be observed for forced-convection flow across a heated tube bank. Interferometric studies have been conducted to determine the point at which eddies are formed in the free-convection boundary layer [1]. and these studies have been used in pre- dicting the start of transition to turbulent flow in free-convection systems. It was mentioned earlier that the velocities in free convection are so small that for most systems they are difficult to measure without influencing the flow field by the insertion of a measuring device. A rough visual indication of the free-convection velocity is given in Fig. 7-6, where a free-convection boundary- ‘History isnot clear on the point, but it appeurs thatthe Grashof number was named for Franz Grashot, a professor of applied mechanics at Karlsre around 1863 and one of the founding directors of Verein deutscher Ingenicure i TASS. He developed some early steam ow formulas, but made no significant contributions to free conveetion [36]930 Naueab convection systems Fig. 7-3. Inteerometer photograph showing tines of constant temperature around a heated vertical fat plate in free convection. (Photograph courtesy E. Soenagen) layer wave resulting from a heat pulse near the leading edge of the plate is presented. It may be noted that the maximum points in the isotherms experience a phase lag and that a line drawn through these maximum points has the approximate shape of the free-convection velocity profile. ‘A number of references treat the various theoretical and empirical aspects of free-convection problems. One of the most extensive discussions is given by Gebhart [13], and the interested reader may wish to consult this reference for additional information. 7-3 EMPIRICAL RELATIONS FOR FREE CONVECTION Over the years it has been found that average free-convection heat-transfer coefficients can be represented in the following functional form for a variety of circumstances:Empitcal relatos for ree conveton 334 Fig. 7-4 Interterometer photograph showing lines ‘of constant temperature around a heated horizontal eyinder in tee convection. (Photograph courtesy €. Soenngen) Nu; = CGr, Pry (7.25) where the subscript f indicates that the properties in the dimensionless groups are evaluated at the film temperature To + Te a T= Fig. 75 Intererometer photograph showing the boundary syerntoraction betwoen IQur heated herlzontal cyinders in {tee convection. (Photograph courtesy E, Soahngen.)1332, Natura-convecon systems The product of the Grashof and Prandt! numbers is called the Rayleigh number: Ra = GrPr (726) The characteristic dimension to be used in the Nusselt and Grashof numbers depends on the geometry of the problem. For a vertical plate it is the height of the plate L; for a hor- izontal cylinder it is the diamter d; and so forth. Experi- mental data for free-convection problems appear in a number of references, with some conflicting results. The purpose of the sections that follow is to give these results in a summary form that may be easily used for calculation purposes. The functional form of Eq. (7-25) is used for many of these pre: sentations, with the values of the constants C and m specified for each case. 7-4 FREE CONVECTION FROM VERTICAL PLANES AND CYLINDERS, 1 Isothermal Surtaces For vertical surfaces, the Nusselt and Grashof numbers are formed with L, the height of the surface as the characteristic dimension. If the boundary-layer thickness is not large com- pared with the diameter of the cylinder, the heat transfer may be calculated with the same relations used for vertical plates. ‘The general criterion is that a vertical cylinder may be treated as a vertical flat plate (13] when D_ 3s > Gn where D is the diameter of the cylinder. For isothermal sur- faces, the values of the constants are given in Table 7-1 with the appropriate references noted for further consultation. The reader's attention is directed to the two sets of constants given for the turbulent case (Gr, Pr; > 10°). Although there may appear to be a decided difference in these constants, a com- parison by Warner and Arpaci [22] of the two relations with experimental data indicates that both sets of constants fit available data, There are some indications from the analytical work of Bayley [16], as well as heat flux measurements of Ref, 22, that the relation 27) Fig. 7-6 interferometer photograph showing isotherms on ahested verical tat, plate resuting from 3 periodic disturbance of the boundary layer. Note phase Shits in maximum points of isotherms. (From Holman, Gartel, and Soehngen aD)Free convection tom verse planes ana oyinsers. 333 Tobie 7-1 Constants for Use with Eq, (7-25) for lsthermal Sutaces. Geometry Ge Pry c = Refs. Vertical planes and cylinders H0-!-10 Use Fig. 7-7 Use Fig. 7-7. 4 ot-10" 0.59 yo 4 10 oer io 10-10 0.10 a Horizontal eylinders o-10-* 04 O24 10-510" Use Fig. 7.8 Use Fig. 728 4 oti 053 ed 0" ob 44 10" 10-2 srs 0038 761 10-7108 1.02 ous 761 0.850 ons 76 0.480 $6 ons b6 Upper surface of heated plates 2 ose 4 sz ‘oF lower surface of cooled plates Upper surface of heated plates 8 x 10*-10" bas 4 ‘F lower surface of cooled plates Lower surface of heated plates 110" on 2 sears ‘oF upper surface of cooled plates Vertical cylinder, toe ons ou 07 height = diameter ‘Characteristic length = diameter Irregular solids, characteristic oI 02 Tength = distance uid particle _tavels in boundary ayer % * Prefered Nu = 0.10 (Gry Pry!” may be preferable. More complicated relations have been provided by Churchill and Chu (71) and are applicable over wider ranges of the Rayleigh number: 0.670 Ra Nu = 0.68 + [E+ (0.492/Pry"'9ye9 for Ra, < 10° (7-28) for 10 = Ra, < 10° (7-29) Equation (7-28) is also a satisfactory representation for constant heat fux. Properties for these equations are evaluated at the film temperature. Constant Heat Flux Extensive experiments have been reported in Refs. 25, 26, and 39 for free convection from vertical and inclined surfaces to water under constant-heat-394 Nata convection systems Fig. 7-7. Free-convection heat ransler correlation for heat trarster frm heated ver tical plates, according to Ret 4 flux conditions. In such experiments, the results are presented in terms of a modified Grashof number, Gr* Grt = Gr, Nu, = 730) where q is the wall heat flux in watts per square meter. The local heat-transfer coefficients were correlated by the following relation for the laminar range: It is to be noted that the criterion for laminar flow expressed in terms is not the same as that expressed in terms of Gr,. Boundary-layrFree convection Hom vonicl planes anc eyindere 338 Fig. 7-8 Free-convecton heat wanster correlation forheat transer om heated hi accorang to Rel 4 zontal cinders, was observed to begin between Gr? Pr = 3 x 10" and 4 x 10! and to end between 2 x 10" and 10". Fully developed turbulent flow was present by Grz Pr = 10", and the experiments were extended up to Grt Pr = 10'°, For the turbulent region, the local heat-transfer coefficients were correlated with Nu, = 0.17 (Gro Pr 2 x 10" < Grt Pr < 10% g, = const (7-32) All properties in Eqs. (7-31) and (7-32) are evaluated at the local film temper- ature. Although these experiments were conducted for water. the resulting correlations are shown to work for air as well. The average heat-transfer coef- ficient for the constant-heat-flux case may not be evaluated from Eq. (7-24) but must be obtained through a separate application of Eq. (7-23). Thus, for the sinar region. using Eq. (7-31) to evaluate f396 Natuatcorvecton systens Tf bea T= ther qe = const ‘At this point we may note the relationship between the correlations in the form of Eq. (7-25) and those just presented in terms of Gr? = Gr, Nu,. Writing Eq. (7-25) as a local heat-transfer form gives Nu, = C (Gr, Pry” (7-33) Inserting Gr, = Gri/Nu, gives Nut" = © Grt Prim G Nu, = ClI+™ Gre Pymrm (734) ‘Thus, when the “‘characteristic”” values of m for laminar and turbulent flow are compared to the exponents on Gr?, we obtain mt Laminar, m = Tecate m__t Turbulent, m = 4: Terie While the Gr* formulation is easier to employ for the constant-heat-fux case, wwe see that the characteristic exponents fit nicely into the scheme which is presented for the isothermal surface correlations. Itis also interesting to note the variation of h, with x in the two characteristic regimes. For the laminar range m = 4, and from Eq. (7-25) rr ‘So when turbulent free convection is encountered, the local heat-transfer coef ficient is essentially constant with x, Churchill and Chu [71] show that Eq. (7-28) may be modified to apply to the constant-heat-flux case if the average Nusselt number is based on the wall heat flux and the temperature difference at the center of the plate (x = 1/2). The result is ae 0.67 (Gre Pr NupNu, - = “ 068) = TS oan where Nu, = q.Lk BT) and BT = 7, at L?2 ~ Te (7-35)Free convection tom vetcl planes and cyinders 337 = EXAMPLE 7-1 In plant location near a furnace, a net radiant energy flux of 800 Wim’ is incident on f vertical metal surface 3.5 m high and 2 m wide. The metal is insulated on the back side and painted black so that all the incoming radiation is lost by free convection t0 the surrounding air at 30°C. What average temperature willbe attained by the plate? Solution We treat this problem as one with constant heat fux on the surface. Since we do not know the surface temperature, we must make an estimate for determining 7; and the air properties. An approximate value of h for free-convection problems is 10 Wim? °C, and so, approximately, 4: swe Then 1-38 50= we = 30K A170 the properies of sr ae ye 208 x o'mis p= b= 27% 0K 1 i, k= 00295 Wim -"CPr= 0. From Eg. (7-30), with x = 3.5 m. — #Baex* _(9.8N2.92 x 10-9800)6.5) kee (0.0295¥2.005 x 10-*)* Gt 2.90 x 10% We may therefore use Eq. (7-32) to evaluate hy: h So met mi 0.0295 35 5.36 Wim? «°C [0.944 Btu/h - f= °F] (0.1712.79 x 10% x 0.7)" In the turbulent heat transfer governed by Eq. (7-32), we note that hx Nu, = = (Gry ~ oy fo h, does not vary with x, and we may take this as the average value. The value of +h = 5.41 Wim + °C is less than the approximate value we used to estimate 7, Recaleu- lating AT. we obtain % ao OL ye are & 149"ur new film temperature would be M9 T,= 30 +P - 14sec ‘At 106.5°C the properties of air are vena cia pe b= 265 «10-1 k= 0.0320 Wim -°C_ Pr = 0.695 (9,812.65 x_10?(800)3.5)* (0.032092.354 x 10-5 and h, is calculated from ort = 1.758 x 10! = Keounors pom (0.032090.17 = CORON (1.758 x 10°(0.695)! = 5.17 Wim? + °C (0.91 Brun - Our new temperature difference is calculated as ge, 80 + AT = (7, ~ Te = & = = 550 ‘The average wall temperature is therefore Teas = 155 +30 = 185°C ‘Another iteration on the value of Tis not warranted by the improved accuracy which would result, w= EXAMPLE 7.2 ‘A large vertical plate 4.0 m high is maintained at 60°C and exposed to atmospheric air at 10°C, Calculate the heat transfer if the plate is 10 m wide. ‘Solution We first determine the film temperature as +10 FO = 3c = 308K ‘The properties of interest are thus 1 ee Bm yg 7325 x 10> k= 0.02685 v= 16.5 x 10-6 Pr=07Fe convection rom horizontal cyinders 339 and (9.89.25 x 10-160 — 10X49 Ore GES = 3.763 x 10" We then may use Eq. (7-29) to obtain Dee (1 + 0.4920.7)""" 7" Ni ‘The heat-transfer coefficient is then (803)0.02685) 40 = 5.39 Wim? = °C The heat transfer is Tar. ~ T.) (5,394)(1060 ~ 10) = 10,781 W ‘As an alternative, we could employ the simpler relation Nu = 0.10(Gr Pry" (0.101.743 x 10 = 720.7 which gives a value about 10 percent lower than Eq. (7-29) 75 FREE CONVECTION FROM HORIZONTAL CYLINDERS ‘The values of the constants C and m are given in Table 7-1 according to Refs. 4and 76. The predictions of Morgan (Ref. 76 in Table 7-1) are the most reliable for Gr Pr of approximately 10~*. A more complicated expression for use over ‘a wider range of Gr Pr is given by Churchill and Chu (70): Gr Pr " —__Gr Pr" ge 10-9 < Gr Pr io. omar} er <108 0360 ‘A simpler equation is available from Ref. 70, but is restricted to the laminar range of 10-* < Gr Pr < 10°: Nui? = 0.60 + oast { 0.518 Gry Pr) Nay = 0.36 + TF S59? on Properties in Eqs. (7-36) and (7-37) are evaluated at the film temperature. ‘Heat transfer from horizontal cylinders to liquid metals may be calculated from Ref. 46: Nuy = 0.53 (Gry Pry" 7-38)340 NattaLeorvacton est EXAMPLE 73 ‘A.2.0-cm-diameter horizontal heater is maintained at a surface temperature of 38°C and. submerged in water at 27°C. Calculate the free-convection heat loss per unit length of the heater. Solution The film temperature is 4227 _ asc From Appendix A the properties of water are k = 0.630 Wim °C, t and the following term is particularly useful in obtaining the Gr Pr product when itis multiplied by a? AT: BB = 2.48 x 10" Wim" “C} ak Gr Pr = (2.48 x 10°38 — 2790.02) = 2.18 10° Using Table 7-1, we get C= 0.53 and m = 4, so that Nu = (0,5312.18 & 10%" = 38,425, _ (38.425(0.63) _ = 1210 Wim? °C h aa 1210 *c ‘The heat transfer is thus 4 5 = amar. - 7.) (1210) (0.02908 ~ 27) = 836.3 Wim EXAMPLE 7-4 AA fine wire having a diameter of 0.02 mm is maintained at a constant temperature of 54°C by an electric current. The wire is exposed to air at 1 atm and O°C. Calculate the ‘electric power necessary to maintain the wire temperature if the length is $0 cm. Solution ‘The film temperature is T; = (S4 + 0V2 = 27°C = 300 K, so the properties are B= 1300 = 0.00333 v= 15.69 x 10-* mils k = 0.02624 Wim - °C Pr = 0.708 ‘The Gr Pr product is then calculated as (9.8(0.00333)(54 — 0X0.02 x 10-3 15.69% 10-4} From Table 7-1 we find C = 0.675 and m = 0.058 so that Nii = (0.675(4.05 x 10-5 = 0.375 Gr Pr = (0.708) = 4.05 x 10-*Free convection tom horizontal cvinders 44 and. = (8) _ 0.375H0.02624) _ 499.6 wim? °C d) * 092 x 107 ‘The heat transfer or power required is then q= A(T. ~ T.) = (492.6)x(0.02 x 10-240.5)54 ~ 0) = 0.836 W EXAMPLE 7-5 ‘A horizontal pipe | ft (0.3048 m) in diameter is maintained at a temperature of 250°C in.a room where the ambient air is at 15°C. Calculate the free-convection heat loss per meter of length. Solution We first determine the Grashof-Prandt! number product and then select the appropriate constants from Table 7-1 for use with Eq, (7-25). The properties of air are evaluated at the film temperature: a = 132.8°C = 405.5 K +. 250415 z k= 0.03406 Wim (9.8)2.47 x 10-9250 ~ 15%0.3048)%0. x LST x 10" From Table 7-1, € = 0.53 and m = 4, so that Nay = 0.53Gr, Pr)! = (0.531.571 108)" = 59.4 KNuy _ (0.03406)59.4) _ te os = 068 Wi C [1.175 Btwhh «ft «°F ‘The heat transfer per unit length is then calculated from (6.63) (0.3048)(250 ~ 15) 49 kWim (1560 Btu/h « fe) a - 4 = neat. ~ 7.) ‘As an alternative, we could employ the more complicated expression. Eq. (7-36), for solution of the problem. The Nusselt number thus would be calculated as pode I+ 5590.67) Narr = 040 + 0307 { Nu = 647 a value about 8 percent higher.342 Natrat-convecton systems 7-6 FREE CONVECTION FROM HORIZONTAL PLATES lsothermal Surfaces ‘The average heat-transfer coefficient from horizontal flat plates is calculated with Eq. (7-25) and the constants given in Table 7-1. The characteristic di mension for use with these relations has traditionally (4) been taken as the length of a side for a square, the mean of the two dimensions for a rectangular surface, and 0.9d for a circular disk. References 52 and 53 indicate that better agreement with experimental data can be achieved by calculating the charac- teristic dimension with LaS (739) where A is the area and P is the perimeter of the surface. This characteristic dimension is also applicable to unsymmetrical planforms. Constant Heat Flux ‘The experiments of Ref. 44 have produced the following correlations for con- stant heat flux on a horizontal plate. For the heated surface facing upward, Ni, = 0.13 Gry Pr)! — for Gr, Pr <2 x 10° (7-40) and Nu, = 0.16 (Gr, Pr)! for2 x 108 < Gr, Pr< 10" (7-41) For the heated surface facing downward, Nu = 0.58 Gry Pr)! for 106 < Gry Pr < 10" (7-42) In these equations all properties except @ are evaluated at a temperature T, defined by T, = T, ~ 0.29(T, - T.) and T, is the average wall temperature related, as before, to the heat flux by Ge i= The Nusselt number is formed as before: aot (1, — Tak Section 7-7 discusses an extension of these equations to inclined surfaces.Free conection rom incned surfaces. 343. CD Irregular Solids There is no general correlation which can be applied to irregular solids. The results of Ref. 77 indicate that Eq. (7-25) may be used with C = 0.775 and ‘m = 0.208 for a vertical cylinder with height equal to diameter. Nusselt and Grashof numbers are evaluated by using the diameter as,characteristic length. Lienhard [78] offers a prescription that takes the characteristic length as the distance a fluid particle travels in the boundary layer and uses values of C = 0.52 and m = tin Eq. (7-25) in the laminar range. This may serve as an estimate for calculating the heat-transfer coefficient in the absence of specific information on a particular geometric shape. Bodies of unity aspect ratio are studied ex- tensively in Ref. 81. = EXAMPLE 76 A-cube, 20 em on a side, is maintained at 60°C and exposed to atmospheric air at 20°C. Calculate the heat transfer. Solution ‘This is an irregular solid so we use the information in the last entry of Table 7-1 in the absence of a specific correlation for this geometry. The properties were evaluated in Ex. 72s B= 325% 10> k= 0.02685 1747 x 10-* Pr = 0.7 ‘The characteristic length is the distance a particle travels in the boundary layer, which is L/2 along the bottom plus Z. along the side plus L/2 on the top, or 21. = 40 cm. The Gr Pr product is thus: _ 98V3.25 % 10-60 = 10K04 3 54 x Gr Pr = azarae (0.7) = 3.34 x 10" From the last entry in Table 7-1 we find C = 0.52 and n = 1/4 and calculate the Nusselt number as Nu = (0,523.34 x 10) = 135.2 & _ (35.240.02685) _ = R= Nu = 9.07 Wien? = °C ‘The cube has six sides so the area is 6(0.2)' = 0.24 m* and the heat transfer is, q = hACT. ~ Ta) = (9.0740.24(60 ~ 10) = 108.8 W 17-7 FREE CONVECTION FROM INCLINED SURFACES Extensive experiments have been conducted by Fujii and Imura (44] for heated plates in water at various angels of inclination. The angle which the plate makes with the vertical is designated 6, with positive angles indicating that the heater‘344, Natura-convecton systems surface faces downward, as shown in Fig. 7-9. For the incl downward with approximately constant heat flux the follo obtained for the average Nusselt number: Nii, = 0.56 (Gr, Pr, cos 9! @ < 88°; 10° < Gr, Pr, cos @< 10" (7-43) In Eq.(7-43) all properties except 8 are evaluated at a reference temperature T, defined by T, - 0.25(T, - T.) (7-44) where T, is the mean wall temperature and T.. is the free-stream temperature; Bis evaluated at a temperature of T= + 0.50(7, ~ T.). For almost-horizontal plates facing downward, that is, 88° < @ < 90°, an additional relation was ‘obtained as Nu, = 0.58 (Gr, Pr)! 10 < Gr, Pr. < 10" (7-45) For an inclined plate with heated surface facing upward the empirical cor- relations become more complicated. For angles between —15 and —75° a suitable correlation is Ni, = 0.14 (Gr, Pro)! — Gre Pre)!] + 0.56 (Gre Pre cos 6)" (7-46) for the range 10° < Gr, Pr, cos 6 < 10". The quantity Gr, is a critical Grashof relation indicating when the Nusselt number starts to separate from the laminar relation of Eq. (7-43) and is given in the following tabulation: degrees Gr. -15 sx 30 2x1 60 108 a15 i For Gr, < Gr, the first term of Eq. (7-46) is dropped out. Additional information is given by Vliet (39] and Pera and Gebhart [45]. There is some evidence to indicate that the above relations may also be applied to constant-temperature surfaces. Experimental measurements with air on constant-heat-flux surfaces [51] have shown that Eq. (7-31) may be employed for the laminar region if we replace Heated surface Hed wernce Fig. 7-9 Coordinate system for inclined plates‘Simpiied equations fra 348 Gr? by Gr? cos 6 for both upward- and downward-facing heated surfaces. In the turbulent region with air, the following empirical correlation was obtained: Nu, = 0.17 (Gr¢ Pr) 10" < Grt Pr < 10" (747) where the Gr? is the same as for the vertical plate when the heated surface faces upward. When the heated surface faces downward, Gr? is replaced by Gr* cos? 6, Equation (7-47) reduces approximately to the relation recommended in Table 7-1 for an isothermal vertical plate. For inclined cylinders the data of Ref. 73 indicate that laminar heat trans- fer under constant-heat-flux conditions may be calculated with the following relation: Nuz, = (0.60 ~ 0.488(sin 6)'] (Gry Pry He "> for Gr, Pr< 2 x 10" (7-48) where is the angle the cylinder makes with the vertical; that is, 0° corresponds to a vertical cylinder. Properties are evaluated at the film temperature except, B, which is evaluated at ambient conditions. Uncertainties still remain in the prediction of free convection from inclined surfaces, and an experimental-data scatter of +20 percent is not unusual for the empirical relations presented above. 7-8 NONNEWTONIAN FLUIDS When the shear-stress viscosity relation of the fluid does not obey the simple newtonian expression of Eq. (5-1), the above equations for free-convection heat transfer do not apply. Extremely viscous polymers and lubricants are ‘examples of fluids with nonnewtonian behavior. Successful analytical and ex- perimental studies have been carried out with such fluids, but the results are very complicated. The interested reader should consult Refs. 48 to 50 for detailed information on this subject. 7-9 SIMPLIFIED EQUATIONS FOR AIR ‘Simplified equations for the heat-transfer coefficient from various surfaces to air at atmospheric pressure and moderate temperatures are given in Table 7-2. These relations may be extended to higher or lower pressures by multi- plying by the following factors: p_\" p (j ni a) for laminar cases 7 G ni =) for turbulent cases ‘where p is the pressure in kilopascals. Due caution should be exercised in the use of these simplified relations because they are only approximations of the more precise equations stated earlierTable 7-2 Simpied Equations or Free Convection trom Various Surtaces to Ar at Atmospheric Pressure, Adapted from Table 7-1 Tarbuet, surface one Vertical plane or elindr iawn a — ~ Horzoma pind i224) = tas Horizontal plat Heated plate facing upward « 1.32 (32)"* Roe or cooled plate facing 1 doanward Heated plate facing downward or cooled plate “acing upward vertical or horizontal dimension. m diameter. m = EXAMPLE 7-7 Compute the heat transfer for the conditions of Example 7-5 using the simplified relations, of Table 7-2 sebon The heat-transfer coetfiient i given by any" 250 = 15)" nam (AZ) = oe QS u9 (7) = (aa) = 6.96 Win 5C : “The heat transfer is then $= (6.96) 200.s0880250 — 15) = 1.57 KWim [Note that the simplified relation gives a value approximately 4 percent higher than Ea 0-35) 7-10 FREE CONVECTION FROM SPHERES Yuge [5] recommends the following empirical relation for free-convection heat transfer from spheres t0 aird Nu, = SS = 2 + 0.392Gr* for 1 < Gry < 10° (7-49) This equation may be modified by the introduction of the Prandtl number to give Nu, = 2 + 0.43 Gr, Pry! (7-50) Properties are evaluated at the film temperature, and it is expected that this relation would be primarily applicable to calculations for free convection in gases. However. in the absence of more specific information it may also be used for liquids. We may note that for very low values of the Grashof-Prandt! umber product the Nusselt number approaches a value of 2.0. This is the value which would be obtained for pure conduction through an infinite stagnant fluid surrounding the sphere For higher ranges of the Rayleigh numbers the experiments of Amato and Tien 79] with water suggest the following correlation: Nu, = 2 + 0.50 (Gr, Pr st) for 3 < 10° < Gr Pr <8 x 108 7-11. FREE CONVECTION IN ENCLOSED SPACES The free-convection flow phenomena inside an enclosed space are interesting examples of very complex fluid systems that may yield to analytical, empirical, and numerical solutions. Consider the system shown in Fig. 7-10, where a fluid is contained between two vertical plates separated by the distance 6. As a temperature difference AT, = T, ~ T:is impressed on the fluid, a heat transfer will be experienced with the approximate flow regions shown in Fig. 7-11, according to MacGregor and Emery [18]. In this figure. the Gi is calculated as shot number 287, ~ T)9” ae Gry 7-52) Fig. 710 Nomenclature tor foe convecton in enclosed vertical spaces,‘948 Natua-convecton systems Bee. ket | ! a Conduction Fig. 7-11. Schematic diagram and flow regimes for the vertical convection layer, ac ‘coding to Ret 18, At very low Grashof numbers, there are very minute free-convection currents and the heat transfer occurs mainly by conduction across the fluid layer. As the Grashof number is increased, different flow regimes are encountered, as shown, with a progressively increasing heat transfer as expressed through the ‘Nusselt number ry Nu, = = k Although some open questions still remain, the experiments of Ref. 18 may be used to predict the heat transfer to a number of liquids under constant-heat- flux conditions. The empirical correlations obtained were: Nus = 0.42 (Gry Pr Pree? (3) au = const (7-83) 10! < Gry Pr< 10” 1 < Pr < 20,000 10

‘sued prwozuo Temas, woot ot ai 20 xy Lhe Oro om er ‘oot aq 1we=009 or=ary ‘coaz> ‘amd eons, pmbry fo $0 Ix tee a zo 101 x Z'¢-0000 nopa ‘wou pore ro 0 = so (001-041 eauomiost ona oot> ‘arm weapon $ wo ow to x 11-000'002 $ wre wi cso ‘00002-0009 reaooet owe ‘oaz> ‘are won, 9 . 2 @ a aD avmoog ma 7 “tr wowr0H ‘7a passnipy sieeve wowmeue9 ‘(09-2) ‘ba 1 UE AR UH SeuneopUG UI LORDeALOD ©eL, 404 sUORBEL OUKWIS 70 KRURING EL OAT‘982 Netual-convecton systems greement between different experimental investigations. The interested reader ‘may wish to consult the specific references for more details. For the annulus space the heat transfer is based on 2nkL aT In (rar) where L is the length of the annulus and the gap spacing is 8 = re — 7. : Extensive correlations for free convection between cylindrical, cubical, and spherical bodies and various enclosure geometries are given by Warrington and Powe [80]. The correlations cover a wide range of fluids. Free convection through vertical plane layers of nonnewtonian fluids is di cussed in Ref. 38, but the results are too complicated to present here. In the absence of more specific design information the heat transfer for inclined enclosures may be calculated by substituting g’ for g in the Grashof number, where (7-64) Fi 8’ = gc0s 6 (765) and @ is the angle which the heater surface makes with the horizontal. This \ transformation may be expected to hold up to inclination angles of 60° and ; applies only to those cases where the hotter surface is facing upward. Further information is available from Hollands et al. (66, 67, 69]. = EXAMPLE 7-8 Air at atmospheric pressure is contained between two 0.5-m-square vertical plates separated by a distance of 15 mm. The temperatures of the plates are 100 and 40°C, respectively, Calculate the free-convection heat transfer across the air space. Solution ‘We evaluate the air properties atthe mean temperature between the two plates: E 1, ME# ~ we = 8K p, 10132 x 108 | : Ra OTN eee fife 773g 7 28S XK 2.043 x 10°F gms k = 0.0295 Wim "C Pr = 0.7 6 » ‘The Grashof-Prandtl number product is now calculated as : (9.8N1.02972.915 10 2) 100 ~ 40)1S < 10-7) 5 5 ' ue 0a e105 1.027 x 10"Free convection n encased spaces 389 ke mayo se Be, (4) 0 caste the eet ema ony, with L 5 mod eS mand he ont olen Tale k re we ( 22) = de = aonqnan x 19 (88) "= 138 ‘The hn ranfer may abe aed wth (9, The tenis 7 = 025m int 8840 25010 (1 28,01295025410 = 40) ayn ss nm exaunie 7 Fenn ites 20m ona sil te Separated y dtc of with a a Pas temperneso! epee forthe loser an a ee ute eat rane actos he ee sun The pets tthe same ven i sample 8 p= 1.02 hem p= 29180 2208 3 10% kgims k= 0085 Wan C mor “The Gr Pr product is evaluated on the basis of the separating distance. so we have oie (2.083 « 10%) oe L «oases (22)" = 146 Example 7-10 Hea! transter across water layer Two $0-cm horizontal square plates are separated by a distance of 1 em. The lower plate is maintained at a constant temperature of 100°F and the upper plate is constant ALSU'E, Water at atmospheric pressure occupies the space between the plates. Calculate the heat lost by the lower plate Solution We evaluate properties at the mean temperature of AFF and obtain. for water, Bec k = 0,623 Wim °c BBR = uk 48 x 10"(384 Naturalconvecton systems ‘The Grashof-Prandtl number product is now evaluated using the plate spacing of 1 cm as the characteristic dimension, Gr Pr = (2.48 x 10°%(0.01)'(100 ~ 80)(5I9) = 2.76 x 10° Now, using Eq. (7-63) and consulting Table 7-3 we obtain ‘Therefore, Eq. (7-63) becomes eo ina x 99? = 55 ‘The effective thermal conductivity is thus k, = (0.623\5.57) = 3.47 Wim «°C and the heat transfer is G.4740.57(100 ~ 80,5/9) _ 9 = KAATIS = wer 964 W 7-12 COMBINED FREE AND FORCED CONVECTION A number of practical situations involve convection heat transfer which is neither “forced” nor “free” in nature. The circumstances arise when a fi is forced over a heated surface at a rather low velocity. Coupled with the forced-flow velocity is a convective velocity which is generated by the buoy- ancy forces resulting from a reduction in fluid density near the heated surface. A summary of combined free- and forced-convection effects in tubes has been given by Metais and Eckert [10], and Fig. 7-13 presents the regimes for combined convection in vertical tubes. Two different combinations are indi- cated in this figure. Aiding flow means that the forced- and free-convection currents are in the same direction, while opposing flow means that they are in the opposite direction. The abbreviation UWT means uniform wall temperature, and the abbreviation UHF indicates data for uniform heat flux. Its fairly easy to anticipate the qualitative results of the figure. A large Reynolds number implies a large forced-flow velocity, and hence less influence of free-convection currents. The larger the value of the Grashof-Prandtl product, the more one would expect free-convection effects to prevail. Figure 7-14 presents the regimes for combined convection in horizontal tubes. In this figure the Graetz number is defined as Gz = Re pr (7-66) The applicable range of Figs. 7-13 and 7-14 is for @ repr(4) <1 10+ < Pr (2)Comoned tree and forced convecton 987 ‘The correlations presented in the figures are for constant wall temperature. All properties are evaluated at the film temperature. Brown and Gauvin [17] have developed a better correlation for the mixed- convection, laminar flow region of Fig. 7-14: Nu = 175 (8) (Gz + 0.012 (Ge Grey" 67) where jy is evaluated at the bulk temperature. This relation is preferred over that shown in Fig. 7-14, Further information i available in Ref. 68. The problem of combined free and forced convection from horizontal cylinders is treated in detail by Fand and Keswani (47) ‘The general notion which is applied in combined-convection analysis is that the predominance of & heat-transfer mode is governed by the fluid velocity associated with that mode. A forced-convection situation involving a fuid ve Tocity of 30 m/s, for example, would be expected to overshadow most free~ convection effects encountered in ordinary gravitational fields because the ve- Tocities of the free-convection currents are small in comparison with 30 mis. (On the other hand, a forced-flow situation at very low velocities (~0.3 mus) might be influenced appreciably by free-convection currents. An order-of- magnitude analysis of the free-convection boundary-layer equations will indi- cate a general criterion for determining whether fee-convection effects dom- inate. The criterion is that when Gr/Re? > 10 (7-68) free convection is of primary importance. This result isin agreement with Figs. 713 and 7-14. EXAMPLE 7-11 Air at | atm and 27°C is forced through a horizontal 25-mm-diameter tube at an average velocity of 30 cm/s, The tube wall is maintained at a constant temperature of 140°C. Calculate the heat-transfer coefficient for this situation if the tube is 0.4 m long. Solution For this calculation we evaluate properties at the film temperature: M0427 _ g3.5°¢ = 356.54 pb | 1.0132 x 10° 0.99 kgim? = Rr GENGI) ~ 0%? He Bad 2805 10K) w= 2337 x 10° kp s 202 x 10% kgim sky = 0.0305 Wim -"C Pr = 0.695 Let us take the bulk temperature as 27°C for evaluating ji then £8462 < 10-* kgm: s288 Natura-convecton systems ‘The significant parameters are calculated as pud _ (0.99V0.3N0.025) He 21D x 108 pid? _ (0.99(9.8N2.805 x 10 140 — 27)0.025" 2.002 x10 3.33 (007 x 10° 0.025 d in Pr f= x 109 ‘2 Gr Pr f= (1.07 x 10390.695) 9925 = 4677 According to Fig. 7-14, the mixed-convection-flow regime is encountered. Thus we must use Eq. (7-67). The Graetz number is calculated as | (8531(0.695)00.025) Gr Re rd 1533 and the numerical calculation for Eg, (7-67) becomes 1.8462) 0% we : 2x07) x 1OyoHs Nu = 1.75 (3352)"" 5.33 + @.o1a|cs.3940.077 x 10yH) 10 ‘The average heat-transfer coefficient is then calculated as (0.0305)7.70 : ac h ors = 240: Wim? °C (1.67 Bwih - fit « °F] Its interesting to compare this value with that which would be obtained for strictly laminar forced convection. The Sieder-Tate relation [Eq. (6-10)] applies. so that momen (2°) L 1.86 G2? (2) 7 He cuoassay (2f ~ B$5H0.0308) _ 5 55 wim? °C (0.977 Blwh «°F Thus there would be an error of —41 percent if the calculation were made strictly on the basis of laminar forced convection, 7-43 SUMMARY By now the reader will have sensed that there is an abundance of empirical relations for natural convection systems. Our purposes in this section are to5-00 6rd 191 995 21295 cL 7s saoegins peam>ay x91 995 Ol > W110 > 5-01 afd 20) 9 = PON “aepns feau9A, elt QO¥O + 1. copy = aa mao iP L8EO eo FL ates oy w poe 9 reasoner 8/50) 9 = ON Baum y women Lamoo5 "suogmyoy 2;9UR1] 190} UopoaALED-2al4 Jo AreWUINg fz SIAL200 Nauwaconvecton systems (1) issue a few words of caution and (2) provide a convenient table to summarize the relations. Most free-convection data are collected under laboratory conditions in still air, still water, etc. A practical free-convection problem might not be so for- tunate and the boundary layer could have a slightly added forced-convection effect. In addition, real surfaces in practice are seldom isothermal or constant heat flux so the correlations developed from laboratory data for these conditions may not strictly apply. The net result, of course, is that the engineer must realize that calculated values of the heat-transfer coefficient can vary +25 percent from what will actually be experienced. To aid the reader in selecting correlations for free convection we have given a summary in Table 7-4. REVIEW QUESTIONS 1 Why is an analytical solution of a free-convection problem more involved than its forced-convection counterpart? 2 Define the Grashof number. What is its physical significance? 3. What is the approximate criterion for transition to turbulence in a free-convection boundary layer? ‘4 What functional form of equation is normally used for correlation of free-convection heat-transfer data? Discuss the problem of combined free and forced convection & What is the approximate criterion dividing pure conduction and free convection in ‘an enclosed space between vertical walls? 7 How is a modified Grashof number defined for a constant-heat-flux condition on @ vertical plate? 74. Suppose the heat-transfer coefficients for forced or free convection over vertical fiat plates are to be compared. Develop an approximate relation between the Reynolds and Grashof numbers such that the heat-transfer coefficients for pure forced convection and pure free convection are equal, Assume laminar flow. 7-2 For a vertical isothermal flat plate at 93°C exposed to air at 20°C and 1 atm, plot the free-convection velocity profiles as a function of distance from the plate surface at x positions of 15, 30, and 45 em. 7-3 Show that B = I/T for an ideal gas having the equation of state p = pRT 7-4 A L-fesquare vertical plate is maintained at 65°C and is exposed to atmospheric air at 15°C. Compare the free-convection heat transfer from this plate with that Which would result from forcing air over the plate at a velocity equal to the ‘maximum velocity which occurs in the free-convection boundary layer. Discuss this comparison.1 18 1 m2 m3 78 716 7 Probiems 961 Plot the free-convection boundary-layer thickness as a function of x fora vertical plate maintained at 80°C and exposed to air at atmospheric pressure and 15°C. Consider the laminar portion only. Derive an expression for the maximum velocity inthe free-convection boundary layer on a vertical lat plate. At what position in the boundary layer does this ‘maximum velocity occur? Two vertical flat plates at 65°C are placed in a tank of water at 25°C. If the plates ‘are 30 em high, what is the minimum spacing which will prevent interference of the free-convection boundary layers? 'A vertical cylinder having a length of 30 cm is maintained at 100°C and exposed to room air at 15°C. Calculate the minimum diameter the cylinder can have in forder to behave as a vertical Nat plate ‘A t-m-square vertical plate is heated to 400°C and placed in room air at 25°C, Calculate the heat loss from one side of the plate. A vertical flat plate is maintained at a constant temperature of 120°F and exposed to atmospheric air at 70°F, At a distance of 14 in from the leading edge of the plate the boundary layer thickness is 1,0in. Estimate the thickness of the boundary layer at a distance of 24 in from the leading edge. ‘Avertival cylinder 1.8 mhigh and 7.5 cmin diameter s maintained ata temperature ‘of 93°C in an atmospheric environment of 30°C. Calculate the heat lost by free convection from this cylinder. For this calculation the cylinder may be treated as a vertical flat plate “The outside wall ofa building 6 m high receives an average radiant heat ux from the sun of 1100 Wim’. Assuming that 98 Wim? is conducted through the wall. ccstimate the outside wall temperature, Assume the atmospheric air on the outside of the building is at 20°C. ‘Assuming that a human may be approximated by a vertical cylinder | fin diameter ‘and 6 ft tall, estimate the free-convection heat loss for a surface temperature of TSF in ambient air at 68°F. ‘A 30-cm-square vertical plate is heated electrically such that a constant-heat-fux condition is maintained with a total heat dissipation of 30 W. The ambient air is at 1 atm and 20°C. Calculate the value of the heat-transfer coefficient at heights Of 1S and 30 em. Also calculate the average heat-transfer coefficient forthe plate. 'A I-fl-square vertical plate is maintained at 120°F and exposed to room air at 1 ‘atm and 65°F. Calculate the heat lost from borh sides of the plate. Calculate the free-convection heat oss from a 2-f-square vertical plate maintained at 100°C and exposed to helium at 20°C and a pressure of 2 atm, A large vertical plate 20 ft high and 4 ft wide is maintained at a constant tem- perature of 135°F and exposed to atmospheric air at 40°F, Caleulate the heat lost by the plate A Lem-square vertical plate is maintained at 120°F and exposed to room air at 40'F. Calculate the heat lost by the plate749 720 721 722 728 724 725 728 v2 728 720 7.30 731 Natural convection systems ‘What vertical distance is necessary to produce a Rayleigh number of 10" in air at standard conditions and AT = 10°C? ‘A 25 by 25 cm vertical plate is fitted with an electric heater which produces constant heat flux of 1000 Wim?. The plat is submerged in water at 15°C. Calculate the heat-transfer coefficient and the average temperature of the plate. How much heat would be lost by an isothermal surface at this average temperature? ‘Assume that one-half of the heat transfer by free convection from a horizontal cylinder occurs on each side of the cylinder because of symmetry considerations. Going by this assumption, compare the heat transfer on each side of the eylinder with that from a vertical flat plate having a height equal to thécircumferential distance from the bottom stagnation point to the top stagnation point on the cylinder. Discuss this comparison, A horizontal cylindrical heater with d = 2 cm is placed in a pool of sodium- potassium mixture with 22 percent sodium. The mixture is at 120°C, and the heater surface is constant at 200°C. Calculate the heat transfer for a heater 40 cm Jong. A horizontal heating rod having a diameter of 3.0 em and a length of I m is placed in a pool of saturated liquid ammonia at 20°C. The heater is maintained at a constant surface temperature of 70°C. Calculate the heat-transfer rate ‘Condensing steam at 120°C is to be used inside a 7.5-cm-diameter horizontal pipe to provide heating for a certain work area where the ambient air temperature is 17°C. The total heating required is 100,000 Btu/h. What length pipe would be required to accomplish this heating? A 10-cm length of platinum wire 0.4 mm in diameter is placed horizontally in a container of water at 38°C and is electrically heated so that the surface temperature is maintained at 93°C. Calculate the heat lost by the wire. Water at the rate of 0.8 kg/s at 90°C flows through a steel pipe with 2.5-cm ID and 3-cm OD. The outside surface temperature of the pipe is 85°C, and the temperature of the surrounding air is 20°C. The room pressure is 1 atm, and the pipe is 15 m long. How much heat is lost by free convection to the room? ‘A horizontal pipe 8.0 cm in diameter is located in a room where atmospheric ai is at 20°C. The surface temperature of the pipe is 140°C. Calculate the free~ convection heat loss per meter of pipe. A horizontal 1.25-cm-OD tube is heated to a surface temperature of 250°C and ‘exposed to air at room temperature of 20°C and | atm. What isthe free-convection hheat transfer per unit length of tube? A horizontal electric heater 2.5 cm in diameter is submerged in a light-oil bath at ‘98°C. The heater surface temperature is maintained at 150°C. Calculate the heat lost per foot of length of the heater. A Ieft-square air-conditioning duct carries air at a temperature such that the ‘outside temperature of the duct is maintained at 60°F and is exposed to room ait at 80°F. Estimate the heat gained by the duct per foot of length, A fine wire having a diameter of 0.001 in (0.0254 mm) is heated by an electric732 133 14 735 7.38 737 738 7.39 740 at a2 143 744 Problems 3863 current and placed horizontally in a chamber containing helium at 3 atm and 10°C. Ifthe surface temperature of the wire is not to exceed 240°C, calculate the electric power to be supplied per unit length. ‘A large circular duct, 3.0 m in diameter, carries hot gases at 250°C. The outside Of the duct is exposed to room air at | atm and 20°C. Estimate the heat loss per unit length of the duct. ‘A2.0-cm diameter is placed ina tank of glycerine at 20°C. The surface temperature of the heater is 60°C, and its length is 60 cm. Calculate the heat transfer. 'A3.5-cm-diameter cylinder contains an electric heater which maintains a constant hheat flux atthe surface of 1500 Wim. Ifthe cylinder is inclined at an angle of 35° with the horizontal and exposed to room air at 20°C, estimate the average surface temperature. 'A.30-cm-diameter horizontal pipe is maintained at a constant temperature of 25°C ‘and placed in room air at 20°C. Calculate the free-convection heat loss from the pipe per unit length. ‘A S.in-diameter duct is maintained at a constant temperature of SO0°F by hot ‘combustion gases inside, The duct is located horizontally in a small warehouse farea having an ambient temperature of 60°F. Calculate the length of the duct necessary to provide 125,000 Btuh of convection heating A horizontal cylinder with diameter of 5 cm and length of 3 m is maintained at 180°F and submerged in water which is at 60°F. Calculate the heat lost by the cylinder. : ‘A 2.0-m-diameter horizontal cylinder is maintained at a constant temperature of 77°C and exposed to a large warehouse space at 27°C. The cylinder is 20 m tong. Calculate the heat lost by the cylinder. Calculate the rate of free-convection heat loss from a 30-cm-diameter sphere maintained at 90°C and exposed to atmospheric air at 20°C. ‘A 2.5-cm-diameter sphere at 32°C is submerged in water at 10°C. Calculate the rate of free-convection heat loss. ‘A spherical balloon gondola 2.4 m in diameter rises to an altitude where the ambient pressure is 1.4 kPa and the ambient temperature is ~ SO°C. The outside Surface of.the sphere is at approximately O°C. Estimate the free-convection heat toss from the outside of the sphere. How does this compare with the forced- ‘convection loss from such a sphere with a low free-stream velocity of approxi- mately 30 cm/s? 'A.2.S-cm diameter sphere is maintained at 38°C and submerged in water at 15°C. Calculate the heat-transfer rate under these conditions. ‘Apply the reasoning pertaining to the last entry of Table 7-1 to free convection from a sphere and compare with Eq. (7-50) Using the information in Table 7-1 and the simplified relations of Table 7-2 devise ‘a simplified relation which may be used as a substitute for Eq. (7-50) to calculate free convection from a sphere to air at | atm.748 748 77 748 749 730 751 782 733 754 755 788 787 Natuia-convecton systoms A circular hot plate, 15 cm in diameter, is maintained at 150°C in atmospheric air at 20°C, Calculate the free-convection heat loss when the plate isin a horizontal position, An engine-oil heater consists of a large vessel with a square-plate electric-heater surface in the bottom of the vessel. The heater plate is 30 by 30 cm and is ‘maintained at a constant temperature of 60°C. Calculate the heat-transfer rate for ‘an oil temperature of 20°C. Small electric strip heaters with a width of 6 mm are oriented in a horizontal position. The strips are maintained at S00°C and exposed to room air at 20°C. ‘Assuming that the strips dissipate heat from both the top and the bottom surfaces, estimate the strip length required to dissipate 2 kW of heat by free convection, ‘The top surface of a 10 by 10 m horizontal plate is maintained at 25°C and exposed to room temperature at 28°C. Estimate the heat transfer. ‘A-4 by 4 m horizontal heater is placed in room air at 15°C. Both the top and the bottom surfaces are heated to 50°C. Estimate the total heat loss by free convection, A horizontal plate, uniform in temperature at 400 K, has the shape of an equilateral triangle 45 em on each side and is exposed to atmospheric air at 300 K. Calculate the heat lost by the plate, AA heated plate, 20 by 20 cm, is inclined at an angle of 60° with the horizontal and placed in water. Approximately constant-heat-fux conditions prevail with a mean plate temperature of 40°C and the heated surface facing downward. ‘The water temperature is 20°C. Calculate the heat lost by the plate. Repeat Prob. 7-51 for the heated plate facing upward. A double plate-glass window is constructed with a 1.25-cm air space. The plate dimensions are 1.2 by 1.8 m. Calculate the free-convection heatransfer rate through the air space for a temperature difference of 30°C and T, = 20°C. A flatplate solar collector is | m square and is inclined at an angle of 20° with the horizontal. The hot surface at 160°C is placed in an enclosure which is evac- uated to a pressure of 0.1 atm. Above the hot surface, and parallel to it, is the transparent window which admits the radiant energy from the sun. The hot surface and window are separated by a distance of 8 cm. Because of convection to the surroundings, the window temperature is maintained at 40°C. Calculate the free. convection heat transfer between the hot surface and the transparent window. A flat plate 1 by 1 m is inclined at 30° with the horizontal and exposed to at- ‘mospheric air at 30°C and 1 atm. The plate receives a net radiant-energy Aux from the sun of 700 Wim?, which then is dissipated to the surroundings by free convection. What average temperature will be attained by the plate? A horizontal cylinder having a diameter of 5 em and an emissivity of 0.5 is placed in a large room, the walls of which are maintained at 35C. The cylinder loses hheat by natural convection with an h of 6.5 Wim? :*C. A sensitive thermocouple placed on the surface of the cylinder measures the temperature as 30°C. What is the temperature of the air in the room? A 10 by 10.cm plate is maintained at 80°C and inclined at 45° with the horizontal Caleutate the heat loss from both sides of the plate to room air at 20°C.758 759 160 781 782 783 764 765 7.68 181 788 789 170 Proves 368 A'S by Sem plate is maintained at S0°C and inclined at 60" with the horizontal. Calculate the heat loss from both sides of the plate to water at 20°C. [Air at | atm and 38°C is forced through a horizontal 6.5-mm-diameter tube at an ‘average velocity of 30 m/s, The tube wall is maintained at $40°C, and the tube is 530 em long, Calculate the average heat-transfer coefficient. Repeat for a velocity of 30 m/s and a tube wall temperature of 800°C. [A small copper block having a square bottom 2.5 by 2.5 cm and a vertical height ‘of Sem cools in room air at 1 atm and 20°C. The block is isothermal at 93°C. Calculate the heat-transfer rate ‘A horizontal plate in the shape of an equilateral triangle 40 cm on a side is maintained at a constant temperature of 55°C and exposed to atmospheric ait at 25°C. Calculate the heat lost by the top surface of the plate A small horizontal heater is in the shape of a circular disk with a diameter of Sem. The disk is maintained at 50°C and exposed to atmospheric air at 30°C. Caleulate the heat loss A hot ceramic block at 400°C has dimensions of 15 by 15 em by 8 em high. Its ‘exposed to room air at 27°C. Calculate the frec-convection heat loss ‘A magnetic amplifier is encased in a cubical box 6 in on a side and must dissipate ‘50 W to surrounding air at 70°F. Estimate the surface temperature of the box A glass thermometer is placed in a large room, the walls of which are maintained at 10°C. The convection coefficient between the thermometer and the room air is $ Wim? -°C, and the thermometer indicates a temperature of 30°C. Determine the temperature of the air in the room. Take ¢= 1.0. ‘A horizontal air-conditioning duct having a horizontal dimension of 30 em and a vertical dimension of 15 cm is maintained at 120°F and exposed to atmospheric air at 70°F. Calculate the heat lost per unit length of duct, A free-convection heuter is to be designed which will dissipate 10.000 ky/h to room air at 300 K, The heater surface temperature must not exceed 350 K. Consider four alternatives: (a) & group of vertical surfaces, (b) a single vertical surface, (c) a single horizontal surface, and (d) a group of horizontal cylindrical surfaces, Examine these alternatives and suggest a esign. “Two 30-cm-square vertical plates are separated by a distance of 1.25 em, and the space between them is filled with water. A constant-heat-flux condition is imposed ‘on the plates such that the average temperature is 38°C for one and 60°C for the “other. Calculate the heat-transfer rate under these conditions. Evaluate properties at the mean temperature “An enclosure contains helium at a pressure of 1.3 atm and has two vertical heating surfaces, which are maintained at 80 and 20°C, respectively. The vertical surfaces are 40 by 40 cm and are separated by a gap of 2.0m. Calculate the free-convection heat transfer between the vertical surfaces. ‘A horizontal annulus with inside and outside diameters of 8 and 10 cm. respec~ tively, contains liquid water, The inside and outside surfaces are maintained at 40 and 20°C, respectively, Calculate the heat transfer across the annulus space per meter of length.mn mm mm 18 776 om 178 r 7-80 781 782 793 794 Natwat-convocton systems ‘Two concentric spheres are arranged to provide storage of brine inside the inner sphere at a temperature of ~ 10°C. The inner-sphere diameter is 2 m, and the gap spacing is 5 em. The outer sphere is maintained at 30°C, and the gap space is evacuated to a pressure of 0.05 atm. Estimate the free-convection heat transfer across the gap space. A large vat used in food processing contains a hot oil at 400°F. Surrounding the vat on the vertical sides isa shell which is cooled to 140°F, The air space separating the vat and the shell is 35 em high and 3 cm thick. Estimate the free-convection loss per square meter of surface area. ‘A special double-pane insulating window glass is to be constructed of two glass plates separated by an air gap. The plates are square, 60 by 60 cm, and are designed to be used with temperatures of ~ 10.and + 20°C on the respective plates. ‘Assuming the air in the gap is at 1 atm, calculate and plot the free convection across the gap as a function of gap spacing for a vertical window. What conclusions can you draw from this plot from a design standpoint? Repeat Prob. 7-73 for a horizontal window with the hot surface on the lower side. ‘Two 30-cm-square vertical plates are separated by a distance of 2.5 em and air at | atm. The two plates are maintained at temperatures of 200 and 90°C. re- spectively. Calculate the heat-transfer rate across the ai space. A horizontal air space is separated by a distance of 1.6 mm, Estimate the heat- transfer rate per unit urew for a temperature difference of 165°C, with one plate temperature ut 90°C Repeat Prob. 7.76 for a hor zontal space filled with water. ‘An atmospheric vertical air space 4.0 ft high has a temperature differential of 20°F it 300 K. Calculate and plot kk and the R value for spacings of 0 to 10 in. At approximately what spacing is the R value maximum? ‘Two vertical plates 50 by S0 em ure separated by a space of 4 em which is filled with water. The plate temperatures are 50 and 20°C. Calculate the heat transfer across the space. Repeat Prob. 7-79 for the plates oriented in a horizontal position with the SO°C surface ay the lower plate Two vertical platey 1.1 by 1-1 m are separated by a 4.0-cm air space. The two surface temperatures are at 300 and 350 K. The heat transfer in the space can be reduced by decreasing the pressure of the air. Calculate the plot kJ/k and the R value as a function of pressure. To what value must the pressure be reduced to make kik = 1.0? Repeut Prob. 7-81 for two horizontal plates with the 350 K surface on the bottom. ‘An air space in a certain building wall is 10 em thick and 2 m high. Estimate the free-convection heat transfer through this space for a temperature difference of, Inc, Develop an expression for the optimum spacing for vertical plates in air in order to achieve minimum heat transfer, assuming that the heat transfer results from796 87 788 eleroncos 367 pure conduction at Grs < 2000. Plot this optimum spacing as a function of tem- perature difference for air at 1 atm. Air at atmospheric pressure is contained between two vertical plates maintained ‘at 100°C and 20°C, respectively. The plates are 1.0 m on a side and spaced 8 cm ‘part, Calculate the convection heat transfer across the air space A special section of insulating glass is constructed of two glass plates 30 em ‘square separated by an air space of cm. Calculate the percent reduction in heat transfer of this arrangement compared to free convection from a vertical plate ‘with a temperature difference of 30°C. ‘One way to reduce the free-convection heat loss in a horizontal solar collector is to reduce the pressure in the space separating the glass admitting the solar ‘energy and the black absorber below. Assume the bottom surface is at 120°C and the top surface is at 20°C. Calculate the pressures that are necessary to eliminate convection for spacings of I, 2, 5. and 10.¢m Air at 20°C and | atm is forced upward through a vertical 2.5-em-diameter tube 30-em long. Calculate the total heat-transfer rate where the tube wall is maintained fat 200°C and the flow velocity is 45 ems, 1 " Eckert. E. R. G.. and , Sochngen: Interferometric Studies on the Stability and ‘Transition to Turbulence of a Free Convection Boundary Layer. Proc. Gen. Dis- cuss. Heat Transfer ASME-IME, London, 1951 Eckert, E. R. G., and E, Sochngen: Studies on Heat Transfer in Laminar Free Convection with the Zehnder-Mach Interferometer, USAF Tech. Rep. S747, De- ccember 1948. Holman, J. P., H. E. Gartrell, and E. E, Soehngen: An Interferometric Method of Studying Boundary Layer Oscillations, J. Heat Transfer. ser. C, vol. 80, August 1960, McAdams, W. H.: “Heat Transmission,” 3d ed., McGraw-Hill Book Company. New York, 1954, Yuge, T-: Experiments on Heat Transfer from Spheres Including Combined Natural land Forced Convection, J. Heut Transfer, ser. C. vol. 82, p. 214, 1960. Jakob, M.: Free Convection through Enclosed Gas Layers, Trans. ASME, vol. 68 p. 189, 1946, Jakob, M.: Heat Transfer.” vol. 1, John Wiley & Sons. Inc., New York, 1949. Globe, S., and D. Dropkin: J. Heat Transfer, February 1959, pp. 24-28, Evans, L. B. and N. B, Stefany: An Experimental Study of Transient Heat Transfer to Liquids in Cylindrical Enclosures. AIChE Pap. 4, Heat Transfer Conf. Los Angeles, August 1965. Metais, B., and E. R. G, Eckert: Forced, Mixed, and Free Convection Regimes, J. Heat Transfer, ser. C. vol. 86. p. 295, 1964, Bishop, E.N..L.R. Mack, andJ. A. Scanlan: Heat Transfer by Natural Convection between Concentric Spheres, Int. J. Heat Mass Transfer, vol. 9, p. 649, 1966308 Naneat-convecton systems 2 “ 7 8 9 2 4 25 6 ea a Dropkin, D., and E. Somerscales: Heat Transfer by Natural Convection in Liquids Confined by Two Parallel Plates Which Are Inclined at Various Angles with Respect to the Horizontal, J. Heat Transfer. vol. 87, p. 71, 1965, Gebhart, B.: “Heat Transfer,” 2d ed., chap. 8, McGraw-Hill Book Company, New York, 1970. Gebhart, B.: Natural Convection Flow, Instability, 69-HT-29, August 1969. Mollendorf, J. C., and B. Gebhart: An Experimental Study of Vigorous Transient Natural Convection, ASME Pap. 70-HT-2, May 1970, Bayley, F. J.: An Analysis of Turbulent Free Convection Heat Transfer, Proc. Inst. Mech. Eng., vol. 169, no. 20, p. 361, 1955, Brown, C. K., and W. H. Gauvin: Combined Free and Forced Convection, 1, Il. Can. J. Chem. Eng., vol. 43, no. 6, pp. 306, 313, 1965, MacGregor, R. K., and A. P. Emery: Free Convection through Vertical Plane Layers: Moderate and High Prandtl Number Fluids, J. Heat Transfer, vol. 91, p. 391, 1968. Newell, M. E., and F. W. Schmidt: Heat Transfer by Laminar Natural Convection within Rectangular Enclosures, J. Heat Transfer, vol. 92, pp. 159-168, 1970. Husar, R. B., and E. M. Sparrow: Patterns of Free Convection Flow Adjacent to Horizontal Heated Surfaces, Int. J. Heat Mass Trans., vol. 11, p. 1206, 1968. Habne, E. W. P.: Heat Transfer and Natural Convection Patterns on a Horizontal Circular Plate, Int. J. Heat Mass Transfer, vol. 12, p. 651, 1969, Warner, C. Y., and V. S. Arpaci: An Experimental Investigation of Turbulent Natural Convection in Air at Low Pressure along a Vertical Heated Flat Plate, Int J. Heat Mass Transfer, vol. 11, p. 397, 1968. Gunness, R. C., Jr., and B. Gebhart: Stability of Transient Convection, Phys. Fluids, vol. 12, p. 1968, 1969. Rotern, Z.. and L. Claassen: Natura! Convection above Unconfined Horizontal Surfaces, J. Fluid Mech., vol. 39, pt. 1, p. 173, 1969. Vliet, G. C.: Natural Convection Local Heat Transfer on Constant Heat Flux Inclined Surfaces, J. Heat Transfer, vol. 91, p. S11, 1968. Vliet, G. C., and C. K. Lin: An Experimental Study of Turbulent Natural Con- vection Boundary Layers, J. Heat Transfer, vol. 91. p. 517, 1969, Ostrach, $.: An Analysis of Laminar-Free-Convection Flow and Heat Transfer about a Flat Plate Parallel to the Direction of the Generating Body Force, NACA Tech, Rep. WAL, 1953. Cheesewright, R.: Turbulent Natural Convection from a Vertical Plane Surface, J. Heat Transfer, vol. 90, p. 1, February 1968 Flack, R. D., and C. L. Witt: Velocity Measurements in Two Natural Convection Air Flows Using a Laser Velocimeter, J. Heat Transfer, vol. 101, p. 256, 1979. Eckert, E. R. G.. and T. W, Jackson: Analysis of Turbulent Free Convection Boundary Layer on a Flat Plate, NACA Rep. 1015, 1951 King, W. J.: The Basic Laws and Data of Heat Transmission, Mech. Eng.. vol. 54, p. 347, 1932. ind Transition, ASME Pap.2 a 2 4 46 a 4 0 Feleronces 369 Sparrow, E, M., and J. L. Gregg: Laminar Free Convection from a Vertical Flat Plate. Trans. ASME, vol. 78, p. 435, 1956. Benard, H.: Les Tourbillons cellulaires dans une nappe liquide transportant de la Chaleur par convection en régime permanent, Ann. Chim. Phys. vol, 23, pp. 62-144, 1901 “Progress in Heat and Mass Transfer.” vol. 2, gamon Press, New York, 1969 Gebhart, B., T. Audunson, and L. Pera: Fourth Int. Heat Transfer Conf, Paris, August 1970. ‘Sanders, C, J.. and J. P. Holman: Franz Grashof and the Grashof Number, Int J Heat Mass Transfer, vol. 15, p. $62, 1972. Clifton, J. V., and A. J. Chapman: Natural Convection on a Finite- Plate, Int. J. Heat Mass Transfer, vol. 12, p. 1573, 1969, Emery, A. F., H. W. Chi, and J. D. Dale: Free Convection through Vertical Plane Layers of Non-Newtonian Power Law Fluids, ASME Pap. 70-WAVHT-1 Vliet, G. C.: Natural Convection Local Heat Transfer on Constant Heat Flux Inclined Surfaces, Trans. ASME, vol. 91C, p. 511, 1968. Bergles, A. E.. and R. R. Simonds: Combined Forced and Free Convection for Laminar Flow in Horizontal Tubes with Uniform Heat Flux, Int, J. Hear Mass Transfer. vol. 14, p. 1989, 1971 ‘Aihara, T.. Y. Yamada, and S, Endo: Free Convection along the Downward-facing Surface of a Heated Horizontal Plate, Int. J. Heat Mass Transfer. vol. 15. p. 2835 1972 ‘Saunders, O. A.. M. Fishenden, and H. D. Mansion: Some Measurement of Con- Vection by an Optical Method. Engineering, p. 483, May 1935, Weber, N.. R. E Rowe, E. H. Bishop, and J. A. Scanlan: Heat Transfer by Natural Convection between Vertically Eccentric Spheres, ASME Pap. 72-WAIHT-2 Fuji, T., and H. Imura; Natural Convection Heat Transfer from a Plate with Arbitrary Inclination, Int. J. Heat Mass Transfer, vol. 1S, p. 735, 1972. Pera, L.. and B. Gebhart: Natural Convection Boundary Layer Flow over Hori Jontal and Slightly Inclined Surfaces, Int. J. Heat Mass Transfer, vol. 16, p. L131. 1973 Hyman, S. C.. C.F. Bonilla, and S. W. Ehrlich: Heat Transfer to Liquid) Metals fram Horizontal Cylinders, AiChE Symp. Heat Transfer, Atlantic City, 1983. 24 Fund, R. M..and K. K. Keswani: Combined Natural and Forced Convection Heat Transfer from Horizontal Cylinders to Water, Int. J. Heat Mass Transfer. vol. 16. p. 175, 1973. Dale, J. D..and A. F. Emery: The Free Convection of Heat from a Vertical Plate te Several Non-Newtonian Pseudoplastic Fluids, ASME Pap. 71-HT-S. Fujii, T,, 0. Miyatake, M. Fujii, H. Tanaka, and K. Murakami: Natural Convective Heat Transfer from a Vertical Isothermal Surface to a Non-Newtonian Sutterby Fluid, Int. J. Heat Mass Transfer, vol, 6. p. 2177, 1973. Sochngen, E. E.: Experimental Studies on Heat Transfer at Very High Prandt! Numbers, Prog. Heat Mass Trunsfer. vol. 2. p. 12S. 1969. ickert Presentation Volume, Per- ize Horizontal{370 Netuskconvecton systems Vliet, G. C., and D. C. Ross: Turbulent Natural Convection on Upward and Down- ward Facing Inclined Constant Heat Flux Surfaces, ASME Pap. 74.WA/HT.32 Ted dR and W. R. Moran: Natural Convection Adjacent to Horizontal Surface of Various Planforms, ASME Pap. 74-WAJHT-66. Goldstein, R. J., E. M. Sparrow, and D. C. Jones: Natural Convection Mass Trans- fer Adjacent to Horizontal Plates, int. J. Heat Mass Transfer, vol. 16. p. 1025, 1973, Holman, J. P., and J. H. Boggs: Heat Transfer to Freon 12 near the Critical State in a Natural Circulation Loop, J. Heat Transfer, vol. 80, p. 221, 1960, ‘Mull, W., and H. Reiher: Der Warmeschutz von Luftschichten, Beih. Gesund. Ing, ser. 1, no. 28, 1930, Krasshold, H.: Warmeabgabe von zylindrischen Flussigkeitsschichten bei nator- ren Konvektion, Forsch, Geb. Ingenieurwes, vol. 2, p. 165, 1931. Beckmann, W.: Die Warmeibertragung in zylindrischen Gasschichten bei natir- licher Konvektion, Forsch. Geb. Ingenieurwes, vol. 2, p. 186, 1931 ‘Schmidt, E.: Free Convection in Horizontal Fluid Spaces Heated from Below, Proc. Int. Heat Transfer Conf., Boulder, Colo., ASME, 1961. Graff, J. G. A., and E. F. M. Van der Held: The Relation between the Heat Transfer ‘and Convection Phenomena in Enclosed Plain Air Players, Appl. Sc. Res., ser. A, vol. 3, p. 393, 1952, Liu, C. Y., W. K. Mueller, and F. Landis: Natural Convection Heat Transfer in Long Horizontal Cylindrical Annuli, Int. Dev. Heat Transfer, pt. 5, pap. 117, p. 976, 1961 Emery, A., and N. C. Chu: Heat Transfer across Vertical Layers, J. Heat Transfer, vol. 87, p. 110, 1965. O'Toole, J., and P. L. Silveston: Correlation of Convective Heat Transfer in Con- fined Horizontal Layers, Chem. Eng. Prog. Symp., vol. 57, n0. 32, p. 81, 1961 Goldstein, R. J., and T. Y. Chu: Thermal Convection in a Horizontal Layer of Ai Prog. Heat Mass Transfer, vol. 2, p. 55, 1969, Singh, S. N., R. C. Birkebak, and R. M. Drake: Laminar Free Convection Heat Transfer from Downward-facing Horizontal Surfaces of Finite Dimensions, Prog. Heat Mass Transfer, vol. 2, p. 87, 1969. ‘McDonald, J. S., and T. J. Connally: Investigation of Natural Convection Heat ‘Transfer in Liquid Sodium, Nucl Sci. Eng., vol. 8, p. 369, 1960. Hollands, K. G. T., G. D. Raithby, and L. Konicek: Correlation Equations for Free Convection Heat Transfer in Horizontal Layers of Air and Water, Int. J. Heat Mass Transfer, vol. 18, p. 879, 1975. Hollands, K. G. T.,T. E. Unny, and G. D. Raithby: Free Convective Heat Transfer ‘across Inclined Air Layers, ASME Pap. 75-HT-5S, August 1975. Depew. C. A.,J. L. Franklin, and C. H. Ito: Combined Free and Forced Convection in Horizontal, Uniformly Heated Tubes, ASME Pap. 75-HT-19, August 1975 Raithby, G. D., and K. G. T. Hollands: A General Method of Obtaining Approx- imate Solutions to Laminar and Turbulent Free Convection Problems, Advances {in Heat Transfer, Academic Press, New York, 1974.7m n n a ” 6 n Felerences 374 Churchill, $. W..and H. H. S. Chu; Correlating Equations for Laminar and ‘Tur- bulent Free Convection from a Horizontal Cylinder, Int. J. Heat Mass Transfer, vol. 18, p. 1049, 1975 Churchill, §, W.. and H. H. S. Chu: Correlating Equations for Laminar and Tur pulent Free Convection from a Vertical Plate, Int. J. Heat Mass Transfer, vol. 18 p. 1323, 1975. Churchill, §. W.: A Comprehensive Correlating Equation for Laminar, Assisting, Forced and Free Convection, AtChE J.. vol. 23, no. 1, p. 10, 1977. AlArabi, M.. and Y. K, Salman: Laminar Natural Convection Heat Transfer from fan Inclined Cylinder. Int. J. Heat Mass Transfer, vol. 23, pp. 45-51, 1980. Holman, J. P.: "Heat Transfer.” 4th ed., McGraw-Hill Book Co... New York, 1976 Hatfield, D. W...and D. K. Edwards: Edge and Aspect Ratio Effects on Naturat Convection from the Horizontal Heated Plate Facing Downwards. Int. J. Heat Mass Transfer, vol. 24, p. 1019, 1981 Morgan, V. T.: “The Overall Convective Heat Transfer from Smooth Circular Cylinders," “*Advanees in Heat Transfer” (T. P. Irvine and J. P. Hartnett, eds.), vol. 11, Academic Press. Inc., New York, 1975. Sparrow, E. M., and M, A. Ansari: A Refutation of King’s Rule for Multi- Dimensional External Natural Convection, Int. J. Heat Mass Transfer, vol. 26. p. 1357, 1983, Lienhard, J. H.: On the Commonality of Equations for Natural Convection from Immersed Bodies, Int. J. Heat Mass Transfer, vol. 16, p. 2121, 1973 ‘Amato, W. S., and C. &.. Tien: Free Convection Heat Transfer from Isothermal Spheres in Water. Int. J. Heat Mass Transfer, vol. 15. p. 327, 1972. Warrington, R.O., and R. F, Powe: The Transfer of Heat by Natural Convection Between Bodies and Their Enclosures, Ine. J. Heat Mass Transfer, vol. 28, p. 319. 1985, Sparrow, E, M., and A. J. Stretton: Natural Convection from Bodies of Unity ‘Aspect Ratio, Int. J. Heat Mass Transfer, vol. 28. p. 741, 1985.RADIATION HEAT TRANSFER = 8-1 INTRODUCTION Preceding chapters have shown how conduction and convection heat transfer may be calculated with the aid of both mathematical analysis and empirical data, We now wish to consider the third mode of heat transfer—thermal ra- diation. Thermal radiation is that electromagnetic radiation emitted by a body as a result of its temperature. In this chapter. we shall first describe the nature of thermal radiaton, its characteristics, and the properties which are used to describe materials insofar as the radiation is concerned. Next, the transfer of radiation through space will be considered. Finally, the overall problem of heat transfer by thermal radiation will be analyzed, including the influence of the material properties and the geometric arrangement of the bodies on the total eneigy which may be exchanged. 6-2 PHYSICAL MECHANISM ‘There are many types of electromagnetic radiation: thermal radiation is only one. Regardless of the type of radiation, we say that it is propagated at the speed of light, 3 x 10* mis. This speed is equal to the product of the wavelength and frequency of the radiation, caay where ¢ = speed of light A= wavelength v = frequency ‘The unit for A may be centimeters, angstroms (1 A = 10 * cm), or micrometers (um = 10m). A portion of the electromagnetic spectrum is shown in Fig. 8-1. Thermal radiation lies in the range from about 0.1 to 100m, while the 373‘374 Racin reat transer = 1 wove i { Tae waves visible? Fig. 61 Electromagnetic spectrum, visible-light portion of the spectrum is very narrow, extending from about 0.35 t0 0.75 um. The propagation of thermal radiation takes place in the form of discrete quanta, each quantum having an energy of E=hv @1 where h is Planck’s constant and has the value h = 6.625 x 10° J +5 A very rough physical picture of the radiation propagation may be obtained by considering each quantum as a particle having energy, mass, and momentum, just as we considered the molecules of a gas. So, in a sense, the radiation might bbe thought of as a “photon gas" which may flow from one place to another. Using the relativistic relation between mass and energy, expressions for the ‘mass and momentum of the “particles” could thus be derived; viz, E= me = hv By considering the radiation as such a gas, the principles of quantum-statistical thermodynamics can be applied to derive an expression for the energy density of radiation per unit volume and per unit wavelength ast Sahca~> 82) + See, for example. J. P. Holman, “Thermodynamics,” 3d ed.. McGraw-Hill Book Company, New York. 1980. p. 350Radiaton properties 378 where & is Boltzmann's constant, 1.38066 x 10°** J/molecule » K, When the enerey density is integrated over all wavelengths, the total energy emitted is proportional to absolute temperature to the fourth power: E, = oT" 83) Equation (8-3) is called the Stefan-Boltzmann law. E, is the energy radiated per unit time and per unit area by the ideal radiator, and o is the Stefan- Boltzmann constant. which has the value of = 5.669 x 10 © Wim? K* [0.1714 x 10 * Bturh «ft R‘| where Ey is in watts per square meter and T is in degrees Kelvin. In the thermodynamic analysis the energy density is related to the energy radiated from a surface per unit time and per unit area. Thus the heated interior surface of an enclosure produces a certain energy density of thermal radiation in the enclosure. We are interested in radiant exchange with surfaces—hence the feason for the expression of radiation from a surface in terms of its temperature. ‘The subscript b in Eq. (8-3) denotes that this is the radiation from a blackbody We call this blackbody radiation because materials which obey this law appear black to the eye; they appear black because they do not reflect any radiation “Thus a blackbody is also considered as one which absorbs all radiation incident upon it. Ey is called the emissive power of a blackbody. it is important to note at this point that the “blackness” of a surface-to- thermal radiation can be quite deceiving insofar as visual observations are ‘concerned. A surface coated with lampblack appears black (o the eye and turns ut to be black for the thermal-radiation spectrum. On the other hand, snow fand ice appear quite bright to the eye but arc essentially “black” for long wavelength thermal radiation. Many white paints are also essentially black for Jong- wavelength radiation. This point will be discussed further in later sections. 8-9 RADIATION PROPERTIES When radiant energy strikes a material surface, part of the radiation is reflected, part is absorbed, and part is transmitted, as shown in Fig, 8-2. We define the reflectivity p as the fraction reflected, the absorptivity @ as the fraction ab- sorbed, and the transmissivity 7 as the fraction transmitted. Thus ptatr=l (8-4) Most solid bodies do not transmit thermal radiation, so that for many applied problems the transmissivity may be taken as zero. Then pta=4 ‘Two types of reflection phenomena may be observed when radiation strikes ‘a surface. Ir the angle of incidence is equal to the angle of reflection, the feflection is called specular, On the other hand, when an incident beam is distributed uniformly in all directions after reflection, the reflection is called376 Raciaton hat ransler Incident radiation Reflection : Fig.8-2 Sketch showing effects of incident radia- Tranemited tion diffuse. These two types of reflection are depicted in Fig. 8-3. Note that a specular reflection presents a mirror image of the source to the observer. No real surface is either specular or diffuse. An ordinary mirror is quite specular for visible light, but would not necessarily be specular over the entire wave- length range of thermal radiation. Ordinarily, a rough surface exhibits diffuse behavior better than a highly polished surface. Similarly, a polished surface is ‘more specular than a rough surface. The influence of surface roughness on thermal-radiation properties of materials is a matter of serious concern and remains a subject for continuing research. ‘The emissive power of a body E is defined as the energy emitted by the body per unit area and per unit time. One may perform a thought experiment to establish a relation between the emissive power of a body and the material properties defined above. Assume that a perfectly black enclosure is available, i.e., one which absorbs all the incident radiation falling upon it, as shown schematically in Fig. 8-4. This enclosure will also emit radiation according to the T* law. Let the radiant flux arriving at some area in the enclosure be 4, Win?. Now suppose that a body is placed inside the enclosure and allowed to come into temperature equilibrium with it. At equilibrium the energy ab- sorbed by the body must be equal to the energy emitted; otherwise there would Source S ofsoune a ® Fig. 83 (1) Specular ( ~ 4) and (2) ate retdyonRediaten prooemies 377 Fig. 6-4 Sketch showing mode! used for deriv Ing Kichnot slaw. be an energy flow into or out of the body which would raise or lower its temperature. At equilibrium we may write EA = qAa @5) If we now replace the body in the. enclosure with a blackbody of the same size and shape and allow it to come to equilibrium with the enclosure at the same temperature, ExA = ql) (86) since the absorptivity of a blackbody is unity. If Eq. (8-5) is divided by Eq. 86), Le Bn * and we find that the ratio of the emissive power of a body to the emissive power of a blackbody at the same temperature is equal to the absorptivity 01 the body. This ratio is defined as the emissivity ¢ of the body, E OD 7) so that ena (88) Equation (8-8) is called Kirchhoff’s identity. At this point we note that the emissivities and absorptivities which have been discussed are the total prop- erties of the particular material; i.e, they represent the integrated behavior of the material over all wavelengths. Real substances emit less radiation than ideal black surfaces as measured by the emissivity of the material. In reality, the emissivity of a material varies with temperature and the wavelength of the radiation.‘378 Radiation heat anster ‘The Gray Body A gray body is defined such that the monochromatic emissivity €, of the body is independent of wavelength. The monochromatic emissivity is defined as the ratio of the monochromatic-emissive power of the body to the monochromatic emissive power of a blackbody at the same wavelength and temperature. Thus Em The total emissivity of the body may be related to the monochromatic emissivity by noting t E ex Em dd so that e= co) where E,, is the emissive power of a blackbody per unit wavelength. If the gray-body condition is imposed, that is, «, = constant, Eq. (8-9) reduces to era (8-10) ‘The emissivities of various substances vary widely with wavelength, tem- perature, and surface condition. Some typical values of the total emissivity of various surfaces are given in Appendix A. A very complete survey of radiation properties is given in Ref. 14 The functional relation for Ey, was derived by Planck by introducing the quantum concept for electromagnetic energy. The derivation is now usually performed by methods of statistical thermodynamics, and E,, is shown to be related to the energy density of Eq. (8-2) by En = 11) or Ess (8-12) eat where A = wavelength, um T = temperature, K C. = 3.743 x 10° W > wm‘? [1.187 x 10" Btu + emfh « £2) Co = 1.4387 10% um = K (2.5896 x 10* wm - *R] A plot of Es, as a function of temperature and wavelength is given in Fig. 8- Sa. Notice that the peak of the curve is shifted to the shorter wavelengths for the higher temperatures. These maximum points in the radiation curves are related by Wien’s displacement law, AmasT = 2897.6 wm -K [5215.6 wm - *R] (8-13)Radiabon properties 370 Monochromatic emisine power. yx X 108 Btu? ym oleats Wavelength em 300 150} Fig. 8.5. (2) Blackbody emssive power as a function of wavelength and Temperature, (b) comparison of emssive power of ideal Blackbocies and ray bodies with tna ofa veal surtace‘200 Racin heat varster Figure 8-5b indicates the relative radiation spectra from a blackbody at 3000°F and a corresponding ideal gray body with emissivity equal to 0.6. Also shown is a curve indicating an approximate behavior for a real surface, which may differ considerably from that of either an ideal blackbody or an ideal gray body. For analysis purposes surfaces are usually considered as gray bodies, with cemissivities taken as the integrated average value. The shift in the maximum point of the radiation curve explains the change in color of a body as it is heated. Since the band of wavelengths visible to the eye lies between about 0.3 and 0.7 um, only a very small portion of the radiant- energy spectrum at low temperatures is detected by the eye. As the body is heated, the maximum intensity is shifted to the shorter wavelengths, and the first visible sign of the increase in temperature of the body is a dark-red color. With further increase in temperature, the color appears as a bright red, then bright yellow, and finally white. The material also appears much brighter at higher temperatures because a larger portion of the total radiation falls within the visible range. We are frequently interested in the amount of energy radiated from a black- body in a certain specified wavelength range. The fraction of the total energy radiated between 0 and A is given by ff tua Fog = (8-14) ff eo an Equation (8-12) may be rewritten by dividing both sides by T*, so that fn & (B15) TOD Now, for any specified temperature, the integrals in Eq. (8-14) may be expressed in terms of the single variable AT. The results have been tabulated by Dunkle 19 oa 06 ag| 1s 20 ar x 10-,6—"R 10 1S Fig. 86 Fraction of olackbody radiation in ram. *K wavelength intervalRaciaton propertes aat Incident radiation Fig. 67. Method of constructing a blackbody enclosure. {2}. The ratio in Eq. (8-14) is plotted in Fig. 8-6 and tabulated in Table 8-1. If the radiant energy emitted between wavelengths A, and Az is desired, then Evy E, ca = Eo, (Et ft Faas ™ Bon (Get ~ Bo) 16) here Enis the total radiation emitted overall wavelengths, Ey... = oT" (17) and is obtained by integrating the Planck distribution formula of Eq, (8-12) over all wavelengths. The concept of a blackbody is an idealization; i-e.. a perfect blackbody does not exist—all surfaces reflect radiation to some extent, however slight. A black- body may be approximated very accurately, however, in the following way. A cavity is constructed, as shown in Fig. 8-7, so that it is very large compared with the size of the opening in the side. An incident ray of energy is reflected many times on the inside before finally escaping from the side opening. With each reflection there isa fraction of the energy absorbed corresponding to the absorptivity of the inside of the cavity. After the many absorptions, practically all the incident radiation at the side opening is absorbed. It should be noted that the cavity of Fig. 8-7 behaves approximately as a blackbody emitter as well as an absorber. EXAMPLE 8-1. Transmission and absorption in lass pate AA glass plate 30 cm square is used to view radiation from a furnace. The transmissivity of the glass is 0.5 from 0.2 to 3.5 ym. The emissivity may be assumed to be 0.3 up to 3.5 jum and 0.9 above that. The transmissivity of the glass is zero, except in the range from 0.2 to 3.5 um. Assuming that the furnace is a blackbody at 2000°C. calculate the energy absorbed in the glass and the energy transmitted, Solution T= 200°C = 2273 K AWT = (0.242273) = 454.6 um: K AsD = (B.9)2279) = 7955.5 wm > K 0.37 = 0.09 m*Table &1 Radiation Functions. ar EwT* Buu w em PRE pm mK am " x0" or Q.000671 0400 x 10-7070 x 10-7 0.0202 0.120 x 10° 0.756 x 10-8 0.204 0.00122 0.106 x 10-+ 1.057 0.00630 0.738 x 18+ 3544 o.o2i11 0.321 x 10- aan 0.08254 o.oo101 1776 0.10567 0.00252 30.686 0.18275 0.00531 an67 0-28081 0.00883 ee 6.336 039505 0.01683 mn 7.047 osiesi 0.03537 ' 108.14 0.64404 0.03677 128.58 0.76578 0.05059 147.56 on7e78 0.08672 16449 0.97963 0.08456 179.04 1.0663 0.10803, 191.05 L378 0.12668 200531 11942 0.14983 20735 12361 0.7337 m3 12645 o.19789 215.06 12808 0.22285 216.00 12864 0.24803 21539 12827 02m 213.46 1273 0.29825, 210.43 12832 0.32300 206.31 1229 034734 201.88 1.2023 o37is 196.69 nina 0.30448 91.09 11380 0.41708 15.18 11029 0.43505 179.08 066s 0.46031 172.86 1.0295 0.48085 166.60 ogni 050066 160.38 0.95499 osis7s 154.16 0813 0.53809 : 148.07 ogg 0.55573 i 12.10 0.84629 0.57267 ' 136.28 o8it6s o'58891 130.63 0.77796 0.60449 125.15 0.74534 o-sisat 119.86 071383 oss371 11476 0.68346 0.64740 109.85 0.65623 0.66051 105.14 0.62617 0.67305 10062 o.sosas 0.68506 96.289 0.57346 0.69655‘Table 61 Radiation Functions (Continued), pm °R 10,200 10,400 10,600 10,800 11,000 11.200 11,800 11,600 11.800 12,000 12.200 12.400 12,600 12,800 13,000 13.200 13.400 13,600 13,800 18,000 14.200 14.400 14.600 14.800 15,000 16,000 17,000 5,666.7 5778 5,888.9 6,000.0 6M 6.22.2 6333.3 6.4444 6555.6 6,666.7 67718 6,888.9 7,000.0 7AM 7202.2 7333.3 7484.4 7585.6 766.7 178 7988.9 8,000.0 SILL 8,222.2 8,333.3 8,888.9 9.4844 10,000.0 10,555.6 int 1,666.7 12,2222 17778 153333 13.8889 tae 1,000.0 151555.6 WIL 16,666.7 2,222.2 7718 33,333.3 38,388.9 444484 0,000.0 0.54877 0.2517 0.50261 0.48107 0.46051 0.44089 0.42218 0.40834 0.38732 omit 0.35565 0.34091 0.32687 031348 0.30071 0.28855 0.27695 0.26589 0.25534 0.26527 0.23567 0.22651 02177 0.20942 0.20145 0.16662 0.13877 0.11635 009817 0.08334 0.07116 0.06109 0.08272 0.04572 0.03982 0.03486 0.03061 0.02699 0.02389 0.02122 0.00758 0.00333 0.00168 0,940 10> 0.564 x 107? 0.359 x 10? 0.239 x 1 0.70754 0.71806 0.72813 0.73777 0.74700 0.75583 0.76829 0.7238 0.78014 0.78757 0.79469 0.80152 0.80806 0.81433 0.82035 0.82612 0.83166 0.83698 0.84209 0.84699 0.85171 0.85624 0.86059 0.86477 0.86880 0.88677 0.90168 oats 0.92462 0.93349 0.94104 0.94751 0.95307 0.95788 0.96207 0,96572 0.96892 0.97174 0.97423, 0.97644 0.98915 0.99414 0.99649 0.99773 0.99845 0.99889 0.99918 303‘304, Ractaton neat warstor From Table 8-1 Faw 9 Fave Fu = 9 Shes = 0.85443 oT* = (5.669 x 10-*92273)" = 1513.3 kWim? ‘Total incident radiation is 0.2 um < A< 3.5 wm = (1.5133 x 10°V0.85443 - 0100.3)" = 116.4KW [3.97 x 10° Bru/h) Total radiation transmitted = (0.5K116.5) = 58.2 kW Radiation _ [(0.34116.4) = 34.92 kW for0 fare measured between a normal to the surface and the line drawn between the area elements r. The projection of dA, on the line between centers is dA, cos dy ‘This may be seen more clearly in the elevation drawing shown in Fig. 8-9. We ‘aysume that the surfaces are diffuse, i.e.. that the intensity of the radiation is the same in all directions, The intensity is the radiation emitted per unit area and per unit of solid angle in a certain specified direction. So. in order to obtain the energy emitted by the element of area dA, in a certain direction, we must ‘multiply the intensity by the projection of dA, in the specified direction. Thus the energy leaving dA, in the direction given by the angle 4 is Fig. 8-9 Elevaton vew of area shown in Fig 88‘308 Raciaton nat ansier In dA, cos dy @ where /, is the blackbody intensity. The radiation arriving at some area element, dA,, at a distance r from A, would be 1 dA, cos 6 Ae ry where dA, is constructed normal to the radius vector. The quantity dA,/P represents the solid angle subtended by the area dA. The intensity may be obtained in terms of the emissive power by integrating expression (b) over a hemisphere enclosing the element of area dA,. In a spherical coordinate system like that in Fig. 8-10, sin g dy dé Then aa, = In dAs ff" sin 6 00s 6 46 dy al, dA, so that Ey = aly (8-20) We may now return to the energy-exchange problem indicated in Fig. 8-8. The area element dA, is given by dA, = cos b: dAr so that the energy leaving dA, which arrives at dA; is, aA dA, Eas £08 dy cs ds ALS dqy-2 ‘That energy leaving dA, and arriving at dA, is dA; dA dqz-1 = Ey €08 62 608 4 ; ly S LI Fig. 8-10 Spherical coordinate system used in derivation of ra- lation shape factor.fagaton shape factor 387 and the net energy exchange is dada: aa 821) das = an ~ Bad ff 008 0s 2054 The integra is ter AsFin of AaFyacoring 1 Ea, 19), To evaluate th integral, the specific geometry of the surfaces Ay and A; must be known. We inca th se omary pba then presen te es TO mor seeped grometes in gaps frm Fo aro he sal ara dA, tothe ft dak A: 85 shown in Fig. 8-11. The element of area dA: is chosen as the circular ring of radius Ths dAy = 2ax de We note that 6; = 62 and apply Eq. (8-21), integrating over the area A> dA Fann = aA [08 6) dads . Making the substitutions r= (Re + 2) and cos dy Performing the integration gives so that Fann (8-22) Pa Fig. 8411 Raciaton tom a smal-area elemento @ dsk1388 Rasiaton heat vansor , 003 002 or o1s02 03 05 1s2 Ratio DX Fig. 6-12 Radiation shape tact for radiation between paral! rectangles Fig. 8-13 Raciation shape factor for radiaion between parallel disks.Racaten she tctor 380 ‘The calculation of shape factors may be extended to more complex geometries, as described in Refs. 3, 5, 24, and 38; 38 gives a very complete catalog of analytical relations and graphs for shape factors. For our purposes we give only the results of a few geometries as shown in Figs. 8-12 through 8-16. Real-Surface Behavior Real surfaces exhibit interesting deviations from the ideal surfaces described in the preceding paragraphs. Real surfaces, for example, are not perfectly diffuse, and hence the intensity of emitted radiation is not constant over all directions. The directional-emittance characteristics of several types of surfaces are shown in Fig. 8-17. These curves illustrate the characteristically different behavior of electric conductors and nonconductors. Conductors emit more energy in a direction having a large azimuth angle. This behavior may be satisfactorily explained with basic electromagnetic wave theory, and is dis. cussed in Ref. 24. As a result of this basic behavior of conductors and non- conductors, we may anticipate the appearance of a sphere which is heated to incandescent temperatures, as shown in Fig. 8-18. An electric conducting sphere will appear bright around the rim since more energy is emitted at large angles 4. A sphere constructed of a nonconducting material will have the opposite behavior and will appear bright in the center and dark around the edge. Fig. 614 Raciation st ron age200 Raciaton het ranster Os 08 06 07 On Fig. 815 _Fadaton shape factors fortwo coneentuc «ytincers of Het Fengtn (a) Outer eyinder to set) outer cylinder to nner evinceRacaion shape lactor 34 02 0304-05 103 3 4 56 810 Ln Fig.8-18 Radiation shape factor fraction between two paral concentnic oisks DF 06 08 OF 0204-08 08 ay Fig. 8-17 Directional emissity of materials 2ocording o Rel. 9: (a) wet ce; (6) ‘W000, (c) gla8s. (} paper, [) cay, (F) copper oxide, (Q) aluminum oxde.‘98 Rachaton neat ranser LS Pg 18 te condone pen Seat ean ope” sac coc » secre nto Reflectance and absorptance of thermal radiation from real surfaces are @ function not only of the surface itself but also of the surroundings. These properties are dependent on the direction and wavelength of the incident ra- diation. But the distribution of the intensity of incident radiation with wave length may be a very complicated function of the temperatures and surface characteristics of all the surfaces which incorporate the surroundings. Let us denote the total incident radiation on a surface per unit time. per unit area, and per unit wavelength as G,. Then the total absorptivity will be given as the ratio of the total energy absorbed to the total energy incident on the surface, [food [rou If we are fortunate enough to have a gray body such that ¢, = € = constant this relation simplifies considerably. It may be shown that KirchofT’s law 1Eq (8-8)] may be written for monochromatic radiation as a 24) or a 82) Therefore, for a gray body. a, = constant, and Eq. (8-23) expresses the result that the total absorptivity is also constant and independent of the wavelength distribution of incident radiation. Furthermore. since the emissivity and ab sorptivity are constant over all wavelengths for a gray body. they must be independent of temperature as well. Unhappily, real surfaces are net alviays “gray” in nature, and significant errors may ensue by assuming gras-body behavior. On the other hand, analysis of radiation exchange using real-surface behavior is so complicated that the ease and simplification of the gray-body assumption is justified by the practical utility it affords. References 10, 11, and 24 present comparisons of heat-transfer calculations based on both gray and nongray analyses.Peistons between shape factors 203 ‘= EXAMPLE 6-2 Heat transfer between black surfaces ‘Two parallel black plates 0.5 by 1.0 m are spaced 0.5 m apart. One plate is maintained at 10H°C and the other at 00°C, What is the net radiant heat exchange between the two plates? souton “The ratios for use with Fig. £12 ae ¥_0S yg X10 Dos boos so that Fi = 0.285, The heat transfer is calculated from Q = AF nlEn ~ En) = AFT ~ Te) = (5.669 10-*N0.540.285)(1273* — 7734) 18.33 KW [62,540 Btw/h] 8.5 RELATIONS BETWEEN SHAPE FACTORS ‘Some useful relations between shape factors may be obtained by considering the system shown in Fig. 8-19. Suppose that the shape factor for radiation from ‘As to the combined area A, is desired. This shape factor must be given very simply as Frags Frit Ba (825) i.e., the total shape tactor is the sum of its parts. We could also write Eq. @ AsFyta = AaFy-1 + ASPs (8-26) Paathathea AsPoiatAaP st Aah Fig. 819 Skeich showing some relations be wa tween shape factors‘304 Racation heat rancor and making use of the reciprocity relations AaFya = AiaFia-s AnPy1 = AVF AsPy-a = AaFr-s the expression could be rewritten AiaFia-s = AiFios + AaFr3 (8-27) which simply states that the total radiation arriving at surface 3 is the sum of the radiations from surfaces | and 2. Suppose we wish to determine the shape factor F;.» for the surfaces in Fig. 8-20 in terms of known shape factors for perpendicular rectangles with a common edge. We may write Fios = Fiat Fa in accordance with Eq, (8-25). Both F, 2, and F,_2 may be determined from Fig. 8-14, so that F_, is easily calculated when the dimensions are known. Now consider the somewhat more complicated situation shown in Fig. 8-21 An expression for the shape factor F_.« is desired in terms of known shape factors for perpendicular rectangles with a common edge. We write AuFiasa = AF + Bane (@) in accordance with Eq. (8-25). Both F,2-14 and F;_..4 can be obtained from Fig. 8-14, and Fy 3.4 may be expressed AlFi-34 = AiFi-3 + AvFi-« ® Also, AuaFia-3 = AiFi-s + AaPe-s 0 Solving for AF» from (c), inserting this in (b), and then inserting the resultant expression for AiFi_s4 in (a) gives AiaFia-sa = AiaPia-a~ AaFaoa + Avia + AnF: (d) Notice that all shape factors except F;.4 may be determined from Fig. 8-14. Fig. 820Relaons between shape factors 308 Thus Fab AiaFia at AaFiey = Aafia = AnFinnd 628) In the foregoing discussion the tacit assumption has been made that the various bodies do not see themselves, i.e., Fu = Fa = Fy =0 To be perfectly general, we must include the possibility of concave curved surfaces, which may then see themselves. The general relation is therefore DF = 10 (8-29) where F, is the fraction of the total energy leaving surface i which arrives at surface j. Thus for a three-surface enclosure we would write Fut Fat Fy = 10 and Fy, represents the fraction of energy leaving surface 1 which strikes itself. A certain amount of care is required in analyzing radiation exchange between curved surfaces. Hamilton and Morgan [5] have presented generalized relations for parallel and perpendicular rectangles in terms of shape factors which may be obtained from Figs. 8-12 and 8-14, The two situations of interest are shown in Figs. 8-22 and 8-23. For the perpendicular rectangles of Fig. 8-22 it can be shown that the following reciprocity relations apply [5]: A\Fiy = AsFur = AvFru = AvFrs 630){396 Radiation hea! vansor Fig. 8-22 Generaizes perpencicularrectan- le arrangement By making use of these reciprocity relations, the radiation shape factor Fix may be expressed by AiFiy = HKo23430° ~ Ka, + Kaser ~ Kuso-wavesey y- Kusser — Karsaso-wser + Kuase-wer + Kasas-wss + Keowee + Kasyasasy + Kase ~ Kons — Kisee ~ Kase ~ Keay + Kod (831) where the K terms are defined by Knon = AnFnn (8-32) Kip = AmE mm (8-33) ---------¥ % Fig. 623 Generaized parale-rectangle ar- rangementFelatons vetwoen shape factors 387 ‘The generalized parallel-rectangle arrangement is depicted in Fig. 8-23. The reciprocity relations which apply to this situation are given in Ref. 5 as AiFiy = AsBa = AgFoy = ArFry (8334) Making use of these relations, it is possible to derive the shape factor F'” as AF = WKazaasaraae ~ Knasecat ~ Kaaasaoe ~ Ki + Kasasy + Kasse ~ Kase — Kiss rer (rer) ~ Kot ~ Kuserase + Kisoree + Kasoe + Kosar ~ Kase + Koel (8-35) The nomenclature for the K terms is the same as given in Eqs, (8-32) and (833), EXAMPLE 8:3. Shape-actor algebra ‘Two concentric cylinders having diameters of 10 and 20 cm have a length of 20 cm. Calculate the shape factor between the open ends of the cylinders. Solunen We use the nomenclature of Fig. 815 for this problem and designate the open ends as surfaces Sand 4. We have Lirs = 2000 = 2.0 and rr. = 0.5; s0 from Fig. 8-15 we obtain Fy = 083 Fy = 033 Using the reciprocity relation [Fq, (8-18)] we have MiB = AaB, and Fre = Addi), = 2010V0.43) = 0.86, For surface 2 we have Fut Fo t Fut nt Fo = RO re 044 Fy + Bet Fu = 10 Sc. iF Fi can be determines, we ean caleulate the desired qu Fos Fut Pu = 0208 Racianon neat vansier and from symmetry Fry = Fi, so that Fu = GX ~ 0.86) = 0.07 Using reciprocity gives AF = AF _ 240020) . Fu = Saar pra 0.07 = 0.187 Then, from Eq. (a) Fy = 1 = 0.187 ~ 0.66 = 0.173 EXAMPLE 8-4. Shape-tactor algebra A truncated cone has top and bottom diameters of 10 and 20 cm and a height of 10 em. Calculate the shape factor between the top surface and the side and also the shape factor between the side and itself. Solution ‘We employ Fig. 8-16 for solution of this problem and take the nomenclature as shown, designating the top as surface 2, the bottom as surface 1, and the side as surface 3. ‘Thus, the desired quantities are F:, and Fy. We have Lir, = 10/10 = 1.0 and rb = SHO = 0.5. Thus, from Fig. 8-16 Fr = 0.12 From reciprocity (Eq. (8-18)) AF: = AF Fay = (20/10N0.12) and Fa=0 sothat Fut Fy = 10 and Fy = 1 ~ 0.28 = 0.76 For surface 3, Fy + Fo + Fu = 10 @ 0 we must find Fy and F,, in order to evaluate F,,. Since Fy, = 0 we have Fy + Fo = 10 and Fy = 1-012 = 0.88 and from reciprocity AF = Fu ‘The surface area of the side is Ay lr, + lle, — ry? + LE = OS + IONS + 105)! = 526.4 em* So, from Eq. (b)Relatons between shape laters 308 (10 Fa = Beg 088 = 0.525 A similar procedure applies with surface 2 so that = 013 Finally, from Eq. (a) Fo = 1 = 0.525 ~ 0.113 = 0.362. EXAMPLE 8-5. Shape-actoralgobra ‘The circular half-eylinder shown has a diameter of 60 cm and a square rod 20 by 20 em placed along the geometric centerline. Both are surrounded by a large enclosure. Find Fis. Fry. and Fy in accordance with the nomenclature in the figure. lot | NA Fig. €x. 05 Soluton From symmetry we have Fa @ In general, Fy, + Fr: + Fy = 1.0. Tovaid in the analysis we create the ficticious surface 4 shown as the dashed line. For this surface, F., = 1.0. Now. all radiation leaving surface | will arrive either at 2orat 3. Likewise, this radiation will arrive atthe imaginery surface 4, so that Fu = Fe t Fo o From reciprocity Aus Adu “The areas are, for unit length, Ay = did = 7 (0.62 = 0.982 Ag = 02 + QMO.0F + 02D? Ay © (4V0.2) = 08400 Raciaton eat ans so that = 0.686 wo so Foe ® = a Combining (b),(c). and (d) gives Fi, = 0.686 ~ 0.425 = 0.261 ( Finally, D Fu = 1 = Fin Fn = 1 - 0425 ~ 0.261 = 0314 ‘This example illustrates how one may make use of clever geometric considerations {0 ‘calculate the radiation shape factors. 8-6 HEAT EXCHANGE BETWEEN NONBLACKBODIES The calculation of the radiation heat transfer between black surfaces is rela- tively easy because all the radiant energy which strikes a surface is absorbed. The main problem is one of determining the geometric shape factor. but once , this is accomplished, the calculation of the heat exchange is very simple. When i nonblackbodies are involved, the situation is much more complex, for all the energy striking a surface will not be absorbed; part will be reflected back to another heat-transfer surface, and part may be reflected out of the system entirely. The problem can become complicated because the radiant enerey can be reflected back and forth between the heat-transfer surfaces several times. The analysis of the problem must take into consideration these multiple re: flections if correct conclusions are to be drawn. We shall assume that all surfaces considered in our analysis are diffuse and uniform in temperature and that the reflective and emissive properties are constant over all the surface. Two new terms may be defined: G = irradiation total radiation incident upon a surface per unit time and per unit area J = radiosity total radiation which leaves u surface per unit time and per unit area Iz addition to the assumptions stated above. we shall also assume that the radiosity and irradiation are uniform over each surface. This assumption is not strictly correct. even for ideal gray diffuse surfaces, but the problems become exceedingly complex when this analytical restriction is not imposed. Sparrow and Cess [10] give a discussion of such problems. The radiosity is the sum ofest exchange between nobiackbodios 404 the energy emitted and the energy reflected when no energy is transmitted, or I= Ey + 0G (836) where € is the emissivity and E;, is the blackbody emissive power. Since the transmissivity is assumed to be zero, the reflectivity may be expressed as p=l-a=l-e so that Jy t= OG (837) The net energy leaving the surface is the difference between the radiosity and the irradiation: -G= y+ - 0G -G Solving for G in terms of J from Eq. (8-37), «A ES aa ) E,-J or Gi preven (8-38) At this point we introduce a very useful interpretation for Eq. (8-38). If the ‘denominator of the right side is considered as the surface resistance to radiation heat transfer, the numerator as a potential difference. and the heat flow as the “current,” then a network element could be drawn as in Fig. 8-24 to represent the physical situation, This is the frst step in the network method of analysis originated by Oppenheim [20] Now consider the exchange of radiant energy by two surfaces A) and A, (Of that total radiation which leaves surface 1, the amount that reaches surface TAF and of that total energy leaving surface 2, the amount that reaches surface 1 is LAFu ‘The net interchange between the two surfaces is gia = JAF ~ SAF, But AiFiz = AF so that Qh-g =U) — JOA 2 = Uy ~ AF or a (839) We may thus construct a network element which represents Eq. (8-39), as4402 Raciaion oat waneer Fig. 8-24 Element representing “surface resistance’ in adiation- network method at Fig. 8-25 Element representing “space resistance” in radiaton-net- aFa work method, shown in Fig. 8-25. The two network elements shown in Figs. 8-24 and 8-25 represent the essentials of the radiation-network method. To construct a net- work for a particular radiation heat-transfer problem we need only connect a “surface resistance” {1 — e)/eA to each surface and a “space resistance” VApFm-» between the radiosity potentials. For example, two surfaces which exchange heat with each other and nothing else would be represented by the network shown in Fig. 8-26. In this case the net heat transfer would be the ‘overall potential difference divided by the sum of the resistances: En ~ Ey Gee — ee + WAFa + = eaeAy _ oll — Ts’) “(1 = ele; + VAF2 + 1 @ileAy A three-body problem is shown in Fig. 8-27. In this case each of the bodies exchanges heat with the other two. The heat exchange between body 1 and body 2 would be (8-40) Anh 03 TAF a Fig. 8-28 Radiation network for two surfaces wich see each other and nothing else. Fig. 8:27 Raciation network for three surtaces which see each other and nothing else.Heat exchange between ronblackbodis 403, lea eg oA naa Fig.8-28 Raciaton network or two surfaces en closed by a third surface which is nonconducting hy but resadiaing and that between body 1 and body 3, hah WAFS To determine the heat flows in a problem of this type. the values of the radi osities must be calculated. This may be accomplished by performing standard methods of analysis used in dc circuit theory. The most convenient method is an application of Kirchhoff"s current law to the circuit, which states that the sum of the currents entering a node is zero. Example 8-5 illustrates the use of, the method for the three-body problem. 'A problem which may be easily solved with the network method is that of, two flat surfaces exchanging heat with one other but connected by a third surface which does not exchange heat, ie., one which is perfectly insulated. This third surface nevertheless influences the heat-transfer process because it absorbs and re-radiates energy to the other two surfaces which exchange heat. ‘The network for this system is shown in Fig. 8-28. Notice that node Jy is not connected to a radiation surface resistance because surface 3 does not exchange energy. Notice also that the values for the space resistances have been written Fy=-F Fr = l= Fx since surface 3 completely surrounds the other two surfaces. For the special case where surfaces 1 and 2 are convex, i-e., they do not see themselves and Fy, ~ Fx: = 0, Fig. 8-28 is a simple series-parallel network which may be solved for the heat flow as Gee © Ay = Ar = 2A, Ar AFL where the reciprocity relation AF = AF. hhas been used to simplify the expression. 11 is 10 be noted again that (8-41) applies only to surfaces which do not see themselves, that is, Fix 0.404, Raciaion neat ranster This network, and others which follow, assume that the only heat exchange ' is by radiation. Conduction and convection are neglected for now. i EXAMPLE 8-6 Hot plates in @ oom i ‘Two parallel plates 0.5 by 1.0 m are spaced 0.5 m apart. One plate is maintained at 100°C and the other at 500°C. The emissivities of the plates are 0.2 and 0.5. respectively. ‘The plates are located in a very large room, the walls of which are maintained at 27°C. ‘The plates exchange heat with each other and with the room. but only the plate surfaces facing each other are to be considered in the analysis. Find the net transfer to each plate and to the room. 17, = 1000re i ae a ( moe ' Room a 27°C Fig. Ex. 8-6 (2) Schematic (b) Network Soluton This is a three-body problem. the two plates and the room, so the radiation network is shown in Fig. 8-27. From the data of the problem Ty = OOC = 273K Ay = A, = 0S mF T= C= mK «= 02 T= C= 30K 4 =05 , Because the area of the room Asis very large, the resistance (I~ e\VevAy may be taken | as zero and we obtain Ey, ~ J. The shape factor was given in Example 8-2: Fay = 0.285 = Fy Fy =~ Fy = 075 Fy 1 Fy = 0715Heat excnange beeen nolackbodies 408, ‘The resistances in the network are calculated as 1-05 -ams7 + wan 72° 1 _1___ jg aaenn 72797 AF ~ @5K0.285) Oso7Is 1 ieee AFy > O510.715) ~ ‘Taking the resistance (I~ «,V/e,A3 as zero, we have the network as shown, To calculate the heat flows at each surface we must determine the radiosities J, and Js. The network is solved by setting the sum of the heat currents entering nodes J, and J; 10 zer0: node Jy @ node co) 20 148.87 kWim? (47,190 Btu/h » £2] 20.241 kWim? (6416 Btu/h - f°) Ey = aT = 0.8592 kWien? (145.6 Buwh - fe] Now Inserting the values of E,,. Es,. and E,, into Eqs. (a) and (b), we have two eq and two unknowns Jy and J; which may be solved simultaneously to give Jy = 33.469 KWim? J, = 15.054 KWim? ‘The total heat last by plate 1 is = 14.425 kW 20.241 — 15.084 ae ap = 2504 kW ‘The total heat received by the room is Anh bah PAF * VAP 33.469 — 0.4592 | 15.054 ~ 0.4592 ~ Bs 04802, 15.054 04522 — 17.020 kW. 158.070 Blu] From an overall-balance standpoint we must have aaata because the net energy lost by both plates must be absorbed by the room.‘400 Ragaton het rancter = EXAMPLE 8-7 Surface in radiant balance Two rectangles 50 by 50 em are placed perpendicularly with a common edge. One surface has T, = 1000 K, ¢, = 0.6, while the other surface is insulated and in radiant balance with a large surrounding room at 300 K. Determine the temperature of the insulated surface and the heat lost by the surface at 1000 K. Solution The radiation network is shown in the accompanying figure where surface 3 is the room and surface 2 is the insulated surface. Note that J, = E,, because the room is large and (1 ~ eVesAs approaches zero. Because surface 2 is insulated it has zero heat transfer and Jz = Ey. Jz “floats” in the network and is determined from the overall radiant balance. From Fig. 8-14 the shape factors are F, Because Fi, = Oand Fy = 0 we have Fat y= 10 and Fy = 1-02 =08 Au = Az = (0.5)? = 0.25 my 02 = Fy ‘The resistances are oO 6x02) L ieee AF a Fa ~ 025N08) oi AFa ~ 02302, 2.667 5.0 200 oO Room st 300 K r= 1000 K ® Fig. Ex. 8-7 (a) Schematic. (b) Networknine paral planes 407 We also have Ex, = (5.669. x 10°*)(1000)* = 5.669 x 10" Wie? Jy = Bay = (5.669 x 10-1300)" ‘The overall circuit isa series-parallel arrangement and the heat transfer is 159.2 Wim? We have 1 Sela era gf CeIOLTS) eee and 56,690 ~ 459.2 @ PET = 8229 KW (28.086 Brw/h) ‘This heat transfer can also be written 4° T= ed, Inserting the values we obtain Jy = 34,745 Wim? ‘The value of J; is determined from proportioning the resistances between J, and J, $0 that and dy = TBI6 = Ey, = oT Finally, we obtain the temperature as ( 7316 599.4 (619°F] When two infinite parallel planes are considered, A, and A: are equal; and the radiation shape factor is unity since all the radiation leaving one plane reaches the other. The network is the same as in Fig. 8-26, and the heat flow per unit area may be obtained from Eq. (8-40) by letting A, = Az and Fi, = 1.0. Thus q_ ott ~T) A We, + Weg - 1 (8-42)‘400 Aclabon heat wander Fig. 8:29 Radiation exchange between two cylindcal surtaces, When two long concentric cylinders as shown in Fig. 8-29 exchange heat we ‘may again apply Eq. (8-40). Rewriting the equation and noting that Fi: = 1.0, gACTS ~ Ts) We, + (AvAsNWe; = 1) vo ‘The area ratio Ay/Az may be replaced by the diameter ratio dy/d: when cylin- drical bodies are concerned. Equation (8-43) is particularly important when applied to the limiting case of a convex object completely enclosed by a very large concave surface. In this instance A,/Ay— 0 and the following simple relation results: q = oA MT ~ Tr) (8-434) This equation is readily applied to calculate the radiation-energy loss from a hot object in a large room. = EXAMPLE 6-8 ‘The 30-cm-diameter hemisphere in the accompanying figure is maintained at a constant temperature of 500°C and insulated on its back side, The surface emissivity is 0.4. The ‘opening exchanges radiant energy with a large enclosure at 30°C. Calculate the net radiant exchange. Encloute arerc Fig. Ex. 08Solution This is an object completely surrounded fy « karge enclosure but the inside surface of the sphere is not convex: ie.. it sees itself, and therefore we are nor permitted t0 use Eq. (43a). In the figure we Lake the inside of the sphere as surface | an the enclosure ‘as surface 2. We also create an imaginary surface 3 covering the opening. We actually hhave a two-surface problem (surfaces | and 2) and therefore may use Eq. 48-40) 10 ccaleulate the heat transfer. Thus, Ey, = aT = (TTA) = 20.241 Wie’ Ey, = oP = (303) = 478 Wim? Ay = 2ar = ym0.15" Laid me Ire 06 6A, ~ OVO Wor Ac sothat Now, at this point we recognize that all of the radiation leaving surface | which will eventually arrive at enclosure 2 will also hit the imaginary surface 3: Le.. Fre = Fy We also recognize that Ay = AE But, Fy = 1.0 s0 that Then WA\Eys = M0.1414V0.5) = 14.14 and we can calculate the heat transfer by inserting the quantities in Eq, (8-40) 20.241 — 478 799 W “Toor + is +0 8.8 RADIATION SHIELDS (One way of reducing radiant heat transfer betwen two particular surfaces is to use materials which are highly reflective. An alternative method is to use ra- diation shields’between the heat-exchange surfaces. These shields do not deliver or remove any heat from the overall system: they only place another resistance in the heat-flow path so that the overall heat transfer is retarded. Consider the two parallel infinite planes shown in Fig. 8-30a. We have shown that the heat exchange between these surfaces may be calculated with Eq. (8-42). Now consider the same two planes, but with a radiation shield placed between them. as in Fig. 8300. The heat transfer will be calculated for this latter case and ‘compared with the heat transfer without the shield,Ragaton naa! ansle wa ws iene 132 Fig. 8:30 Radiation between parallel infinite we rm planes with and without a radiation shila Since the shield does not deliver or remove heat from the system, the heat transfer between plate 1 and the shield must be precisely the same as that between the shield and plate 2, and this is the overall heat transfer. Thus (¢ 2 (4) a Al} is AJy-2 A q_ oT ~T o(Tst - Ts) (8-44) A” We + Wet Vet Ve—t ‘The only unknown in Eq. (8-44) is the temperature of the shield 7;. Once this temperature is obtained, the heat transfer is easily calculated. If the emissivities of all three surfaces are equal, that is, €) = €: = 6, we obtain the simple relation Ts = ATs + T) (8-45) and the heat transfer is q _ dott = 74) A” Ve + We But since ¢; = €2, we observe that this heat flow is just one-half of that which ‘would be experienced if there were no shield present. The radiation network corresponding to the situation in Fig. 8-30b is given in Fig. 8-31 Multiple-radiation-shield problems may be treated in the same manner as that outlined above. When the emissivities of all surfaces are different, the ‘overall heat transfer may be calculated most easily by using a series radiation network with the appropriate number of elements, similar to the one in Fig. Fig. 831 Rachation network fortwo parallel planes separated by one ca dation shield831. If the emissivities of all surfaces are equal, a rather simple relation may be derived for the heat transfer when the surfaces may be considered as infinite parallel planes. Let the number of shields be n. Considering the radiation network for the system, all the “surface resistances"” would be the same since the emissivities are equal. There would be two of these resistances for each shield and one for each heat-transfer surface. There would be n + 1 “space resistances,” and these would all be unity since the radiation shape factors are unity for the infinite parallel planes. The total resistance in the network would thus be R(n shields) = + ins WD) The resistance when no shield is present is 1 R (no shield) = © + 5 ~ -1 We note that the resistance with the shields in place is n + 1 times as large as when the shields are absent. Thus (Sas, = arr (Ame we if the temperatures of the heat-transfer surfaces are maintained the same in both cases. The radiation-network method may also be applied to shield prob- lems involving cylindrical systems. In these cases the proper area relations must be used Notice that the analyses above, dealing with infinite parallel planes, have been carried out on a per-unit-area basis because all areas are the same. EXAMPLE 8-9. Heat-ransfer reduction with shield Two very large parallel planes with emissivites 0.3 and 0.8 exchange heat. Find the percentage reduction in heat transfer when a polished-aluminum radiation shield (€ (0,04) is placed between them. ‘soton The heat transfer without the shield is given by a _ ott = TH) Ao We +e -T = 0.27907 ~ T:4) ‘The radiation network for the problem with the shield in place is shown in Fig. 8-31 The resistances are4442 Raciton nes raster ‘The total resistance is 2.333 + 2)24.0) + (MH) + 0.25 = 52.583 and the heat transfer is 4.001) “ $= AEE ~ pors0204r - 754 so that the heat transfer is reduced by 93.2 percent EXAMPLE 8-10 Open cylindrical shield ‘The two concentric cylinders of Example 8-3 have T, = 1000 K, ¢, = 08, & = 0.2 and are located in a large room at 300 K. The outer cylinder is in radiant balance. Calculate the temperature of the outer cylinder and the total heat lost by the inner cylinder. Solution ‘The network for this problem is shown in the accompanying figure. The room is des- ignated as surface 3 and J, = E,,, because the room is very large. In this problem we ‘must consider the inside and outside of surface 2 and thus have subscripts i and o to designate the respective quantities. The shape factors can be obtained from Example 83s Fi, = 0.86 Fis = (20.07) = 0.14 Fay, = QNO.12) = 0.24 Fay = 1.0 Also, Ay = (0,110.2) = 0.06283 mm? Az = (0,200.2) = 0.12566 m* Ey = (5.669 10-*)(1000}* = 5.669 x 10* Wim? Ey, = (5.669 x 10°*)(300)" = 459.2 Wim? (Gas taaston 419 and the resistances may be calculated as 3183 B16 = 137 AF The network could be solved as a series-parallel circuit to obtain the heat transfer but we will need the radiosities anyway so we set up three nodal equations to solve for J, Jay and J,,, We Sum the currents into each node and set them equal to zero: Fa=h Eni Junk coe yam So? 7 esto ade Iu ad: node. nate, nah 2g ae ate * aren ~ " node Jo at + eats = 0 7.958” (2N31.83) ‘These equations have the solution Jy 50,148 Win 4 Jo = 498 W) ‘The heat transfer is then calculated from En Se $6,690 ~ 50.148 Oe eed 3979) 27,811 Wim? 4 1644 W_ [S611 Btu/h) From the network we see that + doy BN + 3 pb Si + 3498 = 15,685 Wim? z and sassy Tes (Sera) 79K SFI If the outer cylinder had not been in place acting as a “shield” the heat loss from cylinder | could have been calculated from Eq. (8-434) as q = GA(Es, ~ En) = (0,8)0,06283)56,690 ~ 459.2) = 2856 W [9645 Bru/h) 8-9 GAS RADIATION Radiation exchange between a gas and a heat-transfer surface is considerably more complex than the situations described in the preceding sections. Unlike1444 Raciabon heal vanstr most solid bodies, gases are in many cases transparent to radiation. When they absorb and emit radiation, they usually do so only in certain narrow wavelength bands. Some gases, such as Nz, Oz, and others of nonpolar symmetrical mo- lecular structure, are essentially transparent at low temperatures, while CO>, H,0, and various hydrocarbon gases radiate to an appreciable extent. ‘The absorption of radiation in gas layers may be described analytically the following way, considering the system shown in Fig. 8-32. A monochromatic ‘beam of radiation having an intensity /, impinges on the gas layer of thickness dx. The decrease in intensity resulting from absorption in the layers is assumed to be proportional to the thickness of the layer and the intensity of radiation at that point. Thus dl, = ahs dx 47) where the proportionality constant a, is ealled the monochromatic absorption coefficient. Integrating this equation gives 1s dy [ue Ine Ty or by (8-48) Equation (8-48) is called Beer's law and represents the familiar exponential- decay formula experienced in many types of radiation analyses dealing with absorption. In accordance with our definitions in Sec. 8-3, the monochromatic ‘transmissivity will be given as nae 8-49) If the gas is nonreflecting, then nate and a= l- eo (850) ‘As mentioned above, gases frequently absorb only in narrow wavelength bands, as indicated for water vapor in Fig. 8-33. These curves also indicate the effect of thickness of the gas layer on monochromatic absorptivity. The calculation of gas-radiation properties is quite complicated, and Refs. he Ine Lake x0 Fig. 832 Absorption in a gas layerspundsaico Yow w9 062 1d ¥ 2€ 1OAe| Jo SS0UKOHN OH soul wo 2 Kunordde as23 to 25 should be consulted for detailed information. For engineering caleu- ' lations Hotel [23] has presented a simplified procedure which may be used to calculate the emittances of water vapor and carbon dioxide gases. Methods for evaluating the radiant exchange between these gases and enclosures are also available, Mean Beam Length Equations (8-48) and (8-50) describe the intensity variation and absorptivity for a gas layer of thickness x. These are the values we might expect to measure in a laboratory experiment with radiation passing straight through the layer. If we imagine a practical problem of a gas contained between two large parallel plates which emit radiation diffusely, we see that the radiant energy transmitted through the gas travels many distances; the energy transmitted normal to the surface travels a distance equal to the plane spacing: energy emitted at shallow angles is absorbed in the gas over a much, longer distance; and so on. By @ 5); ae oe p 3001000 1S00 2000 2500-3000" 3500 40004500 S000 7? 6 soo ooo 150000380000 Fig. 6-34 Emissivity or carbon doxide accoraing to Ret 22, or tla: pressure alm: 10 kN = 0.09238 atmGas adaton 447 careful correlation of several sources of experimental data Hottel and Egbert [29] were able to present the gas emittances for carbon dioxide and water vapor fas shown in Figs. 8-34 and 8-35. In these figures L, is a characteristic dimension Of the system called the mean beam length. Tabulations of these lengths are presented in Table 8-2 according to Hottel (22] and Eckert and Drake (25]. In the absence of mean-beam-length information for a specific geometry, a sat- isfactory approximation can be obtained from v = 365 : rl B51) where V is the total volume of the gas and A is the total surface area. In Figs. 8-34 and 8-35 the total pressure of the mixture is | atm, and p, and p.. represent the partial pressures of carbon dioxide and water vapor, respectively. For total own 9.019 oo Eo HOO TD 3000 00 ‘Able Temperature. 00085005000, rh sto ae soo 200 250 3000 Fig. £35. Emissiuly for water vapor according to Ref, 22 for tla! pressure = 1 aim SOkNim = 0103238 attTable 82 Mean Equvaient Length L, for Radiation fram Ente Gas Volume, Aecorcing to Fes. 22 ans 25, eS (Characteristic Gas volume dimension Le Volume between two infinite planes Separation distance L 1.8L Circular cylinder with the height Diameter D om diameter, radiation to cent base of Hemisphere, radiation to element in R center of base ‘Sphere, radiation to entre surface 0.6sD Infinite circular cylinder, radiation 0.950 to convex bounding surface Circular eylinder with height Diameter D 0.00 diameter, radiation to entire surface Circular eylinder. semi-infinite Diameter D 0.4sD height, radiation to entire base Cube, radiation to any face Bulge L 0.601 Volume surrounding infinite tube Tube diameter D, bundle, radiation to a single tube distance between tube centers $ Equilaterabtriangle arrangement: Y=. 3.05 = D) S=30 3.85 - D) ‘Squate arrangement 355 — D) Pressures other than 1 atm, correction factors are provided in Figs. 8-36 and 8.37. When both carbon dioxide and water vapor are present, an additional correction Ae from Fig. 8-38 must be subtracted from the total ofthe emittances of the two components; thus the total gas emittance ¢, of the mixture is ex. pressed as dos eeso, e063 EEO LO Seas 1 9k Fig: 838 Correction taco or CO, emssvty according o Ret 22Gasredaton 419 Fig. 637 Correction factor for H:O emissivity according to Ret 22 Heat Exchange between Gas Volume and Black Enclosure Now consider a gas volume at uniform temperature 7, enclosed by a black surface at temperature T,. Because of the band structure of the gas the ab- sorption of energy emitted by the wall at 7. will differ from the energy emitted by the gas at T,. The net heat transfer from the gas to the enclosure is therefore = energy emitted by gas ~ energy from enclosure absorbed by gus ed ToT ~ aT oT 4 (8-53) where «,(7,) is the gas emittance at T, which is evaluated as discussed above and ,(T,) is the gas absorptance for the radiation from the black enclosure at 7, and is a function of both 7.. und T,. For a mixture of carbon dioxide and water vapor an empirical relation for a, is 002 04 06 08 100 0204 0608100 02 04 6 08 10 Fig. 8-38 Correction factor when CO; and H,0 are present in an enclosure ‘according to Ret 2210 atmt = 30.881 kNim420 acaton nest vaster af.) = a + ay ~ ba (834 ns where a= ca: (B) (635) [nye au = Cath () (56) a= de at «57 The values of e: and ei, are evaluated from Figs. 8-34 and 8-35 with an abscissa of T. but with pressure-beam-length parameters of peL«(T./T,) and PuL<(T./T,). respectively. Equation (8-53) was written for the heat exchange between a gas volume and a black enclosure at a uniform temperature. Now let us consider a more complicated case in which the gas volume is enclosed by two black parallel planes at different temperatures 7; and 7;. In this case we must make an energy balance on each plate. For plate 1 the net energy gain is, = EeAy (8-58) Similarly, for plate 2 42 = Graz ~ EysAz 6-59) ‘The irradiations G, and G; must be evaluated in terms of the total energy arriving at each surface. For surface | Irradiation on surface 1 = irradiation from gas + irradiation from surface 2 transmitted through gas or GAs = AF peAT Es, + AvFutdT En (8-60) ‘The transmissivity for the radiation from 7; is obtained from tT) = 1 ~ a(t) where the a, is evaluated as before. Ina similar manner, the irradiation on surface 2 would be Gar = AP p(T PEs, + AFat (TED, sn If we take the gas areas as A, = Ay = A2 for the parallel-plate system, then F,, = Fa, = 1.0. Also, Fin = Fi = 1.0. Once G; and G2 are determined, the heat transfer to each surface can be calculated from Eqs. (8-58) and (8-59), Heat Exchange between Gas Volume and Gray Enclosure Up to this point, the gas-radiation calculation methods discussed were con- cerned only with black surfaces exchanging heat with the gas. In many engi- neering applications the enclosure walls are dirty and sooty, the wall emissivity is very high, and the heat-transfer calculation by Eq. (8-53) may be a reasonableapproximation, An analysis of gray enclosures is very complex because the multiple reflection phenomena must take into account the band-absorption char- acteristics of the gas. For engineering calculations Hott! (22) has show that the net heat transfer to a gray enclosure having an emissivity ¢, may be handled approximately by multiplying Eq, (83) by a factor such that 1 2a Ft fore, > 08 (6-62) Gua 2 For values of ¢, <0.8 much more elaborate procedures must be employed to calculate the heat transfer EXAMPLE 6-11 Gas radiation in a ace A cubical furnace 0.5 m on a side has interior walls that are essentially black. Inside the furnace the gas is 20 percent carbon dioxide by volume and 80 percent nitrogen at ‘a total pressure of 1 atm and a temperature of 1500 K. The walls of the furnace are 10 be maintained at 300°C. Calculate the amount of cooling required to maintain the walls at the specified temperature. ‘sowution For this calculation we make use of Eq. (853) Lo eter - a, ‘ a 4m eT pots ~ afT oT. @ where a7;¢ = (5.669 x 10-*)(1500)* = 287 kWim? {90,980 Btu/h - fe] OT. = (5.669 10-*Y573)* = 6.11 kWin? (1937 Blush - FE) From Table 8-2 the equivalent beam length is L, = (0.6)0.5) = 0.3 m (0.984 ft} ‘and the partial pressure of CO; is 0.2 atm, or p. = (0.2M1.0132 * 107) = 20.26 kPa {2.94 Ibvin® abs} 6.08 kNim, and we enter Fig. 8-34 at 1500 K to obtain © = 0072 = 6(T) ‘There is no correction factor because the total pressure is | atm. To evaluate a,(T.) We use a temperature of 573 K and a pressure-beam-length parameter of ; Te ggg 33. Then p.Le Pelee T 1500 32 kNim From F834 ¢ = 01,6, ~ Lande E80 tain out = a = 0s (32°) «on ‘ (0.072287) ~ (0.1216.11 19.92 kWim?(422, Raciaton heat anster ‘There are six interior sides, and so the total heat transfer is, q = (19-92K1.5) = 29.88 KW (102,000 Btu/h} EXAMPLE 8:12 ‘Two parallel black plates are separated by a distance of 0.7 m and maintained at tem- Peratures of 200 and 500°C. Between the planes is a gas mixture of 20 percent COs, 15 percent water vapor, and 65 percent N; by volume at a total pressure of 3 atm. The {gs temperature is 1000°C. Calculate the heat exchange with each plate. Solution Fox. this problem we must make use of Eqs. (88) and (8-59) which require evaluation of a number of properties. Setting T, = 1000" = 1273 K we have BK Ty = SOC = 773K 148.9 kWim? 2.84 kWim? OT = Ey, = 20.24 Win? Using Table 8-2, we calculate the mean beam length as L, = (1.80.7) = 1.26m For the mixture at 3 atm the partial pressures of CO, and HO are Pe = (0.2013K1.0132 X 10?) = 60.8 kPa (8.82 Ibin® abs) Pa = (0.15)3X1.0132 x 10) = 45.6 kPa (6.61 Ibvin® abs} 1 * (60.8N1.26) = 76.6 kNim = 2.48 atm «ft PL. + Pole = 134.1 KNim = 4.35 atm - ft Also, Gp + px) = (0.53 + 0.45) = 1.725 atm, and aK ‘We must now determine the values of a, at Ty and Ts. AUT = 7, Sy 7 28S KNim 1 vt. (FP) anGas reciaton 493 From the calculation charts © =02% CH1s 2013 Com LIS de = da = 04 Now, from Eqs. (8-55) and (8-56) 1273\° o.= ye. (22) = oa «= 1.9029 (22) = 0 and. fT) = a, + a, ~ Aa = 0.285 + 0.608 ~ 0.02 1AT) = 1 ~ a) = 0.126 0.874 AUT = T, = 73K, Bo 1667 m Plat = 76.6 mm 46.5 kN/m = 1.51 atm - ft Bo 57572. oun = tsa: poke T= 375 7B 349 eNlm = 13 a From the calculation charts 0m Cm 145 017 C= 113 be = da = 0.028 ciasi0.1n (BR) ” = 0.266 my a. = 0.459020 (222) = 90 ‘Then afT) = a. + a, — Sa = 0.266 + 0.436 — 0.028 = 0.674 and ATs) = 1 ~ a,(T;) = 0.326 For the parallel-plate system all the areas are equal and all the shape factors are unity, so Eq. (8-60) becomes G.= elTOEs, + TATDEn (0.44)(148.9) + (0.326N20.24) = 72.1 kWim* [22,860 Btu/h + f°} Similarly, Gs = eo, + ATIEm = (0.44)(148.9) + (0.126)2.84) = 65.9 kWim? Both surfaces are black so the heat gain by each surface is « 4 @ = Ey = TA ~ 2.84 = 69.3 kWim? 65.9 ~ 20.24 = 45.7 kWim? - Es >4424 Radiator heal ansier ‘The net energy lost by the gas is the sum of these two figures, or 115 kWim! of plate area. 8-10 RADIATION NETWORK FOR AN ABSORBING AND TRANSMITTING MEDIUM ‘The foregoing discussions have shown the methods that may be used to cal- culate radiation heat transfer between surfaces separated by a completely trans- parent medium. The radiation-network method is used to great advantage in these types of problems. Many practical problems involve radiation heat transfer through a medium which is both absorbing and transmitting. The various glass substances are one ‘example of this type of medium; gases are another. We have already seen some of the complications which arise with gas radiation. We shall now examine a radiation-network method for analyzing absorbing-transmitting systems, keep- ing in mind the many problems which may be involved with gases. ‘To begin, let us consider a simple case, that of two nontransmitting surfaces which see each other and nothing else. In addition, we let the space between these surfaces be occupied by a transmitting and absorbing medium. The prac- tical problem might be that of two large planes separated by either an absorbing. ‘gas or a transparent sheet of glass or plastic. The situation is shown schemat- ically in Fig. 8-39. The transparent medium is designated by the subscript m. We make the assumption that the medium is nonreflecting and that Kirchoff's identity applies, so that Gn + Tm Sn + Tm (8-63) ‘The assumption that the medium is nonreflecting is a valid one when gases are considered. For glass or plastic plates this is not necessarily true, and reflectivities of the order of 0.1 are common for many glass substances. In addition, the transmissive properties of glasses are usually limited to a narrow wavelength band between about 0.2 and 4 wm. Thus the analysis which follows is highly idealized and serves mainly to furnish a starting point for the solution of problems in which transmission of radiation must be considered. Other complications with gases are mentioned later in the discussion. When both Fig. 8-99 Rlacation system consisting of atransmiting 7 ” 2 medium between two planesRadiation network or an absorbing and vanemiting medium 428 reflection and transmission must be taken into account, the analysis techniques discussed in Sec. 8-12 must be employed. Returning to the analysis, we note that the medium can emit and transmit radiation from one surface.to the other. Our task is to determine the network elements to use in describing these two types of exchange processes. The transmitted energy may be analyzed as follows. The energy leaving surface 1 which is transmitted through the medium and arrives at surface 2 is TAF tm and that which leaves surface 2 and arrives at surface 1 is SAF itm ‘The net exchange in the transmission process is therefore 41 -tomenes = AvFrstalh ~ 59) = AxFlds ~ 4) Las 4 raenet — TALE ML — @) oe and the network element which may be used to describe this process is shown in Fig. 8-40, Now consider the exchange process between surface | and the transmitting ‘medium. Since we have assumed that this medium is nonreflecting, the energy leaving the medium (other than the transmitted energy, which we have already considered) is precisely that enerey which is emitted by the medium Jn = nb on ‘And of the energy leaving the medium, the amount which reaches surface | is AnE mim = AnE mi€mE om Of that energy which leaves surface 1, the quantity which reaches the trans- parent medium is JAF inti = SAF nbn At this point we note that absorption in the medium means that the incident radiation has “reached” the medium. Consistent with the above relations, the net energy exchange between the medium and surface 1 is the difference be- ‘tween the amount emitted by the medium toward surface 1 and that absorbed which emanated from surface 1. Thus AnFmitmE om ~ JAF iném om tae AFa(i eg) Fig. 840. Network element for ransmited radiation trough medium:428 Racin neat vansior Using the reciprocity relation AiFim = AmFm we have n= tne (8-65) This heat-exchange process is represented by the network element shown in Fig. 8-41, The total network for the physical situation of Fig. 8-39 is shown in Fig. 8.42. If the transport medium is maintained at some fixed temperature, then the potential Eun is fixed according to Em = OT! On the other hand, if no net energy is delivered to the medium, then Eom becomes a floating node, and its potential is determined by the other network elements. In reality, the radiation shape factors Fi2, Fim, and Fy» are unity for this example, so that the expression for the heat flow could be simplified to some extent; however, these shape factors are included in the network resis- tances for the sake of generality in the analysis. ‘When the practical problem of heat exchange between gray surfaces through aan absorbing gas is encountered, the major difficulty is that of determining the transmissivity and emissivity of the gas. These properties are functions not only of the temperature of the gas, but also of the thickness of the gas layer; ice., thin gas ayers transmit more radiation than thick layers. The usual practical problem almost always involves more than two heat-transfer surfaces, as in the simple example given above. As a result, the transmissivities between the various heat-transfer surfaces can be quite different, depending on their geo- metric orientation. Since the temperature of the gas will vary, the transmissive and emissive properties will vary with their location in the gas. One way of handling this situation is to divide the gas body into layers and set up a radiation network accordingly, letting the potentials of the various nodes “float,” and thus arriving at the gas-temperature distribution. Even with this procedure, an iterative method must eventually be employed because the radiation properties Of the gas are functions of the unknown “floating potentials.”” Naturally, if the temperature of the gas is uniform, the solution is much easier. We shall not present the solution of a complex-gas-radiation problem since the tedious effort required for such a solution is beyond the scope of our present discussion; however, itis worthwhile to analyze a two-layer transmitting system in order to indicate the general scheme of reasoning which might be applied to more complex problems. Fig. 641 Network element for racalion exchange between medium and surtaceRadiation networ or an absorbing an ranemiting medium 427 Fig. 842 Tota radiation network for systom of Fig. 839, Consider the physical situation shown in Fig. 8-43. Two radiating and ab- sorbing surfaces are separated by two layers of transmitting and absorbing media. These two layers might represent two sheets of transparent media, such as glass, or they might represent the division of a separating gas into two parts for purposes of analysis, We designate the two transmitting and absorbing layers with the subscripts m and n. The energy exchange between surface | and m is given by di = Eom Qi-m = AF imtmd1 — Amb ni mE bmn VA Finn (8-66) and that between surface 2 and 1 is G2-n = A\Fantnd2 ~ AnFnr€nE nn = (8-67) Of that energy leaving surface 1, the amount arriving at surface 2 is qin = AiFititmta = AvProdi(l = Emil = €9) and of that energy leaving surface 2, the amount arriving at surface 1 is, 2-1 = AF datatm = ArFida(l — el — €m) so that the net energy exchange by transmission between surfaces 1 and 2 is Ach = ~ Xl = eh - J) = I 41-aamues = ArFn(l ~ El = =D) = TERT eT Se (8-68) ' 1 mY tl Tet Hob “ ta fel ld 1d nl MoL ORY Fig. 8-43. Raciation system consisting ot wo trans " er iting ayers between two panes.4428 Radiation heat ranster 4 4 ow 8 AFAO=IT=5) Fig. 644 Network element for transmit radiation between planes 4 by ows 5 1 Fig. 6-45 Network element for transmitted radiation for medium 0 DRG ale toplane + bn ew pea ae Fig. 8-46 Network eloment for radiation exchange between two AsFosints transparent layers. Fig. 847 Tota radiation network for system of Fig, 6-43. and the network element representing this transmission is shown in Fig. 8-44. Of that energy leaving surface 1, the amount which is absorbed in n is Qn = AiFicl nbn = AiFindi(l ~ €mEy Also, Gn = AnFnidnton = AnFni€nEba(l ~ &n) since In = Ese The net exchange between surface | and mis therefore Ji = Eon TAF nll — ene. no = ALF = ene, = (8-69)Ragalon nework or a absorbeg and raremiteg meaum 438 and the network element representing this situation is shown in Fig. 8-45. In like manner, the net exchange between surface 2 and mm is hak 4: me = aged a (8-70) Of that radiation leaving m, the amount absorbed in nis, Ginn = SAmFnnQln = AE nsen€ubin and dan = AnFrnb lr Ene so that the net energy exchange between m and n is, Greer = AE nnn Ens ~ Eon) (8-71) and the network element representing this energy transfer is given in Fig, 846, The final network fur the entire heat-transfer process is shown in Fig. 8-47, with the surface resistances added. If the two transmitting layers m and 1 are inaintained at given temperatures. the solution to the network is relatively easy to obtain because only two unknown potentials J, and J: need be determined to establish tive various heat-flow quantities. In this case the two transmitting layers will either absorb or lose a certain quantity of energy. depending on the temperature at which they are maintained When no net energy is delivered to the transmitting layers, nodes E,,,. and Ey, tnust be left “floating” in the analysis: and for this particular system four nodal equations would be required for a solution of the problem, EXAMPLE 8-13. Network for g28 radiation ‘Two large parallel planes are at T, = 800K, €, = 03, T: = 400 K. ¢: = 0.7 and are separated by a gray gas having ¢, ~ 0.2, 1, = 0.8, Calculate the heat-iransfer rate between the two planes and the temperature of the gas using a radiation network. Compare with the heat transfer without presence of the gas. Solution The network shown in Fig. 8-42 applies to this problem. All the shape factors are unity for large planes and the various resistors can be computed on a unit-area basis as Lae 07 «03 2.333 0.4286 23.220 Wim’490 Aavation neat vansir ‘The equivalent resistance of the center “triangle” is L Tas + 50 + 50) ‘The total heat transfer is then R Liaw - = 2820 = 515646 Wigs A RO + Lint + 04286 ~ 5616 W ir here were no gas present the heat transfer would be given by Eq. (42) 23200 - st| ; = FAO US rat win The radiosites may be computed from dee, y(t = sero Wine 4 outs ~ 40 (FE) ) = ssw Which gives Jy = 10.096 Wim? and J. mean of these values 3858 Wim?. For the network Ey, is just the By so that the temperature of the gas is T, = S92.3K 4(10.096 + 3858) = 6977 = oT,* 8-11 RADIATION EXCHANGE WITH SPECULAR SURFACES All the preceding discussions have considered radiation exchange between diffuse surfaces. In fact, the radiation shape factors defined by Eq, (8-21) hold only for diffuse radiation because the radiation was assumed to have no pre- ferred direction in the derivation of this relation. In this section we extend the analysis to take into account some simple geometries containing surfaces that may have a specular type of reflection, No real surface is completely diffuse or completely specular. We shall assume, however, that all the surfaces to be considered emit radiation diffusely but that they may reflect radiation partly in ‘a specular manner and partly in a diffuse manner. We therefore take the re- flectivity to be the sum of a specular component and a diffuse component: p= p+ Bo 8-72) It is still assumed that Kirchhoff’s identity applies so that e=a=1-p 8-73) The net heat lost by a surface is the difference between the energy emitted and absorbed q = A(eE, ~ aG) (8-74)Ragaien excrange wir specuirsurtaces 494 We define the diffuse radiosity J» as the total diffiuse energy leaving the surface per unit area and per unit time, or dy = En + poG (8-75) Solving for the irradiation G from Eq, (8-75) and inverting in Eq. (8-74) gives a a= Se ke + pod ~ Jol Po or, written in a different form, = Jpill = p.) polleAUl — p,)) or a= where I — p, has been substituted for € + pp. It is easy to see that Eq. (8-76) may be represented with the network element shown in Fig. 8-48, A quick inspection will show that this network element reduces to that in Fig. 8-24 for the case of a surface which reflects in only a diffuse manner, i.e... for p, = 0. ‘Now let us compute the radiation exchange between two specular-diffuse surfaces. For the moment, we assume that the surfaces are oriented as shown in Fig. 8-49. In this arrangement any diffuse radiation leaving surface 1 which is specularly reflected by 2 will not be reflected directly back to 1. This is an important point, for in eliminating such reflections we are considering only the direct diffuse exchange between the two surfaces. In subsequent paragraphs we shall show how the specular reflections must be analyzed. For the surfaces in Fig. 8-49 the diffuse exchanges are given by diaz = JiwAiFall ~ px) (8-77) dea = JxpA2Fn( = py) (8-78) Equation (8-77) expresses the diffuse radiation leaving ! which arrives at 2 and Which may contribute to a diffuse radiosity of surface 2. The factor 1 ~ p, represents the fraction absorbed plus the fraction reflected diffusely. The in- ‘clusion of this factor is most important because we are considering only diffuse direct exchange. and thus must leave out the specular-reflection contribution Fig. 8-49Fig. 850 Nonwork element representing Eq (8-79) for now. The net exchange is given by the difference between Eqs, (8-77) and (8-78), according to Ref. 21 Juohl_= pu) = avi = ps0) WA — pial — pal * (8-79) The network element representing Eq. (8-79) is shown in Fig. 8-50. ‘To analyze specular reflections we utilize a technique presented in Refs. 12 and 13. Consider the enclosure with four long surfaces shown in Fig. 8-S1. Surfaces 1, 2, and 4 reflect diffusely, while surface 3 has both a specular and a diffuse component of reflection. The dashed lines represent mirror images of the surfaces 1. 2, and 4 in surface 3. (A specular reflection produces a mirror image.) The nomenclature 23) designates the mirror image of surface 2 in mirror 3. Now consider the radiation leaving 2 which arrives at 1. There is direct diffuse radiation of (Geailitte = nak, (8-80) Part of the diffuse radiation from 2 is specularly reflected in 3 and strikes 1 This specularly reflecied radiation acts like diffuse energy coming from the image surface 2(3). Thus we may write (ay MERE = S2AzoFaoPe 81) ‘The radiation shape factor Fy. is the one between surface 2(3) and surface 1 The reflectivity py, is inverted because only this fraction of the radiation gets to 1. Of course, Ay = Az. We now have = ALF + pFaon) (8-82) unilar reasoning leads to G12 = DAF + pxBron) (883) 1S Seen cn ecg Fig. 851. systensaith one spec uiaraiuse surfaceRagaton exchange wih specular sutaces 43 Fig, 52 Network element for Eq. (8-84). Combining Eqs. (8-82) and (8-83) and making use of the reciprocity relation AIF: = AaFx: gives dah TA: > Foal a ae ‘The network element represented by Eq. (8-84) is shown in Fig, 8-52. Analogous network elements may be developed for radiation between the other surfaces in Fig. 8-51, so that the final complete network becomes as shown in Fig, 8-53. It is to be noted that the elements connecting to Jyp are simple modifications of the one shown in Fig. 8-50 since pis = Pr = Pas = 0. ‘An interesting observation can be made about this network for the case where pi = 0. In this instance surface 3 is completely specular and Jan = Ein so that we are left with only three unknowns, J;, Js, and Js, when surface 3 is completely specular-reflecting. ‘Now let us complicate the problem a step further by letting the enclosure have two specular-diffuse surfaces, as shown in Fig. 8-54. In this case multiple images may be formed as shown. Surface 1(3, 2) represents the image of | after it is viewed first through 3 and then through 2. In other words, it is the image of surface 1(3) in mirror 2. At the same location is surface 1(2, 3), which is the image of surface 1(2) in mirror 3. Fig. 853 Complete raciaton network for system in Fig B57434 Reaton heat taster 23 Specnlwr-ait Fig. 8-54 Systom wih two specular. fuse surfaces This problem is complicated because multiple specular reflections must be considered, Consider the exchange between surfaces | and 4, Diffuse energy leaving 1 can arrive at 4 in five possible ways: direct: Aa reflection in2 only. JAF vad. reflection in3 only. JiAsFuebss reflection first in 2 and then in 3: (8-85) LAvooboF ics reflection first in 3 and then in 2: DApupnFaan The last shape factor Fi,24, is zero because surface 1(3, 2) cannot see surface 4 when looking through mirror 2. On the other hand, Fiz. is not zero because surface 1(2, 3) can see surface 4 when looking through mirror 3. The sum of. the above terms is given as Gros = SAlFua + pxFian + puFian + PubuF vcs) —— (8-86) Ina similar manner, Gant = WAKEa, + PrsFuan + PrFaon + PuPaFas2y) (8-87) Subtracting these two equations and applying the usual reciprocity relations gives the network element shown in Fig. 8-55 Now consider the diffuse exchange between surfaces | and 3, Of the energy leaving 1, the amount which contributes to the diffuse radiosity of surface 3 is Qs = LAF = pu) + SArpa Fan ~ 0) (8-88) The first term represents the direct exchange, and the second term represents the exchange after one specular reflection in mirror 2. As before, the factor Fig. 6-55 Network element represent ing exchange between surtaces 1 and 4 of Fig. 854fadiaton exchange win specular stares 498 Fig. 6-56 Network element representing excnang tween surlaces 1 and ol Fig. 8-54 1 — py is included to leave out of consideration the specular reflection from 3. This reflection, of course, is taken into account in other terms, The diffuse energy going from 3 to 1 is a) = IAF + SvAwpn Fon (8-89) The first term is the direct radiation, and the second term is that which is specularly reflected in mirror 2. Combining Eqs. (8:88) and (8-89) gives the network element shown in Fig. 8-56 The above two elements are typical for the enclosure of Fig. 8-54 and the other elements may be drawn by analogy. Thus the final complete network is given in Fig. 8-57, If both surfaces 2 and 3 are pure specular reflectors, that is, Pan = pw = 0 wehave Jan = GE So = 6.E ns and the network involves only two unknowns, J; and J., under these circum: stances. ‘We could complicate the calculation further by installing the specular sur- faces opposite each other. In this case there would be an infinite number of images. and a series solution would have to be obtained: however, the series BaF? PPP ia Je by CORR Ten GO ad Fig. 857 Complete craton network for system in Fig. 54for such problems usually converge rather rapidly. The reader should consult Ref. 13 for further information on this aspect of radiation exchange between, specular surfaces, 8-12 RADIATION EXCHANGE WITH TRANSMITTING, REFLECTING, AND ABSORBING MEDIA We now consider a simple extension of the presentations in Secs. 8-10 and 8 11 to analyze a medium where reflection. transmission, and absorption modes are all important. As in Sec. 8-10, we shall analyze a system consisting of two parallel diffuse planes with a medium in between which may absorb, transmit. and reflect radiation. For generality we assume that the surface of the trans- mitting medium may have both a specular and a diffuse component of reflection. The system is shown in Fig. 8-58. For the transmitting medium m we have Gm + Pn + Pm + Tm = 1 (8.90) Also, fn = On The diffuse radiosity of a particular surface of the medium is defined by Inv = Enkinn + Pm 91) where G is the irradiation on the particular surface, Note that J,,., no longer represents the total diffuse energy leaving a surface. Now it represents only emission and diffuse reflection. The transmitted energy will be analyzed with additional terms. As before, the heat exchange is written q = AleE, ~ aG) (8-92) Solving for G from Eq. (8-91) and making use of Eq. (8-90) gives Ean = Snell = Fm (8.93) ProllénAn(l Pool The network element representing Eq, (8-93) is shown in Fig. 8-59. This element is quite similar to the one shown in Fig. 8-48, except that here we must take the transmissivity into account f Fig. 858. Physica’ system for anaiy ol ransmaag and retiectgAavioton exchange wih warsmting etecteg, and absorbing media 437 Fig. 859 Network element soning Ea, (6-93) ‘The transmitted heat exchange between surfaces | and 2 is the same as in Sec. 8-10: that is, had: TAFata “The heat exchange between surface | and m is computed in the following way. Of that energy leaving surface 1, the amount which arrives at m and contributes to the diffuse radiosity of m is q (8.94) om = SAF |= Te — Pind 18.95) ‘The diffuse energy leaving m which arrives at 1 is mnt = Imo nFmi (8-96) Subtracting (8-96) from (8-95) and using the reciprocity relation AsFin = AnF ns es Sn = np = to = Pd ei Gn VA F inl = Tn = Pld (8.97) The network element corresponding to Eq. (8-97) is quite similar to the one shown in Fig, 8-50. An equation similar to Eg. (8-97) can be written for the radiation exchange between surface 2 and m. Finally, the complete network may be drawn as in Fig. 8-60. It is to be noted that Jj,o represents the diffuse , J , be, ww a Fig. 680 Ce-rsiete razanon autaoik for Syston: Fig 858radiosity of the left side of m, while Jiao represents the diffuse radiosity of the right side of m If m is maintained at a fixed temperature, then J, and Js must be obtained as a solution to nodal equations for the network. On the other hand, if no net energy is delivered to m, then Em is a floating node, and the network reduces to.a simple series-parallel arrangement. In this latter case the temperature of ‘m must be obtained by solving the network for Em. We may extend the analysis a few steps further by distinguishing between specular and diffuse transmission. A specular transmission is one where the incident radiation goes “straight through" the material, while a diffuse trans mission is encountered when the incident radiation is seattered in passing through the material, so that it emerges from the other side with a random spatial orientation. As with reflected energy, the assumption is made that the trans- missivity may be represented with a specular and a diffuse component + 1D (8-98) The diffuse radiosity is still defined as in Eq. (8-91), and the net energy exchange with a transmitting surface is given by Eq. (8-93). The analysis of transmitted energy exchange with other surfaces must be handled somewhat differently, however. Fig. 8.61 Radiation netwar for nfinte parallel planes separated by a transmit specular-ciluse planeange wih ranemitng Felecbag, and absorbing mesa 438 Consider, for example. the arrangement in Fig, 8-61. The two diffuse opaque surfaces are separated by a specular-diffuse transmitting and reflecting plane. For this example all planes are assumed to be infinite in extent. The specular- transmitted exchange between surfaces 1 and 3 may be calculated immediately with on dads yshgccunr wanted = Ta Byes (8.99) ‘The diffuse-transmitted exchange between | and 3 is a bit more complicated. ‘The energy leaving | which is transmitted diffusely through 2 is JAF aren Of this amount transmitted through 2. the amount which arrives at 3 is (qussinase-seansnines = SVA\Fists0F 2s (8-100) Similarly, the amount leaving 3 which is diffusely transmitted to 1 is (qaidaitase-transmitea = JvAsF ss (8-101) Now, by making use of the reciprocity relations, AyFiz = AxFs) and AF yy AsFoy, subtraction of Eq, (8-101) from Eq. (8-100) gives, dad . ube anemones = TFB (8-102) Making use of Eqs. (8-99) and (8-102) gives the complete network for the system as shown in Fig, 8-61. Of course, all the radiation shape factors in the above network are unity, but they have been included for the sake of generality In this network Jap refers to the diffuse radiosity on the left side of 2, while Jy is the diffuse radiosity on the right side of this surface. It is easy to extend the network method to the case of two finite parallel planes separated by a transmitting plane inside a large enclosure. Such an arrangement is shown in Fig. 8-62. In this network the notation Faz means the radiation shape factor for radiation leaving the right side of surface 2. while Fou. refers to the radiation leaving the left side of surface 2. The notation Fax, means the fraction of energy leaving 1 which arrives at the enclosure 4 after specular transmission through 2. The notation Fig.) has a similar meaning with respect to surface 3. The factors Firoy and Fay designate the radiation shape factors between the image surfaces 1(2) and 3(2) and the enclosure 4. ‘The above networks require information on the specular and diffuse prop- erties of maiezial which are to be analyzed. Unfortunately. such information is quite meager at this time. Some data are available in Refs. 17 to 19.‘440 Fadaton est vanster 1,3,4 Opaque and ditfuse 2 Teanmltingand has Fig, 862 Raciaton network for systom of te specular- K. AT, = (4303) = 1212 wm + K Consulting Table 8-1, we find Ex(0 = 4 wm) = 0.6450E,, = 96,021 Wim? Ex(0 = 4 wm) = 0.002356, = 1.123 Wim? Ex( ~ 2) = (1 ~ 0.6450),, = 52,849 Wim? Ex(4 = 2) = (1 ~ 0.00235)E,, = 476.7 Wi? 4887 10° Wie?(442. Racin hea warster ‘We now apply these numbers to the network for the two wavelength bands, with unit 0 Ful; = fa (8-115a) EXAMPLE 8:15 ‘The geometry of Example 8-5 is used for radiant exchange with a large enclosure, Surface 2 is diffuse with = 0.5 while surface | is perfectly insulated. T, = 1000 K, and T, = 300 K. Calculate the heat lost to the large room per unit length of surface 2, using the numerical formulation. Solution For unit length we have Em = oT; = 5.669 x 10 Ey, = off = 459 A, = 4N0.2) = 0.8 mim Ay = (0.602 = 0.94 mim. We will use the numerical formulation. We find from Example 8-5, using the nomen- lature of the figure, Fu = 0.314, Fir = 0.425, Fiy = 0.261, Fu = 0.5, Fn = 0 Fo = 05, Fy - 0, Fr 0, Fy) + 1.0. We now write the equations. Surface 1 is insulated so we use Eq. (8-113a): Jl ~ 0.314) ~ 0.4254, ~ 0.2614, = 0 ‘Surface 2 is constant temperature so we use Eq. (8-112): JAI = 0) ~ (1 = 0.510.545 + 0.555] = (0.5)(56,690) Because surface 3 is so large, ie 459. Wim?Fomulsion ler pumercal solston 447 Rearranging the equations gives 0.686, ~ 0.425Js 0.254, + which have the solution Jy = 21,070 Wim? Jp = 33,727 Wim? The heat transfer is thus 56.690 — 33.727 ~ 0.5940.514x02) = 18,370 Wim length Because surface | is insulated, J, = Es,. and we could calculate the temperature as (ata) =m T= (5a x 10) EXAMPLE 8:16 Numerical solutons ‘Two I-m-square surfaces are separated by a distance of 1 m with T, = 1000 K, T; 400 K, €; = 0.8, = 0.5. Obtain the numerical solutions for this system when (a) the plates are surrounded by a large room at 300 K and (b) the surfaces are connected by idiating wall perfectly insulated on its outer surface. Solution Consulting Fig. 8-12, we obtain Fy=02 Fy =02 Fy =0=F, Fy=08 Fy = 08 Ay = Ay = mt (surface 3 is the surroundings or insulated surface). For part (a) Eq, = oT! = 56.69 kWim? [17,970 Buu/h » f°] Ex = oT = 1.451 kWim? Ey 459 kWim? Because A, —> », Fy, and Fy: must approach zero since A.Fi AgFr. The nodal equations are writen in the form of Eq. (8-107): surface 1 Jy == @MFids + Fish + Fish) = 0m surface 2: Ja (1 = MEd + Fass + Feds) surface 3: Jy = (= OXF yd + Fish + Fish) = 0m Because Fy, and Fy; approach zero, F must be 1.0 of AsFy and ASF) = @(448 asian heat wontor Inserting the numerical values for the various terms, we have Ki = C= ONO, + 0.22 + (0.85] = (0.8156.69) Jy = (1 = ONO. + Or + (0.85] = 0.5K1.451) Oy Jy = 1 = MO + Or + 1.05] = 610.459) ‘The third equation yields J, = 0.459 kWim? - K. Because the room is so large it acts, like a hoblraum, or blackbody. But it does not have zero heat transfer. Finally, the equations are written in compact form as Jy — 0.04), — 0.16, = 45.352 0.1, + Ja ~ OAL, = 0.7255 © Jy = 0.459 Of course, there only remain two unknowns, Jy and J, in this set. For part (b), As for the enclosing wall is 4.0 m?, and we set Jy 3 is insulated. From reciprocity we have Ey, because surface 1.00.8) _ Ay = Fy Fy = SPO = 02 (1.0K0.8) AiFy = AvP Fn = 0 = 02 ‘Then from Fy) + Fy + Fyy = 1.0 we have Fyy = 06. ‘The set of equations in (a) still applies, so we insert the numerical values to obtain (with Jy = Ey) Je = (= OBO, + 0.2, + 0.8) = (0.8156.69) Ja = (1 = 0.510.211, + OU + 0.8V5] = (0.5K1.451) @ Jo= 0 ~ MOI, + Os + 0.65) = oS Notice that the third equation of set (d) can be written as Jl = 6) = = eNO. + (0.22 + 6] = 0 so that the 1 — 6 term drops out, and we obtain our final set of equations as. Jy O04; = 0.16), = 45.352 Dy + Jy ~ OAS, = 0.7255 © 02 = 024 + 044 = 0 To obtain the heat transfers the set of equations is first solved for the radiosities. For set (0), Jy = 45.644 kim? (14,470 Brush - £2) Jy = 5.474 KWim? Jy = 0459 kWim? ‘The heat transfers are obtained from Eq. (8-110) A _ (1.0108) rT) a (56,69 ~ 45.644) = 44.186 kW [150,760 Btu’)En =~ 3) 19005 4.4) — 54) = -am94W (-730BUI 3 The net heat absorbed by the room is the algebraic sum of q, and qs or Gramates = 44.184 ~ 4.023 = 40.161 KW [137,030 Buu) T-6 For part (b) the solutions to set (e) are Ji = 51,956 KWim? J, = 20.390 KW) ‘The heat transfers are Jy = 36.173 kWim? ~ 51.965) = 18.936 kW 4.00.5) (1.451 ~ 20.390) = = 18.936 kW 10S Of course, these heat transfers should be equal in magnitude with opposite sign because the insulated wall exchanges no heat. The temperature of the insulated wall is obtained from Jn = Ey, = oy! = 36.173 Wien? and T, = 894K (621°C, 50°F] EXAMPLE 8:17 Radian trom a hole ‘To illustrate the radiation formulation for numerical solution we consider the circular hole 2 em in diameter and 3 cm deep, as shown in the accompanying figure. The hole is machined in a large block of metal, which is maintained at 1000°C and has a surface emissivity of 0.6. The temperature of the large surrounding room is 20°C. A simple approach to this problem would assume the radiosity uniform over the entire heated internal surface. In reality, the radiosity varies over the surface, and we break it into ‘segments | (bottom of the hole), 2. 3, and 4 (sides of the hole) for analysis4450 Resaton neat raneter ‘The large room acts like a blackbody at 20°C, so for analysis purposes we can assume the hole is covered by an imaginary black surface 5 at 20°C, We shall set the problem Lup for a numerical solution for the radiosities and then calculate the heat-transfer rates After that, we shall examine an insulated-surface case for this same geometry. All the shape factors can be obtained with the aid of Fig, 8-13 and the imaginary disk surfaces 6 and 7. We have Eq = oT = (5.669 x 10-*KI273}¢ = 1.48874 % 10° Wi? En = Ey = Ew OT = (5.669 x 10-9293) = 417.8 Wim? 7656206 B= 10 Ay = As = aU? = wom? = Ag = Ay Ar = Ay = Ay = 2Q\1) = 29 Fu= Fo 20 Fy =037 Fy 20.175 Fy = 01 Fu = 1 - Fe = 063 = Fu Fry = Fu ~ Fir = 0.195 = Foy = 00075 = Fs Fa = Fy ~ Fay = 0.06 = Fue ‘The equations for the radiosities are now written in the form of Eq. (8-112), noting that Fu = Oand Js = Ey, Jes (1 = GMFiaks + Fish + Fide + Fiske) + 6B 1 T~ Fae)" — i-Fa- 6) 1 Fat ‘When all the numerical values are inserted, we obtai Jy = 0.2525, + 0.078, + 0.03), + 89,341 Ae OM Fade + Fah + Frade + FusEn) + Ee) [0 ~ GME ad) + Fie + Fale + FB) + 6 nl (0 = ey + Fads + Fads + Fan) + EslFoxmulaten fr rumerieal soliton 48 Jy = 0.1479), + 0.102, + 0.02817, + 105,848, Jy = 0.045772, + 0.10214: + 0.1021, + 104,859 Jy = O.OITHLS, + 0.02817 + 0.10215 + 104,902 These equations may be solved to give A ds 1.4003 10° Wie? 4326 x 10° Wim? Jy = 1.3872 % 108 Wim? Ja = 1.2587 10° Wim? ‘The heat transfers can be calculated from Eq. (8-110) 6A, (0.62 10 4 gge7 ~ 1.4003)(10*) = 4.1658 W ooite = 10.4 ‘The total heat transfer is the sum of these four quantities or dou = 40.979 W_ {139.8 Bulb) It is of interest to compare this heat transfer with the value we would obtain by assuming uniform radiosity on the hot surface, We would then have a two-body problem with Ay = 7+ 30m Trem A= Py = 10 & = 06 = 10 ‘The heat transfer is then calculated from Eq. (8-43). with appropriate change of no- smenclature: (En, = Eds (nxt qe Te + (AvAnUe, = 1 = 42.581 W_ [145.3 Bew/h} (1.4887 10° — 417.8) + HU0.6— 1 ‘Thus, the simple assumption of uniform radiosity gives a heat transfer which is 3.9 percent above the value obtained by breaking the hot surface into four parts for the Calculation, This indicates that the uniform-radiosity assumption we have been using is ‘rather good one for engineering calculations. Let us now consider the case where surface 1 is stil radiating at 1000°C with « = 0.6 but the side walls 2, 3, and 4 are insulated. The radiation is still to the large room ‘at 20°C. The nodal equation for J; is the same as before but now the equations for Ja. ‘hand Jemust be written in the form of Eq, (8-113). When that is done and the numerical values are inserted, we obt . + O0TR, + 00M, + 89,341(482 Raciaton net traneter Ja = OSS, + 0.34524, + 0.09524), + 24,869 Sy = ONS, + 0.34524, + 0.34524, + 64.66 Je = 0.059521, + 0.0952J; + 0.34525, + 208.9 When these equations are sulved, we obtain Jy = 1.1532 x 10° Wim? (36,560 Brush - 2) Ja = 0.81019 x 105 Wim? Js = 0.57885 x 10° Wim? Js = 0.34767 x 10° Wim? ‘The heat transfer at surface 1 is A 0.6K x 10-4) Eq ~ 4) = SSS a (1.4887 ~ 1.1532\110) T-« = 15.81 W [53.95 Bh} ‘The temperatures of the insulated surface elements are obtained from Jom Ey = oTt T, = 1093 K = 820°C [1508°F] T, = 1005 K = 732°C (1350°F] Ty = 895K = 612°C (1134°F] Its of interest to compare the heat transfer calculated above with that obtained by assuming surfaces 2, 3, and 4 uniform in temperature and radiosity. Equation (8-41) applies for this case: - AsEn ~ Ex) qo pare, (E-1)+2(E-1) p= Aa * Ay Ae and g = EE IO MAL ABET = 178) ig roy ey (oto4 Bey T zt a- 201), 1 a6 7 FOF In this case the assumption of uniform radiosity atthe insulated surface gives an overall hheat transfer with surface 1 (bottom of hole) that is 18.7 percent too high. EXAMPLE 8-18 Heater with constant heat fx [In the figure shown an electric heater is installed in surface | such that a constant heat flux of 100 kW/m? is generated at the surface. The four surrounding surfaces are in radiant balance with surface I and the large room at 20°C. The surface properties are 4: = 08 and ¢; = 6 = & = « = 0.4. Determine the temperatures of all surfaces. The back side of surface 1 is insulated. Repeat the calculation assuming surfaces 2, 3, 4, and 5 are just one surface uniform in temperature.Formulation fo merical souten 483 Fig. Ex. 018 Solution In reality, surfaces 2, 3, 4, and 5 have rwo surfaces each; an inside and an outside surface. We thus have nine surfaces plus the room, so a 10-body problem is involved Of course, from symmetry we can see that T; = T, and T, = Ts, but we set up the problem in the general numerical formulation. We designate the large room as surface 6 and it behaves as if ¢ = 1.0. So, itis as if the opening were covered with a black surface at 20°C. The shape factors of the inside surfaces are obtained from Figs. 8-12 and 8-14 Fe 0.285 Fy = Fy = 0.28 = Fy = Fy Fy = Fu = OMS Fag = Fay = 0.068. Fy = Fy = 0.285 Fy = Fs = Fue = 0.115 Fay = Fr = Fas = Fo = Fry = Far = Fe = Fe = 0.23 Fut Fas Fy = Fu = Fy = 0 For the outside surfaces. Pig = Fla = Fay = Fin = 1.0 where the primes indicate the outside surfaces. We shall also use primes to designate the radiosities ofthe outside surfaces. For the room, J = Ex, = (5.669 * 10 *)(293) 417.8 Wi. For surface | with constant heat flux, we use Eq. (8-11Sa) and write Sy = Fiads + Fayds + Fide + Figde + Figde) = 10 x 10° @ Because of the radiant balance condition we have and Hy|AB4- Radiaton heat vanster ‘where the prime designates the outside radiosity. A similar relation applies for surfaces 3, 4, and 5, Thus, we can use Eq, (8-112a) for inside surface 2 Da oF + Fabs + Fale Fos + Fuld BGI) 0) and for the outside surface 2 Ji = eFivn = Guns w Equations like (c) and (d) are written for surfaces 3, 4, and 5 also, and with the shape factors and emissivities inserted the following set of equations is obtained: Jy = OSH ~ 0.244, = 0.1154, ~ 0.240, = 1.0012 x 10° 0.138), + 0.8: ~ 0.2/5 ~ O.13RJ, ~ 0.0408, = 0.138), = 57.66 0.2, = 0.80; = ~250.68 = 0.144, ~ 0.0685 + 0.8; ~ 0.245 — 0.069), - 0.054 = 60.16 0.2) - 085 = -250.8 =0.138/, ~ 0.0408/, — 0.138%, + 0.8/4 ~ 0.243 - 0.138 = 7.66 0.24. — O8N, = 7250.68 0.144), ~ 0.0684; ~ 0.0571, - 0.069%. + 0.8, = 0.255 = 60.16 0.245 ~ 0.8; = - 250.68 We thus have nine equations and nine unknowns which may be solved to give Th 127» 1 We resi 9m none Bas The tempers tt comp om Ea. 239151 3.98 po BEE a For surface 1 we observed that so that (1.0 x 1001 - 0.8) O8 and T= DISK 4 1.24887 10° = 1.49887 x 10°We note again that we could have observed the symmetry of the problem and set J; Ju, Ji = Jig and 30 on, By so doing, we could have had only five equations with five unknowns. Surfaces 2, 3, 4, and 5 as one surface ‘We now go back and take surfaces 2, 3, 4, and 5 as one surface which we choose to call surface 7. The shape factors are then Fin = Fa © 0.288 Fy, = 1 = 0.285 = 0.715 Al= 20 Ar = 60 Thus Fay = (0.71518) = 0.2383 = Fre Fry = 1 = 2x0.2883) = 0283 Fh ‘Then for surface 1 we use Eq. (8-115a) to obtain I= Body + Frade) = 1.0 10° Using Es, = Us + J5)2, we have for the inside of surface 7 SAL = Fr — 6) ~~ MF rdi + Fre) = tu +H) while forthe ouside we have J oak = San +49 ‘When the numerical values are inserted, we obtain the set of three equations: J) — 0.718, = 1.0012 x 10 1434, + 0.486J, — 0.2/5 = 59.74 250.68 02d, - 0.805 which have the solution Jy = 1.31084 3 10° Wim? bs Je 29 8.264 ‘The temperatures are then calculated as before: 43.264 + 11,129 : 2 (0 x 0 En = 27,197 Ty = 832.2K PH = 0.8) 5 1 31054 x 105 = 1.65054 « 10° T, = 1306K So, there is about a 30 K temperature difference between the two methods.4486 Radiation neat ranstor Comment With such a small difference between the solutions we may conclude that the extra compleaity of choosing each surface at a different radiosity is probably not worth the effort, particularly when one recognizes the uncertainties which are present in the surface emissivities. This points out that our assumptions of uniform irradiation and radiosity, though strictly not correct, give answers which are quite satisfactory. EXAMPLE 8:19 Combined convection and radiation (nonlinear system) AOS by 0.5 m plate is maintained at 1300 K and exposed to a convection and radiation surrounding at 300 K. Attached to the top are two radiation shields 0.5 by 0.5 m as shown in the accompanying figure. The convection heat-transfer coefficient for all sur faces is 50 Wim? - K, and €, = 0.8, € = 0.3 = 6. Determine the total heat lost from the 1300 K surface and the temperature of the shields. Solution ‘This example illustrates how it is possible to handle convection-radiation problems with the numerical formulation and an iterative computational procedure. Nomenclature is shown in the figure. Using Figs. 8-12 and 8-14, we can evaluate the shape factors as Fa =Fy= 02 Fu=1~ 02-02-06 Fas Fu 002 Fa = Far = 10 Fy = 02 Fae Fu = 06 Fy =o Jaz Jaq from symmetry ‘We now use Eq. (8-112) to obtain a relation for Jy Je 1 = eM iadae + Fda + Fudd + 68s, But Jog = Jou and Fy = Fry s0 that I= (= OVOP hay + Frade) + 6B (@) We use Eq, (8-115) for the overall energy balance on surface 2: 2WT. ~ 7) Ex ~ Son) + En ~ ® Fg En ~ Jun - Jud Equation (8-112) is used for surface Joy Jon = (1 ~ eM Fads + Fast + Funded + En But Jan = Jaz so that 1 SRM — Fads + Fade) + eed (e) TOTS MI - eM Fnds + Faagded + er] ‘“Fomsaton for rumen! soiwen 487 190 x 10! = 680 0x10 600 1502109 L 140x103} 400] Fig. Ex. 819 03 For surface Jy, the equation is Ju = = eM Foal) + GE @ We now have four equations with four unknowns, J.. Jons Jus Eoy With Ts (E,doy". However, Eq, (0) is nonlinear in Ey so we must use a special procedure to4488 Racaton hes vansler solve the set. Before embarking on the procedure, we can insert the various numerical quantities into Eqs. (a) to (d) to obtain the matrix coefficients: de Jon Ju c 1 ~0.08 0 ° 129,584 ° 0.4286 ~ 0.4286 0.8571 30,000 ~ 1007; 0.162791 1 ° 0.348837, ~ 224.2605 o 0 1 03 3214 The computational algorithm which we employ iterates on values of T, until the value of Es, obtained from solving the set of equations agrees with E,, = oT:‘. The sequence of steps is as follows. 1. Assume a reasonable value for Ts 2. Using this value in Eq. (6) solve the matrix for an estimated value of Es, (): 3. Compute E(73) = oT: from the assumed value of 7, 4. Compute AE, = Ex(e) ~ E9fT3) 5. If AEs, > 0, assume a higher value for 7; and go to step 2. If AE;, <0, assume a lower value for T; and go to step 2. 6. Continue until AE,, is suficiently small, at which point the solution to the matrix is the one desired. After two iterations have been completed, interpolation or extrapolation is used to bring ‘AE,, = 0 based on the last two values of AE, to speed the calculation. The results of, the iteration are: Teraso tr Efe ExfT) AEs, 1 300 16,283 459 15,824 2 ‘500 18377 3,543 21,920 3 384 1725.7 1232.6 493.1 4 386.6 1.275 1,266 9 Clearly the method converges very rapidly, and the final solution is J = MBS x 10 Sug = 22,051 Ju = 710 Ey, = 125 Ty ‘The total heat flux lost by surface 1 is 86.6 K Shan ~ 1) + — = 1.7226 x 10° Wim? ©Sols radianon 489 For a 0.5 by 0.5 m surface the heat lost is thus 4, = (17226 % 10°V0.57 = 43,065 W Other cases may be computed, and the influence which h and e; have on the results is shown in the accompanying figure Comment ‘This example illustrates how nonlinear equations with combined convection and ra tion can be solved with a straightforward iterative procedure. Recognizing the uncer- tainties in knowledge of surface emissivity and convection coefiicient, only a few erations are necessary to achieve an acceptable solution, 8-14 SOLAR RADIATION Solar radiation is a form of thermal radiation having a particular wavelength distribution. Its intensity is strongly dependent on atmospheric conditions, time of year, and the angle of incidence for the sun's rays on the surface of the earth, At the outer limit of the atmosphere the total solar irradiation when the earth is at its mean distance from the sun is 1395 Wim. This number is called the solar constant and is subject to modification upon collection of more precise experimental data, Not all the energy expressed by the solar constant reaches the surface of the earth, because of strong absorption by carbon dioxide and water vapor in the atmosphere. The solar radiation incident on the earth's surface is also dependent on the atmospheric content of dust and other pollutants. The max- imum solar energy reaches the surface of the earth when the rays are directly incident on the surface since (1) a larger view area is presented to the incoming solar flux and (2) the solar rays travel a smaller distance through the atmosphere 0 that there is less absorption than there would be for an incident angle tilted from the normal. Figure 8-63 indicates the atmospheric absorption effects for a sea-level location on clear days in a moderately dusty atmosphere with mod- erate water-vapor content. It is quite apparent from Fig. 8-63 that solar radiation which arrives at the surface of the earth does not behave like the radiation from an ideal gray body, while outside the atmosphere the distribution of energy follows more of an ideal pattern. To determine an equivalent blackbody temperature for the solar radiation, we might employ the wavelength at which the maximum in the spec- trum occurs (about 0.5 um, according to Fig. 8-63) and Wien’s displacement law (Eg. (8-13)]. This estimate gives 2987.6 05 5795 K _(10,431°R) The equivalent solar temperature for thermal radiation is therefore about $800 K (10,000°R),(400 Racsaton eat aster eater eee Fig. 8-63. Spectral distribution of solar radiation as functions of atmospheric conditions and angle of incidence according to Re. 15. If all materials exhibited gray-body behavior, solar-radiation analysis would not present a particularly unusual problem; however, since solar radiation is concentrated at short wavelengths, as opposed to much longer wavelengths for most “‘earth-bound” thermal radiation, a particular material may exhibit en- tirely different absorptance and transmittance properties for the two types of radiation. The classic example of this behavior is a greenhouse. Ordinary glass transmits radiation very readily at wavelengths below 2 wm; thus it transmits the large part of solar radiation incident upon it. This glass, however, is es- sentially opaque to long-wavelength radiation above 3 or 4 1m. Practically all the low-temperature radiation emitted by objects in the greenhouse is of such a long-wavelength character that it remains trapped in the greenhouse. Thus the glass allows much more radiation to come in than can escape, thereby producing the familiar heating effect. The solar radiation absorbed by objects in the greenhouse must eventually be dissipated to the surroundings by con- vection from the outside walls of the greenhouse. Similar behavior is observed for the absorptance and reflectance of solar asSole radanon 404 Table 83 Comparisons of Absorpivtes of Various Suraces to Solar and Low-Temperature Thermal Radiation as Compiled from Ret. 14 Absorptvity For tow- For temperature solar radiation face radiation 28 Aluminum. highly polished ous 0.04 Copper, highly polished 0.18, 0.03 ‘Tarnished 06s 0.75 Cast iron 0.94 021 Stainless steel, no. 301 polished: 037 0.60 White marble 0.46, 095 Asphalt 0.90 0.90 Brick, red 0.75 0.93 Gravel 0.29 Os Flat black lacquer 0.96 095 White paints, various types of pigments 0412-016 0.90-0.95 ‘opposed to low-temperature radiation from opaque metal or painted surfaces. Inmany instances, the total absorptivity for solar radiation can be quite different from the absorptivity for blackbody radiation at some moderate temperature like 25°C, Table 8.3 gives a brief comparison of the absorptivities for some typical surfaces for both solar and low-temperature radiation. As will be noted, rather striking differences can occur. This brief discussion of solar radiation is not intended to be comprehensive. Rather, it has the purpose of alerting the reader to some of the property in- formation (like that of Ref. 14) when making calculations for solar radiation. EXAMPLE 8-20 Calculate the radiation equilibrium temperature for a plate exposed to a solar flux of 700 Wim and a surrounding temperature of 25°C ifthe surface is coated with (a) white paint or (b) flat black lacquer, Neglect convection Solution ‘At radiation equilibrium the net energy absorbed from the sun must equal the long- wavelength radiation exchange with the surroundings, or (us For white paint we obtain from Table 8:3 Ave ene OT = Th) @ Bue = 0.12 hon tang = 0.94402. Radation heat tanster so that Eq. (a) becomes (700,012) and T (0.95.69 x 10-(T* ~ 2984) 12.5 K = 395°C [103°F] For flat black lacquer we obtain Gn = 0.96 Ghee emp = 0.95 so that Eq, (a) becomes (7000.96) = (0.9545.669 x 10-47" ~ 2984) and T = 378K = 1048°C [220.66F] We conclude from this example what we may have known from the start, that white surfaces are cooler than black surfaces in the sunlight, EXAMPLE 6-21. A tat plate solar collector A flatplate solar collector is constructed as shown in the figure. A glass plate covers the blackened surface, which is insulated. Solar energy at the rate of 750 Wim is transmitted through the glass and absorbed in the blackened surface. The surface heats up and radiates to the glass and also loses heat by convection across the air gap, which has kJk = 1.4. The outside surface of the glass loses heat by radiation and convection 10 the environment at 30°C with & = 20 Wim? «°C. It is assumed that the glass does ‘not transmit any of the thermal radiation and has € = 0.9. The blackened surface is ‘assumed to have ¢ = 1.0 for all radiation. Determine the temperatures of the glass and inside surface. ae wataion Fig. Ex. 821 Solution This is an interesting example of combined radiation and convection heat-transfer anal- ysis. We designate the black plate as surface 1, the glass as surface 2, and the sur- Toundings as surface 3. We assume no absorption of the solar energy in the glass. For the black plate d= Ex ‘The convection and solar energy delivered to surface 1 isSols ragiaton 463 4] -kin-med IC Eq, (8-118) is now applied, we have &, q where Js is the inside radiosity for the glass. The overall energy balance for surface 2 en b-mekin-wenn-m © : Finally, the radiation lost from the outside of the glass can be written two ways: 4 PERE © ‘The area of the collector is very karge compared to the spacing so Fi; = 1.0. We ‘now have four equations and four unknowns: Ey). En. Jo. and Js, OF course, Ty and T, are expressed in terms of Ey, und E,,. The solution procedure is as follows: 1. Solve Eq. (d) for E.,and Ts by iteration 2. Solve Eg. (e) for Js, Solve Eg. (b) for Js, and substitute in Eq. () giving (with Fr; = 1.0) 2 A af eka 4] J] k ka teat) 0 4. Solve Eq, (by iteration for 7, and Ey When this procedure is followed, with k = 0.03 Wim °C, the results are By = 682.96 Wim? Ty = BIB K Jo, = 662.45 Wim? Ja, = 760.94 Wine Ey, = 1971.68 Wim? T, = 401.4 K (262.5°F] and we see that a rather substantial temperature can be achieved inside the collector. This result is for the insulated case. In normal practice energy will be extracted from the hot surface and lower temperatures will be experienced, depending on the rate of energy removal. Extensive information on solar collectors is given in Ref. 37.4404 Radiator neat ranster 8:15 RADIATION PROPERTIES OF THE ENVIRONMENT We have already described the radiation spectrum of the sun and noted that the major portion of solar energy is concentrated inthe short-wavelength region. Tt was also noted that as a consequence of this spectrum. real surfaces may exhibit substantially different absorption properties for solar radiation than for Jong-wavelength “earthbound™ radiation Meteorologists and hydrologists use the term insolation to deseribe the in- tensity of direct solar radiation incident on a horizontal surface per unit area and per unit time, designated with the symbol J. Although we shall emphasize ‘other units, it will be helpful to mention a unit that appears in the meterological literature Langley (Ly) = 1 calem* [3.687 Btulft"] Insolation and radiation intensity are frequently expressed in langleys per unit, time; e.g., the Stefan-Boltzmann constant would be 0.826 x 10°" Ly/min - K* Radiation heat transfer in the environment is governed by the absorption, scattering, and reflection properties of the atmosphere and natural surfaces. ‘Two types of scattering phenomena occur in the atmosphere. Molecular scat tering is observed because of the interaction of radiation with individual mol- ecules. The blue color of the sky results from the scattering of the violet (short) wavelengths by the air molecules. Particulate scattering in the atmosphere results from the interaction of radiation with the many types of particles that may be suspended in the air. Dust, smog, and water droplets are all important types of particulate scattering centers. The scattering process is governed mainly by the size of the particle in comparison with the wavelength of radiation, Maximum scattering occurs when wavelength and particle size are equal and decreases progressively for longer wavelengths. For wavelengths smaller than the particle size, the radiation tends to be reflected. Reffection phenomena in the atmosphere occur for wavelengths less than the particle size and are fairly independent of wavelength in this region. The term albedo is used to describe the reflective properties of surfaces and is defined by reflected energy incident energy albedo = 417) The albedos of some natural surfaces are given in Table 8-4. ‘The effect of solar incident angle on the albedo of water is shown in Fig. 8-64, where a is the angle which the incoming rays make with the horizontal. ‘The atmosphere absorbs radiation quite selectively in narrow-wavelength bands. The absorption for solar radiation occurs in entirely different bands from the absorption of the radiation from the earth because of the different spectrums for the two types of radiation. In Fig. 8-65 we see the approximate spectrums for solar and earth radiation with some important absorption bandsRadiation propedies ofthe environment. 488 Table 6-4 Albedo of Natural Surtaces, According to Ret. 32. Surface Albedo A Surfa ‘Albedo A Water 0.03-0.40 Spring wheat 0.10-0.25, Black, dry soil od Winter wheat 0.16-0.23 Black, moist soil 0.08 Winter rye 0.18-0.23 Gray, dry soil 0.25-0,30 High, dense grass 0.18-0.20 Gray, moist soil 0.10-0.12 Green grass 0.26 Blue, dry loam 023 Grass dried in sun 0.19 Blue, moist loam 0.16 ‘Tops of oak 08 Desert loam 0.29-0.31 Tops of pine od Yellow sand 0.35 ‘Tops of fir 0.10 White sand 034-040 Cotton 0.20-0.22 River sand 0.43 Rice field 0.12 Bright, fine sand 037 Lettuce 022 Rock 0112-015 Beets 0.18 Densely urbanized areas 0:15-0.25 Potatoes 0.19 Snow 0.40-0.85 Heather 0.10 Sea ice 0.36-0.50 AM Fig. 6464. Effect of solar altitude angle and cloud cover on albedo fora horizontal water surlace, according 10 Ret, 33,406 Racin heat arses 1000 oa vmx 10° TIT Th SF Y 10 100 Fig. 665. Therma-raciation specttums for the sun and earth wth primary absorption bbands indicated by shaded areas. Note diferent scales: 1 ergiem®'s = 1 mWim? superimposed on the diagram. Note the scale differential on the two curves, ‘A quick inspection of these curves will show that the atmosphere transmits most of the short-wavelength radiation while absorbing most of the back diation from the earth. Therefore, the atmosphere acts very much like a green- house, trapping the incoming solar radiation to provide energy and warmth for humans on earth. Some concern is voiced that people may upset the energy budget of the earth through excessive contamination of the atmosphere with pollutants. Such a possibility does exist but is beyond the scope of our dis cussion, The absorption and scattering of radiation may be described with Beer’s law (Eq. (8-48)], which we repeat here for convenience te = ene (8-118) where a, is the monochromatic absorption coefficient and x is the thickness of the layer absorbing the radiation. For a scattering process, we would replace a, by a scattering coefficient ky. The conventional approach in atiospheric problems is to assume that theRacin properties ole enwronment 467 absorption and scattering processes are superimposed on each other and may be expressed in the form of Eq, (8-118) over all wavelengths. The appropriate coefficients are defined as dim, = average molecular scattering coefficient over all wavelengths, dy, ~ average particulate scattering coefficient over all wavelengths @ = average absorption coefficient over all wavelengths aa |. + a = total attenuation cvefficient over all wavelengths With these coefficients, the radiation insolation at the earth's surface is €. pressed as lg mae 119) i Wet where m is the relative thickness of the air mass and rr is defined as the turbidity factor of the air: (8-120) ‘The insolations are defined as 1, = direct. cloudless sky insolation at earth's surface 1, = insolation at outer limits of earth's atmosphere The molecular scattering coefficient for air at atmospheric pressure is given as By din, = 0.128 ~ 0,054 log mt 12D The relative thickness of the air mass is calculated ay the cosecunt of the solar altitude a. The turbidity factor is thus a convenient means of specifying at- mospheric purity and clarity; its walue ranges from about 2.0 for very clear ait to 4 or 5 for very smoggy. industrial environments, The insolation at the outer edge of the atmosphere is expressed in termy of the solar constant By, by 1, = Es sina (8-122) where ais again the angle the rays make with the horizontal, We gave the solar constant as 442.4 Buh 1 2.00 caliem? + min 1398 Wim? ‘An average variation of incident solar radiation for cloudy and cloudless sit- tuations as a function of solar altitude angle is given in Table 8-5(408 Radiation hea! ranster ‘Table 85 Solar radiation (insolation) on a Horizontal Surface Under Average Atmospheric Conditions, According to Re. 23. 25 328 314 30 06 472 3s 417 554.6 0 S47 636.0 45 611 710.4 50 2 Taha o 15 SOLd 0 853981 80 89.7 1083 90 91.4 1063 EXAMPLE 8-22 Accertain smoggy atmosphere has a turbidity of 4.0. Calculate the direct, cloudless-sky insolation for a solar altitude angle of 75°. How much is this reduced from that of a clear sky? is calculation, so that we need to make the following. intermediate calculations. de = Ey, sin a = 1395 sin 75° = 1347 Wim? m = ese 75° = 1.035 Gq = 0.128 ~ 0.054 log m = 0.1272 We have n = 4.0, so from Eq. (8-119), 1, = 1347 exp [~(0.127214.041.035)] = 795.5 Wim? For a very clear day, n = 2.0 and we would have 1, = 1347 exp [= (0.1272)2.0(1.035)) = 1035 Wim? ‘Thus the insolation has been reduced by 23 percent as a result of the smoggy environ- ‘mentRediaton properties of the ervonment 468 “The absorption of radiation in natural bodies of water is an important process because of its influence on evaporation rates and the eventual dispersion of water vapor in the atmosphere. Experimental measurements [34] show that solar radiation is absorbed very rapidly in the top layers of the water followed by an approximately exponential decay with depth in the water. The incident radiation follows a variati (~ ples 6-123) where J, is the intensity at the surface and /, is the intensity at a depth z; B represents the fraction of energy absorbed at the surface, and a is an absorption of extinction coefficient for the material. Beta may be interpreted as a measure of the long-wavelength content of solar radiation, since the shorter wavelengths penetrate the water more readily. This coefficient appears to have the value (0.4 for all lakes for which data are available, and itis assumed to be independent of time. Values of the extinction coefficient can vary considerably, e.g., it has the value of 0.16 m~! for the extremely clear water of Lake Tahoe (35] and a value of 0.89 m~' for the more turbid water of Lake Castle, California [36]. ‘Additional information on radiation in the atmosphere is given by Kondratyev [30] and Geiger [31]. EXAMPLE 8:23 Calculate the heat-generation rate resulting from solar-radiation absorption in a lake with an extinction coefficient of 0.328 m-! and a solar altitude of 90° on a clear day. Perform the calculation for a depth of 1 ft [0.3048 m). solution ‘The heatgeneration rate is obtained by differentiating Eq. (8-123) lee > Soest =F = lal - Be ‘The surface insolation J, is calculated with Eq. (8-119) 1, = Ba, sin = 1395 Wim? m= csc 90" = 1.0 Amy = 0.128 ~ 0.054 log m = 0.128 n=20 1 = 1, = 1395 exp [~(0.128)2.0)1.0)], = 1080 Wim? We also have a = 0.328 and = 0.4, so that (4080)(0.328)(1 — 0.4) exp { ~(0.328N0.3088)], 192.3 Wim? 44470 Racaton neat ranster 8-66 Thormometer element in tow ::.: 16 EFFECT OF RADIATION ON TEMPERATURE MEASUREMENT When a thermometer is placed in a gas-flow stream to measure temperature. the temperature indicated by the sensing element is determined by the overall energy balance on the element. Consider the element shown in Fig. 8.66. The temperature of the gas is T.. the effective radiation surrounding temperature is T,, and the temperature indicated by the thermometer is 7;. Assuming that Tzis greater than T,, energy will be transferred by convection to the thermom- eter and then dissipated by radiation to the surroundings. Thus the energy balance becomes AT. ~ 1) = ARTS = T) (8-124) where A is the surface area of the element and ¢ is its emissivity. Equation (8 124) assumes that the surroundings are either very large or black so that Eq (8-43a) can be applied for the radiation heat transfer. From this energy balance we see that the temperature indicated by the thermometer is nor the true gas temperature but some radiation-convection equilibrium temperature. Very large errors can result in temperature measure- ments if this energy balance is not properly taken into account. Radiation shields are frequently employed to alleviate this difficulty ‘lass thermometer having € = 0.9 hangs in a metal building and indicates a temperature of 20°C. The walls of the building are poorly insulated and have # tem perature of S°C. The value of h for the thermometer may be taken as 8.3 Wim °C Caleulate the true air temperature Solution We employ Eq. (8-124) for the solution: IMT. ~ Ty = ae — 74 Inserting the numerical values, with T, = 20°C = 293 K. 7, = 8°C © 278K. gives (8.3)(T. ~ 293) = (5.669 x 10. "0.94298" ~ 278") and. T. = 3016K = 286°C [83.S°F] In this simple example the thermometer is in error by 8.6°C [15.S°F]!Sunray 474 © 8-17 THE RADIATION HEAT-TRANSFER COEFFICIENT In the development of convection heat transfer in the preceding chapters, we found it convenient to define a heat-transfer coefficient by Aeony = beopvAlT ~ T) Since ratiation heat-transfer problems are often very closely associated with convection problems, and the total heat transfer by both convection and ra: diation is often the objective of an analysis, it is worthwhile to put both pro- cesses on a common basis by defining a radiation heat-transfer coefficient h, das = MAMT, ~ Ta) where T; and 7; are the temperatures of the two bodies exchanging heat by radiation. The total heat transfer is then the sum of the convection and radiation. q = the + ATs ~ Te) (8-125) if we assume that the second-radiation-exchange surface is an enclosure and is at the same temperature as the fluid. For example, the heat loss by free convection and radiation from a hot steam pipe passing through a room could be calculated from Eq. (8-125), In many instances the convection heat-transfer coefficient is not strongly dependent on temperature. However, this is not so with the radiation heat- transfer coefficient, The value of f,, corresponding to Eq. (8-43), could be calculated from 4 ath ~ Ts Ay” Wey + (AVAglle, — D _ oe + TENT + Te) * Tey + AAV — D = hh ~ 1) h (8-126) Obviously, the radiation coefficient is a very strong function of temperature. The reader may recall that we used a concept like Eq. (8-126) to obtain a “radiation resistance” for numerical examples in Chaps. 3 and 4. m 618 SUMMARY In this chapter «© have examined several means for analyzing radiation heat transfer. The gray-body assumption, although not strictly correct, is a viable method for performing heat-transfer calculations. Assumptions of uniform ra- digsity and irreliation over surfaces are also not strictly correct but provide ‘an approxims.tian which is usually well within the accuracy of knowledge of surface propert.c.. In Tuble 8-6, we present a tabular summary of a few formulas which are ofte:, used472. Padation heat tanster Table 86 Summary of Radiation Formulas Energy emitted by a black body Basic radiation shape factor Type. reciprocity relation Net energy lost by a gray surface 4° aA Net radiant exchange between q= cAeITA~ TW convex surface | and large ‘enclosure 2 Radiation-balance equation for J ~ Fu = eh ‘constant surface temperature of ith surface ~ UO SFI, = Raiation-balance equation for Ml = Fy ~ ZF, = 0 surface in radiant balance; 1 e.,q/A = 0 Radiation-balance equation for surface with specified heat flux JA = Fa) = SF, REVIEW QUESTIONS How does thermal radiation differ from other types of electromagnetic radiation? 1 1" 2 2 “ 5 ‘What is the Stefan-Boltzmann law? Distinguish between specular and diffuse surfaces. Define radiation intensity. ‘What is Kirchhoff’s identity? When does it apply?” ‘What is a gray body? ‘What is meant by the radiation shape factor? Define irradiation and radiosity. ‘What is Beer's law? Why do surfaces absorb differently for solar or earthbound radiation? Explain the greenhouse effect Why is the sky blue? Define albedo, What is meant by the atmospheric greenhouse effect? What is meant by the turbidity factor? Equation number (83) 8-180) (638) (8-30) (8-112a) 13a) 1150)Probie 473 = PROBLEMS a a2 33 as 67 29 Fused quartz transmits 90 percent of the incident thermal radiation between 0.2 ‘and 4 um, Suppose a certain heat source is viewed through a quartz window. ‘What heat flux in watts will be transmitted through the material from blackbody radiation sources at (a) 800°C, (6) 550°C. (c) 250°C, and (d) 70°C? Repeat Prob. 8-1 for synthetic sapphire, which has a transmissivity of 0.85 between 0.2 and 5.5 wm. Repeat Prob. 8-1 for cesium iodide, which has « transmissivity of approximately 0,92 between 0.3 and 52 jum Calculate the energy emitted between 4 and 15 ym by a gray body at 100°F with © = 06 ‘A furnace with black interior walls maintained at 1100°C has an opening in the ‘ide covered with a glass window having the following properties 02 p=0 08 p= 02 Assume diffuse behavior and calculate the radiation lost through the window to a large room at 25°C. OSAc3um: 1208 € Bum

You might also like

• Fear: Trump in the White House

  

  From Everand

  Fear: Trump in the White House

  Bob Woodward

  Rating: 3.5 out of 5 stars

  3.5/5 (738)
• The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  

  From Everand

  The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  Viet Thanh Nguyen

  Rating: 4.5 out of 5 stars

  4.5/5 (122)
• A Man Called Ove: A Novel

  

  From Everand

  A Man Called Ove: A Novel

  Fredrik Backman

  Rating: 4.5 out of 5 stars

  4.5/5 (4610)
• • Magazines
  • Podcasts
  • Sheet music
• A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  

  From Everand

  A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  Dave Eggers

  Rating: 3.5 out of 5 stars

  3.5/5 (231)
• Grit: The Power of Passion and Perseverance

  

  From Everand

  Grit: The Power of Passion and Perseverance

  Angela Duckworth

  Rating: 4 out of 5 stars

  4/5 (589)
• The Little Book of Hygge: Danish Secrets to Happy Living

  

  From Everand

  The Little Book of Hygge: Danish Secrets to Happy Living

  Meik Wiking

  Rating: 3.5 out of 5 stars

  3.5/5 (401)
• Rise of ISIS: A Threat We Can't Ignore

  

  From Everand

  Rise of ISIS: A Threat We Can't Ignore

  Jay Sekulow

  Rating: 3.5 out of 5 stars

  3.5/5 (137)
• Principles: Life and Work

  

  From Everand

  Principles: Life and Work

  Ray Dalio

  Rating: 4 out of 5 stars

  4/5 (599)
• Shoe Dog: A Memoir by the Creator of Nike

  

  From Everand

  Shoe Dog: A Memoir by the Creator of Nike

  Phil Knight

  Rating: 4.5 out of 5 stars

  4.5/5 (537)
• Never Split the Difference: Negotiating As If Your Life Depended On It

  

  From Everand

  Never Split the Difference: Negotiating As If Your Life Depended On It

  Chris Voss

  Rating: 4.5 out of 5 stars

  4.5/5 (842)
• Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  

  From Everand

  Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  Margot Lee Shetterly

  Rating: 4 out of 5 stars

  4/5 (897)
• The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  

  From Everand

  The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  Mark Manson

  Rating: 4 out of 5 stars

  4/5 (5807)
• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  

  From Everand

  The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  Ben Horowitz

  Rating: 4.5 out of 5 stars

  4.5/5 (345)
• Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  

  From Everand

  Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  Gilbert King

  Rating: 4.5 out of 5 stars

  4.5/5 (266)
• John Adams

  

  From Everand

  John Adams

  David McCullough

  Rating: 4.5 out of 5 stars

  4.5/5 (2409)
• The Emperor of All Maladies: A Biography of Cancer

  

  From Everand

  The Emperor of All Maladies: A Biography of Cancer

  Siddhartha Mukherjee

  Rating: 4.5 out of 5 stars

  4.5/5 (271)
• Team of Rivals: The Political Genius of Abraham Lincoln

  

  From Everand

  Team of Rivals: The Political Genius of Abraham Lincoln

  Doris Kearns Goodwin

  Rating: 4.5 out of 5 stars

  4.5/5 (234)
• The World Is Flat 3.0: A Brief History of the Twenty-first Century

  

  From Everand

  The World Is Flat 3.0: A Brief History of the Twenty-first Century

  Thomas L. Friedman

  Rating: 3.5 out of 5 stars

  3.5/5 (2259)
• Her Body and Other Parties: Stories

  

  From Everand

  Her Body and Other Parties: Stories

  Carmen Maria Machado

  Rating: 4 out of 5 stars

  4/5 (821)
• Yes Please

  

  From Everand

  Yes Please

  Amy Poehler

  Rating: 4 out of 5 stars

  4/5 (1898)
• The Glass Castle: A Memoir

  

  From Everand

  The Glass Castle: A Memoir

  Jeannette Walls

  Rating: 4.5 out of 5 stars

  4.5/5 (1715)
• Angela's Ashes: A Memoir

  

  From Everand

  Angela's Ashes: A Memoir

  Frank McCourt

  Rating: 4.5 out of 5 stars

  4.5/5 (440)
• The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  

  From Everand

  The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  Brene Brown

  Rating: 4 out of 5 stars

  4/5 (1091)
• The Light Between Oceans: A Novel

  

  From Everand

  The Light Between Oceans: A Novel

  M.L. Stedman

  Rating: 4.5 out of 5 stars

  4.5/5 (789)
• A Tree Grows in Brooklyn

  

  From Everand

  A Tree Grows in Brooklyn

  Betty Smith

  Rating: 4.5 out of 5 stars

  4.5/5 (1929)
• The Art of Racing in the Rain: A Novel

  

  From Everand

  The Art of Racing in the Rain: A Novel

  Garth Stein

  Rating: 4 out of 5 stars

  4/5 (4203)
• The Perks of Being a Wallflower

  

  From Everand

  The Perks of Being a Wallflower

  Stephen Chbosky

  Rating: 4.5 out of 5 stars

  4.5/5 (2104)
• Wolf Hall: A Novel

  

  From Everand

  Wolf Hall: A Novel

  Hilary Mantel

  Rating: 4 out of 5 stars

  4/5 (3811)
• The Woman in Cabin 10

  

  From Everand

  The Woman in Cabin 10

  Ruth Ware

  Rating: 3.5 out of 5 stars

  3.5/5 (2520)
• Sing, Unburied, Sing: A Novel

  

  From Everand

  Sing, Unburied, Sing: A Novel

  Jesmyn Ward

  Rating: 4 out of 5 stars

  4/5 (1104)
• Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  

  From Everand

  Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  Ashlee Vance

  Rating: 4.5 out of 5 stars

  4.5/5 (474)
• On Fire: The (Burning) Case for a Green New Deal

  

  From Everand

  On Fire: The (Burning) Case for a Green New Deal

  Naomi Klein

  Rating: 4 out of 5 stars

  4/5 (74)
• The Outsider: A Novel

  

  From Everand

  The Outsider: A Novel

  Stephen King

  Rating: 4 out of 5 stars

  4/5 (1840)
• The Yellow House: A Memoir (2019 National Book Award Winner)

  

  From Everand

  The Yellow House: A Memoir (2019 National Book Award Winner)

  Sarah M. Broom

  Rating: 4 out of 5 stars

  4/5 (98)
• The Unwinding: An Inner History of the New America

  

  From Everand

  The Unwinding: An Inner History of the New America

  George Packer

  Rating: 4 out of 5 stars

  4/5 (45)
• The Constant Gardener: A Novel

  

  From Everand

  The Constant Gardener: A Novel

  John le Carré

  Rating: 3.5 out of 5 stars

  3.5/5 (104)
• Brooklyn: A Novel

  

  From Everand

  Brooklyn: A Novel

  Colm Tóibín

  Rating: 3.5 out of 5 stars

  3.5/5 (1946)
• Little Women

  

  From Everand

  Little Women

  Louisa May Alcott

  Rating: 4 out of 5 stars

  4/5 (104)
• Steve Jobs

  

  From Everand

  Steve Jobs

  Walter Isaacson

  Rating: 4.5 out of 5 stars

  4.5/5 (806)
• Manhattan Beach: A Novel

  

  From Everand

  Manhattan Beach: A Novel

  Jennifer Egan

  Rating: 3.5 out of 5 stars

  3.5/5 (792)
• Bad Feminist: Essays

  

  From Everand

  Bad Feminist: Essays

  Roxane Gay

  Rating: 4 out of 5 stars

  4/5 (1016)
• Tutorial Aspen Dynamics

  

  Document20 pages

  Tutorial Aspen Dynamics

  Robbie_LOCO

  100% (2)
• Application of The Static Cone Penetration Test

  

  Document19 pages

  Application of The Static Cone Penetration Test

  SD

  100% (1)
• Manufacturing Process-2 PDF

  

  Document290 pages

  Manufacturing Process-2 PDF

  Karthik Rao

  No ratings yet
• Ware House Design

  

  Document8 pages

  Ware House Design

  Jay-Jay Japs

  No ratings yet
• Physics Laboratory Experiment 9

  

  Document4 pages

  Physics Laboratory Experiment 9

  Kristella Draheim

  No ratings yet
• Viscosity Xylitol

  

  Document31 pages

  Viscosity Xylitol

  leo

  No ratings yet
• Design of Low-Cost Noise Measurement Sensor Network Sensor Function Design PDF

  

  Document8 pages

  Design of Low-Cost Noise Measurement Sensor Network Sensor Function Design PDF

  Frew Frew

  No ratings yet
• Standard Solution

  

  Document5 pages

  Standard Solution

  ganesh2gig

  No ratings yet
• DRAFT V1.0 - 2/26/04: Memorandum of Understanding

  

  Document4 pages

  DRAFT V1.0 - 2/26/04: Memorandum of Understanding

  Yudhistira Adi

  No ratings yet
• PHC 13 Section Assessment Answers DRB

  

  Document31 pages

  PHC 13 Section Assessment Answers DRB

  Will Hart

  0% (1)
• Bernard D'espagnat The Quantum Theory and Reality

  

  Document25 pages

  Bernard D'espagnat The Quantum Theory and Reality

  none

  No ratings yet
• Test Ch. 12 (The Gas Laws) Practice

  

  Document4 pages

  Test Ch. 12 (The Gas Laws) Practice

  liza1207

  No ratings yet
• C3 Integration PDF

  

  Document5 pages

  C3 Integration PDF

  Rokonuddin Ahmed

  No ratings yet
• Mold Magic Wax - Brochure

  

  Document2 pages

  Mold Magic Wax - Brochure

  Fer Net

  No ratings yet
• Collimator

  

  Document5 pages

  Collimator

  skc3128

  No ratings yet
• Exercise Chapter6

  

  Document23 pages

  Exercise Chapter6

  Izzat Rawaida

  0% (1)
• 1D Chain Part B

  

  Document6 pages

  1D Chain Part B

  haseeb ahmad

  No ratings yet
• Prosedur Double Regenerasi Demineralizer Plant Train C

  

  Document6 pages

  Prosedur Double Regenerasi Demineralizer Plant Train C

  Ido Aprilian Tarigan

  No ratings yet
• Pitot Tube

  

  Document30 pages

  Pitot Tube

  pushpzala86

  No ratings yet
• Newtons Laws of Motion

  

  Document7 pages

  Newtons Laws of Motion

  Janaka Priyalal

  No ratings yet
• FOREX & Integrated Pitchfork Analysis - Part I

  

  Document7 pages

  FOREX & Integrated Pitchfork Analysis - Part I

  nettide

  No ratings yet
• Question Paper Unit 4723 Core Mathematics 3

  

  Document4 pages

  Question Paper Unit 4723 Core Mathematics 3

  AkatharPurin

  No ratings yet
• Problem Set

  

  Document23 pages

  Problem Set

  Daniel Galindo

  No ratings yet
• Copper Reaction 2

  

  Document8 pages

  Copper Reaction 2

  Rocel Lomeda

  No ratings yet
• DR3M Codigo Fuente PDF

  

  Document205 pages

  DR3M Codigo Fuente PDF

  Robert Marlindo Ramirez Quispe

  No ratings yet
• The Soundpost in The Violin - McLennan

  

  Document32 pages

  The Soundpost in The Violin - McLennan

  AkaneGarcia

  No ratings yet
• Sample Test Using Auto-Cheker

  

  Document3 pages

  Sample Test Using Auto-Cheker

  Jhon Aliangan Mayuyo

  No ratings yet
• International Critical Tables - Google Search

  

  Document2 pages

  International Critical Tables - Google Search

  zoksi

  No ratings yet
• Flow Characteristics in Local Scour

  

  Document16 pages

  Flow Characteristics in Local Scour

  AbhilashJana

  No ratings yet
• Adding Noise To Improve Measurement

  

  Document8 pages

  Adding Noise To Improve Measurement

  Megha Agarwal

  No ratings yet