484 Chapter 7 = External Plow Consider te use of kerosene at 300 K with porous Shere of I-mm diameter. At this temperture the Kerosene has a saturated vapor density of 0.015 kg/m? and a latent hea! of vaporization of 300 idVkg. The mass diffusivity for the vapor-ait mixture 1. 10°> sn. If dy, amosphere air at V= 15 ms and Te 300 K flows over the sphere, what i he minimum ‘mass rate at which kerosene must be supplied oi tain 2 weed surface? For this condition, by how such must Y= actually exceed 7, co maintain the we ted surface a 300 K? 7.138 Consider an air conditioning system composed of a bank of robes aranged normal o air flowing in a duct at amass tale ofr, (kg/). A coolant flowing through the tubes is eble 1 maintain the surface temperature of the tubes at @ constant value of T, < Ty» where Ty, is the inlet air temperature (upstream of the tube bank), ft has been suggested that ait spoling may be enhanced ifs thin, uniform film of water is maintained con the outer surface of each of the tubes, (@) Assuming the water film to be atthe temperature » develop an expression for the ratio of the amount of cooling that occurs with the water film {o the amount of cooling that occurs without the film. The amount of cooling may be defined as Toy Tags Where T,, isthe outlet ai temperature (downstream of the wbe bank). The upstream may be assumed to be dy, andthe diving tials For convection heat and mass tranefer ma approximated as (T,,~ 7.) and p(T), lively. Note: The total rate of heat Loss from air may be expressed as q~ ri,(Ty, Ty Estimate the value of this ratio under cond for which T,, = 35°C and 7, = 10°C Consider a tube bank that is 5 ows deep, with lubes in a row. Each tube is 0.5 m long wit outside diameter of 8mm, and 2 arrangement is wed lor which Sy = 5, Under conditions for which my, = 0.5 kgs, V=B rms, Ty, = 38°C, and T, = 10°C, what she vail (of Ta, if the tabes are wetted? What isthe sei humidity of the air leaving the tube bank? Ina papersdrying process, the paper moves on as veyor belt at 0.2 mis, while dry ai From an aay slot jets (Figure 7.16) impinges nocmal to is sure The nozzle width and pitch are W= 10 mm and $= 0 mm, respectively, and the nozee-to-plate sept tion is H=200 mm. The wet paper is of wid L= 1m and is maintained at 300 K, while he exits the nozzles at a yemperature of 300K asd & Velocity of 20.mvs. In kg/s+m?, shat is the avenge drying rate per unit surface area of the papec?by). The upstream a Ind We driving pot lass transfer may) and Pasa(T), respec bf heat Toss from th indpulTas ~ Ty ftio under conditions $5 rows deep, with [2 5 0 m long, with am and «staggered bS;= 5, = 24m, vn, = 05 kes what isthe value 37 What isthe specific eae bank? aper moves on a cone pir from an array of normal to its surface We 10mm and S= ocee-to-plate separ a paper is of wid 300 K, while the alt use of 300K and a what is the average ofthe paper? CHAPTER 38 Internal Flow180 8.1 Hydrodynamic Considerations Chapter 8 # Internal Flove ving acquired the means t0 compute convection transfer rates for flow, we now consider the convection transfer problem for internal flow. Recall fan external flow is one for which boundary layer development on 2 surf allowed to continue without external constraints as for the flat plate of Figure {In contrast, an internal low, such as flow in a pipe, is one for which the Mui i fined by « surface. Hence the boundary layer is unable to develop without e ally being constrained. The internal flow configuration represents a conved geometry for heating and vooling fluids used in chemical processing, environ control, and energy conversion technologies. ‘Our objectives are to develop an appreciation for the physical phenomena ciated with internal flow and to obtain convection ccefficients for flow conti practical importance. We begin by considering velocity (hydrodynamic) effects ‘ent to internal flows, focusing on certain unique features of boundary ayer ‘ment. Thermal boundary layer effects are considered next, and an overall balance is applied to determine fluid temperature variations in the flow di Finally, correlations for estimating the convection heat transfer coefficient are sented for a variety of internal flow conditions. When considering external flow, itis necessary to ask only whether the flow islam inas or turbulent. However, for an internal flow we must also be concerned with existence of entrance and fully developed regions. 8.1.1 Flow Conditions Consider laminar flow in a circular tube of radius r, (Figure 8.1), where fluid enters the tube with a uniform velocity, We know that when the fluid makes contact with the surface, viscous effects become important, and a boundary layer develops with increasing x. This development occurs at the expense of a shrinking inviseid fw fv Hw epan Foci 8.1 Laminar, hydrodynamic boundary layer development in 8 etcular tube.1 # Hydrodynamic Considerations 487 region and concludes with boundary layer merger atthe centerline. Following this ‘merger, viscous effects extend over the entire cross section and the velocity profile no longer changes with increasing x. The flow is then said to be fully developed, and the distance from the entrance at which this condition is achieved is termed the hnydrodynamic entry length, x4. As shown in Figure 8.1, the fully developed veloc: ity profile is parabolic for laminar flow in a circular tube. For turbulent flow, the profile is flatter due to turbulent mixing in the radial direction, ‘When dealing with internal flows, itis important to be cognizant of the extent of the entry region, which depends on whether the flow is laminar or turbulent, The Reynolds number for low in a circular tube is defined as nD @.) Where up is the mean fluid velocity over the tube cross section and D is the tube diameter. In a fully developed flow, the critical Reynolds number corresponding to the onset of turbulence is Rep, 2300 (8.2) although mach larger Reynolds numbers (Rep ~ 10,000) are needed to achieve fully turbulent conditions. The transition to turbulence is likely to begin in the developing boundary layer of the entrance region. For laminar flow (Rep = 2300), the hydrodynamic entry length may be obtained from an expression ofthe form [1] (53), 00s, ® Reynolds number and that, as a first approximation [2], os (8) <0 4) For the purposes of this text, we shall assume fully developed turbulent flow for (aD) > 10. B.1.2 ‘The Mean Velocity Because the velocity varies over the cross section and there is no well-defined free stream, itis necessary to work with a mean velocity «when dealing with internal flows. This velocity is defined such that, when multiplied by the fluid density p and the cross-sectional area of the tube A,, it provides the rate of mass flow through the tube. Hence t= Ply Ac 5)488 Chapter 8 = Internal Flow For steady, incompressible flow in a tube of uniform cross-sectional area, ‘are constants independent of x. From Equations 8.1 and 8.5 itis evident flow in a circular tube (A, = 2D"), the Reynolds number reduces to in Reo SD Since the mass flow rate may also be expressed as the integral of the mass (pu over the cross section tna Journ, it follows that, for incompressible flow ina circular tbe, ‘The foregoing expression may be used to determine uy at any axial locations knowledge of the velocity profile u(r) at that location, 8.1.3 Velocity Profile in the Fully Developed Region “The form of the velocity profile may readily be determined forthe Jaminarfow an incompressible, constant property fuid inthe fully developed region of ac {ar tube, An important feature of hydrodynamic conditions in the lly devel region is that both the radial velocity component v and the gradient of the velocity component (ulax) are everywhere zero, er) Hence the axial velocity component depends only on 7, u(t, ) = u(r). ‘The radial dependence of the axial velocity may be obtained by solving the appropriate form of the x-momentum equation, This form is determined by fist 1 ‘ognizing that, for the conditions of Equation 8.9, the net momentum ux is ev where zero in the fully developed region. Hence the momentum conservation requirement reduces to a simple balance between shear and pressure forces in the flow. For the annular differential element of Figure 8.2. this force balance may be expressed as (neds) ~ [- Qnrds) + 4 (1, 2mrds)) as} + plorrdn ~ [r2sran +. ponrdn) x} ° which reduces to (8.10) ale — fon) = fon)ctional area, Vitis evident duces to tegral of the mass, “dnaedr axial location x Region ihe laminar ow mined by first rec- tum Rux is every= tum conservation sute forces in the: ve balance may be a} 0 8.1m Hydrodynamic Considerations 489 We 8.2 Force balance ona differential element for laminar fully develope flow ina cireular tube. With y = r, ~ r, Newton's law of viscosity, Equation 65.10, assumes the form eu) ‘anid Equation 8.10 becomes 1) Since the axial pressure gradient is independent of r, Equation 8.12 may be solved by integrating twice ta obtain Cine Cy ‘The integration constants may be determined by invoking the boundary conditions ur) =O and ar Which, respectively, impose the requirements of zero slip at the tube suyface and radial symmetsy about the centerline. Its a simple matter to evaluate the constants, and it follows that 1 (4) ,2)) (2) xe) if (=) ee Hence the fully developed velocity profile is parabolic. Note that the pressure gra- dient must always be negative. ‘The foregoing result may be used to determine the mean velocity of the flow. Substituting Equation 8,13 into Equation 8.8 and integrating, we obtain u(r) (14)490 & & Internal Flow Chapt Substituting this result into Equation 8.13, the velocity profile is then 6. Since «, can be computed from knowledge of the mass flow rate, Equation 8.14 be used to determine the pressure gradient. 8.1.4 Pressure Gradient and Friction Factor in Fully Developed Flow The engineer is frequently interested in the pressure drop needed to susan Jernal flow because this parameter determines pump or fan power re ‘ments. To determine the pressure drop, itis convenient to work with the Me (ot Darcy) friction factor, which is a dimensionless parameter defined as ~(dplds)D und “This quantity isnot to be confused withthe friction coefficient, sometimes the Fanning fiction factor, which is defined as Oa wduldr),.,.it follows from Equation 8.13 that eek au Since 4, \CURE 8.3 Frito Substituting Equations 8.1 and 8.14 into 8.16, it follows that, for fully oped laminar flow, 64 Rep For fully developed turbulent flow, the analysis is much more complicated ad we must ultimately rely on experimental results. Friction factors for @ wide Reynolds number range are preserved in the Moody diagram of Figure 8.3. In addi tion to depending on the Reynolds number, the friction factor is a function ofthe tube surface condition. It is # minimum for smooth surfaces and increases with Increasing surface roughness, e. Correlations that reasonably approximate th smooth surface condition are ofthe form f= 0.316Re5 Rey =2X 10 f= O.184Res" — Rep= 2X 10" Thermal Cons Alternatively, a single correlation that encompasses a large Reynolds number range bas been developed by Petukhov [4] and is of the form. f= (0790InRey— 164? — 3000S Rep=5xI0 (82H)s then te, Equation 8.14 eeded to sustain. fan power reqi jork with the Me r defined as Bl , sometimes call er eu that, for fully devel re complicated, and factors for a wide Figure 8.3. In addi isa function of the and increases. with, ly approximate the (8.20a) (8.206) olds number range sxi —(&21) B.2 = Thermal Considerations 491 ‘Transition i zoe uly rug zo Ht 1.05, 9.08 0.0 jimenial| all I | 02 ors 201, BBle YS aG8 ett routhess, oors| Drawn tubing anf. Gaston 009 p- Garma te Conerete 300-3000 S o00001 ong ELL 8.2 10? 2 2456 alo" 2 3456 810" ‘§ = 0.000.001 2 2456 810 2 3456 a10® afResesio® §=0000,008, Peynalds number, Rep = He WANE B.3 Friction factor for flly developed flow in asreular ube [3]. Used with permission, Note that f, hence dp/dx, is a constant in the fully developed region. From Equation 8.16 the pressure drop \p =p, — pz associated with fully developed flow {rom the axial position x, to.x; may then be expressed as ap= = f° a0= 198 f° a= Bon 39 where f is obtained from Figure 8.3 or from Equation 8.19 for laminar flow and from Equation 8.20 or 8.21 for turbulent fow in smooth tubes. The pump or fan power required to overcome the resistance to flow associated with this pressure drop may be expressed as (8.222) P= (ap (8.22) where the volumetric flow rate W may, in tum, be expressed as V= rifp for an incompressible fluid. Thermal Considerations Having reviewed the fluid mechanics of internal flow, we now consider thermal effects. If uid enters the tube of Figure 8.4 at a uniform temperature T(r, 0) that is492 Chapter 8 = Internal Flow sat Tero) fe 8.4 Thermal boundery layer development ina heated circular tube less than the surface temperature, convection heat transfer occurs and a boundary layer begins to develop. Moreover, if the tube surface condition is bby imposing either a uniform temperature (T, is constant) or a uniform heat Mux is constant), a thermally fully developed condition is eventually reached. The of the fully developed temperature profile Tr, x) differs according to whether a form surface temperature or heat flux is maintained. For both surface con however, tke amount by which fluid temperatures exceed the entrance tem increases with increasing x. For laminar low the stermal entry length may be expressed as (2) (§) sour ‘Comparing Equations 8.3 and 8.23, it is evident that, if Pr > 1. the hydrodyn boundary layer develops more rapidly than the thermal boundary ‘ayer (ray Ty); it decreases with xf the opposite is tue (T, < T,). 8.2.8 Fully Developed Conditions Since the existence of convection heat transfer between the surface and the fluid ictates thatthe fluid temperature must continue to change with x, one might legii- mately question whether fully developed thermal conditions can ever be reached. ‘The situation is censnly different from the hydrodynamic cas, for which (@uldx) = 0 in the fully developed region. In contrast, if there is heat transfer, (4Tq/dx), as well 35 (97/82) at any radius ris not zero. Accordingly, the temperature profile T) i494 Chapter 8 = Internal Flore continuously changing with s, and it would seem that a fully developed could never be reached. This apparent contradiction may be reconciled by w With a dimensionless form ofthe temperature. Analyses may be simplified by working with dimensionless temperature ences, as for transient conduction (Chapter 5) and the energy conservation. {Chapter 6). Introducing a dimensionless temperature difference ofthe form (F,~ (7, ~ Ty), conditions for which this ratio becomes independent of x are known ‘exist [2]. That is, although the temperature profile T(r) continues to change wit the relative shape of the profile no longer changes and the flow is said tobe ‘malty fully developed. The requirement for such a condition is formally states [Te ~Trs) B8| T(0)= Tad Jr cc where Tis the tube surface temperature, Tis the local fluid temperature, and T, the mean temperature ofthe fluid over the cross section of the tube, ‘The condition given by Equation 8.28 is eventually reached in a tube for wi there is either a uniform surface heat fux (is constant) ora uniform surface perature (7, is constant). These surface conditions arise in many engineering tp cations. For example, a constant surface heat flux would exist if te tube wal heated electrically or if the outer surface were uniformly irradiated. In conta, constant surface temperature would exist if a phase change (due to boiling or densation) were occurring at the outer surface. Note that itis impossible to sma neously impose the conditions of constant surface heat flux and constant temperature. If gis constant, T, must vary with x; conversely, if 7, is constant, must vary with x Several important features of thermally developed fiow may be infered Equation 8.28. Since the temperature ratio is independent of x, the derivative of ratio with respect to r must also be independent of x. Evaluating this derivative the tube surface (note that T, and T., are constants insofar as differentiation respect to ris concerned), we then obtain alt=), Substituting for 37/0r from Fourier’s law, which, from Figure 8.4, is ofthe form at a, and for q from Newton's law of cooling, Equation 8.27, we obtain A fe 6 Hence in the thermally fully developed flow of a fluid with constant properties, Local convection caefficien isa constant, independent of x Equation 8.28 is not satisfied in the entrance region, where h varies with x shown in Figure 8.5. Because the thermal boundary layer thickness is zero atthe {ube entrance, the convection coefficient is extremely large at x = 0. However, decays rapidly as the thermat boundary layer develops, until the constant val associated with fully developed conditions is reached,temperature dif sonservation equs of the form (T, ~ of x are kno es to change wit is said to be formally stated as @. perature, and Ty uniform surface ting this derivative differentiation wit 84,is of the form (8.29) want properties, the re varies with x, a8 jokness is zero at the at x= 0, However, h i the constant value 8.2 # Thermal Considerations 495 Fico 8.5, Axil variation ofthe convection heat tarsfer Mas = constant (833) in which case the value of @7/@x depends on the radial coordinate From the foregoing results, itis evident that the mean temperature is a very important variable for internal flows. To describe such flows, its variation with x ‘must be known, This variation may be obtained by applying an overall energy bal- ‘ance tothe flow, as will be shown in the next section, Exanete 8, For flow of aliquid metal through a circular tbe, the velocity and temperature pro files ata particular axial location may be approximated as being uniform and para- bolic, respectively. That is, u(r) = C, and Tr) ~ T, = Coll ~ (r/r,)!], where C, and Cy are constants. What is the value of the Nusselt number Nip at this location”496 Chapter 8 « Internal Flow Souunto: Known: Form of the velocity and temperature profiles at a particular axial tion for low in a circular tube. Find: Nusselt number at the prescribed location Schematic: Energy B 7 4 f woay a 1 tenpoaee otie WT, 6-2 Assumptions: Incompressible, constant property flow. Analysis; ‘The Nusselt number may be obtained by first determining thee tion coefficient, which, from Equation 8,27, is given as From Equation 8.26, the mean temperature is - amar ol emel (2) or, since uy, = C) ftom Equation 8.8, malifeeh HenceB.S & The Energy Balance 497 and _ kD _ (kr) X2r, Nuty = = 4 particular axial Energy Balance J me 8.3.1 General Considerations Because the flow in a tube is completely enclosed, an energy balance may be applied to determine how the mean temperature T,(x) varies with position along the tube and how the total convection heat transfer any is related to the difference in 8) temperatures atthe tube inlet and outlet. Consider the tube flow of Figure 8.6. Fluid moves at a constant flow rate “i, and convection heat «ransfer occurs at the inner surface. Typically, it will be reasonable to make one ofthe four assumptions in Sec- ‘on 13 that leads to the simplified steady-flow thermal energy equation, Equation Le. For example, it is olten the case that viscous dissipation is negligible (see Problem 8.10) and that the fluid can be modeled as either an incompressible liquid or an ideal gas with negligible pressure variation. In addition, it is usually reason- able to neglect heat transfer by conduction in the axial direction, so the heat transfer term in Equation 1.1 e includes only gion. Therefore, Equation 1.1} may be writ- ten in the form Trae ec = MELT — (8.34) fora tube of finite length. This simple overall energy balance relates three important thermal variables (Gene Tye Pj) His 4 general expression that applies irrespee- tive ofthe nature ofthe surface thermal or tube flow conditions ‘Applying Equation 1.1e to the differentia) control volume of Figure 8.6 and recalling thatthe mean temperature is defined stich that ney represents the tre rate of thermal energy (or enthalpy) advection integrated over the cross-section, We obtain so, = He (Tg + AT) ~ Te] 8.35)498, Chapter 8 = Internal Flow ane tly Equation 8.36 may be cast in a convenient form by expressing the rate of section heat transfer to the differential element a8 dai, = 4P dx, where Pi surface perimeter (P = xD for a circular tbe). Substituting from Equation follows that fm _aP de inc, Ewr,- 1) ‘This expression is an extremely useful result, from which the axial variation ray be determined, 107, > Ty, heat is transferred tothe fui and T, increases if T, < T,» the opposite is true. ‘The manner in which quantities on the right-hand side of Equation 837 with x should be noted. Although P may vary with x, most commonly itisa stant (a tube of constant cross-sectional atea), Hence the quantity (Plin) isa stant. Inthe fully developed region, the convection coefficient fi 20 although it varies with x in the entrance reginn (Figure 8.5). Finally, aloo may be constant, T, must always vary with x (except for the trivial case of no| teansfer, T; = T,)- ‘The solution to Equation 8.37 for 7.x) depends on the surface thermal tion, Recall that the two special cases of interest are constant surface heat fix constant surface temperature. I's common to find one of these conditions exis toa reasonable approximation. 8.3. Constant Surface Heat Flux For constant surface heat flux we fist note that i isa simple matter to determine total heat transfer rate dan Since i independent of x, i follows that esas = 94P *L) ‘This expression could be used with Equation 8.34 to determine the fluid ture change, Tro ~ Tr For constant q° italso follows that the middle expression in Equation 8.37% constant independent of x. Hence dy gh dx icy Ff Integrating from x = 0, it follows that Ts) = Tus + Ex gf = constant Accordingly, the mean temperature varies linearly with x along the tube ( 8.Yay, Moreover, rom Equation 8.27 and Figure 85 we als expect the em ditference (7, ~ 7) 10 vay with xa shown in Figare 87a, This ference is ialpressing the rate giP dt, where P g from Equation he axial variation and T, increases of Equation 8.37 commonky it is a antity (Phin) is a surface heat flux se conditions in Equation 8.37 is 83 ng the tube (Figure spect the temperature itference is initially 8.3 © The Energy Balance 499 ‘n-I Hl ro Tf) ¢=emstont & o Fiore 8.7 Avia tempensiore variations for heat eansfer ina tube, (a) Constant surface heat Aus (6) Constant surface temperate small (due to the large value of h sear the entrance) bat inreases with increasing x die tothe decrease in that occurs asthe boundary layer develops. However, in the fully developed region we know that f is independent of x. Hence from Equation 8.27 it follows that (7, ~ T,) mus also be independent of «inthis region. Te should be noted that if the heat fx snot constant but instead, a known function of x, Esuation 8.37 may sill be ineprated to btn the variation Of the incan temperature with x. Sil, the ttl heat rate may be obtained fom the requirement that gan = S$ait0)P de EXAMPLE 8. A system for heating water from an infet temperature of T,, = 20°C to an outlet temperature of T. = 60°C invalves passing the water through a thick-walled tube having inner and outer diameters of 20 and 40 mm. The outer surface ofthe tube is wel insulated, and electrical heating within the wall proves fr a uniform genera: tion rate of 4 = 10° Wan. A. For a water mass flow rate of = 0.1 kg/s, how long must the tube be to achieve the desired outlet teniperature? 2 Ifthe inner surface temperatare of the tube is T, = 70°C at the outlet, what is the local convection heat transfer coetficien a the outlet? SOLUTION Known: Ioteral fow through thick-walled tube having uniform heat generation, Find: 1, Length of tube needed to achieve the desired outlet temperature. 2. Local convection coefficient at the outlet.500 Chapter 8 = Internal Flow Schematic: Assumptions: 1, Steady-state conditions 2. Uniform heat flux 3. Incompressible liquid and negligible viscous dissipation. 4, Constant properties. 5. Adiabatic outer tube surface Properties: Table A‘, water, 13K): 6, = 4179 Ske -K. Analysis: 1. Since the outer surface of the tube is adiabatic, che rate at which energy is erated within the tube wall must equal the rate at which it is convected to water. With ai DIL it follows from Equation 8.34 that ary (Dh = DDL = tne Tay ~ Tos) or 4x Out kets x 4179 Uke °K — —— (60 = 20)°C = {O08 — 0025) mx 10° Win? 79” 177m 2. From Newton's law of cooling, Equation 8.27, the local convection coeffi atthe tube exit is8.8 © The Energy Balance 501 ‘Assuming that uniform heat generation in the wall provides a constant surface heat fax, with of aD La a= wine os 15 X 10" Win? it follows that = ALM 9 i -K < Comments: 1, If conditions are fully developed over the entire tube, the local convection coef: ficient and the temperature difference (T, ~ T,) are independent of x. Hence +h = 1500 Wim? K and (T, ~ T,) = 10°C over the entire tube. The inner sur- face temperature atthe tube inlet is then T,, = 30°C, 2. The required tube length L could have been computed by applying the expres- i. sion for T,(1), Equation 8.40, atx = L. which energy is it is convected to 8.3.3 Constant Surface Temperature Results for the total heat transfer rate sn the axial distribution of the mean temper- ature are entitely different for the constant surface temperature condition. Defining AT 8 T, = Ty, Equation 8.37 may be expressed as ay _ MAT) _P dd key har Separating variables and integrating from the tube inlet to the outlet, JOSS ae he pdt PL Afi ) ‘xr 2b a From the definition of the average convection heat transfer coefficient, Equation 6.13, it follows that 171m q nvection coefficient T,= constant (B4la)502 Chapter 8 Internal low where Fi, o simply fi isthe average value off for the entire tube. Reasranging, constant Had we integrated from the tube inlet to some axial position x within the tubs would have obtained the similar, but more general, resule chat =TeAX)_ (Pez _ 7 FeReaeo( fa) ico where Ji now the average value off from the tube inlet tox. This result thatthe temperature difference (T,~ T,) decays exponentially with distance the tube avs. The axa surface and mean temperature distributions are te shown in Figure 8.76. Determination of an expression for the total heat transfer rate gion is 6 coated by the exponential nature ofthe temperature decay. Expressing Equation inthe form Gea = HET, ~ Tog) ~ (ET) and substituting for fic, from Equation 8.412, we obtain FATy —T, = constant Gea here A, is the tube surface area (A, = P +L) and ATiq isthe log mean teng difference, ang= Bhat we In(QTJOT) Equation 8.43 isa form of Newton's law of cooling forthe entre tube, and Ty the appropiate average ofthe temperature difference over the tube length. Tek atithmie nature of his average temperature difference fin contrat, to anal metic mean temperature difference of the form AT, = (AT, + AT,)/2] is due to exponential nature of the temperature decay Before concluding this section, its important 19 aot tha, in many ep tions, it is the temperature of an external fluid, rather than the tube surface ter tore, that is xed (Figure 88) Jn such cases, iis eal shown that theres this section may stil be used if 7, is replaced by (the fre stream tempera the extemal fund) and Fis replaced by U (dhe average overal et transfer cu cent. For such cases, it follows that andtox. This result ally with distance butions are there! fer rate deans 18 sing Equation 7, a7.) entire ube, and ATi rue length. The I ste, 10 an a 5 + AT] is due to that, in many api be ube surface tem own thatthe results stream temperature Il heat transfer coet 8.3 © The Energy Balance 503 Fict me 8.8 Heat transfer between fuid owing overa tube ‘nd fui passing through the tube ‘The overal eat transfer coefcint is defined in Section 3.3.1, nd for this apple tin it would include contributions de to convection atthe tube nner an cates ss faces Fr thick-walled tbe of small thermal conductivity, it would as the effect of condiction across the tube wal, Note thatthe product UA, 3 the same esl iespective of whether iis defied in ters ofthe iner (0/4, or outer (U,A,,,) Surface areas of the tube (sce Equation 3.32). Also note that (UA,)"' is equivalent to the total thermal resistance between the two fluids, in sshich ease Equations 85a and 6a may be expressed as (8.450) and 4q a (8.46b) A commion variation of the foregoing conditions is.one for which the uniform temperature of an outer surface, 7,,, rather than the free stream temperature OF an external fuid, To, is known, In the foregoing equations, 7. is then replaced by T+ and the total resistance embodies the convection resistance associated with the internal flow, as well as the total resistance due to conduction between the inner sur- face of the tube and the surface comesponding 10 T,, | Exanpie 8.3 100°C, Water flows through the tube at a rate of r = 0.25 kg/s, and its inlet and504 Chapter 8. Internal Flow SOLUTION Known: Flow rate and inlet and outlet temperatures of water flowing tube of prescribed dimensions and surface temperature. Find: Average convection heat transfer coefficient, Schematic: b= 50mm Assumptions: 1. Negligible ube wall conduction resistance 2. Incompressible liquid and negligible viscous dissipation. 3. Constant properties Properties: Table A.6, water (36°C) ¢, 178 Uke K. Analysis: Combining the energy balance, Equation 8.34, with the rate eq Equation 8.43, the average convection coefficient is given by ftp go ~ Tad DL AT From Equation 8.44 T= Tae) ~~) Me Bg. — TMT, To (10051) = (100= 15) Tin = jafO— S7¥100 = 15] ~ VE Hence j= 025kals AITR g K (57 =15°C wx O0Smx6m —-61.6C 155 Wim" K Comments: If conditions were fully developed over the entre tube, the hel convection coefficient would be everywhere equal to 755 Win?» K. —_—with the rate equat 6c yc entire tube, the local IK. 84 « Laminar Flow in Cirenlar Tubes 505 inar Flow in Circular Tubes: Thermal sis and Convection Correlations ‘To use many of the foregoing results, the convection coefficients must be known. In this section we outline the manner in which such coefficients may be obtained theo- retically for laminar flow in a circular tube. In subsequent sections we consider empirical correlations pertinent to turbulent flow in a circular tube, as well as to flows in tubes of noncircular eross section. BA.L The Fully Developed Region Here, the problem of heat transfer in laminar flow of an incompressible, constant property fluid in the fully developed region of a circular tube is treated theoreti- cally. The resulting temperature distribution is used to determine the convection coefficient. ‘A differential equation governing she temperature distribution is determined by applying the simplified, steady-flow, thermal energy equation, Equation 1.11e Lg = tne, (Tg — Tq) (0 the annular differential element of Figure 8.9. If we neglect the effects of net axial conduction, the heat input, q is due only to conduction through the radial surfaces. Since the radial velocity is zero inthe fully developed region, there is no advection of thermal energy through the radial control surfaces, and the only advection isin the axial direction. Thus, Equation |.t1e leads to Equation 8.47, which expresses a balance between radial conduction and axial advection: tna in| (1+ ha) 1] war) (dine, Gear (8.47) eee ne : A re ae ae — Le = a : — aL. Picqut 8.9 Thermal energy halance ona ferential element for laminar, ally developed low ina cirwular tube,506 Chapter 8 Internal Flow ‘The differential mass flow rate in the axial direction is an radial heat transfer rate is g, = ~A@T/Ar)2mrdx, If we assume constant Equation 8.47b becosmes af av) aA" or ‘We will now proceed to solve for the temperature distribution for the constant surface heat flr. In this ease, the assumption of negligible net axial duction is exactly satisfied, that is, (@°T7@x*) = 0. Substituting for the axial ture gradient from Equation 8.32 and for the axial velocity component, u, ‘Equation 8.15, the energy equation, Equation 8.48, reduces 2 Laf,or)_2u(are\[, fr) on lat) e(B)-(E)] eee where Ty(x) varies linearly with x and (2ug/a)dTq/de) is a constant. Se variables and integrating twice, we obtain an expression for the radial temp distribution: ax ox integration may be evaluated by applying appropriate conditions. From the requirement thatthe temperature remain finite at r = 0 it lows that C, = 0. From the requirement that 7%r,) = T., where T, varies with also follows that iin (dT) (313 C= Th) (@) (%) Accordingly, forthe fully developed region with constant surface heat flux, the perature profile is of the form ro-nin E(E Lier elt) He) From knowledge of the temperature profile, all other thermal parameters be determined. For example, if the velocity and temperature profiles, Eau 8,15 and 8.50, respectively, are substituted into Equation 8.26 and the integ over ris performed, the mean temperature is found to be 74) = 14) (e) as pu,(7rD"/4), we then obtain From Equation 8.39, where P= 2D and sit _~ugd a T0) ~ T= ~ 39 ‘Combining Newton's law of cooling, Equation 8.27, and Equation 8.52, i flows a()B.A = Laminar Plove in Circular Tubes 507 in = pucdardr, me constant prof AD, Mug = 9P- 36 gf = constant (633) Hence in a circular tube characterized by uniform surface heat flue and laminar, filly developed conditions the Nusselt number isa constant, independent of Rep, Pr, and axial location. or laminar, fully developed conditions with a constant surface temperanure, the assumption of negligible axial conduction is often reasonable, Substituting for te velocity profile from Equation 8.15 and for the axial temperature gradient from Equation 8.33, the energy equation becomes at) tt 7 \"or |e AA solution t this equation may be obtained by a0 iterative procedure, which involves making successive approximations to the temperature profile. The result- ing profile is not described by a simple algebraic expression, but the resulting bution for the eligible net axial g forthe axial y component, 1, : constant ca im Nisulenunber my be tow 08 Nup= 3.66 T,=constant (8.55) eo propriate tou Noe that in wing Equation £53 o 8.5 © determine the ermal ondutiviy 7, varies with should be evaluated aT Eanes 8.4 sie: Bex Cnr concep ed for soar nergy clleton inves placing a be the ol : pol ayaa efecrand ass lth ke i Insulation AG | 6 imal parameters ; profiles, Equatic and the integra ‘The net effect ofthis arrangement may be approximated as one of creating a condi- tion of uniform heating at the surface of the tube. That is, the resulting heat Mux to the fluid q may be assumed to be a constant along the circumference and axis ofthe tube. Consider operation with tube of diameter D = 60mm on a sunny day for which g! = 2000 Wim? 1. If pressurized water enters the tube at r= 001 kg/s and T,,, = 20°C, what tube length L is required to obtain an exit temperature of 80°C? 2. What is the surface temperature atthe outlet ofthe tube, where fully developed conditions may be assumed (0 exist? then obtain 8.9, it follows thatChapter 8 # Internal Flow SOLUTION Known: Internal flow with uniform sorface heat fux, Find: 1, Length of tube L to achieve required heating 2. Surface temperature 7,(L) at the outlet section, x = L. Schematics vor Perret een Tyy= 2005 be ———4 Assumptions 1. Steady-state conditions. 2. Incompressible liquid and negligible viscous dissipation, +3. Constant properties 4, Fally developed conditions at tube outlet Properties: Table A6, water T= 323 K) 6, = 4181 Ukg-K. Table A6, (Ty = 353K): k = 0.670 Wim > K, w= 352 X LON = sim’, Pr = 2.2. Analysis: |. For constant surface heat flux, Equation 8,38 may be used with the energy ance, Equation 8.34, to obtain nt Toag ~ Tas) Hence G01 kes X4181IMEK 59 ao = 665m 4 0.060 m 2000 Wii 2. The surface temperature at the outlet may be obtained from Newton's law of cooling, Equation 8.27, whereK, Table A.6, 22. 665m 4 re Newton's law of BA # Laminar Flow in Circular Tubes 509 To find the local convection coefficient at the tube outlet, the nature ofthe flow ‘condition must first be established, From Equation 8.6, ‘nie 4X 0001 kes Di. 7% 0,060 m X 352 X 10 N+ sh Hence the flow is Iaminar. With the assumption of fully developed conditions, the appropriate heat transfer correlation is then Rey = 03 D Mp = #2 = 4.36 and k= 4460670 Wim: K 2 n= 4364 = 4.36970 Bim 8.7 Wim? K. “The surface temperature at the tube outlet is then 2000 Wim? + 80°C = 4 = FET Win? -K* APE PEC Comments: For tne conditions given, (xy/D) = 0.0SRepFr = 66.3, while L/D = 110, Hence the assumption of fully developed conditions i justified, Note, however, that with 7, > 100°C, boiling may occur atthe tube surface. —_ L Examen 8.5 |In the human body, blood flows from the heart in a series of branching blood ves- sels having successively smaller diameters. The capillaries are the smallest blood vessels. In developing tke bioheat equation (Section 3.7), Pennes assumed that blood enters the capillaries atthe arterial temperature and exits atthe temperature of the surrounding tissue. This problem tests that assumption [5,6]. The diameters and average blood velocities for three different types of blood vessels are given in the table below. For each of these blood vessels estimate the length required for the mean blood temperature to closely approach the tissue temperature, specifically, to satisfy (T, ~ TyJ(T, ~ Ta.) = 0.05. The heat transfer between the vessel wall and surrounding tissue can be approximated by an effective heat transfer coefficient, 1h, = K/D, where k, = 0.5 Wim K. Vessel Diameter, D (mm) Blood Velocity, (mvs) Large aren) 3 130 Aneriole 0.02 3 Capillary 008 o7 SoLuTION. Known: Blood vessel diameter and average blood velocity. Tissue thermal con- ductivity and effective heat transfer coefficient.510 Chapter 8 = Internat Flow Find: Length of blood vessel needed to satisfy (T, = Ty (T, = Ta) = 0 Schematic: eat Tay Tisoe nt Assumptions: 1. Steady-state conditions 2. Constant properties. 3. Negligible blood vessel wall thermal resistance, 4. Thermal properties of blood can be approximated by those of water, ‘5. Blood is incompressible liquid with negligible viscous dissipation. 16, Tissue temperature is fixed, 7. Effects of pulsation of flow are negligible. Properties: Table A.6, water Fy = 310K) p = 97 = 993 kw’, e, =a ks" K, = 695 % 10° N+ sin’, £ = 0.628 Wien K, Pr 4.02. Analysis: Since the tissue temperature is fixed and heat transfer between blood vessel wall and the tissue can be represented by an effective heat ta coefficient, Equation 8.45: is applicable, with the “tree stream” temperature to the tissue temperature, T, This equation can be used to find the requied length since A, = DL. However, we must first find U, which requires knowledge of heat transfer coefficient for the blood flow, /ty. Taking the large artery as an ex ple, the Reynolds number is given by 993 kg/mm? x 130 X 10? m/s X 3 107m Re, 695 10 N= sin® 587 PgD aa 0 the flow is laminar. Since the other blood vessels have smaller diameters velocities, their lows wall also be laminar. Because we don't yet know the lengh the vessel, we don’t know whether the flow becomes fully developed. However, me will begin by assuming fully developed conditions. Moteover, because the situation is neither one of constant surface temperature nor constant surface heat flux, we wl estimate the required length by approximating the Nusselt number as Nu = 4, in which case h, = 4ki/D. Neglecting the thermal resistance of the vessel wall, forthe large artery D,D___ 3x10 m 3x 107 m ah” T0638 Win KR 3 Wim K = 7.2% 107m KIW= Ta) 1, e of water. pation, 93 kg/m’, cp 462, an transfer between fective heat tra temperature eqt the required length, res knowledge of, eatery as an € om = 957 ft know the length rloped. However, , because the situation ce heat ux, we wa umber as Nu = 4, i vessel wall, for the st siameters a X10" m Swim K BA # Laminar Plow in Circular Tubes SLL o 40 Wim! + K, ‘The length can then be found by solving Equation 8.45a, with r= pilyD/4: 130 x 10°? m/s 3.x 10°? mx 4178 Jike + K 4140 Wim? = K |n(0.05) We can now test the assumption that the flow is hydrodynamically and thermally fully developed, using Equations 8.3 and 8.23: Xan = 0.05Re,,D = 0.05 X 557 X 3X 107m = 0.08 m 0.08 m x 4.62 Xiao = usPr 4m ‘The flow would indeed be fully developed well within the length of 8.7 m. The cal- culations can be repeated for the other two cases and are tabulated below. Vessel Rey U (Wim) Lim) Fay tim) Tuy (tm) Lamgeanery $57 140 a7 0.08 08 Ancriole 0.086 21,000 Box 10% 9x10 4x 107 Capillary 0.0080 $2,000 33107 3x10 x 107 Comments: 1. The blood temperature in the large artery approaches the tissue temperature very slowly. This is due to its relatively large diameter, which leads to a small, overall heat transfer coefficient. Thus, the temperature in large arteries remains close to the inlet arterial blood temperature. In arterioles, the blood temperature comes close to the tissue temperature within a length on the order of 10 um, Since arterioles are on the onder of millimeters in length, the blood temperature exiting them would be essentially the same as the tissue temperature, There would then be no further temperature drop in the capil- laries, which are the next smallest vessel, Thus, itis inthe arterioles and slightly larger vessels in which the blood temperature equilibrates to the tissue tempera ture, not in the capillaries as Pennes described. Despite this flaw, the bioheat ‘equation has proved a useful tool in analyzing heat transfer inthe human body. 3. The properties of blood are moderately close to those of water. The property which differs most is the viscosity, as blood is more viscous than water, How= ‘ever, that change would have no effect on the foregoing calculations. Since the Reynolds number would be even smaller, the flow would still be laminar and the heat transfer would be unaffected. 4, Individual blood cells have dimensions on the order of the capillary diameter. ‘Thus, forthe capillaries, an accurate model of blood flow would account for the individual cells surrounded by plasma512 Chapter 8 «Internal Flow 8.4.2 The Entry Region “The energy equation for the entry region is more complicated than Equation because there would be a radial advection term (Since w # 0 in the enty In addition, both velocity and temperature now depend on x, as well as rand axial temperature gradient 77/0x may no longer be simplified through Ea 8.32 or 8.33, However, two different entry length solutions have been ob “The simplest solution is forthe shermal entry length problem, and itis based assuming that thermal conditions develop in tbe presence of a fully deve velocity profile. Such a situation would exist ifthe location at which heat begins were preceded by an unlcated starting length. Kt could also be asu 2 reasonable approximation for large Prandtl number fluids, such as vis in the absence of an unheated starting length, velocity boundary layer dev ment would occur far more rapidly than thetmel boundary layer development large Prandtl number fuids, and a thermal entry length approximation cos ‘made. In contrast, the combined (thermal and velocity) entry length problem responds to the case for which the temperature and velocity profiles de simultaneously. Solutions have been obtained for both entry length conditions (2), and results ae shown in Figure 8.10. As evident in Figure 8.10a, local Nusselt ou ‘Nu, ae, in principe, infrte at x = 0 and decay to their asymptotic lly de oped) values with increasing x. When plotied against the dimensionless par 2 109 fees Teal ery gh J contin ery res «o| ech T 5 anaes val emu B57 | va 3 5 A 2, sxe 10 66 4} equation 836 2 ‘ean | cea an 1 1 door Toor tar os 4 oor Or os 0008 oo8 moar Vat " ® FiciRe 8.10 Results obtained from entry length soltions for lmsinar flow in a ‘ube: (a) local Nusselt numbers and () average Nusselt numbers 2)which heat tram also be assur such as oils. dary layer devel yer development oximation coul length problem ity profiles devel ons 2], and selecte al Nusselt numb mptotic (Fully BA # Laminar Flow in Circular Tubes 513 xailu,D") = x(DRep Pr, which is the reciprocal of the Graetz number, Gey = (DixyRep Pr, the manner in which Nup varies with Gzp"' is independent of Pr for the thermst entry problem, since the fully developed velocity profile, given by Equation 8.13, is independent of the Muid viscosity. In contrast, for the eombined entry length problem, results depend on the manner in which the velocity distribu- tion develops, which is highly sensitive to the fluid viscosity. Hence, heat transfer results depend on the Prandtl number for the combined entry length case and are presented in Figure 8.10a for Pr = 0.7, which is representative of most gases. At any location within the entry region, Nu, decreases with increasing Pr and approaches the thermal entry length condition as Pr — =, Note that fully developed ‘conditions are reached for [C/D)/Rep Pr] = 0.05. For the constant surface temperature condition, itis desirable to know the average convection coefficient for use with Equation 8.43. Kays [7] presents a cor- relation attributed to Hausen [8], which is of the form. 0.0668(DIL)RepPr Ne 360 TO oat DID Ren PrT™ (8.56) ‘thermal entrance length or ‘combined entrance length with Pr= 5 where Rip = RIK, Because this result is forthe thermal entry fength problem, itis applicable to all situations where the velocity profile is already fully developed For the combined entry length, a suitable correlation for use at moderate Prandtl numbers, due to Sieder and Tate [9 is ofthe form Mip=1 wo Reale) "(a ws? 060< Prss e(Lle< aie (2) 974 Equation 8.57 is recommended for use when 0.60 = Pr < 5, provided Nuip = 3.66 [2, 10) IF Nu falls below this value, i is easonable to use My = 3.66, since fully developed conditions then encompass much ofthe tube. For lager Prandt! numbers (Pr = 5), hydrodynamic conditions develop much faster than thermal conditions, and Equation 8.56 is recommended instead of Equation 8.57 [2], Equations 8.56 and 8.57, along with numerical predictions of Nip versus Vof(D Rep Pr) and Pr, are shown in Figure 8.106. All properties appearing in Equations 8.56 and 8.57, except yz, should be evaluated atthe average value ofthe mean temperature, Ty = (Ty, +Tru)2 ‘The subject of laminar flow in ducts has been studied extensively, and numer- ‘ous results are available fora variety of duct ross sections and surface conditions. ‘These results have been compiled in a monograph by Shah and London [11] and in an updated review by Shab and Ithatt (12514 Chapter 8 w Internal Flow 8.5 Convection Correlations: Turbulent Flow in Circular Tubes Since the analysis of turbulent flow conditions is a good deal more involved, ‘emphasis is placed on determining empirical correlations. A classical express ‘computing the local Nusselt number for fully developed (hydrodynamically thermally) turbulent flow in & smooth circular tube is due to Colburn {131 and bbe obtained from the Chilton-Colburn analogy. Substituting Equation 6.70 Equation 8.18, the analogy is of the form Mio pa Rey Pr!” « Substituting forthe fri factor from Equation 8.21, the Colburn equations Nap = 0.023Re$9 Pr'® ‘The Dittus-Boelter equation (14] isa slightly different and preferred version of above result and is of the form’ Nitp = 0.023Re% Pr” where m = 0.4 for heating (T, > T,) and 0.3 for cooling (F< T,). These €q have been confirmed experimentally for the range of conditions 0.6 < Pr 160] Re = 10,000 koi D ‘The equations may be used for small to moderate temperature differences, T, ~T with al properties evaluated at T,. For flows characterized by large property vt tions, the foliowing equation, du to Sieder and Tate [9], is recommended: = wus pn ( # vs =wome' (2) 0.75 Pr= 16,700] Rey = 10,000 | L fr10 "although it has become common practice 10 refer to Equation 860 as the Dinu!-Rel the original Dittus-Boelter equations are actually of the form 0283 Res* Pe (Heating) 10265 Ref! PP? (Cooling) “The hsterical origin of Equation 8.60 are discussed by Winterton (14)re involved, 25 the Ditus-Boelier 8.5 # Convection Correlations: Turbulent Flow in Cieeular Tubes 315 where all properties except, ate evaluated at Ty. To a good approxination the foregoing correlations may be applied for both the uniform surface temperature and heat fas conditions Although Equations 8,60 and 8.61 are easily applied and are certainly satis factory for the purposes of this text, errors as large as 25% may result from their use, Such errs may be reduced to less than 10% through te use of more recent, but generally more complex, correlations [4, 15]. One correlation, valid over a large Reynolds number range including the transition region, is provided by ‘Gnielinski [16] (fP8)(Rep ~ 1000)Pr 1+ 12.7(718)%(Pr** = 1) up = (8.62) where the friction factor may be obtained from the Moody diagram, or, for smooth tubes, from Equation 8.21. The correlation is valid for 0.5 = Pr = 2000 and 3000 = Rey 5 X 108. In using Equation 8.62, which applies for both uniform surface heat ‘ox and temperature, properties should be evaluated at 7, If temperature differ- ences are large, additional consideration must be given to variable-property effects and available options are reviewed by Kakac [17] ‘We note that, unless specifically developed for the transition region (2300 < ‘Rey < 10"), caution should be exercised when applying a turbulent flow correlation for Rep = 10%, If the comelation was developed for fully turbulent conditions (Rep > 10}, it may be used as a frst approximation at smaller Reynolds numbers, with the understanding that the convection coefficient witt be overpredicted. If a higher level of accuracy is desired, the Gnielinski coreation, Equation 8.62, may bbe used. A comprehensive discussion of heat transfer in the transition region is pro- vided by Ghajar and Tam [18 We also note that Equations 8.59 through 8.62 pertain to smooth tubes. For turbulent flow, the heat transfer coefficient increases with wall roughness, and, as a first approximation, it may be computed by using Equation 8.62 with friction factors obtained from the Moody diagram, Figure 8.3. However, although the general trend is one of increasing / with increasing f, the increase in fis pro- portionately larger, and when fis approximately four cms larger than the corre- sponding value for 9 smooth surface, fh no longer changes with additional {creases in f (19). Procedures for estimating the effect of wall roughness on con- vection heat transfer in fully developed turbulent flow are discussed by Bhatti and Shah [15]. . Since entry lengths for turbulent flow are typically short, 10 = (ty/D) = 60, it is often reasonable to assume thatthe average Nusselt number forthe entire tube is ‘equal to the value associated withthe fully developed region, Nip ~ Mp, However, for shor tubes Ni will exceed Nip,g and may be calculated from an expression of the form Nip 4 ¢ Nupa (Dy (8.63) where Cand m depend on the nature of the inlet (e.g., sharp-edged or nozzle) and entry region (thermal or combined), as well as on the Prandtl and Reynolds num- bers (2, 15, 20]. Typically, errors of less than 15% are associated with assumingChapter 8 Internal Flow ‘ig = Moy for (L/D) > 60, When determining Ni, al fui properties h evaluated at the acthmetic average ofthe mean temperature, Ty = (Ty, + Ty Finally, we note thatthe foregoing correlations do not apply’ © guid (3% O's Pr = 5 X 107), For fully developed turbulent fiow in smooth tubes with constant surface heat fu, Skupinski et al (21] recommend 2c ‘ofthe form Nity = 182 + O.0185PeBh [ies 10° < Rey <9.05 x °] 10° 5 Pey = 10" Similarly, for constant surface temperature Seban and Shimazaki [22] reco the following correlation for Pe = 100: Mip = 50+ 0,025 Pel —T, = constant Extensive data and additional correlations are available in the literature [23] Exampty 8.6 Hoot air flows with a mass rate of sit = 0.050 kg/s through an uninsulated s ‘metal duct of diameter D = 0.15 m, which is in the crawIspace of a house. The air enters at 103°C and, aera distance of L = 5 m, cools to 77°C. The heat coefficient between the duct outer surface and the ambient sir at T = 0°C is tobe h, = 6 Wim? K. 1, Calculate the heat loss (W) from the duct over the length L. 2, Determine the heat flux and the duct surface temperature at x = L, SOLUTION Known: Hot air flowing in a. duct. Find: A, Heat loss from the duct aver the length Z, q (W), 2, Heat flux and surface temperature at.x = L. Sehematie: oie ee [IN Asenption 1, Steady-state conditions. 2. Constant properties.ply t9 liquid terature [23]. of a house, The °C. The heat : uninsulated st 8.5 # Convection Correlations: Turbulent Flow i Circular Pubes 517 3. Ideal gas behavior 4. Negligible viscous dissipation and negligible pressure variations. 'S. Negligible duct wall thermal resistance 66. Uniform convection coefficient at outer surface of duct. Properties: Table A.4, ait (T= 363K): c, = 1010 kg +K. Table A, air (Ta = 350 K): k= 0.039 Wim K, p= 208.2 % 10-7 N= sim’, Pr = 0.70. Analysis: 1. From the energy balance forthe entire tube, Equation 8.34, = ttc, Tina, = Ta) = 0.05 ke/s X 1010 Jkg - K (77 ~103)°C = ~1313 W 4 2. An expression for the heat flux at x= may be inferred from the resistance network Ta Tilt ab A v-9 where h,(C) isthe inside convection heat transfer coefficient atx = L. Hence Tu ~ Te TAD Oh) The inside convection coefficient may be obtained from knowiedge of the Reynolds number. From Equation 8.6 is 4 0.05 kgs ‘nD X 0.13 m X 208.2 X 10 N-Sie (510.15) = 33.3, itis rea- L, Hence from Equation Ge) Reo 0.384 Hence the flow is turbulent. Moreover, with (L/D) sonable t0 assume fully developed conditions at 8.60, with n = 0.3, nwo uy = MEP 003 ft w= 0.02520388)(07099 = 519 1d) Map = 519 20K 6 Hence aoc [Q/11.6)+ (16.0)} m? = KW Referring back to the network, it also follows that gl) = 304.5 Wim? iq) Int To OTD in which case © GL) _ ayog, 3045 Wh = Toa ay Chg win?-K OTE “518 Chapter 8 w Internal Flow Comments: 1, In using the energy balance of part I forthe entire tube, properties (inthis only c,) ate evalusted at 7, = (Fao + Ta)/2. However, in using the tion for @ local heat transfer coefficient, Equation 8.60, properties are e at the local mean temperature, Ty = 77°C. 2. This problem is characterized neither by constant surface temperature ‘constant surface heat fux. It would therefore be erroneous to presume tha total heat loss from the tube is given by 4{(L)mDL = 717 W. This eel substantially less than the actual heat loss of 1313 W because f(x) dex with increasing x. This decrease in 4/(x) is due to reductions in both h(x) [T7,(2) ~ Ta] wits increasing x —_— 8.6 Convection Correlations: Noncircular Tubes and the Concentric Tube Annulus Although we have thus far restricted our consideration to internal flows ofc lar eross section, many engineering applications involve convection transport noncircular tubes. Atleast to fisst approximation, however, many ofthe cir tube results may be applied by using an effective diameter as the chara length. Its termed the hydraulic diameter and is defined as 4A, Pe ‘hese A. and P are the flow cross-sectional area and the weited perimeter tively. Its this diameter that should be used in calculating parameters such as and Nu. For turbulent flow, which still occurs i€ Rep = 2300, i is reasonable 0 us cortlations of Section 8.5 for Pr = 0,7. However, in a noncitcular tube the com tion coefficients vary around the periphery, approaching zero in the comers. Hence in using a circular tube correlation, the coefficient is presumed to be an serge over the perimeter. For laminar fiow, the use of circular tube correlations is less accurate, particu with evss sections characterized by sharp comers, For such cases the Nusselt numb corresponding to fully developed conditions may be obtained from Table 8.1, which is based on solutions of the differential momentum and energy equations for fo through the different duct cross sections. As forthe circular tube, results differ accont ing tothe surface thermal condition. Nusselt numbers tabulated for a uniform surie neat fax presume a constant ux in the axial (flow) direction, but a constant tempere ture around the perimeter at ay cross section. This condition is typical of highly cow: ductive tube wall materials. Results tabulated fora uniform surface temperature apply ‘when the temperature is constant in both the axial and peripheral directions. Although the foregoing procedures are generally satisfactory, exceptions do exist, Detailed discussions of heat transfer in noncircular tubes are provided in sev eral sources [11, 12, 24 cy equations for flow results differ accord- fora uniform surface {a constant tempera- ypica! of highly con- ce temperature apply directions. tory, exceptions do are provided in sev= 8.6 » Convection Correlations: Noncireular Tubes 519 ‘Tante 8.1 Nusselt numbers and friction factors for fully developed laminar flow in tubes of differing cross section ‘Cross Section 10 361 298 3 ° = 143 an 3.08 9 | 20 a 339 @ =a 30 a 396 © Brrr 40 333 aaa B ° mes 80 649 560 2 =e 823 154 9% tates === ~ 539 486 96 “ses A - 3 249 3 ‘Used wih permission from W, M. Kays and M. . Crawfoe, Convection Hea and Mass Tram. eed. MeGriw-Hil, New Yor, 1993. ‘Many internal flow problems involve heat transfer in a concentric mbe annulus (Figure 8.11). Fluid passes through the space (annulus) formed by the concentric tubes, and convection heat transfer may occur to oF from both the inner and outer tube surfaces. It is possible to independently specify the heat flux or temperature, Ficene B21 ‘The concent tbe anmls520 Chapter 8 « Internal Flow that is, the thermal condition, at each of these surfaces. In any case the hea fcom each surface may be computed with expressions ofthe form Gr WT To) ( = hoTag~ Tod t Note that separate convection coefficients are associated with the inner and surfaces, The corresponding Nusselt numbers are of the form For the case of fully developed laminar flow with one surface insulated ang other surface at a constant temperature, Nu, or Ni may be obtained from Table ‘Note that in sueh eases we would be interested only in the convection coeffi associated with the isothermal (nonadiabatic) surface. uniform heat flux conditions exist at both surfaces, the Nusselt numbers ‘be computed from expressions ofthe form Ny a= 8 TH Gilat ‘ Wy Nu, = — (8: giao ‘ ‘The influence coefficients (Nit, Nt... 6°, and @) appearing in these equations be obtained from Table 8.3. Note that q? and 4 may be postive or negang depending on whether heat wane is to et Grom the Mud, respectively. Moreove situations may arise for which the values off, and h, are negative. Such rea Tabi 8.2 Nusselt number for fully developed laminar flow in a circular tbe annulus with one surface insulated and the other at constant temperature DID, Nu, Ne Comments o = 3.66 Sce Equation 8.55, 0.05 1746 4.06 0.10 1156 4n 025 731 423 050 374 493 =100 4.86 486 See Table 81, bir —o (Used with permission fiom W. M, Kays and H.C. Parkin, i W-M. Robsenow and P. anne, 6, Muandbeok of Heat Transfer, Cha. 7, McGraw-Hill New York, 19728.7 «Heat Transfer Enhancement 521 TaBLe 8.3 Influence coefficients for fully developed Jaminer flow in a circular tube annulus with uniform heat flux maintained at both surfaces DID, Nay Nay, or or o = 4364) © o 0.05 rat 4792 28 0294 0.10 not 4839 1383 0.0562 020 8499 4.833 0.905 0.1081 90 6.583 4979 0.603 0.1823 0.60 5912 5.099 0473 02455 0.80 558 524 0401 0.299) 1.00 5.385 5.385 0346 0.346 ‘Use with permission from W. ML Kays and H.C, Peis, in WM. Robin and J.P. Hare, E8, Handbook of Hea Transfer, Chap 7, MGra¥-Hil, New York, 1972 ve insulated and ‘when used with the sign convention implicit in Equations 8.67 and 8.68, reveal the from Table 8, relative magnitudes of 7, and T. vection eoetfci For fully developed turbulent flow, the influence coefficients are a function of the Reynolds ang Prandtl! mumbers [24]. However, toa ist approximation the innec It numbers and outer convection coefficients may be assumed to be equal, and they may be evaluated by using the hydraulic diameser, Equation 8.71, with the Dittus-Boelter equation, Equation 8.60 87 Heat Transfer Enhancement Severs) options are available for enhancing heat transfer associated with intemal flows, Enhancement may be achieved by increasing the convection coefficient andor by increasing the convection surface area, For example, ht may be increased by introduc ing surface roughness to enhance turbulence, as, for example, through machining or insertion ofa coil-spring wire. The wire insert (Figure 8.124) provides a helical zaugh- ness element in contact with the tube inner surface. Alteratvely, the convection coef- ficient may be increased by inducing swe through insertion of a twisted tape (Figure 8.126). The insert consists of a thin strip that is periodically twisted through 360" {nroxuction of tangential velocity component inreases the speed ofthe flow, partic ularly near the tbe wall. The heat transfer area may be increased by manufacturing a tube with @ grooved inner surface Figure 8.12c), while both the convection coeficient ‘and area may be increased by using spiral fins or ribs (Figure 8.124). In evaluating any heat transfer enhancement scheme, attention must also be given to the attendant crease in pressure drop and hence fan or pump power requirements. Comprehensive assessments of enhancement options have been published [25-28], and the Journal of Enhanced teat Transfer provides access to recent development in the field nd P anc Eds, By coiling a tube (Figure 8.13), heat transfer may be enhanced without turbu- lence or additional heat transfer surface area, In this case, centrifugal forces within522 Chapter 8 « Internal Flow cot Twisted seeing ‘ope Fieve £12 Inte flow heat transfer enhancement schemes: a) longitudinal section end view of collspring wiee insert, (6) longitudinal section al eros-sectional view of tape inet fe eu-avay section and end view of fongitudinal fin, and () longitudinal set and end view of helical ibs, the fuid induce a secondary flow consisting ofa pair of longitudinal vortices ta contrast to conditions in a straight tube, can result in highly nonuniform lot transfer coefficients around the periphery of the tube. Hence, local heat coefficients vary with @ as well asx. If constant heat flux conditions are applied, mean fluid temperature, 7,(2), may be calculated using the conservation of e principle, Equation 8.40, For situations where the fud is heated, maximum fi temperatures occur atthe cube wall but calculation ofthe maximnsr local tx ture is not straightforward because of the 6dependence of the heat transfer cient. Therefore, correlations forthe peripherally averaged Nusselt number ae litle use i constant heat flux conditions are applied. In contrast, correlations fe peripherally averaged Nusselt number for constant wall temperature boundary a Gitons are useful, andthe relationships recommended by Shah and Joshi 29] ae provided in the next paragraphs. icin: 8.13 Schematic of helically coiled tube and secondary flow in enlarged cross-sectional view:iui] section a al ores ta nuniform local bi focal heat t ons are applied, ation of It number are 0 correlations for th ure boundary con and Joshi [29] 8.7 4 Heat Transfer Enhancement 523 ‘The secondary flow increases friction losses and heat transfer rates. In addition, the secondary flow decreases entrance iengths and reduces the difference between laminar and turbulent heat transfer rates, relative to the straight tube cases consi ered previously in this chapter. Pressure drops and heat transfer rates exhibit little dependence on the coil pitch, S. The critical Reynolds number corresponding to the ‘onset of turbulence for the helical tube, Rep iS Reoer= Repgll + 12ADIO"] 7) where Rep. is given in Equation 8.2 and C is defined in Figure 8.13. Strong sec~ ‘ondery flow associated with tightly wound helically coiled tubes delays the transi- tion to turbulence. For fully developed laminar flow with C/D = 3, the frietion factor is om I> Reg RefDIO" 530 619) 21 (DIEM 30 Rep(DIC)"* = 300 Rae 5 Reo) = (8.758) 22 (icy 3005 Rey(DIC)"® 8.750) mas 300 = Rep(DIC) (8.750) For cases where C/D 3, recommendations provided in [29] should be followed. ‘The heat transfer coefficient for use in Equation 8.27 may be evaluated from a cor- relation of the form Nuy= [(s6 +838) + 1s9(O yey “exp where (<0) a1 4040 . (1-220) ays pat array 0.05 = Pr 1500 15 Rep(DIC)"? = 1000. Friction factor correlations for turbulent flow are based on limited data. Fur- thermore, heat transfer augmentation duc to the secondary flow is minor when the flow is turbulent and is less than 10% for C/D = 20. As such, augmentation by using helically coiled tubes is typically employed only for laminar flow situations. In laminar flow, the entrance length is 20% (o 50% shorter than that of a straight tube, while the flow becomes fully developed within the fist half-turn of the feli- cally coiled tube under turbulent conditions. Therefore, the entrance region may be neglected in most enginoering calculations. ‘When a gas oF liquid is heated in a straight tube, a fluid parcel that enters near the centerline ofthe tube will exit the tube faster, and always be cooler than, afuid parcel524 Chapter 8 = Internal Flow that enters near the tube wall, Hence, the time-temperature histories of ind fluid parcels, processed in the same heating tube, can be dramatically difere adgition to augmenting heat transfer. the secondary flow associated with the Coiled tube serves to mix the luid relative to laminar flow in a stright tube, in similar time-temperature histories for all the fluid parcels. It is for this reason coiled mbes are routinely used to process and manufacture highly viscous, high ‘added fluids, such as pharmaceuticals, cosmeties, and personal care prod. 8. Microscale Internal Flow ‘The tubes and channels considered so far have been characterized by fyi diameters of conventional size. Many new technologies involve microscale i flows where the hydraulic diameters are in the range D, = 100 jum. An im ‘motivation for developing various microfluidic devices is readily evident by ing the correlations for the Nusselt numbers associated with laminas, Fully d oped flow. Specifically, heat transfer coefficiems are inversely proportional to hydraulic diameter. Therefore, as the size of the flow channel is decreased sii cantly. @ dramatic increase in heat transfer coefficients will occur [30]. 8.8.1 Flow Conditions in Microscale Internal Flow In Section 2.2.1, we saw that heat transfer in gases ean be affected significant the characteristic length scale of the material volume, L, is reduced to the ‘order of magnitude as the mean free path Of the gas, Ayiy. At very small values the hydraulic diameter Dy, therefore, the gas will no longer behave as a conti Because the continuum assumption has been made throughout our discussion ‘convection so far, the results of Chapters 6 through 8 are not expected to apply £0808 when Dy/Ajg 5 10, and recourse test be made to advanced methods to imate the velocity distribution and pressure drops associated with internal flow (3 For liquids, experiments have shown that, for laminar flow in circular tl ‘with diameters as small as 50 jum, Equations 8.19 and 8.22a successfully predict pressure drop over the tube length [32]. These equations are expected to be valid most liquids for hydraulic diameters a8 small as | sum (32, 33) ‘From ihe discussion in Section 6.3.1 one might anticipate that, since tubal is characterized by the motion of relatively large parcels of fuid im convestina sized devices, Bauations 8.20 and 8.21 will not be applicable for flow in mic idic devices because the volume of the fluid parcel is restricted by the hyd diameter of the channel. One may similarly anticipate tht the criterion forthe ons of turbulence, Equation 8.2, will be different for microscale internal flows. Carel ‘measurements using various liquids have shown that the critical Reynolds number ‘corresponding to the onset of turbulence is, i fact, well described by Equation 82 for liquid low in tubes with diameters as small as 50 jum [32). Since extremely high mean velocities are necessary to achieve Rep = 2300, one is seldom concerned about turbulent microscale convection.Proportional to s decreased si smell values: We 88 & continu ‘our discussion ed to apply ‘methods to es jnternal flow (31), w in circular cal esfully predict ted t0 be Valid since turbulence in conventional= low in microfia d by the hydraulic terion for the onset mat flows. Careful | Reynolds number ed by Equation 8,2 nce extremely high seldom concerned 8.8 & Microseale Internal Flow 525 8.8.2 Thermal Considerations in Microseale Internal Flow Convective heat transfer in microscale internal flows continues to be the subject of ongoing research. The analytical results and convection correlations of Chapters 6 and 8 shout not be used for gases when Dy/Anz 10 and shouldbe used with cau- tion for liquids when Dy 5 1 um. EXAMPLE 8, Combinatorial chemistry and biology ar’ used in the pharmaceutical and biotech. nology industries to reduce the cme and cost associated with producing new drugs. Scientists desire 1o create large populations of molecules, or libraries, that can be sub> sequently screened en masse. Producing vast libraries increases the probability that novel compounds of significant therapeutic valve willbe discovered, A crucial variable in producing new compounds is the temperature at which the reactants are processed. A microreactor chip is fabricated by first coating a I-mam-thick glass microscope slide with a photoresist material. Lines are subsequently etched into the photoresist and a second glass plate is attached to the top of the structure, resulting in multiple parallel channels of rectangular cross section that are a = 40 pm deep, b = 160 jm wide, and L = 29 mm long. The spacing between channels is s = 40 jam, so that N= Zw + 8) = 100 channels are present within the 20 mm 20 mm microreac- tor. A mixture of two reactants, both initially at T,, = 5°C, is imroduced into each channel, and the edges of the chip are maintained at temperatures 7, = 125°C and 7, = 29°C so that the reactants in each channel are subject £6 4 unique processing temperature. Flow is induced through the struciure by applying an overall pressure difference of AP = 500 kPa, The reactants and the product of reaction have thermo- physical properties similar co ethylene glycol, Estimate the time that elapses for the ‘entering reactants to come within I°C of the desired processing temperature, SOLUTION Known: Dimensions and operating conditions for flow of reactants and product of reaction in a microreactor. Find: Time needed to bring the reactants to within °C of the processing temperature Schematie: oles * = 160 um ae \s ry Microckanne! aay aces Photorsst, Rercms donnie TneChapter 8 = Internal Flow Assumptions: 1. Laminar flow. 2. Linear temperature distribution across the width of the microreactor. 3. Steady-state conditions. 4. Incompressible liquid with constant properties, 5. Negligible viscous dissipation. Properties: Table A.5, ethylene glycol (T, = 288 K): p = 11202 kel 2359 Mkg*K, j2 = 2.82 X 107 Nesim’, k= 247 x 10°3 WimK, Pr GF, = 338 K): p = 1085 kg/m’, ¢, = 2583 Ikg*K, = 0.427 X 10°, Ne 261 x 107 Wim: K, Pr = 45.2 Analysis: We will bracket the heat transfer and fluid low behavior by eva the performance at the extreme processing temperatures. ‘The flow of reactants is induced by the applied pressure difference betwee inlet and outlet of the microreactor. Because of the large variation of the vi ‘with temperature, we expect the flow rate that is associated with the highest cessing temperature to be the largest. ‘The perimeter of each microchannel is P= 2a+2b=2% 40% 10m +2 160 10 m=0.4 X 10m and the hydraulic diameter of each microchannel is found from Equation 8.6638 X40 x 10°F m x 160 x 10"* 04x 10m 64x 10% m We begin by assuming a relatively short entrance length, to be verified late, the flow rate may be estimated by using the friction factor for fully developed ditions, From Table 8.1 tor bla = 4, f = 73/Rep,. Substituting this expression i Equation 8,22, rearranging terms, and using properties at T = 125°C (inthis oq tion and those following) results in Hence, the Reynolds number is 0.657 mis 64 X 10° m X 1085 kg/m! a 7 # 0427 10 N= sin 0 Rep, and the flow is deep in the laminar regime. Equation 8.3 may be used to deterigg the hydrodynarmic entrance length, which is, ign = 0.05D,Rey = 0.05 X 64 % 10% m X 10.7 = 34.2% 10°% m,ion 8.66 as, 10-6 verified later, developed expression i °C (in this = 0.657 mis 8.8 » Microscale Internal Flow 527 and the thermal entrance length may be obtained from Equation 8.23, yielding, Niay = Saar = 34.2% 10 mx 45.2 = 1.55% 107 Both entrance lengths occupy less than 10% of the total microchannel Fegth, L 20 mm, Therefore, use of fully developed values of fare justified, andthe mass flow tate for the T= 125°C microchannel is fin = pAty = publi, = 6085 kg/tn? X 40 X 10% m X 160 30°* mn X 0.657 mls F456 X 10° pis Equation 842 may now be used to determine the distance from the entance of the microchannel 10 the location, a, Where Ty = 124°C, chats, witha PC ofthe surface temperature, The average hea transfer coefficient, Fi, i replaced by the fully developed value of the heat transfer coefficient, h, becalse ofthe relatively sbort thermal entrance length From Table 8.1, We se that fr lo = 4, Mug ~ ADA = 4,44 Therefor, td > 0261 Wim -K = 58h 56 10 Win? = K 64 10° m Bah = Nigh = Fok Mop. As expected from our discussion of microscale lows, the convection coefficient is extremely large. Rearranging Equation 8.42 yields they gf Ex Tas] __456-X 10-taphs x 2583.hhg*K_ | [ (125 — SC Pu LT, Tee} OA X17 m X LBL X IO" Wiss" K") 025 D257 = 7.79 10m | Therefore ete eed fo the ect © ek & me epee | stn 1° othe processing engeaies 179 10> m/O.687 s/s = 0.012 5 4 fe = Kelty Repeat cessing temperature of 25°C yields 14, = 0.0995 mm/s, Rep 107? m, yo, = 0.218 LO m, b= 171% 108 Wim? + Kx, 4. = 0007s, ie catcutations for the microchannel associated with the smallest pro- 0.253. siyg = 8:09 73 x 107m, and Comments | 1s The toa thickness of the glass (2 mm) is 50 times greater than the depth of each microchannel, while the abesrya) conductivity of the 1885, yas 14 Wim K (Table A.3), is 5 times greater than that of the fluid. The presence of sucita smait amount of fluid is expected to have a negligible effect on the linear temperature distribution that is established across the chip. The temperature dif- ference across the bottom or top surface of each channel is approximately AV (Dy = Tbe = 25 ~ 25°C & 240 X 10-* aK % 10 m) = 48C.528 8.9 Convection Mass Transfer Chapter 8 # Internal Flose 2, Solving Equation 8.42 over the range 0 3.66, use of Equation 857 is valid. Finally. Dd, 4.22 X 0.28 x 10™ m/s Ty = Tro Sone O12 mus Comments: From Equation 8.23, x4, ~ (0.05Re, Sc)D = 0.58 m, and flly coped conditions exis over approximately 40% ofthe tube length. An assumption fully developed conditions over the entire tube would provide a value of Shy 3.66, which is 13% less than the above result. —the air are at piouconsnn 2 ms mn, and fully devele An assumption of 8 value of Shy = 531 In this chapter we have considered forced convection heat and mass transfer for an important class of problems involving internat flow. Such flows are encountered in numerous applications, and you should be able to perform engineering calculations that involve an energy balance and appropriate convection correlations, The methodology involves determining whether the flow is laminar or turbulent and establishing the length of the entry region. After deciding whether you are inter- ested in local conditions (at a particular axial location) or in average conditions (for the entire tube), the convection correlation may be selected and used with the appro- priate form of the energy balance to solve the problem. A summaty of the correla- tions is provided in Table 8.4 You should test your understanding of related concepts by addressing the fol- lowing questions. + What are the salient features ofaydrodynamic entry region? A thermal entry region? Are hydrodynamic and thermal entry lengths equivalent? If not, on what do the relative lengths depend? ‘What athe salient ydrodynamic features of fally developed flow? How isthe ‘ction factor for fully developed flow affected by wall roughness? To what important characteristic of an internal flow isthe mean or bulk rem- perature linked? ‘What re the salient thermal features of flly developed flow? If fluid enters a tube at a uniform temperature and there is heat transfer or from the surface of the tube, how does the convection coefiient vary with distance along the tube? For uid low through a tube with a uniform surface heat flux, how does the ‘mean temperature ofthe fuid vary with distance from the tube entrance in (2) the entrance region and (b) the fully desefoped region? How does the surface tem perature vary with distance i the entrance and fully developed regions? For heat transfer to or fom a fd flowing through a tube with a uniform sur- face temperature, how does the mean temperature ofthe fuid vary with dis- tance fom the entrance? How does the surface heat flux vary with distance fiom the entrance? Why isa log mean temperature difference, rather than an arithmetic mean tem- perature difference, used to calculate the total rate of heat transfer to or from & fig lowing through a tube with a constant surface temperature? ‘What two equations may be used to calculate the total heat rate toa fluid flow- ing through a tube with a uniform surface heat fux? What two equations may ‘be used 10 caleulate the total heat rate or from fluid flowing through a tube with » uniform surface temperature? ‘Under what conditions is the Nusselt number associated with intemal flow equal to a constant value, independent of Reynolds number and Prank number? +s the average Nusselt number associated with flow through a tube larger than, {equal to, or es than tke Nusselt number for fully developed conditions? Why? + How is the characteristic length defined fora noncircular tube? + Whar are the salient features of a concentration entry region?532 Chapter 8 # Internal Flow TABLE 8.4 Summary of convection correlations for flow in a circular tuby Correlation Conditions (819) (853) Laminar, fully developed, uniform ¢” (8.55) Laminar, fully developed, uniform 7, 8.0668(012)Rey Pr 1+ 004((DiL)Re, Pr (8.56) Laminar, thermal exis or combined entry with Pr= 5), uniform 7, o(Be)" (a) (5) Lai cote. 06 P= 804 “ 9.75, uniform T, = 03316Re5" 8.200)" Turbulent, fully developed, Rep = 2 % 10" = 0.184Re"* 6.2007 Turbulent, flly developed Rep = 2 x 10" " yee oom $= (0790 In Rey ~ 1.68) (8.21 Turbulent, fully developed, 3000 Rep 5 x 10" p Moody, L. Fa k Petukhoy. B.S, Map = 0.023Re§ Pr* (8.60) Turbulent, flly developed. 0.6 5 Pr 160, Advances in H Rep = 10,000, (L/D) = 10,n = Ot for T, > Ty [New York, 19% and = 03 foe T, < Ty Chen, MM or 235,137, 1980 Ning = B82ReS Pr'”| (2) (8.61 Turbulem, Fully developed, 0.7 = Pr = 16,700, Chato, J.C, 1.5 7 " Rep = 10,000, L/D = 10 J. Kays, W.M, Tr Hausen, HZ V ten — 10003 “H, May = LOR 1000) (8.62)! Tubal, fully developed, 0.5 = Pr 2000, Sieder, EN. an Ve 1278) 1) 3000 = Reg = 5 % HP L/D) = 10 tose Mig = 482 + O.0185(Rep Pr)?” (8.64) Liquid metals, urbulent, fully developed, uniform Whitaker, 8. Al 3.6 X 10° 5 Rey $9.05 X 10°, 10° = Pep = 10" + Shah, R: Ke an — Convection nD 50-4 0025(Rep Prj!* (8.65) Liquid’ metals, twrbveat, fully developed, Shah, R.K.. a uniform T, Peo = 100 and W. Aung Heat Transfer, 1987. ‘Colburn, AP, |4. Winterton, RH 1998. IS. Bhat, M.S. a and W. Aung fave Heat Tran ‘York, 1987. Grielinski,V. Kakac, Sin S Handbook of 5 Chap. 18, Wiley The ans vansfer colons may be obtained by replacing i and Pry Sh and Se, respectively. ‘Properties in Equations 8.53, 8.55, 8.60, 861, 862,864, and 8.65 are based on T, popetes in Equations 8.19, 8.20, and 8.2) ae se T, + Tay properties in Equations 8 6 and 8.87 ae based oT, = (Ta, + Ta “Eiqutions 8.30 and 8.21 pein o smooth abe. For ough aes, Eaton 862 sould he used with he ests of Figute 83, “Asa fest approximation, Bgtions 860,861, or § 62 may be used o evaluate the erage Naselt amber Nip over the emir tbe eng if (1/0) = 10- The propeties shou then he eva at the average of the mean temperature, T= (Tq, Tu 32 ‘For tubes of noncrela ros section, Rep = Dyeyl,Dy = 4A.JP- and a» mfp, Ress for fully developed laminar Row are roi Table 81. Foe urbulnt ow, Equation 8 60 maybe wed a8 &Ind 8.21 are Wased on be 8.3 peenibe length, btow are provided in 1 References 533 + What are the salient features of fully developed flow for mass transfer? + How may convection mass transfer correlations be inferred Several features that complicate internal flows have not been considered in this chapter. For example, a situation may exist for which there is a prescribed axial vari- ation in 7, org’, rather than uniform surface conditions. Among other things, such a variation would preclude the existence of a fully developed region. There may also exist surface roughness effects, circumferential heat flux or temperature variations, ‘widely varying fluid properties, or transition flow conditions. For a complete ais- cussion of these effects, the literature should be consulted (11, 12, 15, 17, 24) J. Langhaar, H, LJ Appl Mech. 64, A-S5, 1942. 2 Kays, W. M, and ME. Crawford. Convective Heat and Mass Transfer, 3t4 ed, MeGraw-Hill, New York, 1993 3. Moody. LF, Trans. ASME, 66, 671, 1944, Petukhov, B.S., in T.F. Irvine and J.P. Hartnett, Eds, Advances in Heat Transfer, Vol. 6, Academic Press. New York, 1970, Chen, M. M., and K.R, Holmes, Ann N.Y. Acad. Sci. 2335, 137, 1980, 6. Chato,J.C..J. Biomech. Eng., 102, 110, 1980. 1. Kays, WM, Trans. ASME, 77,1265, 1955, A Hausen, H.,Z VDI Beih. Verfahrenstech, 4, 91,1943 9 Sieder,E.N., and G. E, Tate Ind Eng. Chem, 28, 1429, 1936, 10, Whitaker. S., AICHE J.,18, 361, 1972. Il, Shah, RK. and A. L. London, Laminar Flow Forced Convection in Ducts, Neademie Press, New York, 1978, 12. Shah, R.K., and M.S. Bhat, in $. Kakac, RK, Shah, and W. Aung, Bd, Handbook of Single Phase Convective Heat Transfer, Chap. 3, Wiley-Intescience, New York, 1987, Colburn, A.P. Trans. AICHE, 29, 174, 1933 Winterton, RH. S., Int. J. Heat Mass Transfer, 41, 809, 1998, Bhatt, M.S. and R.K. Shah, in S. Kakac, R. K. Shah, and W. Aung, Eds, Handbook of Single-Phase Convec: tive Heat Transfer, Chap. 4, Wiley-Inerscience, New York, 1987, Gaielinski, Vt. Chem Eng, 16,389, 1996. Kakac, Sin $. Kakuc, R.K. Shah, and W. Aung, Eds, Handbook of Single-Phase Convective Heat Transfer, Chap, 18, Wiey-Inerscience, New York, 1987 Ghajar, A. J, and LM. Tam, Exp. Thermal and Fluid Science, 8,79. 1994, Noms, R HL, in A.E, Berples and R.L. Webb, Eds.. Augmentation of Convective Heat and Mass Transfer, ‘ASME, New York, 1970, 20, Molki, M., and E.M, Sparrow, J. Heat Transfer, 108 482, 1986. 21, Skupinski, ES., J. Tontel, and L. Vautrey, Int J. Heat Mass Transfer. 8,937, 1965 22, Seban, R.A. and T.T. Shimszaki, Trans: ASME, 73, 803, 1951 23, Reed, C.B.,in S. Kakac, R. K Shah, and W. Aung, Eds. m4, 25 26, 2, Handbook of Single-Phase Convective Heat Transfer, ‘Chap. 8, Wiley-Interscience, New York, 1987. Kays, WM. and H.C. Perkins, in W.M. Rohsenow, J.P. Hartnett, and E. N. Ganic, Eds, Handbook of Heat Transfer, Fundamentals, Chap. 7, MeGraw-Hill, New York, 1985, Berges, A. “Principles of Heat Transfer Augmenta- fon,” Heat Exchangers, Thermal-Hydraulic Fundam: ‘ais and Design, Hemisphere Publishing. New York, 1981, pp. 819-842, Webb, R.L., in S. Kakae, RK. Shah, and W. Aung, ds,, Handbook of Single-Phase Convective Heat Trans: fer, Chap. 17, Wiley-Interscience, New York, 1987 Webb, RL. Principles of Enhanced Heat Transfer Wiley, New York, 1983 Manglik, RM, ard A. E, Berles. in J, P, Hartnett TF. lIvine, Y. 1. Cho, and R. E. Greene, Bds., Adwances in Heat Transfer, ol. 3, Academic Press, New York, 2002 Shah, R. K., and S, D. Jodi in Handbook of Single Phase Convective Heat Transfer, Chap. 3, Wiley-Imerscience, New York, 1987534 Chapter 8 = Internal Flow 30. Jensen, K.F., Chem. Eng, Sci, $6, 293, 2001 31. Koviany, M.. Principles of Convective Heat Transfer, Springer-Verlag, New York, 1994, Problems Hydrodynamic Considerations 8.1 Fall develope conitions are Known to exit for water flowing though 2-minmeter be a 00! ky and 27°C Whats the maximum velocity ofthe Wo in he ube? What ste pressure Baden sociated ih the tow? {42 What i be pesure dopassocsted with water at 27°C Rowing witha mean velocity of 02 mf dough +60: Jong es iron pie of 0.5m inside ameter? 83 Water at 27°C flows wih a mean veociy of Imi though enon pip a0 2m se ameter, (a) Detemine he pressure op ovr the pipe engin and the cresponding pump power regent if he pipe surges smooth (0) Ifthe pipe is mae of cast ron and is suc is clea, determin the presture drop and pp power eltement [Ga] Fer the smooth pipe condition, generate plot of presue dtop and pump. power requiement for trea veces nthe ange fom 0501.5 ms 84 An engine oil cooler consists of a bundle of 25 smooth tubes, each of length Z = 2.5 m and diameter D = 10 man, (4) IF oil at 300 K and a total ow rate of 24 ks is in fully developed flow through the tubes, what i the pressure drop and the pump power requirement? (©) Compote and plot the pressure drop and pump power requirement asa funetion of low rate For 10 = ri == 30 kes. 85 For fully developed laminar flow through a paralle-plate channel, the x-momentum equation hes the Form Pu) _dp . (2) ay constant “The purpose ofthis problem isto develop expressions for the velocity distribution and pressure gradiem analo- 0usto those forte circular tbe in Seation 81 (a) Show that the velocity profit, (s,s parabolic and of the form wl 32, Sharp, K. V.,and R. J, Adrian, Exp, Fluids, 36,74, 33, Travis, K.P., B.D. Todd, and D. J. Evans, Phys 58, 4288, 1997 here a is the mean velocity a (2) Ta \ae, and ~dplde = ApiL, where Ap isthe pressue across the channel of length J. () Write an expression defining the friction ft ‘using the hydraulic diameter D, as the char Jength. What is the hydraulic diameter for palle-plate channel? () The friction factor is estimated from the expe Fo CiRe, where C depends upon the Now ‘ection, a8 shown in Table 8,1. What is the cient C forthe paallel-pate channel? (@) Air flow in a parallel-plate chanel with ase ‘of S mm and a length of 200 mm experiences sute drop of Sp = 375INin?. Calculate the velocity and the Reynolds number for air a spheric presure and 300 K. Is the assumption f developed flow reasonable for this application rot, whats the effect on the estimate for u,? ‘Thermal Entry Length and Energy Balance Considerations 86 Consider pressurized water, engine oil (unused, NaK (224%/78%) flowing in& 20-mm diameter tbe, (@) Determine the mean velocity, the hydrody entry length and the thermal entry length for ah the fluids when the fluid temperature is 366 K the flow rate is 0.01 kgs. ) Deternine t entry Fength and engine o of 0.02 is ‘with units of mis responding valu a this axial posit At 2 panicular units of mis (ua Determine cone | toy and mean (0 ity and tempera and T,, appear steady-flow ene trans-Alaska pip Jong distances c reasonable to qu nifleant. Conside eter D= 1.2m, properties are p 0.765 Nin work, and the te JO When viscous ¢ (cnultiplied by g a Pe This problem ex + pation. The cond fully developed Equation 8.15,= Problems (b) Determine the mass flow sate, the hydrodynamic ‘entry length, and the thermal ety Feng for water tnd engine oil at 300 and 400 K and a mean velocity 0f 0.02 ms Velocity and temperature profiles for Taminar flow in a Fluids, 36,741, 2 1 Evans, Phys, tube of rads r, = 10 mm have the form We) = 0.11 = h,] Te) = 3448 + 75.0¢Ir) ~ 18.80 ith units of avs and K, respectively. Determine the cor responding value of the meun (or bulk) temperatute, this axial position. §8 Ar a particular axial station, velocity and temperature profiles for laminar low in a parallel plate channel have the form O75[1 = (yy) = 50+ 95.6609," HO) nT 47831090 ith units of vs and °C, respectively. Determine corresponding values of the mean velocity ‘ug and mean (or bulk) femperatute, Ty, Plot the veloe= ity and temperature distsbuions. Do your values of uy, and T, appear reasonable? 9 In Chapter 1 it was stated that for incompressible liq vids, low work could usually be neglected in the ‘seadycflowe energy equation (Equation 111d). Inthe ‘eans-Alaska pipeline, he high viscosity ofthe oil and long distances cause significant pressure drops, and itis reasonable to question wether flow work would be si. nificant. Consider a L = 10G km length of pipe of diam: cae¢ D = 1.2m, with ol flow rate rn = $00 kgs. The oil properties are p = 900 key’. c= 2000 kg» K. (0.265 N-sim?. Calculate the pressure drop, the flow work, an dhe temperature rise caused by the flow work. 410 When viscous dissipation is included, Equation $48 {amalipied by p,) becomes 2,97), (ax «ales () Tis pln cos einer ef i ds me nds ir ncn fully developed flow in a circular pipe, with w given by Equation 8.15, nergy Balance af oil (unused), PM am-diameter tbe 535 () By integrating the left-hand side over 8 section of @ pipe of length L and radius ,, show that this term {yields the right-hand side of Equation 8.34 (6) Integrate the viscous dissipation term over the same volume (6) Find the temperature rise caused by viscous disi- pation by equating the ewo terms calculated above Use the same conditions as in Problem 8.9. BIL Water enters a tube at 27°C with a Mow rate of 450 kg/h. The heat transfer from the tube wall 10 the Aud i piven as g! (Wim) = ax, where the coeficient a 20 Wim’ and «(mis the axial distance from the tube entrance (2) Beginning with a properly defined differential con trol volume in the tube, derive an expression forthe ‘emperature disteibution T,(x) of the water. (b) What isthe outlet temperature of the water for a heated section 30 m long? (6) Sketch the mesa fuid temperature, T(x), and the tube wall temperature, 7,09, a8 a function of dis tance along the tube for fully developed and devel ‘oping flow conditions. (a) What value of a uniform wail heat flux, q (instead of g] = av), would provide the same fiid outlet emperatute as dhat determined in pact (by? For this type of heating, sketch the temperiture distribu tions requested in pat (e) 8.12 Consider flow ina circular tube. Within the tes section Jength (between 1 and 2) a constant heat flux q’ is ‘maintained (2) For the following «wo eases, sketch the surface ‘emperature 7x) and the fluid mean temperature T.As) 88.4 fonction of distance along the test see- tion x. In case A, flow is hydrodynamically and thermally fully developed. In case B, flow is not developed, 4b) Assuming that the sueface flux g! and the inlet mean temperature Tare identical for both cases, will the exit mean temperature Tay for case A be areater than, equal to, or less than 7, for case BY Briefly explain why.813 upto, D Chapter 8 = Internal Flow Consider a cylindrical nuclear fuel rad of length and diameter D thats encased in concentric tbe. Pressur- dined water lows through the annular egion between the ‘od and the tbe ata rate and the outer surace of the tbe is well insulate. Heat generation occurs within the fel rd. and the volumetric generation rat is known sinusoidally with distance along the rod. Tati 4 sin (xa), where (Wh?) is a constant. A Unifoees convection coetcient & may be assumed to ‘exist between the suirace ofthe rod and the water. a G2 dsintewn a4 81s (a) Obtain expressions for the local heat tux 4°(x) and the total heat transfer g rom the fuel rod to the water (©) Obtain an expression for the variation ofthe mean femperature T4(x) of the water with distance long the tube (©) Obtain an expression forthe variation ofthe ro sur face temperature 7.2) with distance x along the tbe. Develop an expression for the x location at which this temperature is maximized Ina particular application involving fuid flow at a rate Fn through a circular tube of length L and diameter D, the surface heat flux is known to havea sinusoidal vari- ation with x, which is of the Form q/(x) = dl Sit (UL), The maximum Aux, gq. is a known constant, and the fluid enters the tube at a known temperature, Ty. Assuming the convection coefficient 10 be con- stant, how do the mean temperature ofthe uid and the sunfacetemperatute vary with x? A fa plate solar collector is used to heat atmospheric air flowing through a rectangular channel. The bottom surface ofthe channel is well insulated, while the top surface is subjected to a uniform heat flux 4, which is ‘due to the net effect of solar radiation absorption and heat exchange betwen the absorber and cover plates (@) Beginning with an appropriate differential contol volume, obeain an equation that could be used to determine the mean air temperature T() a8 a func- ‘ion of distance along the channel, Solve this equation to obtain an expression forthe mean temperature of the ar leaving the collector. 87 water flow Transpren coer te Asrber ate Rectanguat Ae oe a) Write a (0) With ar inlet condions of in = 01g to) wie a SUC. wht the ai ote coetci hm, and = 700 Win The Sew Sit he fais = 1008 ka" nase va 16 Atmosphere air eters he heated eton of represen tute a Now eof 003 Kas anda ene coe a 20°C. The tbe is of ameter DSO nm, td (0) Deter Geveloed onto ith h = 25 Win cut the emt length of L = 3 m. BB stp fow i. (a) For he ease of uniform sutace heat xa the sets 1000 Win’ determine the teal est eter te cross tnd the mean temperate of te i eng sina ni {tbe Tue What te value ofthe sae the fal de tue a be be inlet 7, and out 7,” See tril vation of 7 and On he same 1 Superimpsi ts shech (quately) evel varton a one and T,, forthe more realistic casein which te tion 8.88a ‘convection coefficient varies with x (b) If the surface heat fox varies linearly with, ‘An experi sists of alo that gf (Wim*) = S00x (m), what are the vals ‘and length L Tap Ta 7? Ske he al vaio Sina td, Gn the same ie so skh ive the aul vara fT and Ty oe teal ce in wich he lol conven Gent ae ih were it (© Foe the two Retin onions of pa (a) Fata ple mean fd and sre temper ae meol find Ts especie, at incon of di ‘along the tube. What effect will a fourfold ine “expressions inthe convection client have on the rg (0) thea tre dssons? Mei (qd) For each type of heating process, what heat fu Cb) the fui ae eed aie an ai ot tegen © thea oF 125°C? Plot the temperature distributions Water at 300 K and a flow rate of 5 kg/s enters a bck, thinwalled tube, which passes through a large fume ‘whose walls and air areata temperature of 700 K. Th iameter and length of the tube are 0.25 m and 8m respectively. Convection coefficients associated i= Problems ater flow through the tube and ait flow over the ube tue 300 Wins» K and 50 Wim?» K,eespectively. {) Write an expression for the linearized radiation coefficient correspor jon exchange between the outer surface of the pipe and the fur- nace walls. Explain how to calculate this coeff- cient if the surface temperature of the tube Is represented by the arithmetic mean of it inlet and outlet values. (6) Determine the outiet temperature ofthe water, Tx 700 Wim? The 3 25 Wim" K exist, Slug ow is an ideatized tube low condition for whieh the velocity #8 assumed (0 be uniform over the entire lube cross section. For the ease of laminar slug owe wih a uniform surface beat ux, deere the form of the fully deseloped temperature distbution Tl) and the Nusselt nmber Np face heat ux at oul eat transfer ‘of the air leaving: ofthe surface te outlet 7,,? Sketc Ta On the same fig the axial variation o 119 Superimposing a control volume thats differential in x ‘on the tube flow conditions of Figure 88, derive Eq ‘ion 8.45a, jes with x An experimental nuclear core simulation apparatus con: sists of a long thin-walled metallic tube of diameter D and length L, which is electrically heated to produce the sinusoidal heat flux distribution sico=atsin(®) ere is te dace meaie om tebe let Fluid at an inlet temperature T,, flows through the tbe flare af Assuming he feta al fovlped oer he fe gi fe abe lp capes fr (oe ol tof eat ase, tei (te i out eper. Tas (c) the axial distribution of the wall temperature, T,(x); = linearly with x, what are the value the axial variation of local convection coe fons of pars (a) and. i wil a fourfold inet have on the temp from the ube tothe (a) te magni and poston of the highest wal temper [ilcomider» dOmntiameter tbe of 4. length Si seseta oat fx dstbuion for sich 10000 Wh Fu pasing through he tbe pe are 0.25 m and & cients associated with thas a flow rate of 0.025 ky/s, a specific heat of 4180 Jtke*K, an entrance temperature of 25°C, land a convection coefficient of 1000 Win’ + K. Plot the mean fuid and surface temperatures as 3 function of distance along the tube. Identify impor: tant Features of the distributions, Explore the effect of £25% changes inthe convection coefficient and ‘the heat Aux on the distribution, 8.21 Water at 20°C and a low rate of 0.1 ks enters a heated, thin-walled tube with a diameter of 15 mm and length of 2m. The wall heat x provided by the heat ing elements depends on the wall temperature acord- ing tothe relation HOO Gall Fal, ~ Td] where qi, = 10¢ Wim?, a = 0.2 K"!, Tar = 20°C, and 7s the wall temperature in *C_ Assume fully devel ‘oped flow and thermal conditions with a convection oeicient of 3000 Wri? + K (@) Begining witha propery defined ferential conto ‘volume inthe tube, derive expressions fo Ue varie Aion of the water Tate), and the wall, 3) temper tures asa fueton of distance fom the tube ik D]Using a numerical integration scheme, calculate and plot the temperature distributions, 7,(1) and ,09.90 the same graph, Mentify and comment on the main features of the distributions. Hint: The IMT integral function DER(T,-t) can be used 10 perform the integration along the length of the tube () Caleta the total ae of beat transfer 10 the water Heat Transfer Correlations: Circular Tubes 8.22 Engine oil is heated by owing through a circular ube of diameter D=S0mm and length L= 28m and ‘whose surface is maintained at 150°C. (2) I the low rate and inlet temperature of the oi are (05 kgls and 20°C, what is the outlet temperature 2 What isthe total heat transfer rate q for the te? For flow rates in the range 0.5 = n= 2.0 kes ‘compute and plot the variations of T,, and with ‘in, For what flow rate(s) are q and T,., maximized? Explain your results ro 8.23 Engine oil lows through a 25-mm-diameter ube ata eof 0.5 kgs. The oll enters the tube ata temperature fof 25°C, while the tube surface temperature i main- tained at LOO°C. (@) Determine the oil outlet temperature fora S-m and or 100-m long tube, For each ease, compare theChapter 8 Internal Flow Jog mean temperature ditferenee fo the arithmetic mean temperature difference. [t)]For 5 = £ = 100m, compute and plot the average ‘Nusselt number Nu and the ol outlet temperature asa function of 8.24 An oil preheater consists of a single tube of 10mm tiameter and Sm length, with its sueface maintained at 175°C by suiting combustion gases. The engine oil (ne) enters at 75°C. What flow rate must be supplied to intain an oil outlet temperature of 100°C ? What is the corresponding heat transfer rate? 8.25 Engine oil flows at a rate of | kgs through 3 Ssmm- iameter staight tube. The oil has an inlet temperature ‘of 45°C and itis desired to heat the oil to a mean tem- perature of 80°C atthe exit of the tube. The surface of the tube is maintained at 150°C. Determine the required length of the tube, Hint: Calculate the Reynolds num- bers atthe enteance and exit of the sube before proceed- {ng with your analysis. 8.26 Ethylene glyco! flaws at 0.01 kpls through a 3-mm- ameter, thin-walled tube. The tube is coiled and sub- ‘merged in a wellsired water bath maintained at 25°C. the Aid enters the ube 285°C, what hea rate and tube length are required forthe fluid t0Teave at 35°C? [Neglect heat transfer enhancement associated with the coiling. 8.27 In the Final stages of production, a pharmaceutical is sterilized by heating it from 25 to 75°C as it moves at (0.2 ms through a straight thin-walled stainless steel tube ‘of 12.7-mn diamete. A uniform heat Nu is maintained by a electic esistance heater wrapped around the outer surface of the tbe. If the tube is 10/m long what isthe required beat Sux? If fluid enters the tube with a fully developed velocity profile and a uniform temperature profile. what isthe surface temperature atthe tube exit and ata distance of 0.5 m from the entrance? Fluid prop- eties may be approximated as p= 1000 ky/m’, 4000 kg K, e= 2% 10" kalo, and Pr= 10. 8.28 Baich processes are often used in chemical and pharma- ceutical operations fo achieve a desired chemical com: position for the final product. Related heat tansfer processes are typically transient, involving a liquid of Fixed volume that may be heated from room tempera- ture to a desired process temperature, or cooled from the process temperature to oom temperature, Consider a batch process for which a pharmaceutical (the cold Fluid, c) is poured into an insulated, highly agitated ves: sel (8 stirved reactor) and heated by passing 2 hot uid (4) through a submerged heat exchanger coil of thin ‘walled bing and surface area A, The low rate, iy, ican ie pent, Te and se a tow te tui akon, ae th tl oy ears Tas Tau the volume, V,, mass density, p. and heat, cof the pharmaceutical, Heat transfer fom ‘The mold that is hhot fluid to the pharmaceutical is governed by an ‘constructed of m all heat transfer coefficient U. ‘The mold, tobe termes betes Gud 30°C is eval aa at 30 mm 10 assy oan coe anbeptoede te mle — Teommes Ta ‘The total mas on eo Containment verse GV = 5 long pa - the nol stator | faster and, in tu eS ‘be manufacture (> Sang om si ees, ete eg ine 08 tac teed terme the sans cir and T;,, With time during the heating process. eh To Sons maybe een fre Te eng a oe rma ov fe tage ate tog tn tegen lees ade ine wate oo aco) tne for ow fhe ha nee troup th ube: oe thee apes Schone Scene 9 8 unsion f E a wie Conte Sebel pies, Ue expen fe sett tnd ery lane fer fo oh et ie ee Sih cep ace fora con solo me sigh lamer aan an esi Oo ao for T0. tube eng (b) Consider @ pharmaceutical of volume V, = 1m, ‘ture of Ty, density p.=1100kg/m', specific heat, a tank of | 2 ig ‘sl an “na emp coms 228. ci og = 0 mae ‘comp Seed = Sonmancol dance €- 50m raat Stoned wes bth nee ‘eine slings, = 20 nbn = 2a gh ec : venta sii at eu eg Colt ty te ronald = 100 Wm ead es th fd pets ee = 200 Jn" ie Oita Pr we Port feng conti compte sod pot me phumacesia put ane ee imc pete Tras anton oF ome oe ee ome Sr Sactie How tng dus ie ech bach sempre of Ter? The pees to sola ‘operator may contol the heating time by varying Jin. For 1 = fy = Skpls, explore the effect of he compression a plastic pipe= Problems flow rate on the time f required to reach a value of ‘The mold that is used in an injection molding process is onstucted of metal (p = 7800 kg/m’, c = 450 kg.“ K). “The mold, to be heated to 190°C prior to injection ofthe thermoplastic material, must be subsequently cooled before ejection ofthe finished part. Pressutized water at WFC is available for cooling. The mold has dimensions 50 mn x 100 mm X 40 mm and the mold designer must specify inclusion of ' cooling passages, of viame= ter 5 mm, to be machined into the mold. Tone passage «an be placed every 19 mim along the length or width of the mold, she designer can therefore speciy either five {00-mm-long passages or ten S0-mm-long passages. “The total mass flow rate of water, divided equally among the channels, is 0.02 kg/S. Which configuration (W= 5 long passages or N = 10 short passages) should the mold designer specify in order to coo! the mold faster and, in ur, increase the numberof pats that can be manufactured daily? What is the initial rate of cool- ing of the mold (°Cis)? The velocity profile in each channel is fully developed prior (0 entering the bot mold. Neplect the mass ofthe thermoplastic part Tao, aan vse revatio Uetive express the variation of ing process for the rate of ‘The evaporator section ofa heat pump is installed in a large tank of water, whichis used as a heat source dur ing the winter. As energy is extracted from the water, it begins to freeze, creating an ice/nater bath at O°C, which may be used for airconditioning during the sum mer. Consider summer cooling conditions for which ait is pussed through an array of copper tubes, each of inside diameter D = 50 mm, submerged in the bath (@) If aie enters each tube at @ mean temperature of Ty, = 24°C and a ow rate of r= 0101 kws, what lube length Lis needed to provide an exit tempera luce of Ty, = 14°C? With 10 tubes passing through 8 tank of total volume V= 10m’, which initially ‘contins 80% ie by volume, how long would it take to completely melt the ice? The density and latent heat of fusion of ioe are 920 kg/m? and 3.34 x 10° sg, respectively. 1 air outlet temperature may be regulated by adjusting the tube mass flow rate, For the mbe 0 wi os length determined in part (2), compute and plot T., nae as a function of rt for 0.005 =n = 0.05 kpls, It sand Pr the duellingcaoted by this system requires approx- imately 0.05 ks of air at 16°C, what design and ‘operating conditions should be preseribed for the system? ABIL To cool a summer home without using a vapor compression refrigeration eycle, air is routed through 2 plastic pipe (= 0.15 Wim °K, D,=0.1Sm, D, = by vary lore the effect of 539 (0.17 mi) that is submerged in an adjoining body of water “The water temperature is nominally aT. = 17°C, and a convection coelicent of h,~ 1500 Wim? »K is main= tained athe outer surface ofthe pipe I air from the home enters the pipe at a temperature of T,, = 28°C and a volumetric flow rate of V, (06025 m/s, what pipe length Lis needed 10 provide a Aischarge temperature of Tg = 21°C? What i the fan power required to move the ar through this fengdh of Pipe if is inner surface is smooth? 8232 Water owing 42 kl through a 40-mm-iamter abe is ta be Beate from 2510 15°C by maintaining the tbe surface temperate at 100%. (a) Whats te equied abe engi for thse contions? [a]tn order 16 desig a water hein syste, we wish to consider sing tbe diameters nthe range fom 300 0mm, What are the requted tbe Tens foe water ow ces of I, 2, nd 3s? Represent this design information graphical. [Go te presse gradient as a function of tube ter forthe tee flow rats, Assume the ube wall sooth, 183 Consider the conditions associated with the hot water pipe of Problem 7.56, but now account forthe convec- tion resistance associated with water flow at 8 mean velocity of uy = 05 mis in the pipe. What isthe core- sponding daily cost of heat loss per meter of the uninsu- lated pipe? 834 A thick-walled, stainless steel (AISI 316) pipe of inside and outside diameters D, = 20 mm and D, = 40 mm is ‘heated electrically to provide a uniform heat generation rate of = 108 Win. The outer surface of the pipe is insulated, while water flows through the pipe at rate of i= OL gh (a) Ifthe water inlet temperature is Ty, = 20°C and the desired outlet temperate is Ty, 40°C, what is the required pipe length?