= TRANSFER METHODS HEAT TRANSFER BY CONDUCTION, 87 geal to the ratio of the wave series of important deductions concerning practical problems to be derived Bena (uss case iat errs any con Sung hen tems of he nts J thermal diffusivity formula, ‘Equation (144) becomes on | it Fe (1442) a p= Equation (139) permits the It follows from the above, for example, that the coefficient Fo remains constant in a certain transient conduction process. Therefore, heating or cooling (ie. the temperature field) does not vary when the thermal con- ductivity is doubled and the time halved. In other words, heating time Seeough a surface into a body fa) Bea ft (142) Fime is obtained by dividing fs inversely proportional to thermal conductivity. Furthermore, it follows Som Equation (144) that a layer of double the thickness necessitates four eh ate times as long a period, ie, heating time is proportional to the square of layer thickness. Example: A rolling mill operator proposes to push ingots through the ‘sith increasing thermal con- furnace in a double layer instead of the usual single layer, in order to om fof the swings. secs during onc period at a double the soaking time and promote more thorough heating. However, from Equation (130) fom Equations (144) or (83) it follows that equal heating of a double ‘wall ~ hickness’,Z, takes four times as long. Consequently the proposal may be Bs a the euriace anc at 8 ejected without any further ealelations. It should be noted that this holds fod only when the surface temperature ofthe ingot is suddenly inereased {Eruation 83), and in practice the greater the heat transfer to the ingot, the Teore valid che relationship. If, and this is almost always the cae, the ingot Ge plates treated ina heat transfer medium, Equations (98)-(to0) ae to be sed the theory of simile Replied, and therefore acurate values may be derived. sich together form a dimes "The Fourie number has a significant role amongst the dimensionless i the Reynolds number. (See Eeefficients, since the determination of this coefficient is a necessary and ple, particularly where thee Teicient condition for the equality of two temperature errors. AS ex- analyse systems using various {phined later on p. 8r, the identity of coefficients in the theory of similarity & of kg-m-h, or the Es ‘necessary but insufficient condition for the flow to be analogous. For Teese, equal Reynolds’ numbers of 3000, 5000 or some other value may msfer equations quoted above Heerespond cither to turbulent or laminar flows (see p. 63), or to a system has always appeared. In the Eeeolving a disturbed stream. ‘This kind of deviation does not occur in the IE L represents half of the wall ‘of the Fourier number. This coefficient is likely to be used more often fare. ~ E. Schmidt's approximate Method for Calculating any Arbitrary Heat Flor in a Plate ‘preceding exact calculations are, unfortunately, limited to a few J formulae, otherwise they would necessitate a large amount of ized work preparing curves and tables, which has not yet been mber (Fo), Bs) also contains this Fourier taken on its own, permet= TRANSFER METHODS HEAT TRANSFER BY CONDUCTION 37 wal t0 the ratio of the wave- series of important deductions concerning practical problems to be derived eto Equation (115), caus without recourse to any equation. Substituting the individual terms of the sbody of infinite extent is hermal diffusivity formula, Equation (144) becomes » Gan) trae oman (448) Hs Equation (139) permits the It follows from the above, for example, that the coefficient Fo remains constant in a certain transient conduction process. Therefore, heating or cooling (ie. the temperature field) does not vary when the thermal con- fa) Bru/t (42) ductivity is doubled and the time halved. In other words, heating time is inversely proportional to thermal conductivity. Furthermore, it follows from Equation (144) that a layer of double the thickness necessitates four A times as long a period, i. heating time is proportional to the square of = (040) isjer ticks, Example: A rolling mill operator proposes to push ingots through the furnace in a double layer instead of the usual single layer, in order to rough a surface into a body Fime is obtained by dividing ‘with increasing thermal con- on ty of the swings. ssee during one period at a double the soaking time and promote more thorough heating. However, from Equation (139) From Equations (144) or (83) it follows that equal heating of a double ‘wall ~ thickness’ L, takes four times as long. Consequently the proposal may be fangs at the surfuce and at © ejected without any further calculations. It should be noted that this holds 0d only when the surface temperature of the ingot is suddenly increased {Esquation 83), and in practice the greater the heat transfer to the ingot, the = Sore valid che relationship. If, and this is almost always the cae, the ingot Ge plates treated ina heat transfer medium, Equations (98)-(too) ae to be and the theory of similarity Replied, and therefore accurate values may be derived. sthich together form a dimen- "The Fourier number has a significant role amongst the dimensionless i the Reynolds number. (See Meefficients, since the determination of this coefficient is a necessary and ple, particularly where thee BeScient condition for the equality of two temperature errors. AS © analyse systems using varions Dhsned later on p. 81, the identity of coefcients in the theory of similar BE eccessary but insufcient condition for the flow to be analogous. For Teepe, equal Reynolds’ numbers of 3000, 5000 or some other value may isfer equations quoted abowe. Eeerespond either to turbulent or laminar flows (see p. 63), oF to a system Bias always appeared. In the Sving a disturbed stream, This kindof deviation does not oceut in the If L represents half of the wall ‘of the Fourier number. This coefficient is likely to be used more often fore. = of kg-m-h, or the Engl E Schwidts approximate Method for Calculating any Arbitrary Heat ater (Fo). Flos in a Plate 3) also contains this Fourie preceding exact calculations ae, unfortunately, Limited to a few formula, otherwise they would necessitate a large amount of ized work preparing curves and tables, which has not yet been taken on its own, permedauaitken. This is particularly true in cases where the initial temperature distribution ofthe periodically varying surface tempera hee an effec, srrntte, With given constant surface temperature, the ined temperature distibution inthe body isnot linea, In addition, sola available sau {or few types of surface temperature fluctuation, and anes do not for a number of eases cannot be completely sclved. I is expedient, there- fore in these eases to apply the approximate method ates iench tt based on the calculation of diferenoce On substigte of the ee acral claves 25, 00 ete of he hex Seaucibe ea coe 5: (Gee p. 32) by the finite values of Ax, A and ‘on, Equation (5 becomes Ar acoest ("Os cost Bee age (x45) int {2m Afs-ont signifies the increment of temperature 0° within the Sime interval At hours at a constant distance of x fe from chy {hicks and, similarly, to divide the required time inte sores ‘umber of intervals A. Then Equation (145) provides 4G on+1)ae ++ Onat¢m-r93) °F 7) Ai the above, is the numberof time intervals concerned, and m is the Peon iver. The time interval At houra cannot be freely selected, but is related tothe length of section (layer) Ay an siven by the equation ara, rasa mise It is necessary to accustom oneself to the large number of indices in Equation (£46), since its use depends on the inencdin’ recognition by The length of sane mA under the surface, and after (n + s)Ar hens [he length of section Ax canbe arbitrarily selected The ac the value of Ax chosen, the more accurate Equation (134) will he Generally adequate to divide the wall under consideration nto foes ane sections, cach Ax ft long, SE. Schmidt: A. Poppet? Conmemor. 89, on Day paper (Bein, 1924, J. Spring HEAT TRANSFER “3 Equation (146) would be the simplest face temperature. In these circumstanceg surface temperature during the entire time: hours is permitted, and does not compl erature field in any way, This shows the: ‘compared with accurate analytical proceds the surface temperatures is known, but the ‘medium (which is an independent variably erature is then determined by the approai TAs, + ag, = where h is the heat transfer coefficient of the wal, in Bru/f h °F As is the length unit (fr example, 3 9, _is the temperature of the gas (or k is the thermal conductivity of thet Gst,ax is the temperature after nAt hour and Ax fect under the surface, in Bl DERIVATION The temperature change with the time ¢ Ax ft under the surface (Fig. 19) Se Adeamae = Ons 3 45. Designation of the temperatures nt hours from Be change of temperature with the position BFime mAs is $B em 40 in Equation (145) signifies a [Fm berween two differences It is MP moat — (Caas,in+ x — Saat, max) — aSSFER METHODS es where the initial temperature face temperature has an effect, the initial temperature solutions are available = fluctuation, and these do not ce. The differential equations y solved. It is expedient, there- mate method of E. Schmidt," eences. On substitution of the Best conduction Equation (55) WB and so on, Equation (53) = (45) fof temperature 0° within the ge of ft from the surface. In gary to assume that the wall gn number of layers, each Ax [Sime into a certain number of EP asscm-vas)"F (146) ervals concerned, and m is the pars cannot be frecly selected ide, a8 given by the equation : G be large number of indices in the immediate recognition b feertain index. For example the plate under consideration i process, and at a distance of bess max Fepresents the t ge, and after (n +. 1)A¢ hous selected. The smaller the val (234) will be. Generally, ite gion into four or five sect Peper (Berlin, 1924, J. Spin ' HEAT TRANSFER BY CONDUCTION 39 Equation (146) would be the simplest application for a known wall sur- face temperature, In these circumstances, any arbitrary change in the surface temperature during the entire time being considered of (n + 1)At hours is permitted, and does not complicate the calculation of the tem- perature field in any way. This shows the advantage of the above method ‘compared with accurate analytical procedures, When none or only one of the surface temperatures is known, but the temperature of the surrounding medium (which is an independent variable) is specified, the surface tem- perature is then determined by the approximate equation KAxDy + AO, Cae = ET nee OF (048) where h is the heat transfer coefficient of the gas (or liquid) in respect of the wall, in Btu/f® h °F Ax is the length unit (for example, 1/5 of the wall thickness) in feet G is the temperature of the gas (or liquid) in °F is the thermal conductivity of the wall, in Beu/fth °F the temperature after mA hours from the start of the process, and AAs feet under the surface, in °F. (See example on p. 374.) DERIVATION The temperature change with the time (¥ = constant) in a layer which is Be f& under the surface (Fig. 13) is Mya Ona, max ~ Ooat,max F (49) Sraes.,| - nat Spat mde Srat.szl,| |W} bg Zan Faster. Fe 3. Designation of the temperatures in a sub-divided wall at a time At hours from the start. Fie henge of temperature with the position (¢ — constant) after a lapse ie iis Or-nat = Oaas,¢misise — Oaatynas F (250) Fem 829 in Equation (145) signifies a second difference, ie, the differ~ He Becrcen two differences. Tt is BP enas = (Pass, m-rax — Yast, mas) — Baas, mas — Oaatonni80)60 DETAILED CONSIDERATION OF HEAT TRANSFER METHODS Anat = Onst ims nax — Maat,ma + Onstan—nax (1S) ‘Substituting Equation (149) to (152) in Equation (145), we obtain Gorriaemx — Onatimax _ Maat iminiar ~ 20st, max + Prat, tm—nae ae a8 ‘ast (ssa) ‘This equation generally relates 01) At, ie. the temperature after (n — 1) ‘hours, and the temperature Ouas which existed previously At hours earlier. ‘The changes of the temperature field with time can therefore be computed. ‘The equation obtained is greatly simplified by introducing 2a At = (As)? as in Equation (147). Under these conditions, Equation (152) is trans- formed into Equation (146). An example calculation demonstrates. that Equation (146) is inauficient for the determination of surface temperatures 34° The latter may be computed in various ways from the temperature of the surrounding medium @,, and the heat transfer coefficient h. ‘The heat transferred is given by Equation (2) =a[2], pene» oso Similarly, we have a= he The unknown surface temperatu subtraction nats) (158) is derived from these wo equations after (ss) at the surface. Normally, these are uinown; however, the amount of he being transfered, q i known more frequently. Therefore, the roquced surface texmperatce may be given by Poste = (36) In the majority of instances, neither g nor 29/@x are known. However, the temperature Ogat,ax at a distance of Ax feet from the surface may be always computed from Equation (146). ‘The average temperature gradient betweea this point and the surface is then 20 _ Grats — nst,4x op Fn me ae one Gs7 Substituting from above the value of 89/2 in Equation (15), we obtain HOaat.s — Oat, ae Oaata = Og — KAS op 139) = bg Gs ‘The unknown surface temperature is deduced after solving the above fr aay — Me + Bas ae op Re hs = surface temperature, and has in fact 1+ HE ‘T TRANSFER 4] General Nota The character of heat taster by 3. This form of het transfer “3 Ioton ofthe hest-traeting parce ofthe medium ino coo bee, Depending upon whether dil i externally generated presse Botcr), « dsineon is made b tonto beat tranfr by con ston. Amor all uc, and the to thatthe hea transfered by cm Beh tat eanaered by conve Mite vellAnown expe ofa Bled wit water may be quoted bea the water a he top ofthe tube hse, Thien boca the bot wall Ibe density. Theetore heat may Bite by conduction oly a ed this aroun of ext iat ngs completely ifthe tx tube Bowed to tke up any arbiter tic, The nested wate ied the ice by convection, vo tha Sena ets maybe achieved inthe be Free convection due to heat = nother inporat example of Beep cen inte inwaliaon ofa Bee be at ground level rather tha rection currents, Never Ie cas of heat transfer by com Brenda thin boundary layer tne meightourhod of tou i igoas, tnd moves in Rebel tis layer by condSEANSFER METHODS HEAT TRANSFER BY CONyECTION a Ecchiidt used a somewhat diferent method for the determination of ss eer temperature, and has in fac deveiene ny ‘staphical method for its Pscomnax (151) determination, Es), we obtain BE Ela innse 5 7 Spersture after (n — 1) Ssicoaly Ae hours caer, Stherefore be computed. aducing motion of the he Ration (152) is trans- Particles of the ion demonstrates that face, Depending of sartice temperatures an externally gen. Feoeficient h, The heat ee): at-transferring medium, medium into contact wit upon whether this inte erated Which continuously brings new th the heating or cooling nal motion is fore (ase Se fwo equations after ample of a picce of ice at the hott ty be quoted here. 1 tom of a test tube test tube is heated atthe top, a long time without melting 8 at the top & the amount of heat ities, the scroll Sate completely if the teat tube is heated ect take up any arbitrary postion, fr fon Be arc®, The heated water immediately movie S the ice by convection, so that the nce, floating on the ipwards and transfers heat latter melts within a short time, Another important example of cond; luction compared wit $y be scen in the installation of hot 1. disregarded even to the flow theory of pes of stream in the is ayer is always more heat to be transferred ly then may convectiveSEANSFER METHODS HEAT TRANSFER BY CONyECTION a Ecchiidt used a somewhat diferent method for the determination of ss eer temperature, and has in fac deveiene ny ‘staphical method for its Pscomnax (151) determination, Es), we obtain BE Ela innse 5 7 Spersture after (n — 1) Ssicoaly Ae hours caer, Stherefore be computed. aducing motion of the he Ration (152) is trans- Particles of the ion demonstrates that face, Depending of sartice temperatures an externally gen. Feoeficient h, The heat ee): at-transferring medium, medium into contact wit upon whether this inte erated Which continuously brings new th the heating or cooling nal motion is fore (ase Se fwo equations after ample of a picce of ice at the hott ty be quoted here. 1 tom of a test tube test tube is heated atthe top, a long time without melting 8 at the top & the amount of heat ities, the scroll Sate completely if the teat tube is heated ect take up any arbitrary postion, fr fon Be arc®, The heated water immediately movie S the ice by convection, so that the nce, floating on the ipwards and transfers heat latter melts within a short time, Another important example of cond; luction compared wit $y be scen in the installation of hot 1. disregarded even to the flow theory of pes of stream in the is ayer is always more heat to be transferred ly then may convective