cae 6 Seer Boiling and Condensation a a =~ Many heat transfer activites in industrial practice involve a change of phase lik sy oF liquid to solid, and liquid to vapour or vapour to liquid. The examples are numess,! of a solid is common to many polymer processing operations, where the polymer beagy ly before moulding. Freezing of water in an ice plant or conversion of drops of urea-mely tl in a urea prilling tower involves a change of phase from liquid to solid. A change pe ty the liquid to the vapour or from the vapour to the liquid is encountered much more p= For example, a distillation column is provided with a reboiler at the bottom and Sha, the top. In many chemical processes the feed liquid is vaporized before it enters ee example, liquid ammonia is vaporized before it is mixed with air and fed to te wt oxidation reactor in a nitric acid plant). Vaporization of water to produce high pressures condensation of steam leaving the turbine ae the two major heat transfer operation ns power plant. md Itis important to study the fundamentals ofthe phenomena of boiling and condensaig to Jearing the design methodologies of related heat transfer equipment. The rate of beat eon during boiling or condensation is generally much faster than the rate of conduction ot iciey heat transfer. This is because in conduction or convection, transport of heat occurs fom a ss (er an interface) to the bulk of a medium. But in boiling or condensation, the hex vie phenomenon is offen limited to a very narrow region near the surface. Change of phar example, from liquid to vapour or from vapour to liquid, occurs within this region, Neverieles convection plays a major role in heat transfer during boiling or condensation. 6.1 THE BOILING PHENOMENON When some water is taken in a beaker and placed on a heater, heat transfer to the water fom te hot bottom of the beaker occurs by free convection at the beginning. Wher the wate ss considerably heated, evaporation from the free surface becomes visible. As heating proceeds, ie water eventually reaches its boiling point and the formation of small vapour bubbles sats However, at this stage most of the tiny bubbles which are formed at the bottom surface colle before reaching the top. This means that although the water at the bottom is at its boiling pats the temperature of the upper layers is still below the boiling point. The cooling of the upper 65 occurs by evaporation heat loss at the free surface. However, this situation does not come it Jong, and soon the liquid starts boiling. Vapour bubbles are generated at the bottom, grow i $0 and rise up vigorously through the liquid. This is a simple and brief picture of the differen st that appear during heating of a liquid from the cold to the boiling condition. As 2 mato the temperature of a boiling liquid remains above its true boiling point, that is, best superheat. The amount of energy a liquid element possesses because of the superheat ¥° geOILING AND CONDINS \ TON: im when a vapour bubble in contact with the clement grows in size and draws the heat of Vaporization from the superheated liquid element The course of heating of a liquid first leading to evaporation and then to boiling can be followed quantitatively if the temperature of the hot surface, and consequently the heat flux. is varied systematically over a wide range. There are quite a few classical works on the boiling phenomenon (e.g. Nukiyama, 1934; Cichelli and Bonilla, 1945; Farber and Scorah, 1948; Bromley, 1950). An electrically heated surface, for example. a wire was used by many workers because of the ease with which the rate of heat input and the temperature can be controlled. The experimental arrangement including the results of Farber and Scorah (1948) is described here briefly. A schematic of their experimental set-up is shown in Fig. 6.1. An electrically heated wire immersed in the liquid supplies the heat required for boiling. The rate of heat input is To voltage source Fig. 6.1 Schematic of a set-up for experimental study of boiling heat transfer. calculated from the measured values of the applied voltage and the current. This measurement also gives the resistance of the wire and hence the wire temperature T,. The surface heat flux q, can be calculated from the rate of electrical heat input and the area of the wire. A set of temperature-flux data can be generated by changing the applied voltage over a reasonably wide re. ine boiling liquid, the temperature of the hot surface must be higher than the boiling point of the liquid, i.e. 7, > Tian Where T,y is the temperature of the saturated liquid or the boiling point of the liquid. The temperature driving force for the boiling liquid which is also called the “excess temperature’ is defined as follows. Excess temperature, 7, = T, — Tuy (6.1) A plot of the logarithm of the heat transfer coefficient at the wire surface (A) vs. the excess temperature (T,) typically looks like Fig. 6.2(a). The modes or regimes of boiling over different ranges of excess temperature are identified in this figure. It is to be noted that these results are characteristic of poo! boiling. Pool boiling refers to boiling of a quiescent liquid (that is, a liquid pool) in which the motion is caused by free convection and by the formation, growth, detachment and rise of the bubbles, Pool boiling of the liquid occurs in the experimental arrangement shown in Fig. 6.1. There is another kind of boiling, as distinguished from pool boiling, called forced convection boiling in which motion in the boiling medium is caused by an external means (like @ pump) in addition to the factors involved in pool boiling. agAz HEAT TRANSPER-PRINCIPLES AND APPLICATIONS on 1.0 10 190 1000 T.°F 9 ———> Fig. 6.2(a) Analysis of che boiling curve (Farber and Seorah, 1948). [Zone I: Interfacial eva ‘Nucleate boiling starts, most of the bubbles collapse before leaving the liquid, pie ih boiling, bubbles do not collapse in the liquid; IV: Partial nucleate boiling, unst ae boiling; V: Stable film boiling; VI: Radiation becomes important } He film Over the section 4—B (regime 1) in Fig. 6.2(a), the wire temperature is slightly above saturation temperature of the liquid (7, < 2°C). The rate of vaporization of the liquid oF the ry of formation of vapour bubbles is pretty small, Motion in the liquid medium is caused prnciay by free convection. The hot liquid vaporizes only at the free top surface. This regime is called he interfacial evaporation regime. The section B-D is characterized by nucleate boiling. Here the excess temperature varies from T. = 2°C at B to T. = 35°C at D. The nucleate boiling regime may be subdivided inte two smaller regimes. Over the section B-C (T, ~ 2 to 6°C), representing regime Ul, isolated bubbles are formed at the surface of the wire, but most of them collapse before reaching the free suritce of the liquid. Because bubble formation at a reasonable rate starts at B, this point indicates the onset of nucleate boiling. Over the remaining section, C-D, which is regime Ill, 2 large runber of vapour bubbles are generated that vigorously rise through the liquid and escape at the free surface. Near the point D, the vapour bubbles rise as jets or columns and then form bigger bubble slugs. Breakage and coalescence of vapour bubbles also occur because of the intense motion generated in the liquid When the excess temperature exceeds that at the point D, the mode of boiling gradually changes from nucleate to film boiling. Here the bubbles are formed in so large numibers that they coalesce right on the heating surface to form a film of vapour. This greatly reduces the beat transfer coefficient. The film boiling regime may be subdivided into regimes IV, V, and VI In regime IV (section D-E), the formation of the vapour film on the wire surface is not complet, and the film is discontinuous. The film may even disappear momentarily and then reappear. Ths regime represents partial nucleate boiling or transition boiling. Beyond the point E, stats the regime V (section E-F) where stable film boiling occurs and a continuous vapour film alvays blankets the heating surface. The surface heat flux decreases significantly although the excess temperature may be as high as a few hundred degrees. After the point F, the wire temperate etlBOILING AND CONDENSATION as increases so much that radiative heat transfer becomes important. This is characteristic of regime ‘VI (section F-G). If the excess temperature 7, is too high, the heating wire may even melt. This. in fact, occurred in the boiling heat transfer experiments of Nukiyama (1934) who used a nichrome heating wire that was burt out at a high temperature, But Farber and Scorah (1948) used a platinum wire and could reach a higher temperature and heat flux without burn-out of the wire. The melting or burn-out of the heat transfer surface is sometimes called the boiling crisis. In some cases, a direct shift from the point D to the point G may occur as the temperature increases. The excess temperature and the heat flux in some clectrical heating devices or in a nuclear reactor may be so large that it may lead to boiling crisis. This possibility should be taken into account at the design stage. The boiling phenomena in the different regimes are also shown in Fig. 6.2(b) AT=11-65°F : CH oF 6 AT'= 400-1000°F AT >1000°F Fig. 6.2(b) Schematic of the 1948). 1g phenomenon in different boiling regimes (after Farber and Scorah, It is pertinent at this point to mention the so-called Leidenfrost phenomenon. When water droplets fall on a sufficiently hot plate, the droplets bounce up and down a few times before they disappear by vaporization. This phenomenon is common in the kitchen where it occurs when a little water falls un a hot oily pan. If some liquid air or liquid nitrogen spills on the Moor of a laboratory, the droplets move so briskly that they appear to be dancing before they disappear The above phenomenon was first observed by Leidentrost in 1756. The mechanism is related to film boiling. When a water droplet lands on a hot surface, a film of steam immediately forns eeei 174 HEAT TRANSFER—PRINCIPLES AND APPLICATIONS between the droplet and the hot surface, Because the thickness of this vapour film increase, quickly by further vaporization of the liquid, the droplet experiences an upward thrus,.. bounces. It falls back on the surface and bounces again by the same mechanism and reduce, * size at every touch with the surface. The droplet disappears eventually. So the Leigat® phenomenon is caused by film boiling and occurs for evaporating droplets on a hot sup corresponding to regime V of the boiling curve. a 6.2 HYSTERESIS IN THE BOILING CURVE This phenomenon was first reported by Nukiyama (1934) in his experiments on pool boitig Wvates using a hot nichrome wire. Nukiyama reported a boiling curve shown in Fig, 63 in gig the surface heat flux is plotted vs. the excess temperature on the logarithmic scale. Although te log 9, Excess temperature, 7, Fig. 6.3. Hysteresis in a boiling curve. did not identify the different regimes as distinctly as Farber and Scorah (1948) did, he observed a hysteresis phenomenon in the boiling curve. As the excess temperature 7, increased, the surface heat flux q, followed the path A~B-C reaching a maximum at C, On further increase in the power input to the wire, the excess temperature suddenly jumped to a very high value at £ following te path C-E, When the power input to the wire was reduced, the drop in 7, and the flux q, followed the path E-D-B-A. Had there been no jump in the temperature after the point C during heating, the boiling curve should have been like 4-B-C-D-E, although C-D represents an unstble regime. The point C in Fig. 6.3 corresponds to the maximum heat flux or the critical heat flux. The Leidenfrost phenomenon of an evaporating droplet also occurs at the point C. 6.3 THE MECHANISM OF NUCLEATE BOILING Nucleate boiling proceeds through the formation of bubble nuclei in a superheated liquid (6 * has been stated earlier, a boiling liquid always has a degree of superheat), growth of the bubbles and their escape from the liquid. There are quite a few questions associated with this phenoment® Does a bubble form in the bulk of the superheated liquid or on the hot surface with which be sie aROILING AND CONDENSATION: 175 jiquid isin contact? How fo determine the rate of bubble growth? After it is formed on the hot gorface, when does the bubble get detached from if? Not all of these questions will be addressed here, Our discussion will rather be confined to the phenomenon of formation of a bubble nucleus Formation of a tiny bubble in a liquid is called nucleation. A supetheated liquid is said to be ina metastable condition, i.c. it cannot exist in this condition indefinitely and any disturbance in the system tends to bring it to the thermodynamic equilibrium state, A bubble itself in a superheated liquid may be considered to be a source of disturbance, If a bubble nucleus is formed in the bulk of a superheated liquid, it is called homogeneous nucleation. On the other hand, 1f a aucleus is formed on a hot surface or on a solid particle suspended in the superheated liquid, we call it heterogeneous nucleation. ‘The condition of mechanical equilibrium at a liquid-vapour (or a liquid-gas) interface 1s given by the Gibbs equation, n-n-ot+d] (6.2) A 4% where p, and p, are the pressures exerted by the vapour and the liquid phases at the interface, js the surface tension of the liquid, and r; and r; are the principal radii of curvature of the interface. If the interface is spherical (i.e. if we consider a spherical vapour bubble in a liquid), r= r= F (say), then Eq. (6.2) reduces to Py Pr (6.3) It is clear from the above equation that the pressure inside the bubble is higher than that in the neighbouring liquid. The pressure difference, p, — pj, can be related to the liquid superheat by using the Clausius~Clapeyron equation assuming that the vapour phase behaves like an ideal gas. Thus, we have dip __, ar RT 9) where L, is the molar heat of vaporization and R the gas constant. Let us assume that 7) is the temperature of the superheated liquid, p; is the pressure in the superheated liquid, and Ty is the corresponding boiling point of the liquid, The vapour in the bubble is in thermal equilibrium with the liquid, ie. the vapour temperature is also 7). As the Vapour is saturated, the pressure in the bubble p, is equal to the vapour pressure of the liquid at the temperature 7). Let us integrate Eq. (6.4) within the appropriate limits. Thus, we have 2 - Api a7 , ane ra PE Because6 TIEAT TRANSFER PRINCIPLES AND APPLICATIONS Therefore, 2 m R Substituting for p, — p; from Eq, (6.3), we get 25, Ri. 25 Ra TT ry by TB by (63, The radius of a vapour bubble in mechanical equilibrium in a superheated liquid can be calc from the above equation (Rohsenow,, 1966). For example, if we consider 2 pool of wat, 10°C superheat at atmospheric pressure, we have ih T= 110°C, T= 100°C, py = 1.013 bar, = 0.063 N/m, by = 4.07 x 10" kieg-mol, R= 53, JIIK kg-mol ‘Therefore, we have from Eq. (6.5) (2)(0.063)_ (8.314)(373)" Hla = WC = 10) 407 x 10" or = 3.55 x10 m What happens if'a bubble nucleus of a radius smaller than above, say 10-* m, is generated in ty liquid by some way? This nucleus cannot remain in mechanical equilibrium if the supers only 10°C. It is sure to collapse as per the theory. Equation (6.5) thus explains why some tay bubbles collapse in the liquid if the degree of superheat is insufficient. Equation (6.5) also tells us that a high degree of superheat is necessary for the generation ot a tiny nucleus in the bulk liquid. However, our experience shows that a liquid boils vigorouiy even at a superheat of several degrees. The reason is that bubble nuclei are formed by heterogeneous nucleation at the minute cavities and crevices present at the heating surface Ths is moze favoured on a rough surface than it is on a smooth one. Bumping of boiling water in glass beaker occurs because it has 2 smooth surface. This docs not happen when water bab ix ‘an aluminium pan. Hans and Griffith (1965) developed an expression for the critical sizeof 1 nucleation site (ie. a cavity or a crevice on the surface) that will produce a bubble aucleus at given degree of supetheat. Theoretical and experimental studies on bubble growth and detachment have been reported by a large number of researchers, The different stages of nucleation and bubble growth at a cavity are shown schematically in Fig. 6.4. Before detachment um Fig. 6.4 Formation of bubbles over cavities.HOILING AND CONDENSATION 7 Example 61 Consider mucleate poot ty ‘ulate oiling of ethyl acetate at 1 atm pressu am (a) If the superheat in the liquid is S°C, calculate the dames: of @ cavity on the belling autiny sikh roduces a bubble nucleus that docs not collapse on cade of 1.0 jm, what degree of sopethent gue evar siren acai es ater dtchrrent fom the essary $0 that a bubble nucleus grows in size SOLUTION The radius of a bul ibble nucleus in can be determined from Eq, (6.5 mechanical equilibrium in a superheated liquid ). This is our . aie working equation for this problem. 2 Ty = Ty, = 20 Rl rm oh, @ bubble nucleus is equal t on ae equal to the radius of the cavity at the (2) Boiling point of ethyl acetate at 1 atm pressure is, Tw = 350 K (77°C), The following physical data are used in the calculation [for the correlations used for estimation of physical Properties, see Reid et al. (1988)], Liquid temperature, 7; = 350 + 5 = 355K ise the liquid has 5°C superheat), — The vapour pressure of ethyl acetate at this temperature can be calculated by using the Antoine equation 124495 log p, = 4.22658 — 124495 — + 21788 here p is the vapour pressure in bar, and the temperature in °C. At O= 77 + 5 = 82°C, we get p= 1.189 bar. The surface tension of the liquid is given by & = 26,29 — 0.11618 dynelem At 0 = 82°C, we get = 16.8 dyne/em = 1.68 x 102 Nim Molar heat of vaporization, L, = 33,605 ki/kg-mol = 3.3605 x 107 The universal gas constant, R = 0.08314 m? bar/kg-mol K. Substituting the values of the various quantities in the working equation, we get Ikg-mol (355 ~ 350) k — 2X68 x 10°? N/m)(0.08314 m? bar/ kg-mol K)3S0 x 350 Ky (rm)(.189 bar)(33605 x 10” J/kg-mol) or r=1713 x 106m =[1713 pm ‘Therefore, a bubble nucleus larger than 1.713 jum that has been detached from a cavity will ot collapse in the liquid. (b) In this case, r TK = (7,- um (given) = 10°* m. It is required to calculate 7). At the temperature 273)°C, the vapour pressure and surface tension of ethyl acetate are given by 1244.95 = = be tte ren 42 aea ‘The values of Tay, Ly» and R are as in part (a) above. Putting all these quent, expressions in the working equation, [Eq. (6.5)], we get a transcendental equation in the yay, 7, This equation, given below, can be solved by any suitable technique (for example, the y<°™ Rhapson method). Moy, (2{0.02629 ~ 1161 x 107%(7) ~ 273)K0.08314)(350)? (10-)(3.3605 x 10° )[108228° — AIT 279 + 217A 178 HEAT TRANSFER—PRINCIPLES AND APPLICATIONS. = 0.02629 - 1.161 10+ (7) ~ 273), Nim T,- 350= The solution is 7, = 358 K or 85°C; —_superheat of the liquid ~ 85 - 77 = [®C ‘Therefore, if the liquid has a superheat of 8°C, ‘a bubble nucleus detached from a 1 um cavity will not collapse. It will rather grow in size after detachment. Tad, 6.4 CORRELATIONS FOR POOL BOILING HEAT TRANSFER While the preceding discussion targetted the physical mechanism of boiling phenomenon, not help us in calculating the rate of heat transfer in a boiling liquid. The second aspects nee important in process calculations and in the design of liquid boiling equipment. A large minis of empirical and semi-empirical correlations available in the literature can be used for i, purpose. The more important and practically useful correlations are given below. tes 6.4.1 Nucleate Boiling Liquid boiling equipment in the process industry is designed to operate in the nucleate being regime because it is in this regime that heat transfer occurs more effectively. The equigmey configuration is mostly of the shell-and-tube type (see Chapter 8). The liquid boils either in the shell or in the tube, and the heating medium is steam or any other hot fluid. Therefore, correlation, on boiling in or outside a horizontal or vertical tube are most important for the purpose of desig. Also, because the rate of boiling, and for that matter the rate of heat transfer, depends upon quie a few major factors like the characteristics of the boiling surface, physical properties of the liquid (0. cp KM, o, ete.) pressure, surface temperature (7,), etc., no single correlation can be satisfactory for all situations. Further, it is difficult to quantitatively take into account the fis factor (i.e. surface characteristics, which include the size distribution of cavities on the surfise) Even if this factor is included in a correlation, it cannot be used unless the surface characteristics of the material used in the boiler are known. Rohsenow (1952) proposed the following correlation for heat flux in nucleate pool boiling based on his experimental data on boiling of quite a few liquids. aye r vt 4 = ty | EA = 8 2] ee 69 s Cy Ly (Pr)) where Hy is the liquid viscosity (Pa s) , is the enthalpy of vaporization of the liquid (J/kg)BOILING AND CONDENSATION 19 g is the acceleration due to gravity (mis) p is the density (kg/m') @ is the surface tension (N/m) 6p is the specific heat (kg °C) T, is the excess temperature = 7, — 7. (°C) T, is the temperature of the boiling «12 E 1G surface (K) Pr, is the liquid Prandtl number g, is the heat flux (W/m?) the subscripts mean: 1 = liquid, and y = vapour cris s the constant C, which depend upon the surface characteristics, ar listed in Table or a few materials. ‘The value of the exponent n is: n = 1 for water, and n = 1.7 for ‘most organic liquids. The liquid properties should be taken at the saturation temperature Ty. Ertor in the prediction is pretty high, and may be even 50% in some eases, Table 6.1 Values of Cy in Eq, (6.6) for several fluid-surface combinations Fluid-suface G n Water-stainless steel mechanically polished 0,013 10 ground and polished 0.006 10 chemically etched. 0.013 10 Water-brass 0.006 10 ‘Water~copper 0.013 1.0 Benzene-chromium: 0.010 LT -Pentane-chromium 0.018 10 Ethyl alcohol—chromium 0.0027 Lr iso-Propyl alcohol-copper 0.0025 10 Carbon tetrachloride-copper 0.013 Lo ‘Aqueous K,CO, (35%}-copper 90.0054 10. Another useful correlation for the heat flux proposed by Mostinski has been cited by Kakac (1981). A conservative form of his correlation is hy = 0.00341 (P.)° (TY (P,) Wim? °C (6.7) where P. is the critical pressure in bar P, is the reduced pressure T= T= Tox in °C 642 Critical Heat Flux ‘The estimation of critical heat flux is very important in the design of a boiling equipment. If the temperature of the heating fluid is much higher than the boiling of the liquid, film boiling of the liquid may occur. So the critical heat flux corresponding to the point C in Fig. 6.3 is calculated first, and the temperature of the heating fluid may be selected accordingly so as to maintain boiling in the nucleate regime, We will cite here two important correlations for the critical heat flux.*, (i) A simple correlation for the critical heat MUX rx WAS developed by Lienhard et a) Ag; by fitting their experimental data to a theoretical equation suggested by Zuber (195g) "7 a8 HEAT TRANSFER—PRINCIPLES: "AND APPLICATIONS 4 s BCE) = P| yyjeet Fax = 9149 wal we A (64, Where the terms have significances and units as in Eq. (6.6) (ii) Peak boiling heat flux on a horizontal cylinder can be estimated by using the Flog correlation of Sun and Lienhard (1970). dix Toa [089 + 227 exp (-44V""] for r*> 01s (69, we = dimensionless radius = [ee= 2) (6 where Guax, the peak heat flux on an infinite horizontal plate, is given by Ws Ginx = 0.131, Yp,[oe(A -,)]* Wen? 61 6.4.3 Stable Film Boiling ‘The arm E-F of the boiling curve, Fig, 6.2(a), represents stable film boiling, sometimes called ‘film pool boiling’. The correlation of Bromley (1950) can be used to calculate the boiling ten transfer coefficient in this regime. Bale ~ adel + 04%) | hy = 0.62 : du, Te Wim? °C ry where the subscript v denotes the vapour phase, and d is the tube diameter. The above equation is applicable to film boiling outside horizontal tubes. However, this equation can be used fora vertical plate after replacing d by L (plate height) and the constant 0.62 by 0.7. If the surface temperature is high enough to make the contribution of radiative heat transfer important, the ‘total heat transfer’ coefficient may be calculated by using the following equation A= hgfh)? + hy (683) where A, is the radiation heat transfer coefficient, and Ay is given by Eq. (6.12). Example 6.2 Calculate the rate of boiling of water in a 0.35 m diameter stainless steel pan at 1 atm pressure if the bottom of the pan is maintained at 115°C. SOLUTION It is expected that water will boil in the nucleate poo! boiling regime at an excess temperature 7, = 115 ~ 100 = 15°C. We will first calculate the rate of boiling using the Rohsenow’s correlation, (Eq, (6.6)], and then check the result by comparing with the critical het flux. The various physical properties involved are taken at the saturation temperature (ie. 100°C) under the given conditions.> 18s BOILING AND CONDENSATION For water at 100°C: fip= 2.79 10 Nm’ ey = 4.22 kuleg °C; py = 958 kgin'; L,* 2257 uieg: > 0.089 Nim; Pry» 1.76 For saturated steam: p, “ 0.5985 kg/m’, Py For the stainless steel pan, take Cy = 0.013, and m= | (Table 6.1). Also, g ~ 98 7's Putting these values in Eq. (6.6), we have . 4 (9.84958 - 05955)? (422,09) RRM vas 0059 | (0013922571176) = 462.5 kWim? , 462.5 kW/m? ‘ The rate of boiling = FEF g ~ 0.205 Kym's ‘Area of the pan = (72/4)(0.35)? = 0.0962 m? The total rate of boiling = (0.205)(0,0962) = 0.01972 kg/s = [71 kg/h Let us now calculate the critical heat flux using the Lienhard’s correlation, [Eq. (6.8)} 14 dn = 0149 Lp, [aaa] PB v4 = (0.149)(2257)(0.5955) [memes tse = 1259 kWim? (05955) ‘The actual heat flux is well below the critical heat flux value. So it is reasonable to assume that nucleate pool boiling of water will occur under the given conditions. The boiling curve [see Fig, 6.2(b)] also indicates that pool boiling will occur at an excess temperature, T, = 15°C It will be interesting to calculate the boiling heat flux using the simple correlation of Mostinski [Eq. (6.7)] and to compare it with the value predicted by the Rohsenow’s equation. hy = 0.00341(P.)(7.) (P,)* 013 bar. For water, P,= 221.2 bar, and the reduced pressure, ‘The given pressure = 1 atm = P, = 1.013/221.2 = 0.00458. ‘Therefore, rg = (0.00341)(221.2)?°(15)?°(0,00458)%* = 22 kW/m? °C Heat flux, g, = hy (7; — Tex) = hsTs = (22)(15) = 330 kWh? This value compares reasonably well with the value of 462.5 kW/m? predicted by Rohsenow’s equation. Example 63 Formaldehyde is one of the raw materials for the manufacture of a variety of resins like PF (phenol-formaldehyde), UF (urea-formaldehyde), MF (melamine-formaldehyde), etc. The technical method of production of formaldehyde is catalytic air oxidation of methanol (finely divided silver is one of the catalysts used).1a HIBAT TRANSEDR PRINCIPLES AND APPLICATIONS: | Ita formaldehyde plant, « hot stream of flue yas is used to venetie at 18 btn shell-and-tube vaporizer at a gauge pressure of 0,8 bar so that etal ie sate heat my recovered, The vaporized methanol is then mixed with air and fed fobs oe Inthe va the hot gas flows through the tubes and the methanol boils cr I The tubes ae 14 fe 12 BWG (AISI 3161), and the surface temperature is 200°C. Calculate the rate of boiling SOLUTION The problem does not indicate the boiling regime; it does not provide any Pheg Property data, We will follow the following steps in order to solve the problem, () Calculate the boiling point of methanol at 0.8 bar gauge (1.813 bar absolute) press th Antoine equation with the constants taken from the literature may be used, (i) Calculate the temperature excess, 7) (ii) If the experimentel boiling curve fora particular liquid is not available, prediction ty boiling regime (nucleate or film) becomes difficult. As we do not have the boiling Curve for methanol, we do the following. (%) Calculate (a) the heat fx sssuming nucleate poo! boiling and (b) the critical fay, Critical heat flux is larger, then the assumption of nucleate boiling is reasonable ad 5g further calculation is necessary, (©) If the calculated critical heat flux appears to be less than that obtained by assum nucleate boiling, obviously the system is not operating in the pool boiling regime Therefore, @ suitable boiling correlation should be used to calculate the heat fig and the boiling rate. (vi) Physical property data for many common fluids are available in Reid et al. (1988), {If the data necessary for a particular problem are not available, these can be estimated by using the available correlations. Reid et al. (1988) provide an excellent documentation and assessment of correlations for the estimation of physical properties of pure fluids and mixtures. The use of a few such correlations is illustrated in this example.) Estimation of liquid and vapour properties necessary for the given problem ‘To calculate the boiling point of methanol at 1.813 bar, we use the Antoine equation Inp, =4 ~ 3 ~ 15673. 42346 FE T where p, is the vapour pressure in bar and "is the absolute temperature, Putting p, = 1.813 bar, we get T= 353.7 K = saturation temperature, Ta, Tube wall temperature = 200°C = 473 K ‘Temperature excess, 7, = 473 — 353.7 = 119.3°C Mean vapour temperature = 473 * 3537 _ 414 5 2i | HOLING AND CON, NSA TION The relevant th 3 ies ( evaluated atthe saturation temperature FOF the calculation of film boiling heat fy ni Pool boiling heat fhix) ate 'X) ae estimated at the tom eet {required forthe calculation ofthe Liquid properties (Pr ¢ 'mperatire uid propery ‘ackett technique (Reid et al P. 67), Molat volume of methanol at the boiling point, p Rhy yre-ay i) tetas ‘ For methanol: ~ critical temperature — S126 = 0.69; Z,, 56 ~ 0.08775, yo Ks Te = reduced temperature = .290: Constant = 0.08314 m? ph; : TT, = 353.7) ‘critica nt ctor = 0.556, R = the fl . ar/g-mol K; P= critical pressure = 80.9 bar. ee Substituting for the different quantities in Eq (i), we get . Molar volume, ¥, = 0.0461 m’/kg-mol methanol Molecular Weight of = 32; Saturated liquid density, p, Temperature, K. Contribution of i (I/g-mol °C) CH, wm oi gap 348 45.8 617 1075 373 48.3 Tl 119.4 By linear interpolation at 353.7 K, we have 53.7 — 341 ep = 107.5 + (119.4 ~ 107.5) 3537 = 348 sa 373 — 34g ~ 110.2 Kikke-mol °C = 3.444 kiike °¢ (©) Surface tension (@) at 353.7°C: Data available (Reid et al., p. 635) are Temp, K: 313333 @ (dyne/em): 20.96 19.4 By linear extrapolation, at 353.7°C, o = 17.8 dynelem = 1.78 x 102 Nim (4) Liquid viseosity (14): If the viscosit ty of a liquid is available at one temperature, that at another temperature can be calculated usin; i the following equation ‘Reid et al., p. 439). : ai TT [OP = (yr) D0 if . , ively. id 1 are the viscosities (in cP) at temperatures 7 and 1%, respectivel ea at 25°C (298 K) is u’ = 0.55 cP (ftom Reid et al.). Therefore, at 353.7 K, we get 3947 - 298 233 021 (0.55).os it NEAT TRANSEER- PRINCIPLES AND APPLICATIONS or p= 0.274 oP That is, Liquid viscosity, jy = 0.274 cP = 2,74 x 104 kg/m 5 (€) Prandtl number: 1 and cy have been obtained before. The thermal Conductiy, methanol is calculated by using the following equation (Reid et al, Table 10.5, P54), Ny k= a+ bT+ eT Wim °C; a= 0.3225, b =~ 4,785 x 10-4 1 E= 1.168 x 19 At T = 353.7 K, we have Sot _ (3494)2.74 x 104) _ hy 0168 “a6 (f) Heat of vaporization (,): The heat of vaporization of methanol at its normal boiling (337.7 K) is 1100 ki/kg. Because the boiling temperature of methanol at the given pressure =" much different from its normal boiling point, we use the above value of the heat of Vaporiat ite. L, = 1100 ki/kg. ", 4, = 0.168 W/m °C; and Pr, = Properties of methanol vapour (ky Py Gy and) at the mean vapour temperature, 413 oe (2) Density of the vapour (p,): We assume ideal gas behaviour (which is Teasonable consideration of the given pressure), and put a P= 1.813 bar, T= 413.3 K, R = 0.08314 m? bar/kg-mol K RT Molar volume, V = >" 18.95 m/kg-mol Density of the vapour, p, = Ton -688 kg/m? (mol. wt. = 32) (©) Thermal conductivity of the vapour (k,): We use the equation, K=a+ ort el? + dT Wim °C (Reid et al, Table 10.3, p. $16) @=~1.797 x 107, b = 4.167 x 10%, ¢ = 1214 107, Putting T= 413.3 K, we get d =-5,184 x 10" 4, = 0.0265 Wim °C (c) Heat capacity of the vapour (cj): We use the equation, Sy = 9+ bT + eT? + dT? keirkg-mol °C where @=21.15, b=7.092x 107, ¢=2.589 x 10%, d = -2.852 x 10° (Reid et al, p. 617), ALT= 4133 K, cy, ~ 52.86 ki/kg-mol K = 1652 kg K (4) Viscosity of vapour (y,): The following data are available (Reid et al) Temperature, °C; 67 177 Viscosity, uP: 112 132 ‘We use linear extrapolation (for simplicity, and because the temperature range involved is m* large) to caleulate the viscosity at 413.3 K (140.3°C). Thus, Hi, = 1.364 x 10° kg/m s. The necessary physical properties of the liquid and of the vapour having been estimated, "¢ now proceed to calculate the heat flux.ROILING AND CONDENSATION 195 If nucleate pool boiling occurs (it does not appear likely here in view of the large excess temperature of 119.3°C), the Rohsenow’s equation, Eq. (6.6), may be used. Take Cy ~ 0.0027 and n= 1.7 (sce Table 6.1; in the absence of specific data for the methanol stainless steel combination, we use the data for ethanol-stainless steel). Putting the values of the various quantities in Eq. (6.6), we get ((9.8)(694 1.78 x aay? (3.444)119.3) iil (0.0027)(1100)(5.6)'7 q-Q74x 11100 = 75.3 MWim? To calculate the critical heat flux, we use the Sun and Lienhard correlation (Eqs. (6.9) (.10)}. Gun = 1131, TOR - AI) = (0.131)(1100)(1.688)'7{(1,78 x 10)(9.8)(694 — 1.688)]"* = 620.5 kW/m? V2 (em [ee = a] o 2 = (0.016) as, = 23) = 9.88 178 x 107 (r = outer radius of 1-1/4 inch 12 BWG tube = 0.016 m) ax = Gras[089 + 2.27 exp (34407) = (620.5)[0.89 + 2.27 exp{(-3.44)(9.88)!"}] = 552.2 kW/m? It appears that the calculated heat flux for nucleate pool boiling exceeds the critical heat flux which is not acceptable. So the assumption of pool boiling is not correct. There will be film boiling of methanol instead. The film boiling heat transfer coefficient can be calculated using the Bromley’s correlation, Eq, (6.12). That is, p,(P) - By) + 04c,,7. i voy HER oe ME Neds pi S | Wim? °C = 0,62] (0.0265) (1.688)(694 — 1.688)(9.8){(11 « 10°) + (04)(1652)(119.3)} ‘i - (0.032)(1.364 x 10°)(119.3) = 163.4 W/m? °C (Note that the heat transfer coefficient value is pretty low.) The heat flux, gy = (hy\7,) = (163.4)(119.3) = 19.49 kW/m? 19.49 kW/m? aa Pipe : The boiling rate 1100 ki/ke 0.01772 kg/m? s = | 63.79 kg/m*hHEAT TRANSFER PRINCIPLES AND APPLICATIONS 6.5 FORCED CONVECTION BOILING the motion in the liquid is created by the rising vapour bubbles and aleg bre conve Fored eoecion toe ane ster head eaten sith ow of along with the vapour bubbles driven by an externally imposed meena a ent, oF is by some kind of natural circulation, Six low regimes or zones that can es identified in fan, convection boiling in a vertical tube are shown in Fig. 6.5. Sel tpflew (ie. upton st Hiquid with vapour bubbles dispersed in it) occurs in most part of the tui 186 Vapour out M2 Single phase vapour flow v Mist ow ‘Two-phase forced wv Annular flow convection um Slug fow Saturated nucleate " Bubbly flow boiling Subcooled boiling Ht __'Subcooled? liquid in Fig. 6:8 Flow regimes in forced convection boiling in a vertical tube. In zone I of Fig. 6.5, the liquid is still below its boiling point and subcooled boiling occurs at the wall. Most of the tiny bubbles formed at the wall Collapse in the bulk liquid (see Section 63). In zone I, the liquid has sufficient supetheat to nurtase the vapour bubbles; the superheated liquid flows up along with the bubbles (subbiy flow). Because of growth and coalescence, the bubbles gradually become large in size and Vapour slugs are formed in zone III (slug flow. In one IV (annular flow), the liquid adheres to the wall as a film surrounding the inner vapour core ‘The inner core also contains tiny suspended liquid Particles carried over from the zone below (these liquid particles are formed during vigorous bubble breakage). Over zone V, there is very hitle liquid on the wall but there are minute liquid droplets in the flowing vapour (mist flow) There is nothing else than Vapour in zone VI, The rates of heat transfer are very high in zones HL I, and 1V, but very low over zones V and VI where heat transfer occurs predominantly through the vapour.HONING AND CONDENS ATION 167 Extensive research work has been carried out on forced convection boiling heat transfer, and excellent reviews are available, Here we shall cite the correlation by Chen (1966) that covers a Considerably wide range of flow rates and regimes--both saturated nucleate boiling, and two- ase forced convection boiling. Chen considered the heat transfer coefficient to consist of two parts—a nucleate boiling part (/,), and a forced convection boiling part (H,). Thus, hehth (6.14) h, = 0023 (Re,)"™ ray) Wim? K, Re, = maw (6.15) 1 9p, AC el "(eon)" (on . 07s 2 hy = 1218 x 10° ET.) (Pa) SWiK (6.16) 2 Ora Ea Oem) a BB Hh Tas Wow (6.17 Taw s=fi+oire]"; for Re< 325 (6.18) = [+042 "YP"; rs 70 (6.20) Re = 10° Re, FS (6.21) 0736 1 F=10 fort 0.10 (6.22) x x x 09 pS 1 us [a] [2] [4] es x Vw, Co A where Gis the fluid mass velocity, kg/m” s; w is the mass fraction vapour (or ‘quality’) in the fluid; Vy = vy — ¥y m/kg (v = specific volume) subscripts / and v denote the liquid and the vapour phase, respectively T, is the surface temperature ‘Apu (Nim?) is the difference of vapour pressure between the superheated liquid (at surface temperature 7,) and the saturated liquid (at temperature T,.) d is the inner diameter of the tube @ is the surface tension of the liquid (N/m) at the bulk temperature other terms have significances as stated before and are in SI units. In Eq, (6.15), Re, is the Reynolds number based on the liquid flow rate only (the mass of vapour is subtracted from the total flow). The factor S in Eq, (6.16) is called the boiling suppression factor. The factor F in Eq. (6.21) modifies the “liquid only Reynolds number’, Re, toa ‘two-phase’ Reynolds number.| 138 HEAT TRANSFER—PRINCIPLES AND APPLICATIONS In the liquid preheating region, if any, the Dittus-Boelter equation, Eq. (4.9), mn, calculate the heat transfer coefficient. TAY be used ig Example 6.4 A mixture of benzene and toluene, taken a cut from a reformate dist in a refinery, has to be separated further in another distillation column. A reboiier 1°? ling Partially boil out the bottom liquid of the column. The vapour generated flows up yet column. The bottom liquid contains a litle benzene, but for the purpose of yer wt calculations the physical properties of the liquid can be assumed to be the same sc toluene. So far as the boiling calculations are concemed, we formulate the problem tase Toluene, as a saturated liquid, enters the bottom of a 14 BWG 25 mm vertical tube at Ay gauge pressure at the rate of 200 ky/b, 30% of which is vaporized per pass. The wie = temperature is maintained at 160°C (433 K) by saturated steam condensing outside the ape bar (gauge). It is desired to calculate the required length of the tube. The following das cyt? the given conditions are supplied. Boiling point of toluene at the given pressure (0.3 bar gauge), 7,.. = 120°C (393 K), heat of heat o vaporization, L, = 3.63 x 10° J/kg. Other properties of the liquid: &y = 0.112 Wan °C, cy = 1968 Ske °C, p, = 753 by = 1.66 x 107 Nim, pt) = 2.31 x 10° kg/m s. For toluene vapour: p, = 3.7 kg/m’, viscosity of toluene vapour, j1, ~ 10 kg/m 5. SOLUTION Here forced convection boiling occurs in the tube. We can use the correlations of Chen [Eqs. (6.14) -(6.23)]. The boiling heat transfer coefficient under forced convection depends upon the ‘quality’ of the vapour-liquid mixture, ie, the mass fraction of vapour w in the mixture. However, the nature of such dependence is not very simple. Here the quality, w, varies from 0 to 0.3. We adopt a stepwise calculation procedure. In each step (starting from w = 0) we increase the value of w by 0.05. Heat transfer coefficients and the area are calculated for each step taking the mean valie of w jin the interval. The total area required can then be obtained by adding the areas required for the individual steps. Given below are the detailed calculations for the interval w= 0.1 to w=0.15, the mean value being w = 0.125. Calculation of h, and hy for w = 0.125 d= id. of 14 BWG 25 mm tube = 0.0211'm 200 ~ ap 7 159 kein? Mass flow mate, O~ Toy ea?Tay g/m? s ey = GUL. _ (1591 = 11250621) _ 7 sot R - Hi 231% 107 eptltr _ (1968)(2.31 x 10) Py Er. Ce) aaig Ee oll2 4 From Eg. (6.23), we have 1 a r By [we]. [os2sp? frsapf__10* =| = 181 x U-wl lad Lay 0.875} [3.7] [2.31 x 107BOILING AND CONDENSATION 199 ‘ 0736 From Eq, (6.22), F = 2.35 [: + os] =3973 From Eq. (6.21), Re = 10 Re, P25 = (10-)(1.271 x 10*(3.973)!25 = 7.13 From Eq. (6.18), $= [1 + 0.12 ®e"J" = 0.47 From Eq. (6.15), the forced convection heat transfer coefficient, O12 he = (0.023(1.271 x 104)° (4.06) 973) = (0.023)( x 10°)"* (4.06) (= ssn 1630 Wim? °C From Eq, (6.16), the boiling heat transfer coefficient can be calculated. Excess temperature, T, = 160 — 120 = 40°C = 1.374 x 10° N/m? = Fb, _ _(40)(3.63 x 10°) Ne ae ety 37 753 Therefore, ky” (e)™*%p hy = 1.218 x 107 ain om (pad? 8 = (1.218 x 10 (0.112) (1968)945(753)"* (40)°2*(1.374 x 108)°% “ (2.66 x 1077)°(2.31 x 104)? (3.63 x 10°) (3.7) = 4077 Wim* °C (0.47) Total heat transfer coefficient, h = h, + hy = 1630 + 4077 = 5707 Wim? °C Calculation of the required heat transfer area Rate of vaporization in order to change the quality w by 5% = (200)(0.05) = 10 kg/h = 2.778 x 107 ke/s Heat load, Q = (2.778 x 10°)(L,) = (2.778 x 107) 3.63 x 109) = 1008 W If A is the area of heat transfer, Q = h A AT = (S707)(A\(160 - 120) or A = 0.004416 m? If the required length of the tube is J, A _ 2004416 ad (0.0211) 1.066 66 m = 6.66 cmHEA'T TRANSFER—PRINCIPLES AND APPLICATIONS, 190 Es uA idered. The res The calculations have to be done for all the intervals of ia A Ter LIS a shy in Table 6.2 (the calculations are not shown and are left as Table 6.2 Example 6.4 = het ; Suep size w Average w wh is “ey Hh ioe) Twi : of ‘ “i a enn 0-005 0.025 736 6364 7100 1008 0.0535 05-010 0.075 1235 6872 6162 1008 0.0617 010-015 012s 1630 4077 $707 1008 0.0566 15-020 0.175 1970 3504 5474 1008 0.0695 020-025 0.225 2073 3088 5361 1008 0.0709 025-030 0.275 2546 2776 $322 008 ona Total tube length = [0393 m Commens. Ics seen that the total heat transfer coefficient is very high, and a rather small length of the tube (0.393 m) offers the necessary heat transfer area. But does it happen in actual practice? No, In realty, the overall heat transfer coefficient gets drastically reduced by the ‘dirt factor (see Chapter 8). So a much larger tube length will be necessary in practice to achieve the desied boiling rate. For boiling of organics, an overall heat transfer coefficient in the range 500-1009 W/m? °C (100-200 Bru/h f? °F) is used in the design of reboilers (Ludwig, 1983; Walas, 1990; also see Chapter 8), 6.6 THE CONDENSATION PHENOMENON Condensttion means the change of phase from the vapour tothe liquid. IF the temperature of 3 vapour is reduced below its saturation temperature, condensation occurs, A liquid at is boiling point is called a saturated tiguid and the vapour in equilibrium with this liquid is called sanandl ceaay A ‘auld or vapour above the saturation temperature is called superheated. Like boling. Condensation is also an important heat transfer operation, A distillation column is always equipped with a condenser to condense the vapour coming out of the top of the column, Prods coming out of 8 reactor are often obtained as a vapour, which is then required to be condemed ftBOILING AND CONDENSATION — Vapour, NS Condensate Condensate... dro film Surface directly exposed to the vapour () ©) ~cold wall Fig. 6.6 - Modes of condensation: (a) film condensation and (b) dropwise condensation. In film condensation, the latent heat is transferred through the liquid film (which blankets the surface on which condensation occurs) and then conducted through the wall to the cooling fluid on the other side of the wall. The condensate film offers considerable heat transfer resistance. In dropwise condensation, on the other hand, a part of the surface is covered by liquid drops and the rest is directly exposed to the vapour. As a result, heat transfer coefficient for film condensation is considerably lower than that for dropwise condensation. A vapour, that normally condenses as a film on a particular surface, can be made to condense in the dropwise mode by applying a suitable coating on the surface. But this technique has not met with much success in practice. In dropwise condensation, 2 drop nucleus first forms on the surface and then grows in size. The approximate practical ranges of film-coefficient for condensation are: condensation of steam (dropwise)—5,500-16,000 W/m? °C (1000-3000 Btu/h ft? °F); organic vapour (film condensation)—1000-3000 W/m? °C (150-550 Brush ft? °F). These values may be much higher for very clean surfaces—around 5500 W/m? °C (1000 Btu/h f° °F) for film condensation of organics, and $5,000-450,000 W/m? °C (10,000-80,000 Brwh ft? *F) for dropwise condensation of steam. Extensive experimental and theoretical research work has been done with a view to fundamental understanding of the phenomenon of condensation and also with a view to developing equations and correlations for the calculation of the heat transfer rate in condensation. A few simple approaches and results are described below. 6.7 FILM CONDENSATION ON A VERTICAL SURFACE A theoretical analysis of laminar film condensation of a vapour on a vertical plate was given by Nusselt in 1916, This analysis is based on the following major assumptions: (i) The condensate film is in ‘locally’ fully developed laminar flow with zero interfacial shear and constant liquid properties.SS 192 HEAT TRANSFER—PRINCIPLES AND APPLICATIONS: (ii) The vapour is saturated. Heat transfer through the condensate film occurs by conduction only, and the, profile in the film is linear. emPete Let us refer to Fig. 6.7 in which a vertical section of the film is shown. In order tp the velocity profile in the film, we make 2 force balance on a liquid element or ‘contys of sides dx and (5 — y), where 6 = 6(x) is the ‘local’ thickness of the condensare p,*°lne’ sate depends upon the distance x from the top of the wall. The film is assumed to be x on Which z-direction (the z-direction is normal to the plane of the paper), and therefore the ite becomes virtually two-dimensional. For the purpose of force balance, a unit breadth oe film, in the z-direction is considered. Vepour temperature Fig. 6.7 Control volume in a condensate film in laminar flow. The three forces , F), F), and F;, representing the shear force, gravity force and the buoyancy force, acting on the control volume are also shown in Fig. 6.7. By a force balance, we have FL=F,-Fy or a “em = prt — yde ~ pyg(S ~ yd Integrating, we get, x= (2 BIE bs = #) + CC’ is the integration constant. H ‘Using the boundary condition u = 0 at y = 0 (i. ‘no-slip condition’ at the wall), we have C’ = 0. Therefore, y= AE fg _ Fe 62) # 2 alFe BOILING AND CONDENSATION 193 ves the velocity profile in the freely flowing, film. The profile is half-parabolic. 4) gir sgustion (6.24) Bee condensate (per unit breadth of the film) at any location x is obtained by f flow of ee of this equation.” ine 3 = fp Bde (ys 2) 4 = Ble Pded? mt =f) ula tdpy = fy PL | 98 - I by ” (6.25) stated before, 6 is the ‘local thickness’ of the condensate film. #e 06 SMereases downwards, the film thickness also increases because of condensation of more a ‘vapour. The fate of condensation on an ‘elementary surface’ of size dr x 1 exposed to and me obtained from the rate of heat transfer through this area. The temperature profile can be te ape ena lm is assumed to be inar and we may, therfore, wrt in Het The rate of heat transfer = aS (ax) and T» are the vapour and the wall temperature, respectively.** The change in the rate low over a distance dr is dm’ Here Ty of condensate substituting for m’ from Eq. (6.25), differentiating with respect to x and rearranging, we get PAP: ~ PE L-%, Hy Integrating the above equation and using the appropriate boundary condition (at x = 0, 5 = 0) it is easy to get an expression for the local thickness of the condensate film. That is, 8 dd ky a [Auk za" 4 pele, (529) If A is the heat transfer coefficient for the condensate film, heat flux through the film at any location x is |= h= Using the expression for 8 = 8(x) given by Eq. (6.26), we get ma the breadth of the film (i.e. its dimension in the z-direction) is unity, the elementary area of flow normal ee pF Editection (ic. the direction of flow) is | dy ~ in 7 poet lemperaiure is considered uniform throughout, and the temperature atthe vapourcondensate interface The temperature drop across the condensate film is T, ~ Ty. Pe194 HEAT TRANSFER—PRINCIPLES AND APPLICATIONS 4uyx(Z, — T,) (62) The local Nusselt number is Nu, 62) The average heat transfer coefficient over a length L is Lye 1) glpta ~ ade] (4 B= At ide = 4) ole = dir (Se ch if 4ui(T, — T,) | 3)” i ay = 0.943] S4A(A — p.) 4m L(T, ~ T,) (625) While using Eq. (6.29) the liquid properties should be taken at the mean film temperatire : ' ies at (F, + T,)?2. If condensation occurs on an inclined surface, g in the above equation should replaced by gcos @, where @ is the inclination of the surface to the vertical, Nusselt’s analysis of film condensation on a vertical surface has several limitation, Thee are: (i) itis grossly approximate to assume that the flow in the condensate film is one-dimensional and that the velocity profile bas a half-parabolic form at each location along the plate, (i) nox consideration of convection effects and assumption of a linear temperature distribution inthe fm are far from adequate. A more refined theoretical treatment of the problem was given by Sparow and Gregg (1959) on the basis of two-dimensional boundary layer approximation for flow and temperature fields, It is, however, interesting to note that, despite its limitations, the Nussel’s equation (6.28) predicts the condensation heat transfer coefficient on a flat vertical (or inclined) surface with an error of only a few per cent in many situations of practical importance. The accuracy of the Nusselt’s equation can be improved by incorporating a modification suggested by Rohsenow et al. (1985). At any position on the plate, the liquid-film temperature changes from T, at the free surface to 7, at the plate surface. This means that some amount of sensible heat is also removed in addition to the latent heat of condensation. To take this into account, Rohsenow et al. (1985) suggested the use of a modified latent heat term L’,= L,(1 + 0.68la) in place of the heat of vaporization, L,. The quantity Ja is called the ‘Jacob number’, Ja = Gy (7, - TyWLy | | | 6.8 TURBULENT FILM CONDENSATION If the rate of condensation is low or the height of the condensing surface is small, ie thickness of the condensate film remains small and the flow remains laminar. The free surface 2 film in laminar flow appears smooth and ripple-free. The nature of flow is determine’ My the film Reynolds number, Rep The local average liquid velocity in the film can be obtained Eq. (6.24). —