Documente Academic
Documente Profesional
Documente Cultură
Address correspondence to Professor T. F. Lin, Department of Mechanical Engineering, National Chiao Tung University, Hsinchu, Taiwan
30050, ROC.
Experimental Thermal and Fluid Science 1995; 11:61-71
© Elsevier Science Inc., 1995 0894-1777/95/$9.50
655 Avenue of the Americas, New York, NY 10010 SSDI 0894-1777(94)00112-L
62 Y.L. Tsay and T. F. Lin
*o
qw :
-~ ¢/--
Intirface
r- lnlorfoc~ in the liquid film can be described by the following equa-
tions.
X-momentum equation:
--III
9 Energy equation:
The interface-matching conditions specified at y---- It is noted that in the above formulation the thermo-
6L(X) are the following. physical properties of the liquid film and gas mixture are
1. The continuity of velocity, temperature, and shear stress considered to be variables depending on temperature and
is expressed as mixture composition. They are calculated from the pure
component data by means of the mixing rules [17]. The
UL, i = UG, i, ZL, i = TG, i, pure component data are available in [18, 19]. Addition-
ally, we note that the wave motion of the liquid/gas
( /£L OUL
Oy I]L,i= ( ]£G --~'-Y
(8) interface and the entrainment of the liquid in the form of
OuG) G, f droplets into the gas stream may be important when the
relative motion of the gas flow against the liquid film is
2. The transverse velocity of the air-vapor mixture, de- large. These processes are rather complex and are ne-
duced by assuming that the interface is semipermeable glected in the present study because of a lack of adequate
- - t h a t is, the solubility of air in liquid water is negligi- information.
bly small and the air velocity in the y direction is zero
at the interface, is SOLUTION M E T H O D
Since the flows governed by Eqs. (1)-(6) are parabolic in
1 ~ Oy ]G,i" (9) x, the finite-difference solution for these equations can be
marched in the downstream direction. A fully implicit
3. Assuming that the interface is at thermodynamic equi- numerical scheme, in which the longitudinal convection
librium and the air-water vapor mixture is an ideal gas terms are approximated by the upstream difference and
mixture, the mass fraction of the water vapor can be the transverse convection and diffusion terms by the cen-
calculated as [8] tral difference, is employed to transform the governing
MvPv, i equations into finite-difference equations. Each system of
the finite-difference equations forms a tridiagonal matrix
Wv'i = ma(P~ - By, i) + mvev, i ' (10)
equation that can be efficiently solved by the Thomas
where Pv, i is the saturation pressure of water at the algorithm [20]. For a given flow and thermal condition a
interface temperature. brief outline of the solution procedures is described as
4. The energy balance at the interface is follows.
1. For a given longitudinal location x j, assume a film
= - AG -~y + m"v,ihLG , (11) thickness 6 L j.
- L - Y/L,i O,i 2. Solve the fi~iite-difference forms of Eqs. (1) and (4)
simultaneously for u L and u 6.
where the vaporizing flux of the water vapor is calcu-
3. Numerically integrate Eq. (3) to find vG.
lated by using the equation
4. Solve the finite-difference forms of Eqs. (2) and (5)
f PG,iD OWv ] .
(12)
together for T L and TG.
m"'i = - 1 5. Solve the finite-difference form of Eq. (6) for Wv.
vi,, y,oi 6. Check whether the relative error between two consecu-
Equation (11) indicates that energy exchange between tive iterations n - 1 and n is small enough, that is,
the gas stream and liquid film depends on two related [(ion -- ~ n - l[//l~n[max < 10-3 for all nodal points, where
factors: the interface temperature gradient at the gas side, q~ represents variables UG, UL, TG, TL, or Wv. If not,
resulting in sensible heat transfer, and the rate of mass repeat steps 2-6.
transfer, resulting in latent heat transfer. The total heat 7. Check the mass conservation of the liquid water film by
transfer from the interface into the gas stream can then examining the satisfaction of the inequality
be expressed as
qi' = q's',i + qe',i
(rhL,j_ 1 -- rhL,j) -- (m"vi,j Axj) < 10 -4, (16)
rhL,j_ 1 -- FhL, j
--"~-y)G,i--( pG'iDhL'G (13)
where
is a rectangular channel 220 x 160 mm in cross section The electric power input to the heated steel plate is
and 500 mm long, with three sides made of acrylic plates determined from the measured voltage drop across the
and another made of stainless steel plate. The stainless plate and the current passing through it,
steel plate is 50 cm long, 21 cm wide, and 1 mm thick.
Fourteen calibrated copper-constantan thermocouples are atot = / v = 12 pel (18)
glued onto the back surface of the plate to measure its ,~wZ
temperature. To prevent heat loss from the plate to the
surroundings, the plate is insulated with plastic and wood. where I and V are electric current and voltage, and Pe is
At the buffer end, a variable-speed blower provides air the electrical resistivity of the plate. To accurately mea-
with velocities up to 10 m / s through the vertical channel sure the electric current, a shunt is arranged between the
in the test section. The speed of the blower is controlled cables.
by an ac motor speed controller with an output frequency
of 0-60 Hz and full load current of 4 A. To reduce the Measuring Probes and Data Acquisition System
influence of vibration from the blower, a flexible connec-
tor made of cloth is employed to connect the blower and In the experiments, temperature, liquid, and gas flow rates
the buffer end. and the power applied to the steel plate are measured.
Plate Heating Facilities Two copper conductors are Flow Rate Measurement The flow rate of the liquid film
screwed to the top and bottom of the stainless steel plate is measured by a variable-area flowmeter (M-type). All the
(50 cm long, 21 cm wide, and 1 mm thick). Direct electric flowmeters are calibrated by the static weighting proce-
current, generated by a 10-kW dc power supply, is passed dure, as described by Mattingly [23], at several liquid
through the steel plate to provide the required heating. temperatures to account for the property variations with
66 Y.L. Tsay and T. F. Lin
temperature. The accuracy of the flow rate measurement Table 1. Uncertainties in the Basic Parameters
is believed to be within 0.009987 kg/min. and Nusselt Numbers
The gas flow rate is evaluated by integrating the velocity Basic Parameter Uncertainty
profile obtained by measuring the velocity at several posi-
tions in the cross section near x = 10 cm. The gas velocity x distance, m + 0.0005
is measured by a Pitot tube 67 cm long and 0.03 cm in I, m + 0.0005
diameter (Model PDA-24-F-22KL) and a high-precision z, m 5:0.0005
manometer (Model P-600). The manometer allows cali- TL, in, °C -+ 0.15
bration and measurement of differential, absolute, or Tw, °C _+0.15
gauge pressure within the range of 0-1 psi with 98.5% Tw - Tt.,i., °C -+0.21
accuracy. I, A -+1%
V,V +1%
Uncertainty Analysis for Experimental Results 0tot' W -+ 1.4%
tt
The purpose of performing an uncertainty analysis is to qw, W/m 2 -+3.5%
estimate the uncertainty levels in the experimentally de- h L , W/(m2 °C) -+ 1%
termined Nusselt numbers. The procedures of the uncer- /')L' m2/s -+ 1%
tainty analysis proposed by Kline and McClintock [24] and Calculated uncertainty in N u L 3.9-14.3%
Moffat [25] were used. The uncertainties in the basic
parameters and Nusselt number are presented in Table 1.
It is noted that, except in the region near the leading edge Shown in Fig. 3 are the predicted temperature distribu-
(x < 0.1 m), the uncertainty in N u L is 3.8-14.3%. tions in the liquid film and gas stream for cases with
TL, in = 3 0 ° C , UG,~ = 2 m / s , and mL, in = 0.02 k g / ( m s )
RESULTS AND DISCUSSION under various heating conditions. The inserted plots in
these figures provide a close look at the heat transfer
The problem under investigation is governed by six physi- process near the interface. The dashed line in each insert
cal parameters: the inlet liquid temperature TL, in , inlet marks the location of the liquid/gas interface. Compari-
liquid mass flow rate rhL,i,, free-stream water vapor con- son of the corresponding curves in Figs. 3a and 3b at the
centration Wv,=, free-stream temperature To, ~, free-stream same x indicates that an increase in q~ results in a higher
velocity u c = , and imposed wall heat flux q~. In view of temperature rise in the liquid film and gas stream. The
the large number of parameters, a fully parametric explo- concentration distributions of the water vapor in the gas
ration is unrealistic. Rather, the parameters were varied stream, given in Fig. 4, show that the water vapor in the
systematically in order to examine the key trends in the interface region increases in the flow direction and more
results. Particular attention was paid to investigating the water vapor exists in the gas stream when q~, is higher.
effects of q", / ' h L , in , TL, in , and u~,= on the heat and mass The trend is similar to the temperature development in
transfer characteristics in the flow. The ranges of these Fig. 3 since thermodynamic equilibrium was assumed at
parameters are as follows: q~, 1000-4000 w/m2; /'hE, in , the interface, Eq. (10).
0.01-0.04 k g / ( m s); T L in, 3 0 - 5 0 ° C , UG ~, 0.5--4 m / s ; with Presented in Figs. 5 - 7 are the measured and predicted
TG = fixed at 30°C and'Wv = at 0.013 (this corresponds to a wall temperature distributions along the plate for various
relative humidity of 50%)~ cases. First, it is observed that the measured and predicted
, ! ,! 8 i |
(b) a LI' G
I
6 i -i
. \ k - k - - - 0.38
, Y-k-k--- 0.25
2
.4
files for rhLi . = 0.02 kg/(ms), TLi n = 30°C,
u G~ = 2 m/s, and (a) q PP =2000 w / m ,2 (b)
q,4; = 1500 w / m 2. y (cm) Y (cm)
Evaporation of a Heated Falling Liquid Film 67
3 i I m 1 i ! w
(b)
~
I
2.25 2.25
v
~ 1.5 0.5m 1.5
0.38 ~0~.10"5m
X
I oils I
~ 0.75 0.n
wall temperatures are in good agreement, which lends higher, which in turn causes a larger amount of energy to
strong support to the theoretical model employed in the be transported to the interface to sustain the required
computation. The effects of the imposed wall heat flux are latent heat of vaporization (m"v ihLG ). Thus, the fraction
illustrated in Fig. 5. As expected, a larger temperature rise of the heat input to the system t~rough the plate that goes
results for a system with a larger input heat flux. Careful to heat up the liquid film decreases with an increase in the
scrutiny of the curves in this figure reveals that the wall inlet liquid temperature, and consequently the tempera-
temperature rise per unit heat flux input, ATw/q", is ture rise of the liquid film is smaller. It is interesting to
higher for the case with a larger q". This suggests that for note that for TL, in = 50°C the wall temperature decreases
a system with a larger q ' , a smaller part of the input heat in the upstream region although the plate is being heated.
transfers across the liquid film to cause the film vaporiza- This is due to the fact that the energy required to sustain
tion and a larger part of the input heat goes to heat up the the latent heat of film vaporization is much larger than
liquid film. the heat input from the plate in this region, as can be
The effects of the inlet liquid temperature on the wall clearly seen later in Fig. 8. To maintain energy conserva-
temperature distributions are shown in Fig. 6. The results tion, the energy required for film vaporization is also
indicate that the temperature rise is less for the case with supplied by the liquid film through the loss of sensible
higher inlet liquid temperature at the same heating condi- heat, and thus the liquid film is cooled. The wall tempera-
tion. This is readily understood by realizing that the film
vaporization is stronger when the liquid temperature is
8 ! i ! !
prediction
•, •, • experiment ~ ,
81 prediction ' ' 1
/ • ' " " experiment ~ 6 j -
O ~ , , , I -I
0 0.I 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5
X(m) x (m)
Figure 5. Effects of wall heat flux on the local wall tempera- Figure 6. Effects of inlet liquid temperature on the local wall
ture distributions for m E , in = 0 . 0 2 k g / ( m s ) , T L , in = 3 0 ° C , temperature distributions for q" = 2000 w / m E, /~/L, in = 0.02
and u o ~ = 2 m/s. kg/(m s), and uG.~ = 2 m/s.
68 Y.L. Tsay and T. F. Lin
8 1.5 . . . . (o)
?rediction ' ' .,fl
• , , , • experiment .~ ~
1.(1
/
0"-:-7 ,
: E in = 50"C
% 0.1 0.2 0.3 0.4 0.5 o- 0.05
40"C
X(m)
Figure 7. Effects of inlet liquid mass flow rate and free-stream
o : Ic)
velocity on the local wall temperature distributions for q" =
2000 w / m 2 and TL, in = 30°C.
:~ 1.0
0.2 i i ! 0.2
prediction l 'm,,io = 6.o . kglr .s,Uo,i= m,s
l
•
experiment
i
.J
e:
i#
=.
z
0.1
0.05 O.OE
OI 0
• ,.,= experiment
I I I I
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5
X(m) X(m)
Figure 9. Effects of wall heat flux on the local Nusselt Figure 11. Effects of inlet liquid mass flow rate and free-
number distributions for mE, in = 0.02 kg/(m s), TL, i, = 30°C, stream velocity on the local Nusselt number distributions for
and uc, ~ = 2 m/s. q itw = 2 0 0 0 w / m , T2 L i n = 3 0 Co .
teristics in a laminar gas stream flowing over an evaporat- Re L Reynolds number of liquid film = (4mL//~L),
ing liquid film falling along a vertical plate subjected to a dimensionless
constant wall heat flux. The influences of thermal condi- S parameter = PGDhLo(Wv, r -- Wv,~), W//m
tion, inlet liquid mass flow rate, and gas stream velocity on T temperature, K
the heat and mass transfer in the liquid film and the gas
stream are investigated in detail. The predicted wall tem-
rw, measured back surface temperature of heated
plate, K
perature and Nusselt number distributions compared fa-
U longitudinal velocity, m / s
vorably with the measured data. What follows is a brief
summary of the major results. V Voltage drop across a heated plate, V
U transverse velocity, m / s
1. Heat transfer between the liquid film and gas stream is
dominated by the transport of latent heat in conjunc- Wv mass fraction of water vapor, dimensionless
tion with the evaporation of the liquid film. Wv.r saturated mass fraction of water vapor at Tw and
2. The evaporative latent heat transfer depends largely on Po, dimensionless
the inlet liquid temperature TL, in and the wall heat flux x, y longitudinal and transverse coordinates, m
q'. Ax, Ay longitudinal and transverse grid sizes, m
3. In the ranges of the experimental conditions, the Nus- Z plate width, m
selt number for heat transfer from the heated wall to
the liquid film was correlated by a simple equation. Greek Symbols
t~ Lliquid film thickness, m
It has been realized during the course of the study that
when TL, in is much higher than TG ~ or when the plate is 6w plate wall thickness, m
subjected to a high wall heat flux,'the upward combined A conductivity, W / ( m K)
buoyancy force may result in flow reversal of the gas /z dynamic viscosity, Pa- s
stream in certain parts of the flow field for small values of P kinematic viscosity, mZ/s
uG,~. In this situation, a more complicated elliptic flow P density, k g / m 3
analysis must be performed.
Pe electrical resistivity, ~2m
The financial support of this study by the engineering division of the Subscripts
National Science Council, Taiwan, ROC, through contract NSC79- a of air
0401-E009-12 is greatly appreciated.
G of gas stream
i at interface
NOMENCLATURE in at inlet
Cp specific heat, k J / ( k g K) j, k the jth longitudinal and kth transverse grid
points
D mass diffusivity, m 2 / s
L of liquid
g gravitational acceleration, m / s 2
v of water vapor
hL heat transfer coefficient [ = q " / ( T w - TL, io)],
W / ( m 2 K) w of wall
hLG latent heat of vaporization, k J / k g free-stream value
I electric current, A
l length of heated plate, m
REFERENCES
M molecular weight, k g / ( k g mole)
/'h L m a s s flow rate of liquid water per unit heated 1. Chun, K. R., and Seban, R. A., Heat Transfer to Evaporating
plate width, k g / ( m s) Liquid Film, J. Heat Transfer 93(3), 391-396, 1971.
mrv, i interface mass flux of water vapor, Eq. (12), 2. Seban, R. A., and Faghri, A., Evaporation and Heating with
k g / ( m 2 s) Turbulent Falling Liquid Films, J. Heat Transfer 98(2), 315-318,
1976.
NU L Nusselt number = h L ( 1 , ' 2 / g ) l / 3 / h L ,
3. Gill, W. N., Casal, E. D., and Zeh, D. W., Binary Diffusion and
dimensionless Heat Transfer in Laminar Free Convection Boundary Layers on
P pressure, Pa a Vertical Plate, Int. J. Heat Mass Transfer 8(8), 1135-1151,
0tot total electrical dissipated heat rate in heated 1965.
plate, W 4. Saville, D. A., and Churchill, S. W., Simultaneous Heat and Mass
q~.i interface latent heat flux = m"v,ihLG , W / m 2 Transfer in Free Convection Boundary Layers, AIChE J. 16(2),
qg heat generation per unit volume in heated plate, 268-273, 1970.
W/m" 5. Bottemanne, F. A., Theoretical Solution of Simultaneous Heat
qi' overall interface heat flux = q",i -k- q"s,i, W/m2 and Mass Transfer by Free Convection About a Vertical Flat
Plate, Appl. Sci. Res. 25(1/2), 137-149, 1971.
q'(ss heat flux through insulation to surroundings, 6. Gebhart, B., and Pera, L., The Nature of Vertical Natural
W/m 3
Convection Flows Resulting from the Combined Buoyancy Ef-
q'sri interface sensible heat flux in gas side = fects of Thermal and Mass Diffusion, Int. J. Heat Mass Transfer
- (A G 0 T o / 3 y ) a , i, W ~ me 14(12), 2025-2050, 1971.
q" wall heat flux, W / m 2 7. Chen, T. S., and Yuh, C. F., Combined Heat and Mass Transfer
Re G Reynolds number of gas stream = in Natural Convection on Inclined Surfaces, Numer. Heat Trans-
uG ' ~( ~,2/ g )1/3 / vG ' dimensionless fer 2(2), 233-250, 1979.
Evaporation of a H e a t e d Falling Liquid Film 71
8. Chang, C. J., Lin, T. F., and Yan, W. M., Natural Convection Two-Component Film Evaporation in a Vertical Tube, 8th Int.
Flows in a Vertical Open Tube Resulting from Combined Buoy- Heat Transfer Conf., Vol. 4, pp. 1843-1848, 1986.
ancy Effects of Thermal and Mass Diffusion, Int. J. Heat Mass 17. Bird, R. B., Stewart, W. E., and Lightfoot, E. N., Transport
Transfer 29(10), 1543-1552, 1986. Phenomena, Wiley, New York, 1960.
9. Chandra, V., and Savery, C. W., Forced Convective Heat and 18. Yano, T., and Uchida, H., Thermophysical Properties of Fluids,
Mass Transfer from a Failing Film to a Laminar External Bound- JSME Data Book, JSME, Tokyo, 1983.
ary Layer, Int. J. Heat Mass Transfer 17(12), 1549-1557, 1974. 19. Fuji, T., Kato, Y., and Mihara, K., Expressions of Transport and
10. Chandra, V., Mass, Momentum and Heat Transfer from a Falling Thermodynamic Properties of Air, Steam and Water, Sei San Ka
Film to a Countercurrent Air Stream, Ph.D. thesis, Drexel Univ., Gaku Ken Kyu Jo, Report No. 66, Kyu Shu Univ., Kyu Shu,
Philadelphia, PA, 1975. Japan, 1977.
11. Schroppel, J., and Thiele, F., On the Calculation of Momentum, 20. Anderson, D. A., Tannehill, J. C., and Pletcher, R. H., Computa-
Heat and Mass Transfer in Laminar and Turbulent Boundary tional Fluid Mechanics and Heat Transfer, McGraw-Hill, New
Layer Flows Along a Vaporizing Liquid Film, Numer. Heat York, 1984.
Transfer 6(4), 475-496, 1983. 21. Szewczyk, A. A., Combined Forced and Free-Convection Lami-
12. Chow, L. C., and Chung, J. N., Evaporation of Water into a nar Flow, J. Heat Transfer 86(4), 501-507, 1964.
Laminar Stream of Air and Superheated Steam, Int. J. Heat 22. Holman, J. P., Heat Transfer, 5th ed., McGraw-Hill, New York,
Mass Transfer 26(3), 373-380, 1983. 1981.
13. Chow, L. C., and Chung, J. N., Water Evaporation into a Turbu- 23. Mattingly, G. E., Volume Flow Measurements, in Fluid Mechan-
lent Stream of Air, Humid Air or Superheated Steam, 21st ics Measurements, R. J. Goldstein, Ed., Chap. 6, Hemisphere,
A S M E / A I C H E Natl. Heat Transfer Conf., Seattle, WA, ASME Washington, DC, 1983.
Paper 83-HT-2, 1983. 24. Kline, S. J., and McClintock, F. A., Describing Uncertainties in
14. Haji, M., and Chow, L. C., Experimental Measurement of Water Single-Sample Experiments, Mech. Eng. 75, 3-8, 1953.
Evaporation Rates into Air and Superheated Steam, J. Heat 25. Moffat, R. J., Contributions to the Theory of Single-Sample
Transfer 110(1), 237-242, 1988. Uncertainty Analysis, J. Fluid Eng. 104(2), 250-260, 1982,
15. Shembharkar, T. R., and Pai, B. R., Prediction of Film Cooling
with a Liquid Coolant, Int. J. Heat Mass Transfer 29(6), 899-908,
1986.
16. Baumann, W. W., and Thiele, F., Heat and Mass Transfer in Received September 1, 1993; revised November 21, 1994