Академический Документы
Профессиональный Документы
Культура Документы
increases, subl~tio~ chemical reaction, dissociation, and Research, U.S. Air Force for support under Contract
ionization occur. In order to find the e9uilibrium com- 33(657)-7380, Project 7063, Task 7063-03.
position, the method used is that of minimizing the free
energy. The thermodynamic functions of the equilibrium
REFERENCES
mixture based on 1 mol of initial reactant are calculated by 1
F. Bo&+~a~ovrC, Modern Developments in Heat Transfer,
the equations for an ideal mixture.
edited by W. -E. IBELE, p. 305. Academic Press, New
The enthalpy composition charts are then constructed York (1963).
with the calculated vales of H(P, T, +h). Taking H as 2. D. B. SPALDING, Modem Developments in Heat Transfer,
the ordinate and $ as the abscissa, the constant temperature edited by W. E. IBELE, p. 65. Academic Press, New York
and entropy lines are shown in the charts The carbon (1963).
saturation line is also indicated, having been determined 3. K. T. SHIH, W. E. IBELE,E. R. F. WINTEZand E. R. G.
for given JI by the temperature at which C, disappears in ECKERT,Thermodynamic properties of carbon-nitrogen
the equilibria composition. For the pressures O-001, 4 mixtures at high temperature, ARL 64-149 (1964).
W. FICKETTand R. D. COWAN, Values of the~~yna~c
Ml, @lo, 1.410, and l~atmospheres, enthalpy~om~sition *
functions to 12000 K for several substances, Los Alamos
charts are given in Figs. 1 through 6, respectively.
Report LA-1727 (1954). Also J. C&em. P&s. 23, 1349
(1955).
ACKNOWLEDGEMENT 5. W. B. WHITE, S. M. JOHNSON and G. B. DANTZIG,
The authors wish to express their appreciation to the Chemical equilibrium in complex mixtures, J. Chem,
Aerospace Research Laboratories, Office of Aerospace P&s. 28, 751 (1958).
hf. J. Hear Mars Trawfer. Vol. 10, pp. 1457-1462. Pergamoo Press Ltd. 1967. Printed in Gnat Britain.
hot fluid (fluid with the highest inlet tem- for the physical problem: (a) There is no heat exchange
perature); between the exchanger and the surroundings. (b) Steady
intermediate fluid. state prevails. (c) The physical properties of the fluids are
essentially constant. (d) There is no longitudinal conduction
along the separating surfaces. (e) Perfect mixing occurs in
COMMONheat exchanger designs involve transfer of energy each flow passage.
as heat between pairs of fluid streams; however, heat In writing the expressions for the heat exchanger effective-
exchange between three fluids is required in such applica- ness, two cases must be treated : (a) C, and Ci less than C,,,
tions as air separation systems, ammonia gas synthesis and (b) C,, less than Ci and C,. In the first case, the maximum
systems, etc. The purpose of this communication is to heat transfer occurs for an infinite-area heat exchanger in
present analytical relationships between the design variables which the outlet temperatures of the intermediate and cold
(the NTU-effectiveness relationships) for a general three- streams are equal to the inlet temperature of the hot stream.
fluid heat exchanger, in which all three streams are in
thermal communication. e,,, = Ci(Th, - Z,) + C&1 - T,,). (1)
The NTU-effectiveness design method [l], commonly The actual heat-transfer rate is
used in two-fluid heat exchanger design, was extended by
Sorlie [2] to the restricted case of three-fluid heat exchangers e actual= CiKz - T,) + C,(T,, - T,,). (2)
of the concentric-tube and plate-fin types, in which the Using these two expressions in the delinition for heat
intermediate and cold streams were thermally isolated exchanger elfectiveness and expressing the result in dimen-
from each other. The present investigation was an extension sionless form, the following expression is obtained.
of Sorlies work by treating the case in which all three
streams were in thermal communication, as in many three- Qactual (GKcki x + e,
(3)
fluid heat exchangers used in cryogenic systems. t = ~Qnlax = (C,/C,)X + 1
A schematic of the three-fluid heat exchanger considered For the second case, in which the capacity rate of the hot
in this investigation is shown in Fig. 1. The cold and inter- stream is smallest, the hot fluid will leave an infinite-area
mediate streams flow in parallel, while the hot stream flows heat exchanger at a temperature Go, given by
in the opposite direction. The analysis is also valid for the
situation where the cold stream flows counter to the hot T = UWs)K, + T,
ho (4)
and intermediate streams. The following mathematical (RJR,) + 1
Imodel (set of assumptions or idealizations) was formulated
The maximum possible heat exchange for the second case is
e,., = C*(T,, - Z)
so that the expression for the heat exchanger effectiveness
becomes
t = ~WJGJe,+ G/GkiXIC(R~I&)
+ 11
(5)
(R,IRJX + 1
(6)
Hot
[33. The final results for the temperature ratios e, and ei (CJW3 cc
B4 = --
are as follows. 1 + (R&)(WG) Ci
83 (I?, - 1) - (1 - X)(B, - 1)
ei = x (B, - l)(E, - 1) - (B2 - l)(B, - 1)I
(16)
x (exp CrAl - 1)+ *
4
The quantities r2 and r3 are the roots of the characteristic
x 1_ x _ (B, - 1X(4 - 1) - (1 - X)(& - Ul equation for the system of differential equations (a), (7),
(B,- W, - 1)- (& - W3 - 1) and (8), given by
x (expCr3fhI- 1) (9)
(84 - 1) - (1 - X)(B, - 1)
ec = [ (B, - l)(B, - 1) - (B* - l)(B, - 1)I
x (evCb%l- l)+{~--~~)
(B4- 1) - (1 - X)(B, - 1)
x [ (B 1 - 1w-h - 1) - (4 - w3 II - 1)
x(exp[r3A11
-1)(10)
where : Although the expressions for the temperature ratios
1ev
e, and e, are rather involved, the complexity of the expressions
&
B1 = [ 1 + (R,/R,)(CJC,)
C@J (11) is not as serious a disadvantage as it would seem at fast
glance, since the expressions are easily handled by a digital
1exp
computer. A typical computer program is given in [3].
Some representative curves for the temperature ratios are
Cr341
s3
(12) presented in Figs. 24. More complete tabulated data are
= [ 1 + (RIIWWG)
given in [S] also.
For the special case when CJC, + CJC, - 1 = 0, the
(WC& Cc expressions for the temperature ratios become indeterminate.
B, = (13)
1 + (RdRd(CdCi) - c, In this case the solution is found to be the following.
-a?
9;
, 0.25
/
/
-0.2 \
/
- o.40
2 3 4 5 0 I 2 3 4 5
Nf 3
FIG. 2. Temperature ratios. RJR, = 0.50;RJR, = 0.25; FIG. 4.Temperature ratios. RJR, = 0.75;RI/R3 = 2.00;
cc/c,= 0.50;CJC, = 0.75;~ cold stream, e,; CJC, = 1.00; CJCI, = 1.00; ~ t?,; - - ~ ei.
ei.
0.6
0.25
I I I
-0.2
111
_,.r
-0.2
0 2 3 4 5
-,,+ 0 2 3 4 5
4,
Ntu FIG. 5. Temperature ratios for the special case (CJC, +
FIG. 3. Temperature ratios. RI/R,= 1GO: RJR, = 2-00; CJC, = 1). RJR, = 0.75;RJR, = 1.50;CJCh = 0.25;
cc/c,= 0.5O;CJC*= 0.75;---- t?,;-- -ei. C,JC,= 0.75;--- e,;- - -ei.
SHORTER COMMUNICATIONS 1461
so that the mixed-mean temperature was determined. were paired, then the cold and intermediate, and finally
Hot water at approximately 140F was mixed with cold the hot and intermediate streams were paired to obtain
water at approximately 70F to produce the intermediate- RI, R,, and R, respectively. This procedure is described in
temperature stream. The inlet temperature of the intermediate detail in 131.
stream was varied by varying the ratio of hot and cold Since all fluid temperatures were measured, the tempera-
water mixed_ Mass flow rates were measured by collecting ture ratios and heat exchanger effectiveness could be
the liquid in weighing vessels over a measured length of calculated directly from the experimental measurements.
time. The thermal resistances between the streams was The final experimental results are presented in Table 1.
obtained by making auxiliary runs at the same mass flow Detailed experimental data is recorded in [3].
rates with only two fluids flowing. The hot and cold streams It was noted from both the analytical and experimental
0.756 0.197 3.68 1.066 0.383 0.682 0.055 0.044 0.205 0.243 0.118 0.128
0.28 1 0.176 3.08 0.743 0.938 1.579 -0.194 - 0.424 0.177 0.363 0.058 0111
0.068 0121 4.14 1.106 1.723 1.890 - 1.210 - 3.593 0.197 0.370 0.339 0.324
0486 0151 1.16 0.757 0.953 0.760 - 0.068 - 0.054 0.169 0.192 0.054 0.084
0.47 1 0.136 0.487 0.707 1.238 1.061 0.051 0.014 0.146 0.143 0.264 0.236
Note: The Reynolds number range for the experimental data is as follows :
Cold stream (in annular passage) -2200; Intermediate stream (in $-in tube) 1500&28ooO. Hot stream (in &in tube)
11m39000. Equivalent diameter for cold stream in D, = 4(cross sectional area)/(heated perimeter).
results that negative values were obtained for the temperature or he may use the results presented in this work by analyzing
ratio e, for small values of the inlet temperature ratio X. the heat exchanger in sections in which the fluid properties
From its definition, a negative value of ei implies that the do not vary widely. An approximate method for analysis
intermediate stream leaves the exchanger at a lower tem- of a three-fluid heat exchanger when variation of fluid
perature than that at which it entered, or the cold stream is, properties is significant is outlined by Lenfestey 151.
in effect, cooling both the hot stream and the intermediate The assumption of negligible longitudinal heat conduction
stream. Since the intermediate stream temperature never along the separating surfaces of the streams is generally
falls below the cold stream temperature, no violation of the valid except when the heat exchanger effectiveness is high
Second Law of Thermodynamics is involved, and the (above about 090) or when the length of the flow passage
occurence of negative ei is a physical reality. The fact that is short (actually when the parameter kAJC,L is large).
negative ei occurs for small inlet temperature ratios X is This aspect of the problem is one which deserves further
to be expected, since a small value of X implies that the investigation, since heat exchanger effectivenesses of 0.95
intermediate stream enters at a temperature near that of or higher are not uncommon in cryogenic systems.
the hot stream at inlet; therefore, one would expect that the As shown in Table 1, the agreement between analytical
intermediate stream would be more likely to be cooled and experimental values for the heat exchanger effectiveness
than heated under this condition. is satisfactory. Since a rather small laboratory model was
There are two points in the mathematical model upon used in the experiments, the N, values were small (N,,
which the analytical solution is based that require examina- gives a measure of the sire of the exchanger). Note that
tion when applying the solutions in design : (a) the assump- negative values of ei were measured experimentally, in
tion of constant fluid properties, and (b) the assumption agreement with the analytical predictions.
of negligible longitudinal conduction. When wide tempera- It was felt that the experimental determination of the
ture ranges are involved, the assumption of constant fluid thermal resistances was responsible for the major differences
properties is weak, although this assumption is also used between the experimental and analytical results, since
in two-fluid heat exchanger design [4]. For the case in experimental temperature and flow rate measurements
which fluid property variations cannot be ignored (such as were accurate within +2 per cent. When all three fluids
when the state of the fluid passes through the vicinity of were flowing in the primary runs, the circumferential
the critical point), the designer must solve the specific conduction through the tubes could have been somewhat
problem numerically (no general solution is possible), different from the case when only two fluids were flowing
1462 SHORTER COMMUNICATIONS
in the secondary runs used to measure the thermal resistances. 2. T. SORLIE, Three-fluid heat exchanger design theory,
A difference in circumferential conduction would result in counter- and parallel-flow, Dept. of Mech. Engng Tech.
a difference in the thermal resistances. This is a second Rep. No. 54, Stanford University (1962)
aspect of the heat exchanger design problem which deserves 3. D. D. AULDS,An analytical method for the design of a
three-channel heat exchanger for cryogenic applications,
further investigation.
M.Sc. Thesis, Louisiana Polytechnic Institute, Ruston,
La. (1966).
4. M. JACOB,Heat Transfer, Vol. 2, p. 210. John Wiley,
REFERENCES New York (1957).
I.W. M. KAYS and A. L. LONDON, Compact Heat Ex- 5. A. G. LENFESTEY, in Progress in Cryogenics, edited by
changers, 2nd edn., pp. l&63. McGraw-Hill, New York K. MENDELSSOHN, Vol. 3, pp. 2547. Academic Press,
(I 964). New York (1961).