Professional Documents
Culture Documents
507517, 1998
1998 Elsevier Science Ltd and IIR
All rights reserved. Printed in Great Britain
PII:S0140-7007(98)00031-0 01407007/98/$19.00+00
507
508 M. Abu Madi et al.
NOMENCLATURE
Aa air-side heat transfer area Nt number of tube
Ac free flow area Ntu number of transfer unit
Af face area Pr Prandtl number
Ai tube inside heat transfer area R 1R 9 geometric parameters
Ap heat transfer primary area Re Reynolds number
As heat transfer secondary area Sf fin spacing
A t,m tube wall heat transfer area at middle Sr row spacing
distance St tube spacing
C min minimum heat capacity St Stanton number
dh air-side hydraulic diameter U all overall heat transfer coefficient
di tube inside diameter Va face velocity
do tube outside diameter y fin parameter
z fin parameter
f friction factor
Ga air mass velocity Greek symbols
ha air-side heat transfer coefficient
hi tube-side heat transfer coefficient
j Colburn factor b fin parameter
j4 Colburn factor for four-row heat df fin thickness
exchanger dt tube wall thickness
j Nr Colburn factor for N r number of rows Dp pressure drop
kf fin thermal conductivity ha heat exchanger surface effectiveness
kt tube thermal conductivity hf fin effectiveness
l fin length from root to centre J fin parameter
L flow length ma air viscosity
Lt tube length ra air density
m fin effectiveness parameter j porosity
Nr number of tube rows t fin parameter
Introduction determine j and f factors for round tube and plate fin heat
exchangers with staggered rows from the air-side
Traditional round tube and plate fin heat exchangers are Reynolds number based on the tube outside diameter.
widely used in applications where pressure drops for the Geometrical effects were included as the ratio of the total
internal fluid flow is a critical design factor, notably in air-side area to the tube outside area. The equation was
automotive applications for engine cooling and air derived from practical tests on heat exchangers with four
conditioning. The authors have developed software rows of tubes. The effect of the number of tube rows was
packages for the design of the compact heat exchangers included by McQuiston through the use of the experi-
used for condensation and evaporation in automotive air mental data of Rich6. El Mahdy and Biggs7 used
conditioning systems14. The heat transfer and pressure previously published data for heat exchangers with
loss correlations used on the internal tube surfaces round tubes and staggered tube arrangements and
provide adequate accuracy. Since the air-side thermal circular or continuous flat fins to develop a correlation
resistance is significantly higher than that of the tube for the calculation of the j factor from Reynolds number,
internal wall, small errors in predicting the air-side heat based on the hydraulic diameter. The geometric effects
transfer performance lead to significant errors in were included as ratios of fin thickness to fin length,
predicting the overall thermal performance of heat hydraulic diameter to fin thickness and fin spacing to fin
exchangers. The accuracy of a heat exchanger model thickness. Gray and Webb8 collected published data on
may, therefore, be judged by the ability to predict the samples of four row heat exchangers with continuous flat
air-side heat transfers and pressure losses accurately. fins and round tubes in a staggered arrangement to
Many experimental studies have been carried out on develop a correlation of j factor and Reynolds number
this type of compact heat exchanger to improve the based on the tube diameter. Geometrical effects were
understanding of heat transfer performance and pressure included as the ratios of tube spacing to row spacing and
losses over a wide range of air on velocities. Correlations fin spacing to tube outside diameter. A multiplier was
have been developed to predict the Colburn factor, j, and used to account for the effect of the number of tube rows.
friction factor, f, from experimental results for the wide The effect of the number of tube rows above five was
range of fin and tube geometries used in these found to be small. They also derived a correlation for f
applications. McQuiston5 developed a correlation to factor based on the superposition of friction over the
Round tube and plate nned heat exchangers 509
tubes and that on the fins. Geometric effects were where the overall heat transfer coefficient, U all, is given
included as the ratio of tube spacing to tube diameter. by;
The tubes friction factor was calculated in this correlation 1
from the work of Zukanskas9. Wang et al.10 studied the Uall (2)
1 Aa d t A
performance of 15 samples of plate fin heat exchangers a
ha ha At, m kt Ai hi
with different geometrical parameters, including the
Parameters, R n, may be introduced to identify the effects
number of tube rows, fin spacing and fin thickness in an
of the heat exchanger geometry on the thermal perfor-
open circuit wind tunnel. The results were presented in the
mance and pressure loss performance in the heat transfer
form of j and f factors against Reynolds number based on
coefficient and friction factor correlations.
the tube collar diameter. The results for two, four and six
The heat exchanger fin surface effectiveness, h a, in
tube rows were correlated using a multiple linear
eqn (2) is taken into consideration through the geometric
regression technique in a partial range of the experimental
parameter:
data (800 Re 7500) based on the previously published
1
correlation by Briggs and Young11. Although Wang et al R1 (3)
suggested a negligible effect of the fin thickness on the ha
thermal performance and pressure drop, the proposed
correlations for j and f factors contained the number of tube
rows and the ratios of the fin thickness and fin pitch to the
tube collar diameter. Ninety-seven percent of the experi-
mental data of j factor were correlated within 10%
while 88% of the f factor data were correlated within 10%.
The published correlations, described above, are
restricted to flat finned heat exchangers and to a limited
number of geometrical configurations. None of the above
authors reported any plate fin with modified surfaces
such as corrugated and sine wave. Use of these
correlations in the models developed by the authors for
automotive air conditioning heat exchangers introduced
errors as many of the configurations used were outside
the range of the correlations which were not applicable to
modified fins. The authors, therefore, undertook a
comprehensive experimental programme with tests on
28 heat exchangers in a thermal wind tunnel. The
samples included staggered tube arrangements, one, two
and four rows, fin spacing from 4.233 mm (6 fins/inch)
to 1.587 mm (16 fins/inch), row spacing of 22, 16.5 and
16 mm, tube external diameter of 9.956 mm, and tube
spacing of 25.4 and 19 mm. Two types of fin surfaces
were tested, flat and corrugated, with fin thicknesses
ranging from 0.2 to 0.12 mm. Figure 1 shows a sketch of Figure 1 Sketch of the corrugated fin
the corrugated fin used in this study. Figure 1 Representation de lailette onduleee
A regression analysis of the experimental data was
used to derive correlations based on Colburn factor or
friction factor. The Reynolds number was based on the
hydraulic diameter. The geometrical parameters were
included in the correlations by consideration of their
significance and adoption of a fundamental approach to
heat transfer to establish the effects of these geometric
factors on the thermal and frictional performance. Unlike
the correlations published previously, the geometrical
parameters considered in this study included ratios derived
from the thermal and frictional performance and ratios of
the fin and tube geometries.
Analysis
The thermal performance of a compact heat exchanger may
be expressed in terms of the number of transfer units, Ntu, as:
AU Figure 2 Hexagonal fin array
Ntu a all (1)
Cmin Figure 2 Disposition hexagonale dailettes
510 M. Abu Madi et al.
Table 1 Heat exchanger geometry for flat fin configuration with staggered tube
Tableau 1 Geometrie de lechangeur de chaleur : configuration a ailette plate et a tubes en quinconce
Sample no. No. of rows (N r) Tube spacing Row spacing Fin spacing Fin thickness
(S t) mm (S r) mm (S f) mm (d f) mm
a 4 25.4 22 1.847 0.13
b 4 25.4 16 1.615 0.12
c 4 19 16.5 1.843 0.12
d 4 19 16.5 2.646 0.12
e 4 19 16.5 2.545 0.13
f 2 25.4 22 2.620 0.13
g 2 25.4 22 1.833 0.13
h 2 25.4 16 2.543 0.12
i 2 19 16.5 1.640 0.12
j 2 19 16.5 2.593 0.12
k 2 19 16.5 2.541 0.13
l 2 19 16.5 1.915 0.13
m 1 25.4 16 1.859 0.12
n 1 25.4 16 1.635 0.12
o 1 25.4 16 2.586 0.12
512 M. Abu Madi et al.
drop across the heat exchanger was measured by means is reported on corrugated fins in the literature. The
of 2 4 junction thermopile. The air-side pressure drop physical dimensions of the compact heat exchanger
across the sample was measured from pressure tappings samples tested with the two types of fin are listed in
upstream and downstream, which were connected to a Tables 1 and 2 below.
pressure transducer with an operating range of 025 mm Each sample was tested with the air-on velocity varied
H 2O gauge. Water was heated in a 450-l reservoir with from 1 to 20 m/s. Separate tests were undertaken to
three 7.5-kW electric heater elements. A thyristor determine the heat transfer and pressure losses. In
controller maintained the water temperature within the case of heat transfer the water flow rate was constant
0.2 K. A fixed high water flow rate of 1.8 kg/s was at 1.8 kg/s with the inlet temperature of 84C. Pressure
supplied by a centrifugal pump to ensure a low water- loss tests were conducted isothermally to eliminate
side thermal resistance. The detailed description of the inertia terms and changes in the fluid transport
test equipment is reported by Achaichia and Cowell14 properties.
and Yu15. The experimental results are typified by Figure 4 of
heat dissipation and air pressure losses against air
velocity for sample h.
Experimental results
Figures 59 show the value of Colburn factor and
Two types of fin were considered in the experimental friction factor determined experimentally for the heat
programme; flat fin and corrugated fin. Little information exchangers used in the correlation.
Table 2 Heat exchanger geometry for corrugated fin configuration with staggered tube
Tableau 2 Geometrie de lechangeur de chaleur : configuration a ailette ondulee et a tubes en quinconce
Sample no. No. of rows (N r) Tube spacing Row spacing Fin spacing Fin thickness
(S t) mm (S r) mm (S f) mm (d f) mm
p 4 25.4 22 4.129 0.2
q 4 25.4 22 2.548 0.12
r 4 25.4 22 1.820 0.12
s 4 25.4 22 2.567 0.13
t 4 25.4 22 1.853 0.13
u 4 25.4 16 2.586 0.13
v 4 25.4 16 2.122 0.13
w 2 25.4 22 4.018 0.2
x 2 25.4 22 2.543 0.12
y 2 25.4 22 1.838 0.12
z 2 25.4 22 2.612 0.13
A 2 25.4 22 1.827 0.13
B 2 25.4 16 2.589 0.13
Geometry effect The effect of fin thickness on the heat transfer and
pressure drop was studied by comparing the performance of
Figure 10 shows the effect of fin type on the thermal 12 samples with flat and corrugated fins. Two fin thickness
performance and friction factor of the round tube and of 0.12 and 0.13 mm were considered in this comparison.
plate fin heat exchanger. The results show an increase in Figure 5 includes the performance of samples d and e.
the j and f factors for corrugated fins over the flat fins. The results show an increase in the j factor with smaller
This may be due to the fact that corrugated fins fin thickness. The fin thickness effect on the friction factor
accelerate the air flow and intensify the turbulence of was found to be very small. The same results were found
the air flow between them, thereby enhancing the heat for samples with different fin spacing and number of tube
transfer rate. This results in an increase in pressure drop rows, samples j and k in Figure 7, samples q, s, r and t in
and corresponding increase in the friction factor. Figure 6 and samples x, z, y and A in Figure 8.
Figure 5 Colburn factor and friction factor for four-row, flat fin heat exchangers
Figure 5 Coefficients de Colburn et de frottement pour des echangeurs de chaleur a ailette plate a 4 rangees
Figure 6 Colburn factor and friction factor for four-row, corrugated fin exchangers
Figure 6 Coefficients de Colburn et de frottement pour des echangeurs de chaleur a ailette ondulee a 4 rangees
514 M. Abu Madi et al.
Figure 7 Colburn factor and friction factor for two-row, flat fin exchangers
Figure 7 Coefficients de Colburn et de frottement pour des echangeurs de chaleur a ailette plate a 2 rangees
Figure 8 Colburn factor and friction factor for two-row, corrugated fin exchangers
Figure 8 Coefficients de Colburn et de frottement pour des echangeurs de chaleur a plaque ondulee a 2 rangees
Round tube and plate nned heat exchangers 515
and 93.9% with an rms error of 1.93.6%. The friction data of McQuiston16 for a similar geometry. The sample
factor for both types of fin was based on a similar had a tube spacing of 25.4 mm, fin spacing of 1.814 mm,
correlation used for j factor with coefficient of row spacing of 22 mm, fin thickness of 0.1524 mm, tube
determination varying between 98.1 and 92.7% with outside diameter of 9.956 mm and four tube rows. The
rms errors of 1.2 and 2.8% for flat and corrugated fins, predicted j and f factors were plotted against the experi-
respectively. mentally determined ones in Figure 11. This showed an
overall agreement of 11% and 5% for the j factor,
f Rec1 Rc2 c3 c4 c5
4 R5, 1 R8 R9 (29) while the f factor was predicted within 11% and 8%.
Figure 9 Colburn factor and friction factor for one-row, flat fin exchangers
Figure 9 Coefficients de Colburn et de frottement pour des echangeurs de chaleur a ailette plate a 1 rangee
factor to account for the effect of a different number of are shown in Table 5. The limits of eqn (30) are shown in
rows in the correlation. One heat exchanger configura- Table 6. The results show that the tube row effect is
tion was used, circular flat fins on round tubes. In the influenced by the heat exchanger geometries as well as
above studies the only factor reported as influencing the the Reynolds number. A regression analysis of the
thermal performance through the effect of the number of experimental data showed that the number of tube
rows of tubes was the Reynolds number. The number of rows had little effect on the friction factor.
rows had little influence on friction factor.
In this study, 10 samples with five different geome-
tries of corrugated fins and 14 samples with six different Conclusions
geometries of flat fin were used to establish the influence Experimental results are presented for a number of round
of the number of tube rows on thermal performance. A tube and plate fin heat exchangers of different geometric
regression analysis was used to establish the influence of configurations with both flat and corrugated fins.
the number of tube rows on Colburn factor from the Correlation equations for the j and f factors are presented
experimental data. as functions of Reynolds number, based on the hydraulic
j4 diameter, and geometric parameters of the heat exchan-
a0 a1Rea2 Nra3 Ra4 a5
3 R5, 1 (30) gers. This equation is believed to be more accurate than
jN r
earlier correlations with a broader scope of applications.
The values of the constant and indices (an) for eqn (30) A novel approach was developed for deriving the
Round tube and plate nned heat exchangers 517