International Journal of Heat and Mass Transfer 120 (2018) 534–539

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ijhmt

Thermo-hydraulic analysis of multi-row cross-flow heat exchangers

Cheen Su An a, Man-Hoe Kim b,⇑
a
Lasernics Co. Ltd, KAIST ICC, 193 Munji-ro, Daejeon 34051, Republic of Korea
b
School of Mechanical Engineering & IEDT, Kyungpook National University, Daegu 41566, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: This paper presents thermal hydraulic analysis of the cross-flow finned tube heat exchangers for an out-
Received 16 October 2017 door unit in residential air-conditioning and heat pump applications. Performance of heat exchangers
Received in revised form 15 December 2017 affect significantly the system energy efficiency and size of the air-conditioning and heat pumps. The
Accepted 17 December 2017
Navier-Stokes equations and the energy equation are solved for the three dimensional computation
domain that encompasses multiple rows of the fin-tube heat exchangers. Rather than solving the flow
and temperature fields for the outdoor heat exchanger directly, the fin-tube array has been approximated
Keywords:
by the porous medium of equivalent permeability, which is estimated from a three dimensional finite
Heat exchanger
Wave fin
volume solution for the periodic fin element. This information is essential and time-effective in carrying
Tube-by-tube method out the global flow field calculation which, in turn, provides the face velocity for the microscopic
Vapor compression cycle temperature-field calculation of the heat exchanger. The flow field and associated heat transfer for a wide
Heat pump range of face velocity and fin-tube arrangements are examined and the results are presented compared
Air-conditioning with experimental data. The predicted pressure drop and heat transfer rate for various inlet velocities are
in excellent agreement with the measured data.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction refrigerant circuit optimization study and they reported that the
capacity of heat exchanger and system with optimized circuit
The fin and tube heat exchangers are widely used in residential can be improved by 7.9% and 2.2%, respectively. Kim and Bullard
air-conditioning and heat pump applications because of its highly [6] investigated system performance for a window room air-
desirable properties such as compactness and manufacturing easi- conditioning system with microchannel condenser and compared
ness. The performance of heat exchangers (condenser and evapora- the results with the conventional system with fin and tube con-
tor) for vapor compression system affect significantly the efficiency denser. An et al. [7], and An and Choi [8] conducted numerical
and size of the heat pump system. Many investigators have con- study on thermal hydraulic performance of fin-tube heat exchang-
ducted various experimental and numerical works on heat ers under dry and wet conditions. Effect of heat exchanger config-
exchangers to improve system performance [1–12]. Kim et al. [1] uration on refrigeration cycle performance was also investigated
conducted a critical review of numerical and experimental studies by Saboya et al. [9] and Klein and Reindl [10]. A great deal of efforts
on the thermal hydraulic performance of louvered fin heat have been put into obtain heat-transfer correlations involving, for
exchangers. Saleem and Kim [2] investigated numerically the effect instance, Reynolds numbers, Colburn j-factor, friction f-factors, etc.
of louver pitch variation on the air-side thermal hydraulic perfor- This approach of predicting the heat exchanger performance would
mance of microchannel heat exchanger in Reynolds number of be economic, when successful, as it avoids time-consuming exper-
50–450. Brignoli et al. [3] evaluated the effect of transport proper- iments or computations. However, the empirical correlations, in
ties of fluid and refrigerant temperatures in heat exchangers. They general, lack generality and cannot readily be modified for chang-
developed a model which simulates the refrigeration cycle with ing geometries. Furthermore the air-side flow profile of heat
inlet temperature profiles. The model is also capable of optimizing exchanger affects significantly the thermo-hydraulic performance
the refrigerant mass flux in order to improve the system perfor- of the heat exchangers in air-conditioning and heat pump applica-
mance. Yashar et al. [4] developed heat exchanger circuit design tions [5,11,12], however it is very complicated to get the accurate
method with genetic algorithm. Yashar et al. [5] conducted the face velocity profile data from experimental or numerical investi-
gations. A more accurate and yet practical approach would be
⇑ Corresponding author. approximate the heat exchanger as a porous medium to obtain
E-mail address: [email protected] (M.-H. Kim). the global flow field. The heat-transfer characteristics may then

https://doi.org/10.1016/j.ijheatmasstransfer.2017.12.088
0017-9310/Ó 2017 Elsevier Ltd. All rights reserved.
 C.S. An, M.-H. Kim / International Journal of Heat and Mass Transfer 120 (2018) 534–539 535

be calculated by the tube-by-tube method using the flow field

already obtained.
The purpose of this paper is to develop such method and apply
it to an outdoor unit of a heat-pump system. The details of the
analysis procedure and the results are presented in the following
sections.

2. Analysis and modeling

The schematic of a heat exchanger under consideration in an air

conditioning system is shown in Fig. 1.
The tube arrangement is staggered and the fins are of wave type
(see Fig. 2). Rather than solving the flow through the heat exchan-
ger directly with any finite arithmetic, which requires prohibitively Fig. 2. Wave fin-tube heat exchanger.
high computational efforts, the heat exchanger is modeled as a
porous medium. The effective permeability and the inertial resis-
tance factor of the medium, needed to calculate the global flow and symmetric condition is imposed at the periodic and symmetric
field for an outdoor unit, are estimated from an accurate three- boundary. All the boundary conditions are presented as mathemat-
dimensional finite-volume calculation for a single periodic module ical forms in Eqs. (4)–(9).
of the given heat exchanger depicted in Fig. 3. For the computa- Velocity inlet boundary condition is used at the inlet face:
tional domain shown in Fig. 3, the following steady incompressible u ¼ uin v¼w¼0 T ¼ T in ð4Þ
flow equations are solved for the velocity and temperature fields
by using a commercial code, FLUENT [13]: No slip and constant temperature conditions at the wall:
Continuity equation: u ¼ v ¼ w ¼ 0 T ¼ Tw ð5Þ
@
ðquj Þ ¼ 0 ð1Þ Periodic boundary condition on both sides of the fins:
@xj
@u @ v @T
Momentum equations: ¼ ¼0 w¼0 ¼0 ð6Þ
@z @z @z
   
@ @p @ @u @u 2 @u Symmetry boundary condition is used on center between fins:
ðquj ui Þ ¼  þ l i þ j  l dij þ F i ð2Þ
@xj @xi @xj @xj @xi 3 @x
@u @w @T
¼ ¼0 v¼0 ¼0 ð7Þ
Energy equation: @y @y @y
@ @2T Pressure outlet boundary condition is applied at the outlet of
qC p ðuj TÞ ¼ k 2 ð3Þ
@xj @xj the fluid domain:

The flow is assumed laminar as the Reynolds number based on @u @ v @w @T

¼ ¼ ¼ ¼0 ð8Þ
the fin pitch of 1.7 mm is less than 700 [14]. The upstream and @x @x @x @x
downstream boundaries are placed sufficiently far away from the For fluid-solid interface:
fin, from 3 to 6 times of tube diameter (do), so that a uniform
@T s @T f
velocity distribution and the parabolic condition can be applied u ¼ v ¼ w ¼ 0 Ts ¼ Tf ks ¼ kf ð9Þ
at the respective boundaries. The no-slip and the constant temper- @n @n
ature conditions are prescribed at the solid wall while the periodic Once the relationship between the pressure drop and the inlet
velocity has been established from this calculation, the permeabil-
ity K and the inertial factor C of the equivalent porous medium can
be determined. The Forchheimer formulation [15] is given as:
 
Dp l 1
¼ u þ qCu2 ð10Þ
l K 2
K: permeability of porous medium
C: inertial factor
U: air velocity.

The first and second terms in Eq. (10) indicate the viscous and
inertial characteristics of porous media flow [16], respectively.
The heat transfer from the solid of constant temperature to the
cooling air is computed from the enthalpy change of the air. Taking
the contact resistance between the fin and the tube wall and the fin
efficiency into account, the total heat transfer rate of the heat
exchanger unit is then calculated from the tube–by-tube method
[17] described below for the velocity distribution at the fin inlet
obtained in the global flow analysis.
For the global flow calculation, Eqs. (1), (2) and (10) are solved
for the outdoor unit (see Fig. 8) of a heat pump system. The heat
exchanger segment of the domain is approximated by the porous
Fig. 1. Schematic of a heat exchanger. medium, and Eq. (10) is solved in that region in place of Eq. (2).
536 C.S. An, M.-H. Kim / International Journal of Heat and Mass Transfer 120 (2018) 534–539

Fig. 3. Schematic of a periodic module.

 
Using the heat-transfer results for the periodic module, the
C a UA=C a
T r;out ¼ T a;in þ T r;in  T a;in exp ðe  1Þ ð14Þ
overall heat-transfer rate for the whole heat exchanger, in which Cr
the temperature varies along the tube, may be estimated by the
Following the procedure described below, the temperature at
tube-by-tube method. The overall heat transfer coefficient, U, of a
each point along the heat exchanger can be determined.
row of the heat exchanger is defined as
Q ¼ U  A  DT lm ð11Þ 1. Assume the air temperature at the exit every tube.
2. Using Eqs. (13) and (14), calculate the water temperature at the
where DTlm is the log-mean temperature difference and A is the
exit of each tube for the given air temperature.
heat transfer area.
3. Calculate the air temperature by using the water temperature
T a;in  T a;out given and Eqs. (13) and (14).
DT lm ¼ ð12Þ
ln½ðT w  T a;out Þ=ðT w  T a;in Þ 4. Repeat Steps 2 and 3 until the air and water temperatures
converge.
The net heat-transfer rate Q is estimated by multiplying the fin
efficiency to the heat flux calculated for a single periodic module A major drawback of the above procedure may be that it relies
above. The fin efficiency is obtained from the empirical formula on the fin efficiency which is empirical to determine the actual
of Schmidt [18]. heat flux. One can avoid this by solving the entire heat-transfer
When the UA of the heat exchanger, the inlet temperatures of chain simultaneously, from coolant inside the tube to the outside
the air and the water are known, the outlet temperatures can be air. The energy equation needs to be solved for a domain that
determined by using Eqs. (13) and (14) which are derived from includes the tube and fin element.
the energy balance relation of the heat exchanger [19]. The thermal resistance for the system may be broken down to
  
Cr
C a UA=C a
five stages (see Fig. 4): convection from the coolant to the tube
T a;out ¼ T a;in þ T r;in  T a;in 1  exp e 1 ð13Þ (1/(hrAr)), conduction through the solid (tt/(ktAr)), tube and fin col-
Ca Cr
lar (1/(hcAr)), contact resistance between the tube wall and the fin
 C.S. An, M.-H. Kim / International Journal of Heat and Mass Transfer 120 (2018) 534–539 537

Table 2
Permeability and inertial resistance factor for various row arrangements.

Number of rows K (m2) C (m1)

1 Experiment 9.5725e8 1.1236e2
Calculation 2.2944e7 1.7794e2
2 Experiment 1.7118e7 1.6988e2
Calculation 1.9416e7 1.6181e2
3 Experiment 1.2964e7 1.6845e2
Calculation 1.9915e7 1.6210e2

Fig. 4. Detailed view of fin-tube assembly.

Table 1
Relative thermal resistance range for refrigerant air fin-tube heat exchangers.

1/(hrAr) tt/(ktAr) 1/(hcAr) tf/(kfAr) 1/(haAags)

0.08–0.18 0.0008–0.0018 0.12–0.27 0.0005–0.0011 0.55–0.80

Fig. 7. Overall heat transfer rate versus various face velocities.

Fig. 5. Sample grid.

Fig. 6. Pressure drops for various face velocities. Fig. 8. Computational grid for an outdoor unit.
538 C.S. An, M.-H. Kim / International Journal of Heat and Mass Transfer 120 (2018) 534–539

collar (tf/(kfAr)), and convection from the fin surface to the air (1/
(haAags)). A general order of relative thermal resistance for each
phase is given in Table 1 [20].
It can be seen from the table that the resistance due to conduc-
tion through the tube wall or fin collar is negligible. The thermal
resistance between the fin surface and the cooling air is reflected
through the use of fin efficiency. The effects of the convection
inside the tube and the contact resistance are determined empiri-
cally for this fin shape and given in Youn [20]. Corresponding val-
ues for hr and hc are 1.2340  104 W/m2 K and 6.791  103 W/m2
K, respectively, and are used in the present analysis.

3. Results and discussion

The fin pitch of the heat exchanger unit under consideration is

1.7 mm and the number of rows is 1, 2 or 3. Fig. 5 shows a sample
grid about a single tube, which is non-uniformly distributed to
place more grids near the solid surface. Only half of the unit fin
module has been considered using periodicity and symmetry
boundary conditions. Approximately 200,000 grid cells are used
to fit the domain of each row after grid independence study. Addi-
tional 10,200 and 17,000 grid cells are distributed for the upstream
and downstream regions of the fin, respectively [8]. Fig. 6 shows
the pressure drop versus the inlet flow velocity. Comparing calcu-
lation pressure drop with the test data has the largest error of
18.2% for 3-row heat exchanger partly due to the laminar flow
assumptions. Using 2nd order polynomial curve fitting, K and C
are determined and are presented in Table 2. The overall heat
transfer rates calculated by the tube by tube method is shown in
Fig. 7, and calculated values are good agreement with test data
in the error range of 0.5–2.5%.
Fig. 8 shows the computational grid for the outdoor unit of the
heat-pump system. Using the model coefficients in Table 2 as
inputs for the analysis of the heat exchanger, the flow field for a
given pressure drop is calculated. Fig. 9 shows the streamlines
and pressure distribution in the outdoor unit while the resulting
inlet velocity profile for 2- or 3-row tube array is depicted in
Fig. 10.
Fig. 9. Streamlines (a) and pressure distribution (b) for the outdoor unit of an air- It is seen from Fig. 10 that the cooling air flows more readily
conditioning system. through the upper part of the heat exchanger than the lower part,
especially for the single-row case, since the length of the flow pas-
sage is shorter there. However, the velocity profile becomes more

Fig. 10. Velocity distribution at the inlet for a given pressure drop.
 C.S. An, M.-H. Kim / International Journal of Heat and Mass Transfer 120 (2018) 534–539 539

Acknowledgement

This work is supported by the Ministry of Trade, Industry &

Energy (MOTIE, Korea) under Industrial Technology Innovation
Program (No. 10070117).

References

[1] M.-H. Kim, S.Y. Lee, S.S. Mehendale, R.L. Webb, Microchannel heat exchanger
design for evaporator and condenser applications, Adv. Heat Transf. 37 (2003)
297–429.
[2] A. Saleem, M.-H. Kim, Air-side thermal hydraulic performance of microchannel
heat exchangers with different fin configurations, Appl. Therm. Eng. 125
(2017) 780–789.
[3] R. Brignoli, J.S. Brown, H.M. Skye, P.A. Domanski, Refrigerant performance
evaluation including effects of transport properties and optimized heat
exchangers, Int. J. Refrig. 80 (2017) 52–65.
Fig. 11. Overall heat transfer rate for number of rows.
[4] D.A. Yashar, J. Wojusiak, K. Kaufman, P.A. Domanski, Evolutionary computation
approach to heat exchanger design, in: ASHRAE Annual Conference, 2010.
[5] D.A. Yashar, S. Lee, P.A. Domanski, A rooftop air-conditioning unit performance
uniform as the number of row increases. This is because the rela- improvment using refrigerant circuitry optimization, Appl. Therm. Eng. 83
tive variation in the total pressure loss from the inlet to the exit (2015) 81–87.
[6] M.-H. Kim, C. Bullard, Performance evaluation of a window room air
becomes less as the pressure drop across the heat exchanger conditioner with microchannel condensers, J. Energy Resour. Technol 124
increases. The velocity profiles at the inlet plane can be used to (2002) 47–55.
repeat the tube-by-tube analysis for more realistic heat-transfer [7] C. An, D.H. Choi, M.-H. Kim, Flow and heat-transfer analysis for cross-flow fin-
tube heat exchangers in an air-conditioning system, in: Proceedings of 4th
analysis; the results are presented in Fig. 11. For a given pressure ICCHMT, Paris-Cachan, France, 2005.
loss through the outdoor unit, the two-row heat exchange appears [8] C. An, D.H. Choi, Analysis of heat-transfer performance of cross-flow fin-tube
to perform best in the study. heat exchangers under dry and wet conditions, Int. J. Heat Mass Trans 55
(2012) 1496–1504.
[9] F.E.M. Saboya, C.E.S.M. da Costa, Minimum irreversibility criteria for heat
4. Conclusions exchanger configurations, J. Energy Resour. Technol 121 (1999) 241–246.
[10] S.A. Klein, D.T. Reindl, The relationship of optimum heat exchanger allocation
and minimum entropy generation rate for refrigeration cycles, J. Energy
An efficient and accurate heat-transfer analysis procedure for a Resour. Technol. 120 (1998) 172–178.
multi-row cross-flow heat exchanger of an air-conditioning system [11] J.N. Mao, H.X. Chen, H. Jia, Y.Z. Wang, H.M. Hu, Effect of air-side flow
has been developed. The steady state 3-dimensional continuity, maldistribution on thermalehydraulic performance of the multi-louvered fin
and tube heat exchanger, Int. J. Therm. Sci. 73 (2013) 46–57.
momentum, and energy equations are solved by using the finite [12] D. Wang, C. Liu, D. Yu, J. Chen, Influence factors of flow distribution and a
volume method to obtain the thermal-hydraulic characteristics of feeder tube compensation method in multi-circuit evaporators, Int. J. Refrig.
multi-row fin and tube heat exchangers. Approximating the heat 73 (2017) 11–23.
[13] ANSYS, ANSYS-Fluent user manual, 2016.
exchanger by a porous medium, of which the effective permeabil- [14] S. Fukui, M. Sakamoto, Some experimental results on heat transfer
ity and the inertial resistance factor are estimated by a separate characteristics of air cooled heat exchangers for air conditioning condenser,
three-dimensional finite-volume calculation, the overall flow field Bulletin of JSME 11 (44) (1968) 303–311.
[15] P. Forchheimer, Wasserbewegung durch Boden, Zeitschrift des Vereines
can be calculated fairly efficiently. By coupling the three-
Deutscher Ingenieuer, 45 ed., 1901.
dimensional heat-transfer results for a periodic domain with the [16] D.A. Nield, A. Bejan, Convection in Porous Media, third ed., Springer, 2006.
tube-by-tube method, the total heat transfer characteristics for [17] P.A. Domanski, Computer Modeling and Prediction of Performance of An Air
Source Heat Pump With A Capillary Tube (Ph.D. Dissertation), The Catholic
the whole heat exchanger can be obtained. If the appropriate tur-
University of America, Washington D.C., 1982, pp. 62–93.
bulent model is adapted in the study, the results will be better than [18] T.E. Schmidt, Heat transfer calculations for extended surfaces, J. ASRE, Refrig.
laminar model. Eng. 4 (1949) 351–357.
[19] J.L. Threlkeld, Thermal Environmental Engineering, Prentice-Hall Inc, New
York, 1970.
Conflict of interest [20] B. Youn, Private Communication, Samsung Electronics Co., 2004.

The authors declare that there is no conflict of interests.