Applied Thermal Engineering 125 (2017) 780–789

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Air-side thermal hydraulic performance of microchannel heat

exchangers with different fin configurations
Arslan Saleem, Man-Hoe Kim ⇑
School of Mechanical Engineering, Kyungpook National University, Daegu 41566, Republic of Korea

h i g h l i g h t s

Investigate air-side thermal hydraulic performance for microchannel heat exchangers.

Local air-side heat transfer coefficient and pressure drop are addressed.

Fin configuration with louver pitch of 1.0 mm has the maximum Nusslet number.

Provide the effect of fin height on air-side thermal hydraulic performance.

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

Article history: This paper presents three-dimensional steady state analysis of compact multi-louvered fin and flat tube
Received 7 April 2017 heat exchangers applicable in domestic and mobile air conditioning systems. Conjugate heat transfer
Revised 5 June 2017 analysis has been conducted for five geometric configurations of the heat exchanger based on the louver
Accepted 10 July 2017
pitch values (0.8, 1.0, 1.4, 1.7 and 2.0 mm) in the laminar range of Reynolds number 50–450. For all geo-
Available online 12 July 2017
metric configurations, local air-side heat transfer coefficient and pressure drop are presented in terms of
Colburn j factor and Fanning friction f factor, respectively. In addition, the effect of fin height on thermal
Keywords:
hydraulic performance has also been addressed based on local Nusselt number distribution along the fin
Microchannel heat exchanger
Louver pitch
height. The results reveal that geometric configuration with the louver pitch of l.0 mm showed the high-
Conjugate heat transfer est average Nusselt number of 123% at the penalty of 116% pressure drop compared to louver pitch of
Nusselt number 2.0 mm at the Reynolds number of 50.
Friction factor Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction ger geometry based on their results. The effect of louver angle vari-
ation (15–29°) on air-side heat transfer at high flow velocities has
The louvered fin and flat multi-channel heat exchanger have been studied by Kim and Bullard [4], they concluded that flow
enormous applications in residential and automotive air condition- depth is also one of important parameter responsible for pressure
ing systems because of their compactness, high-pressure sustain- drop. Kim and Cho [5] conducted a detailed experimental study on
ability and reduced refrigerant charge. Multiple researchers prediction of critical Reynolds number for optimum heat transfer
performed experimental and numerical investigations in order to coefficient as well as the sensitivity of critical Reynolds number
explore air-side thermal and hydraulic performance of microchan- towards louver pitch to fin pitch ratio. Moreover, they also
nel louvered fin heat exchangers. developed new correlations of j and f factors. Kang and Jun [6]
Kim et al. [1] performed a critical review of numerical and studied fourteen heat exchanger geometries with the purpose of
experimental studies on the thermal hydraulic performance of flat investigating the effects of louver pitch and angle variations on
tube and louvered fin heat exchangers. Chang and Wang [2] and the air-side heat transfer and pressure drop characteristics for
Chang et al. [3] conducted an experimental study on the air-side automotive applications. They also proposed j and f factor correla-
heat transfer and pressure drop characteristics and established tions as a function of average louver pitch, louver angle, fin pitch,
generalized j and f factor correlations for louvered fin heat exchan- and Reynolds number (based on louver pitch).
Some researchers [7–9] opted for numerical study of the
louvered fin heat exchangers. Tafti et al. [7] conducted
⇑ Corresponding author. two-dimensional numerical studies on flow transitions in multi-
E-mail addresses: [email protected] (A. Saleem), [email protected] louvered fin array and reported instability in the upstream flow
(M.-H. Kim).

http://dx.doi.org/10.1016/j.applthermaleng.2017.07.082
1359-4311/Ó 2017 Elsevier Ltd. All rights reserved.
 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789 781

Nomenclature

A surface area ½m2  Q heat transfer rate ½W

E energy ½J T temperature ½K
f fanning friction factor ½ V velocity magnitude ½m s1 
Fd flow depth ½m v overall velocity vector ½m s1 
Fh fin height ½m l dynamic viscosity ½Pa s
Fp fin pitch ½m q density ½Kg m3 
h airside convective heat transfer coefficient, ½W m2 K1  s wall shear stress ½N m2 
1
i enthalpy, ½kJ kg 
I unit matrix ½ Subscripts
j Colburn j-factor ½ a,in inlet air
2
J diffusion flux ½kg m s1  a,out outlet air
k thermal conductivity ½W m1 K1  B bottom
Lfirst first louver ½ T top
Lp louver pitch ½m f fin
Llast last louver ½ w water
p pressure ½Pa

with increasing Reynolds number. Later on, Atkinson et al. [8] con- reported. Conjugate heat transfer simulations have been per-
ducted two and three-dimensional transient numerical studies on formed for low Reynolds range of 50–450 reflecting operation con-
flat tube louvered fin heat exchanger and reported that 3D model dition of microchannel heat exchangers in residential air
estimate accurate results in comparison to 2D model. However; conditioning and heat pump applications.
Perrotin and Clodic [9], in their numerical and experimental study
on flat tube louvered fin heat exchangers, pointed towards under 2. Numerical methodology
prediction of results for three-dimensional numerical analysis
and attributed computational limitations for it. 2.1. Physical and computational model
Several experimental and numerical studies have been con-
ducted in an effort to improve the thermal hydraulic performance Literature review reveals that most of the numerical studies
of microchannel heat exchangers by implementing various tech- used laminar flow model with the assumption of infinite fin
niques. Ryu et al. [10] carried out design optimization study on a cor- thermal conductivity and isothermal tube wall temperature for
rugated louvered fin geometry and recommended an optimal design the computational analysis, however, some of the studies asserted
for corrugated louvered fin based on three geometric parameters. that the flow appeared to be laminar only up to louver pitch based
Along with louvered fin heat exchangers, Li et al. [11] numerically Reynolds number value of 400 and at higher Reynolds the
investigated the air-side performance of a microchannel heat unsteadiness dominates the flow regime. In this study, a steady
exchanger with 3-D integrated fins, and reported improved perfor- state conjugate heat transfer analysis has been conducted to imple-
mance of louvered fin heat exchangers in comparison to integrated ment the effects of fin conduction and convection. The flow was
and wavy fins. Li et al. [12] suggested generalized j and f factor cor- assumed laminar considering the Reynolds number range
relations based on their experimental test data obtained from a sam- (50–450). Mass, momentum and energy conservation equations
ple of 11 heat exchangers with integrated fin. Li and Wang [13] for a steady state analysis are described in Eqs. (1)–(3).
experimentally examined the thermal and hydraulic performance Continuity equation:
of six multi-region geometrical configurations of louvered fins for
!
a range of Reynolds number 400–1600 and subsequently developed r  ðq v Þ ¼ 0 ð1Þ
j and f correlations. Jang and Chen [14] performed variable louver
angle optimization study and reported the initial and variable louver Momentum equation:
angle values for nine face velocities. Liang et al. [15] conducted an ! ! ! ! 2 ! !
experimental and numerical study on thermal hydraulic perfor- r  ðq v Þ v ¼ rp þ r  ðl½ðr v þrv T Þ  r v IÞ þ q g ð2Þ
3
mance using 24 automotive heat exchanger samples. Recently
Shinde and Lin [16] conducted an experimental study using 26 heat And the energy equation for steady state analysis is given as:
exchanger samples with different louver fin configurations and pro- !
r  ðv ðqE þ pÞÞ ¼ r
posed low and high Reynolds number region correlations, respec-     
! ! ! 2 ! !
tively for both j and f factors.  krT  h J þ l ðr v þrv T Þ  r v I v
In light of the literature study, it has been observed that despite 3
previous louver fin parametric studies for compact heat exchang- ð3Þ
ers there is still a considerable gap of knowledge regarding the
influence of louver pitch on thermal and hydraulic performance where specific energy, E is defined by Eq. (4).
specifically for the case of microchannel heat exchangers. There-
p v2
fore, the present work focuses on the effect of louver pitch varia- E¼i þ ð4Þ
q 2
tion on the air-side thermal hydraulic performance of
microchannel heat exchangers. Fig. 1 and Table 1 show geometrical
parameters of louver geometry considered in the study. 3-D 2.2. Computational domain and boundary conditions
numerical analysis has been conducted with five different louver
pitches (0.8, 1.0, 1.4, 1.7 and 2 mm) as shown in Table 1. And air- For the current study, the computational domain comprises of
side local heat transfer coefficient and friction factor has been two domains, i.e. outer fluid domain Xf that encloses the solid
782 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789

(a) Geometry parameters

Fin domain N

F H

N
Downstream
domain
F

Upstream
domain
(b) Computational domain distribution

(c) Boundary conditions

Fig. 1. Geometry parameters, computational domain and boundary conditions.

Table 1
Geometrical parameters.

Geometries Fin pitch Flow depth Louver angle Louver pitch Louver length Fin height Fin thickness Number of
½mm ½mm ½deg ½mm ½mm ½mm ½mm louvers
Fp Fd La Lp Ll H df Nl

G1 1.4 20 23 0.8 6.4 8.15 0.1 22

G2 1.4 20 23 1.0 6.4 8.15 0.1 18
G3 1.4 20 23 1.4 6.4 8.15 0.1 12
G4 1.4 20 23 1.7 6.4 8.15 0.1 10
G5 1.4 20 23 2.0 6.4 8.15 0.1 8

inner domain Xs within. The rectangular fluid domain is subdi- five times and fifteen times the value of fin pitch respectively.
vided into three parts, the upstream domain, fin domain and the Whereas, the length of fin domain is kept equivalent to louvered
downstream domain as shown in Fig. 1. The length of the upstream fin parameter, flow depth. Width of the whole fluid domain is half
and downstream domain is a function of the fin pitch and equals of fin height since half fin is modeled for the computational
 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789 783

analysis and depth of the fluid domain is kept same as the fin pitch
parameter.
The boundary conditions used for the simulations are described
below in Eqs. (5)–(10):
Velocity inlet boundary condition is used at the inlet face:
u ¼ uin v ¼w¼0 T ¼ T in ð5Þ
Periodic boundary condition on both sides of the fluid domain:
@u @ v @T
¼ ¼0 w¼0 ¼0 ð6Þ
@z @z @z
Pressure outlet boundary condition is applied at the outlet face
of the fluid domain:
(a) Blocking for structured mesh

@u @ v @w @T
¼ ¼ ¼ ¼0 ð7Þ
@x @x @x @x
Symmetry boundary condition is used with the fluid domain’s
top and bottom faces:
@u @w @T
¼ ¼0 v¼0 ¼0 ð8Þ
@y @y @y
At the constant temperature (tube region):
u ¼ v ¼ w ¼ 0 T ¼ Tw ð9Þ (b) O-grid around louvered fins

For fluid-solid interface: Fig. 2. Mesh topology.

@T s @T f
u ¼ v ¼ w ¼ 0 Ts ¼ Tf ksl ¼ kfl ð10Þ nodes along the flow depth and fin height were increased at 25%
@n @n
and 20% respectively. Steady state simulations were conducted for
all structured meshes at frontal air velocity of 2 m/s. Colburn j factor
2.3. Numerical solution and the computational cost parameters were documented and com-
pared for all structured meshes (M1–M4). Colburn j factor against
In order to solve continuity, momentum and energy equations the grid size of hexahedral meshes is shown in Fig. 4. Computational
over the individual cells in the computational domain finite vol- results show that the variation in j factor for M2, M3 and M4 is
ume technique is opted, using commercial software ANSYS- around 1% while the computational cost that includes the computa-
FLUENT. SIMPLE scheme is used for pressure-velocity coupling tion time and allocated memory is several times higher for M3 and
and least square cell based discretization scheme is used to solve M4. Therefore, on the basis of the aforementioned reasons, M2
the three-dimensional steady state model. Under relaxation factors was selected for carrying out all the simulations in the present study.
of 1, 0.7 and 0.3 were fixed for energy, momentum and pressure
respectively. And residuals for convergence criteria of energy,
3. Results and discussions
momentum and continuity equations were set to be 1e7, 1e3
and 1e3 respectively.
3.1. Validation of the computational model
Fluid and stationary both domains were discretized using hex-
ahedral elements to ensure conformal mesh interface between the
Numerical analysis on airside thermal hydraulic performance
two domains. The hexahedral mesh was adopted for both domains
was conducted against louver pitch based Reynolds number, corre-
in order to avoid the interpolation losses. Mesh topology for the
sponding to the thermal conditions i.e. at the air inlet and water
computational domain is such that O-grids were created around
inlet temperatures of 21 °C and 45 °C respectively. Fin conduction
each louver. In order to resolve the velocity and temperature gra-
and effect of air convection were both included by setting up a con-
dients, fifteen nodes were established within the O-grid. Blocking
jugate heat transfer simulation. The tube wall temperature was
was created using ANSYS-ICEM CFD to generate the structured
considered constant which points towards the assumption that
grid. The five geometrical configurations G1 to G5 have 453, 382,
tube wall has infinite conduction coefficient. Moreover, water-
262, 227, and 191 blocks in total, respectively. The detailed mesh
side was not modeled to predict airside heat transfer coefficient
topology and blocking are shown in Fig. 2. Grid skewness and qual-
values accurately. Log mean temperature difference method was
ity were made certain to be higher than 0.5 across the whole com-
implemented to predict the airside heat transfer coefficient. Math-
putational domain of hexahedral elements. The structured mesh of
ematical formulation of airside heat transfer coefficient and modi-
the computational domain is shown in Fig. 3.
fied expression of LMTD for cross flow heat exchanger is stated in
Eqs. (11) and (12) respectively.
2.4. Grid independence study
Q
h¼ ð11Þ
The structured mesh of four different mesh sizes was generated A:LMTD
in order to undergo the simulations, whereas clustering was used
to capture the boundary layer near the louvers. Mesh details for Ta;out  Ta;in
LMTD ¼   ð12Þ
the computational domain for all four structured mesh are shown T Ta;in
ln TwwTa;out
in Table 2. Number of nodes were kept constant in all grid sizes,
upstream and downstream of the louvered fin i.e., along N uf and For validation of the used numerical model sample 2 from the
N df . Along the fin pitch and within the O-grids around the louvers, experimental study carried out by Shinde and Lin [16] is modeled.
the number of nodes were also kept constant. While, number of Sample 2 has following geometric parameters; fin pitch 1.25 mm,
784 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789

Fig. 3. Mesh of computational domain.

Table 2
Mesh properties.

Mesh No. of No. of nodes No. of nodes No. of nodes No. of nodes Nodes within Near wall Maximum Total Computation
nodes alongside flow downstream alongside fin alongside fin the boundary element element number of time
upstream depth pitch height layer size size elements
Nuf Nf Ndf Fp H O Enw ½m Emax ½m Nt t ½s

M1 50 200 100 50 64 15 1.7E05 0.0001852 1153287 197

M2 50 250 100 50 85 15 4.1E06 0.0001001 2498579 478
M3 50 300 100 50 100 15 3.0E06 0.0000845 3565820 864
M4 50 320 100 50 120 15 2.8E06 0.0000823 4120965 925

late the airside heat transfer coefficient. The formulations used for
finding out local heat transfer coefficient are shown below in Eqs.
(13)–(15).

q_ wall
T fin wall ¼ þ T wall adj: ð13Þ
hwall

q_ wall
hlocal ¼ ð14Þ
T fin wall  T bulk

hlocal  Lp
Nulocal ¼ ð15Þ
k
where, T bulk is the user defined function containing the bulk temper-
ature data away from the wall. Local Nusselt number is calculated
at top and bottom surfaces of fin at different fin heights.Fig. 6 shows
Fig. 4. Variation of Colburn j factor for different grid sizes. the change in airside local Nusselt number corresponding to three
Reynolds values. The local Nusselt number is plotted along the flow
in the x-axis direction. The x-axis values are non-dimensionalized
fin thickness 0.1 mm, louver pitch 1.14 mm, louver angle 25, fin by dividing over the whole flow depth. The local Nu is plotted on
height 10 and fin depth 25 mm. Half of the fin was modeled using fin wall surface, in the first half array of louvered fins the heat flux
the symmetry boundary condition. The results are reported as Col- is higher at the top surface while the opposite is true for the second
burn j factor as shown in Fig. 5, the experimental and numerical half of the louvered fin array. The highest values of local heat trans-
results are found in good agreement. fer coefficient are noted at the leading edges of the louvered fins [9].
The heat transfer coefficient generally seems to drop at the trailing
3.2. Local heat transfer coefficient edges of the louvers with an exception of the redirection louver
where local heat transfer increases due to the flow acceleration.
The simulations were carried out on five heat exchanger config- The flow is more pronounced in the second half of the louver array
urations with different louver pitch values and the airside heat resulting in higher heat transfer coefficient values at bottom surface
transfer coefficient is reported as local Nusselt number. Commer- in comparison with those at top of fin surfaces in the first half.
cial software over predicts the local Nusselt number since the bulk Moreover, the second louver’s leading edge shows the highest value
temperature data gathered from nodes adjacent to the wall is of the local Nusselt number and that may be because of the vortex
almost equivalent to the wall temperature due to no-slip boundary generation in between first and second louver. This phenomenon is
condition on fin walls. In order to deal with this problem, bulk tem- noticed at higher Reynolds numbers and is absent in the case of low
perature data is collected from nodes away from the wall to calcu- Reynolds value of 50.
 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789 785

Fig. 5. Numerical results validation with the experimental results.

Fig. 6. Variation of local Nusselt number along the flow direction at different Reynolds number.

Heat transfer coefficient variation in different heat exchanger wakes are noticed. This increased air flow velocity is due to the
geometries based on louver pitch is also observed at different Rey- flow acceleration since the redirection louver directs the flow into
nolds values. Fig. 7 shows at Reynolds number 450, the trend seen the bottom surface of the fin. A thin thermal boundary layer is
in the local Nusselt number curves is same for different louver observed on the bottom face of the redirection louver compared
pitch geometric configurations with some noticeable differences. to its top surface, consequently leading to high local Nusselt num-
The highest value of the local heat transfer is detected for the ber values on the bottom face.
geometry with 0.8 mm louver pitch i.e. in the latter half of the lou- Wakes and flow acceleration phenomenon are also seen at
ver array. While for geometry with louver pitch values of 2.0 mm the bottom and top surface respectively of the first and last lou-
the highest heat transfer is noticed for the second louver due to ver. At the first louver, higher heat transfer coefficient is noticed
stationary vortex generation. This flow blockage phenomenon at both top and bottom surface of the first louver that suddenly
due to stationary vortex generation is noticed to be more frequent decreases. The local Nusselt number values are small around the
in 0.8 mm and 1.4 mm louver pitch geometries where the vortex is inner edge of both first and last louver, while the contrary is
observed in almost every alternate louver pair. observed on the outer surfaces. Fig. 9 displays that Nusselt num-
Local heat transfer coefficient at redirection louver surface is ber curves intercept each other right after the bend where the
plotted against the nondimensionalized length as shown in flow impinges itself more on the inner surface of the first and
Fig. 8. The flow velocity around the outer edges of the redirection last louver and the bottom surface shows higher values than
louvers is more pronounced compared to the inner edges where the top surface.
786 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789

Fig. 7. Variation of local Nusselt number in different HX configurations for Reynolds number 450.

to one-half and one-fourth times fin height is negligible. Fig. 10

reports the Nusselt number values at one-half and one-eighth of
fin height. It can be clearly noticed that there is no difference
between the two in the first half of louver array, while in the sec-
ond half higher values of local Nusselt number are observed at one-
eighth of fin height. Since no significant change is seen for local
Nusselt number belonging to half and quarter of fin height, this
raises the point of reducing the fin height. Higher thermal hydrau-
lic performance can be achieved by reducing the fin height since it
can accomplish same thermal performance with a high percentage
of volume reduction.

3.4. Average heat transfer coefficient and pressure drop

Local heat transfer coefficient is averaged over the whole fin

surface to calculate the average airside heat transfer coefficient.
The results of average airside local Nusselt number and pressure
drop for different louver fin configurations are shown in Figs. 11
and 12. An exponential increase in average Nusselt number is
noticed with the decrease of louver pitch value. While in the case
of pressure drop an exponential decrease is seen with louver pitch
increase for obvious reason. As the louver pitch decreases the num-
ber of louvers in a louvered fin array increase and the friction drag
Fig. 8. Local Nusselt number variation in different HX configurations at redirection increases multiple times. At low Reynolds number i.e. Re < 100, fin
louver for Reynolds number 400. geometry with louver pitch 1.0 mm showed maximum heat trans-
fer however at Re = 150, fins geometries with louver pitch 0.8 mm
and 1.0 mm showed identical values of the air-side local Nusselt
3.3. Effect of fin height number. With the further increase in Reynolds number fin geome-
try with louver pitch 0.8 mm dominates over the one with 1.0 mm
Effect of fin height is also investigated by calculating the local louver pitch. This is because the number of louvers increases with a
airside heat transfer coefficient along the fin height. Local Nu is decrease in louver pitch consequently resulting in increased ther-
reported at one-half, one-fourth and one-eighth of the fin height mal boundary layer regeneration phenomenon. Louvered fins with
value. The offset between the local Nusselt values corresponding louver pitch 1.7 mm and 2.0 mm showed approximate results with
 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789 787

Fig. 9. Local Nusselt number variation in different HX configurations at first and last louver for Reynolds number 400.

Fig. 10. Local Nusselt number variation along fin height for HX configurations with louver pitch 1.7 mm.

slightly improved results for 2.0 mm fin geometry. Indicating that has also been observed that with increasing Reynolds number
further increase in the louver pitch, would not be significantly the difference between calculated Nusselt number for all louver
affecting Nusselt number under similar boundary conditions. It pitch values becomes more significant. At Reynolds number 450,
788 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789

Fig. 11. Average Nusselt number for louver pitch based HX configurations.

Fig. 13. Colburn j factor vs. Reynolds number for different geometric
configurations.

Fig. 12. Pressure drop variation in louver pitch based HX configurations.

the percentage increase of Nusselt number calculated for fin

geometries with louver pitch 1.0, 1.4, 1.7 and 2.0 mm with respect
to 0.8 mm louver pitch fin geometry is 5.1%, 62.1%, 116.1% and
106.3% respectively.
As expected for all geometric configurations with louvered fins,
the pressure drop increases with increasing Reynolds number as
the pressure gradient at the back and forth of louvers increases Fig. 14. Fanning f factor vs. Reynolds number for different geometric
with increasing velocity. The difference between all pressure drop configurations.
values is very minute at low Reynolds number but with the
increasing Reynolds number, this difference becomes remarkable. where A; Ac ; h; qm and U are airside surface area, minimum airside
Hence it can be suggested that fin geometry with 1.0 mm louver free flow area, airside heat transfer coefficient, mean air density
pitch provides optimum results under the operating conditions and face air velocity respectively. Results show that both f and j fac-
investigated in this study, for both low and high Reynolds number tor exhibit decreasing trend with increase in Reynolds number as
range. Since at low Reynolds number it shows highest Nusselt reported by multiple researchers in the literature [9,11,12,16]. The
number values and almost similar pressure drop as compared to f and j factor plots vary from the average Nusselt number and pres-
other louver pitch configurations while, at higher Reynolds num- sure drop results since the face velocity parameter varies for each
ber although the fin geometry with louver pitch 0.8 mm provides louver pitch geometry configuration, to keep the flow Reynolds
improved heat transfer results but the penalty in pressure drop is number similar. The peak j factor values are seen for the geometric
remarkably high i.e. at Re = 450, the percentage increase of Nusselt configurations with louver pitch 1.0 mm and 2.0 mm respectively.
number and pressure drop for fin geometry with louver pitch The heat exchanger geometry configuration with louver pitch
0.8 mm as compared to 1.0 mm louver pitch fin is 5% and 73.6% 1.0 mm showed highest Colburn j factor curve, an increase of
respectively. 123–96% for average Nusselt number and 116–267% increase in
The results have also been represented in terms of f and j factor pressure drop is noticed when compared with 2.0 mm louver pitch
in Figs. 13 and 14, and the expressions used for calculation of fac- geometry for Reynolds number range of 50–450. A wide j factor off-
tors are given in Eqs. (16) and (17). set is visible between all geometric configurations at low Reynolds
number and it reduces with the Reynolds number increase.
h
j¼ Pr2=3 ð16Þ
qm Ucp
4. Conclusion
2DP Ac
f ¼  ð17Þ The focus of the study was to numerically investigate the effect
qm U 2 A of louver pitch variation on the air-side thermal hydraulic
 A. Saleem, M.-H. Kim / Applied Thermal Engineering 125 (2017) 780–789 789

performance of microchannel compact heat exchanger. In this Industry & Energy (MOTIE) of the Republic of Korea (No.
regard, 3-D steady state numerical simulations have been con- 10052926).
ducted on louvered fin geometric configurations with five different
louver pitch values. The geometric configurations and Reynolds References
number range (50–450) were selected in correspondence with
the domestic and mobile air conditioning applications. Air-side [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)
local heat transfer coefficients were analyzed for better under- 297–429.
standing of undergoing thermo-physical phenomenon. Moreover, [2] Y.J. Chang, C.C. Wang, A generalized heat transfer correlation for louver fin
the simulations were carried out to highlight the effect of fin height geometry, Int. J Heat Mass Transf. 40 (3) (1997) 533–544.
[3] Y.J. Chang, K.C. Hsu, Y.T. Lin, C.C. Wang, A generalized friction correlation for
on the performance of compact microchannel heat exchanger. The louver fin geometry, Int. J Heat Mass Transf. 43 (2000) 2237–2243.
following conclusions are deduced from the study; [4] M.-H. Kim, C.W. Bullard, Air-side thermal hydraulic performance of multi-
louvered fin aluminium heat exchangers, Int. J. Refrig. 25 (3) (2002) 390–400.
[5] N.H. Kim, J.P. Cho, Air-side performance of louver-finned flat aluminum heat

Reducing the fin height parameter can improve the thermal exchangers at a low velocity region, Heat Mass Transf. 44 (2008) 1127–1139.
hydraulic performance of microchannel heat exchangers [6] H.C. Kang, W.J. Gil, Heat transfer and flow resistance characteristics of louver
significantly by giving high values of volume reduction while fin geometry for automobile applications, J. Heat Transf. 133 (10) (2011),
pp.101802-1-101802-6.
maintaining same thermal performance.
[7] D.K. Tafti, G. Wang, W. Lin, Flow transition in a multi-louvered fin array, Int. J.

The results show that the redirection louver provides higher Heat Mass Transf. 43 (2000) 901–919.
values of local heat transfer coefficient by accelerating the flow. [8] K.N. Atkinson, R. Drakulic, M.R. Heikal, T.A. Cowell, Two and three-dimensional
numerical models of flow and heat transfer over louvered fin arrays in compact
It is suggested that inclusion of more redirection louvers in the
heat exchangers, Int. J. Heat Mass Transf. 41 (1998) 4063–4080.
louver array of the fin may result in the enhanced thermal per- [9] T. Perrotin, D. Clodic, Thermal hydraulic CFD study in louvered fin and flat tube
formance of heat exchangers. heat exchangers, Int. J. Refrig. 27 (2004) 422–432.

Higher local heat transfer coefficients are observed for small [10] K. Ryu, S. Yook, K. Lee, Optimal design of a corrugated louvered fin, Appl.
Therm. Eng. 68 (1–2) (2014) 76–79.
louver pitch fin geometries but the all-inclusive heat transfer [11] J. Li, S. Wang, W. Cai, W. Zhang, Numerical study on air-side performance of an
coefficient results may vary based on the louver pitch value. integrated fin and micro-channel heat exchanger, Appl. Therm. Eng. 30 (17)
However, the penalty in pressure drop gets remarkably high (2010) 2738–2745.
[12] J. Li, S. Wang, W. Zhang, Air-side thermal hydraulic performance of an
as the louver pitch is decreased. integrated fin and micro-channel heat exchanger, Energy Convers. Manage. 52

The geometric configuration with the louver pitch of 0.8 mm (2) (2011) 983–989.
showed the highest percentage increase in the average Nusselt [13] W. Li, X. Wang, Heat transfer and pressure drop correlations for compact heat
exchangers with multi-region louver fins, Int. J. Heat Mass Transf. 53 (15)
number of 97–116.1% compared with 1.7 mm louver pitch (2010) 2955–2962.
geometry, however, a pressure drop penalty of 75–406% was [14] J.-Y. Jang, C.-C. Chen, Optimization of louvered-fin heat exchanger with
observed. variable louver angles, Appl. Therm. Eng. 91 (2015) 138–150.
[15] Y.Y. Liang, C.C. Liu, C.Z. Li, J.P. Chen, Experimental and simulation study on the

It is suggested that fin geometry with louver pitch value of
air side thermal hydraulic performance of automotive heat exchangers, Appl.
1.0 mm provided optimum results in terms of heat transfer Therm. Eng. 87 (2015) 305–315.
and pressure drop characteristics. [16] P. Shinde, C.-X. Lin, A heat transfer and friction factor correlation for low air-
side Reynolds number applications of compact heat exchangers (1535-RP), Sci.
Technol. Built. Environ. (2016) 1–19.

Acknowledgement

This work was supported by the Korea Institute of Energy Tech-

nology Evaluation and Planning (KETEP) and the Ministry of Trade,