Anda di halaman 1dari 10

Journal of King Saud University Engineering Sciences (2014) 26, 122131

King Saud University

Journal of King Saud University Engineering Sciences


www.ksu.edu.sa
www.sciencedirect.com

ORIGINAL ARTICLE

Study of the axial heat conduction in parallel ow


microchannel heat exchanger
Mushtaq Ismael Hasan
a
b

a,*

, Hayder Mohammed Hasan b, Ghassan Adnan Abid

Mechanical Engineering Department, College of Engineering, Thi-Qar University, Iraq


Machines and Equipments Department, Shatra Technical Institute, Iraq

Received 21 August 2012; accepted 18 December 2012


Available online 28 December 2012

KEYWORDS
Microchannel heat exchanger;
Parallel ow;
Numerical solution;
Axial heat conduction;
Effectiveness

Abstract In this paper the axial heat conduction in an isosceles right triangular microchannel heat
exchanger is numerically investigated, for laminar, 3D, incompressible, single-phase, steady state
ow. The behaviour of axial heat conduction in the separating wall under different conditions is
studied. The solution was obtained by solving the continuity and NavierStokes equations for
the hot and cold uids by using the pressure-correction method to obtain the velocity distribution,
and then the energy equations were solved for the two uids and the separating wall simultaneously
to obtain the temperature distribution. The governing equations are discretized using nite-volume
and the hybrid differencing scheme with FORTRAN code was used. Various parameters that can
have effect on the axial heat conduction were investigated.
The results showed that, the axial heat conduction plays an important role in a parallel ow
microchannel heat exchanger and the factors affecting the local and average axial heat conduction
are; Reynolds number (Re), thermal conductivity ratio (Kr), hydraulic diameter (Dh), thickness of
separating wall (ts) and channel volume. Increasing of Re, Kr and ts leads to an increase in the axial
heat conduction while increasing of Dh and channel volume leads to a decrease in the axial heat
conduction.
2013 Production and hosting by Elsevier B.V. on behalf of King Saud University.

1. Introduction
The rapid development of modern electronics industry has
necessitated effective cooling techniques which are capable of
dissipating ultra-high heat ux of about 100 W/cm2 from the
highly integrated microelectronics systems to ensure a stable
* Corresponding author. Tel.: +964 7801140094.
E-mail address: mushtaq76h@yahoo.com (M.I. Hasan).
Peer review under responsibility of King Saud University.

Production and hosting by Elsevier

and optimal operation (Liu and Lee, 2003). A variety of novel


cooling schemes have been proposed and investigated to meet
this demand requirement, among which, the microchannel heat
exchanger is especially promising for its superior thermal performance. Compared with conventional heat exchangers, the
main advantage of the microchannel heat exchangers is their
extremely high heat transfer area per unit volume. As a result,
the overall heat transfer coefcient per unit volume can be as
greater than 100 MW/(m3 K), which is much higher than that
for conventional heat exchangers (12 orders of magnitude)
(Jian et al., 2001). In the conventional heat exchanger the solid
thickness is comparatively small to the hydraulic diameter;
therefore the axial heat conduction may be neglected. This

1018-3639 2013 Production and hosting by Elsevier B.V. on behalf of King Saud University.
http://dx.doi.org/10.1016/j.jksues.2012.12.004

Study of the axial heat conduction in parallel ow microchannel heat exchanger

123

Nomenclature
Symbol
vin
x
D
L
u
v
w
Cc
Ch
q
z
Dh
m
qmax
P
ts
Cp
T
k
y
V

Quantity, Unit (SI)


average velocity, m/s
axial coordinate, m
channel height and base, m
channel length, m
uid x-component velocity, m/s
uid y-component velocity, m/s
uid z-component velocity, m/s
heat capacity of cold uid, W/K
heat capacity of hot uid, W/K
heat transfer rate, W
horizontal coordinate, m
hydraulic diameter, m
mass ow rate, kg/s
maximum heat transfer rate, W
pressure, Pa
separating wall thickness, m
specic heat at constant pressure, J/(kg K)
temperature, K
thermal conductivity, W/m K
vertical coordinate, m
volumetric ow rate, m3/s

means that, the performance of heat exchanger is primarily


dependent on the ow in the ducts (uid properties and mass
ow rate). But, for a microchannel heat exchanger, the solid
thickness is comparatively large to the hydraulic diameter,
therefore the axial heat conduction in the separating wall (solid)
is considerable and the microchannel heat exchanger effectiveness may decrease due to the effect of axial heat conduction.
Kroeger (1967) solved the governing equations of a two
stream counter ow heat exchanger, taking into account the effect of axial heat conduction, he presented a closed form solution for nding the effectiveness of a balanced ow heat
exchanger using a simplied one dimensional plate separating
the two uids. He showed that, the effectiveness is largely
dependent on the axial heat conduction in the separating wall.
Wang and Shyu (1991) experimentally studied the effect of
channel size and wall thermal conductivity in micro heat
exchangers for a hot/cold water test loop. They showed that
the channel size and wall material have a strong inuence on
the heat transfer capability of a micro heat exchanger.
Yin and Bau (1992) studied the ow between innite parallel plates and circular pipes to study the effect of axial heat
conduction on the performance of microchannel heat exchangers. They used a fully developed velocity eld and analytically
solved for the temperature elds in the channel and solid wall.
They found that, the axial conduction plays an important role
at the entrance region.
Stief et al. (1999) numerically investigated the effect of solid
thermal conductivity in micro heat exchangers. They showed
that the reduction of conductivity of the wall material can improve the heat transfer efciency of the exchanger due to the
inuence of axial heat conduction in the separation wall.
Vekatarathanam and Narayanan (1999) also found that the
performance of the heat exchanger is largely dependent on the

Greek symbols
q
density, kg/m3
l
dynamic viscosity, N s/m2
e
heat exchanger effectiveness,
Subscripts
c
cold uid
h
hot uid
I
inlet
max
maximum value
o
outlet
s
solid
Dimensionless groups
Pe = RePr Peclet number
P
Pr lC
kf Prandtl number
Re = quh,in Dh/l Reynolds number
Kr = ks/kf thermal conductivity ratio

heat conduction that takes place through the walls of the heat
exchangers used in miniature refrigerators. They used a two
dimensional energy balance to account for the conduction
through wall and convection through the uid. They assumed
the temperature of the uid to be uniform at any ow cross
section.
Al-bakhit and Fakheri (2005) numerically investigated
the parallel ow microchannel heat exchanger with rectangular ducts. They showed that using a high conductive material will not have an effect on increasing the heat
exchanger effectiveness since the heat exchanger effectiveness
will be independent of the wall thermal conductivity for Kr
above 40.
Al-bakhit and Fakheri (2006) numerically investigated the
parallel ow microchannel heat exchanger with rectangular
ducts. They showed that the overall heat transfer coefcient
is rapidly changed for x/DhPe (Graetz number) below 0.03,
and therefore the assumption of constant overall heat transfer
coefcient is not valid if the Graetz number based on the heat
exchanger length is of the order of 0.03. Also, the accurate results can be obtained by solving the thermally developing energy equation using fully developed velocity proles.
Al-Nimr et al. (2009) numerically investigated the hydrodynamics and thermal behaviours of the laminar, 2D, fully developed, slip ow inside an insulated parallel-plate microchannel
heat exchanger. They showed that both the velocity slip and
the temperature jump at the walls increased with increasing
Knudsen number.
Hasan (2009) made numerical investigation to study the
counter ow microchannel heat exchanger with different channel geometries and working uids. He studied the effect of axial heat conduction on the performance of counter ow
microchannel heat exchanger with square shaped channel

124

M.I. Hasan et al.

and he found that the existence of axial heat conduction leads


to a reduction in the effectiveness of this heat exchanger.

Based on these assumptions, the governing equations and


boundary conditions in Cartesian coordinates are written as
below (Hasan, 2009):

2. Mathematical formulation
Simulation of full continuity, momentum and energy equations for different uid ows arises in various engineering
problems. Various different algorithms have been proposed
and developed by various researchers. But an approach that
is fully robust from the point of view of numerical and modelling accuracy as well as efciency has yet to be developed.
The conguration shown in Fig. 1 represents a unit cell of
microchannel arrays. It includes two isosceles right triangular
channels with a separating wall between them where, L is
channel length, D is height and base of channel, ts is solid
thickness and S is projected distance of ts. In this gure, the
hot uid enters the lower channel with a uniform velocity uh,in
and uniform temperature Th,in, while the cold uid enters the
upper channel at uc,in and Tc,in. Heat is transferred from the
hot uid to the cold uid through the inclined wall of thickness
ts. However, the governing equations for the present model are
based on the following physical and geometrical assumptions:
 The ow is laminar and steady.
 The Knudsen number is small enough so that, the uid is a
continuous medium (no slip).
 The uids are incompressible, Newtonian with constant
properties; in this case the water is used as working uid.
 There is no heat transfer to/from the ambient.
 The energy dissipation is negligible.
 The pressure gradient is in axial direction only.
 Three dimensional is the ow and heat transfer.

outlet

x-Momentum equation:


 2

@ui
@ui
@ui
@p
@ ui @ 2 ui @ 2 ui
qi ui
vi
wi
 i li
1

@x
@y
@z
@x
@x2 @y2 @z2
y-Momentum equation:


 2

@vi
@vi
@vi
@ vi @ 2 vi @ 2 vi
vi
wi
li

qi ui
@x
@y
@z
@x2 @y2 @z2
z-Momentum equation:


 2

@wi
@wi
@wi
@ wi @ 2 wi @ 2 wi
vi
wi
li
2 2
qi ui
@x
@y
@z
@x2
@y
@z

c
h

c
h
c

c
h
c

h
c
h

h
c
h

h
c

h
c

h
c
h

h
c
h

@ui @vi @wi

0
@x @y
@z

Energy equation for uid:


 2



@Ti
@Ti
@Ti
@ Ti @ 2 Ti @ 2 Ti
5

vi
wi
ki
qi CPi ui
@x
@y
@z
@x2
@y2
@z2
Conduction equation for solid:
@ 2 Ts @ 2 Ts @ 2 Ts

2 0
@x2
@y2
@z

 Lower channel (hot uid)

uh uh;in ; vh wh 0; Th Th;in
At x L

h
h c
h c c h
h c
c
h
c
h
c
c
h
h
c
c
h

@uh @vh @wh


@Th

0;
0
@x
@x
@x
@x

inlet

At y 0; 0 6 z 6 D

separating wall (solid)

uh vh wh 0;

chamfered
S

@Th
0
@y

At y D  z tan45; 0 6 z 6 D

S
cold fluid
D

uh vh wh 0

ts
hot fluid

S
D

y x

L
chamfered

kh

@Th
@Ts
ks
; Th Ts
@n
@n

At z 0; 0 6 y 6 D

Figure 1 Schematic of parallel ow microchannel heat


exchanger.

where i = h and c refers to the hot and cold uids, respectively.


As mentioned in the previous assumptions the pressure gradii
i
@p
0.
ents in y and z directions are negligible therefore, @p
@y
@z
The boundary conditions are:

h
c

Continuity equation:

At x 0
h

uh vh wh 0;

@Th
0
@z

Study of the axial heat conduction in parallel ow microchannel heat exchanger


At

z D  y tan 45; 0 6 y 6 D

uh vh wh 0
@Th
@Ts
ks
; Th Ts
kh
@n
@n
 Upper channel (cold uid)

At

x0

uc uc;in vc wc 0; Tc Tc;in
At

xL

@uc @vc @wc


@Tc

0;
0
@x
@x
@x
@x
At

y D  z tan45 2S; S 6 z 6 D S

uc vc wc 0

kh

@Th
@Ts
ks
; Th Ts
@n
@n

At y D  z tan45 2S; S 6 z 6 D S
kc

@Tc
@Ts
ks
; Tc Ts
@n
@n

At z D  y tan45; 0 6 y 6 D
kh

@Th
@Ts
ks
; Th Ts
@n
@n

At z y tan45; S 6 y 6 D S
kc

@Tc
@Ts
ks
; Tc Ts
@n
@n

At z 0; D 6 y 6 D S
@Ts
0
@z
At z D S; 0 6 y 6 S

kc

@Tc
@Ts
ks
; Tc Ts
@n
@n

@Ts
0
@z

At

yDS

At y D S; 0 6 z 6 S

uc vc wc 0

@Ts
0
@y

@Tc
0
@y

At y 0; D 6 z 6 D S

At

z y tan45; S 6 y 6 D S

uc vc wc 0
kc

@Tc
@Ts
ks
; Tc Ts
@n
@n

At

zDS

uc vc wc 0
@Tc
0
@z
 Separating wall (solid)

At

x0

@Ts
0
@x
At

xL

@Ts
0
@x
At

y D  z tan45; 0 6 z 6 D

125

@Ts
0
@y
where, n is the normal to the iso surfaces (isotherm lines). After
establishing the governing equations and boundary conditions,
the nite-volume method with FORTRAN code will be used
to obtain a numerical solution to the problem.
The nite-volume method has become popular in Computational Fluid Dynamics CFD, this is because, rst, it ensures
that the discretization is conservative, i.e., mass, momentum,
and energy are conserved in a discrete sense. Second, the
nite-volume method does not require a coordinate transformation in order to be applied for irregular meshes. This exibility can be used to great advantages in generation grids about
arbitrary geometries (Lomax, 1999). Moreover, the nite volume method combines the simplicity of the nite differences
method with the local accuracy of the nite element method;
also, at the same dimension of the discretized problem, the
accuracy is higher than with nite differences and nearly the
same as with nite elements (Petrila and Trif, 2005; Versteeg
and Malalasekera, 1995).
After discretization of the governing equations and boundary conditions, a FORTRAN program is developed to perform the numerical solution to the problem. Firstly the
distribution of velocity, pressure and temperature for both
hot and cold liquids is obtained in addition to the distribution
of temperature in solid wall. Then, the parameters such as the
axial heat conduction q00 in the separating wall, the amount of
heat transferred between two uids and the effectiveness e

126

M.I. Hasan et al.

which is the ratio of actual heat transfer to the maximum


possible heat that can be transferred can be calculated from
Ashman and Kandlikar (2006):
e qactual =qmax: possible

qactual m_ c cc Tc;out  Tc;in m_ h ch Th;in  Th;out

1
0.8

Tm

0.6

For convenience, the ow rates and specic heats are lumped


together, and the term product of the capacity rates is

0.4

Cc m_ c cc

0.2

_ h
and Ch mc

For Ch < Cc,

qmax Ch Th;in  Tc;in

10

Otherwise
qmax Cc Th;in  Tc;in

11

Thus
qmax Cmin Th;in  Tc;in

12

Then the effectiveness is


e

Cc Tc;out  Tc;in
Ch Th;in  Th;out

Cmin Th;in  Tc;in Cmin Th;in  Tc;in

For the axial heat conduction in section x




Ts; i  Ts; i1
Q00x ks
Dx
x

13

14

where, Ts; i is the average solid temperature in certain sections


and Ts; i1 at the next section as shown in Fig. 2.
3. Results and discussion
The behaviour of axial heat conduction in the separating wall in
a parallel ow microchannel heat exchanger under different
conditions has been studied. The dimensions of triangular
channels are 2 cm, 200 lm and 200 lm in x, y and z directions.
Both hot and cold uids are water with properties taken based
on the average bulk temperature (viscosity = 0.000598 N s/m2,
density = 981.3 kg/m3, thermal conductivity = 0.643 W/m K,
specic heat = 4189 J/(kg K). The inlet temperatures of hot
and cold uids are 80 C and 20 C, respectively. The material
of solid wall is variable; therefore the conductivity of material
will be studied through the ratio kr = ks/kf. The grid was 100
nodes in x direction and 50 nodes in each of y and z directions.
In order to verify the accuracy of the present numerical
model, a comparison is made between the results of present
model for rectangular microchannel heat exchanger and that
in literature. Fig. 3 represents a comparison between present
numerical model and data of Al-bakhit and Fakheri (2006).
The gure illustrates the variation of dimensionless mean tem-

0.01

0.02

0.03

0.04

0.05

x+

0.06

0.07

0.08

0.09

Figure 3 Comparison of the dimensionless mean temperature of


the hot and cold uids for present model and data of Al-bakhit
and Fakheri (2006).

lar microchannel heat exchanger with the dimensionless axial


x
distance x Dh Re
at Re = 100, and kkwsDtsh 100. From this
Pr
gure it can be seen that, there is a good agreement between
the numerical results of present model and that for Al-bakhit
and Fakheri (2006) and the maximum percentage error was
2.01%.
Fig. 4 represents the distribution of central axial velocity
along the direction of ow for different values of Reynolds
number at Dh = 117 lm. From this gure it can be noted that,
the velocity increased in the entrance region to reach its
maximum value in the fully developed region. Also it can be
seen that, the velocity increased with increasing Reynolds
number.
Fig. 5 shows the distribution of dimensionless axial velocity
in y and z directions, respectively for Re = 100 and
Dh = 117 lm at different locations along the x-axis (direction
of ow). From this gure one can see the parabolic shape of
the velocity, also the velocity increased along the ow direction
until it reaches its maximum value in the fully developed ow
region.
Fig. 6 represents the distribution of temperature for hot uids, cold uids and solid wall along the heat exchanger at
Re = 100, Dh = 117 lm, ts = 50 lm and kr = 1. This gure
indicates that, the temperature of cold uid increased and

T T

perature Tm Th;ini Tc;inc;in for the hot and cold uids in a rectangu-

flow

i+1
"
i qx
__

T s ,i +1
__

T s ,i

Figure 2

Schematic of separating wall.

Figure 4 Distribution of axial central velocity along ow


direction for different values of Re.

Study of the axial heat conduction in parallel ow microchannel heat exchanger

Distribution of axial velocity along y and z directions in different locations along x axis.

Figure 6 Distribution of temperature for hot uids, cold uids


and solid wall along heat exchanger.

the temperature of hot uid decreased along the ow direction


due to heat transfer from hot to the cold uid and the temperature of separating solid wall is in between them.
To show the importance of axial heat conduction in a
microchannel heat exchanger. Fig. 7 indicates the distribution
of axial heat conduction ux and the ux of heat transferred
between two uids at Re = 100 and ts = 50 lm Kr = 1,
Dh = 117 lm along the heat exchanger (direction of ow).
This gure shows that, the amount of heat conducted in the
axial direction of the solid wall separating the uids cannot
be ignored due to its considerable value.
Fig. 8 shows the distribution of longitudinal axial heat conduction ux q00x in the separating wall along the heat exchanger for different Reynolds numbers at Dh = 117 lm, Kr = 1
and ts = 50 lm. The gure shows that, the value of q00x for all
Reynolds number is high in the entrance region, due to the effect at entrance region the maximum heat transfer occurs in
the entrance region and as a result a maximum value of axial
heat conduction is created in this region. Also it can be seen
from this gure that, the axial heat conduction ux increased
with increasing Re. Due to increasing amount of heat transferred and increasing effect of entrance region with Re, the
fraction of axial heat conduction is increased.

Figure 7 Distribution of the axial heat conduction and heat


conducted between two uids along ow direction.

2000
1800

Re=100

1600

q "x ( kW m 2 )

Figure 5

127

Re=200

1400

Re=400

1200
1000
800
600
400
200
0

0.004

0.008

0.012

0.016

0.02

x (m)
Figure 8 Longitudinal variation of the axial heat conduction
ux with axial distance x for different Reynolds numbers.

Fig. 9 shows the longitudinal distribution of the axial heat


conduction ux in the separating wall q00x along the heat exchanger for different hydraulic diameters at Re = 100,
Kr = 1 and ts = 50 lm. The gure shows that, the q00x for all

128

M.I. Hasan et al.


1200

1200
1000
Dh=58

800

q "x ( kW m 2 )

q "x ( kW m 2 )

1000
Dh=117
Dh=234

600
400

ts=0.00005 m
ts=0.0001 m

800

ts= 0.0002 m

600
400
200

200
0
0

0.004

0.008

0.012

0.016

0.004

x (m)
Figure 9 Longitudinal distribution of the axial heat conduction
ux with axial distance x for different hydraulic diameters.

0.008

0.012

0.016

0.02

x (m)

0.02

Figure 11 Distribution of the axial heat conduction ux with


axial distance x for different wall thicknesses.

2500

0.025

ts=50

0.02

q "x (W )

ts=100
ts=200

0.015
0.01

0.005

0.004

0.008

0.012

0.016

0.02

x (m)
Figure 10 Longitudinal distribution of the axial heat conduction
with axial distance x for different wall thicknesses.

Kr=1

2000

q "x ( kW m 2 )

hydraulic diameters is high in the entrance region as explained


before. Also, the q00x decreased with increasing hydraulic diameter due to decreasing effect of wall thickness compared to the
channel dimensions.
Fig. 10 illustrates the longitudinal distribution of the axial
heat conduction in the separating wall along the heat exchanger for different values of wall thicknesses ts at Re = 100,
Kr = 1 and Dh = 117 lm. From this gure one can see that,
the value of axial heat conduction increased with increasing
wall thickness due to an increase in the cross-sectional area
in which the axial heat conduction transferred also due to
increasing conduction resistance for heat transferred to the
uid and as a result increasing the fraction of heat transferred
in the longitudinal direction (axial heat conduction). While
Fig. 11 shows the decreasing of the axial heat conduction ux
with increasing wall thickness due to an increase in the cross
sectional area of wall.
Fig. 12 shows the longitudinal distribution of the axial heat
conduction ux q00x with axial distance x for different thermal
conductivity ratios Kr at Re = 100, ts = 50 lm and Dh =
117 lm. For all values of thermal conductivity ratio, the results
indicate that the q00x is high in the entrance region and decreased

Kr=5
Kr=10

1500

Kr=40
Kr=100

1000

500

0
0

0.004

0.008

0.012

0.016

0.02

x (m)
Figure 12 Longitudinal variation of the axial heat conduction
ux with axial distance x for different values of Kr.

towards the fully developed region. Also, it can be seen that,


the axial heat conduction ux increased with increasing Kr
up to Kr = 10, after this value the axial conduction becomes
zero. This is due to that, the heat transferred from hot uid
to cold uid increased with increasing Kr up to Kr = 10, after
Kr = 10, which is considered as an optimum value for Kr in
this case there is no increase in heat transfer. The numerical results of Al-bakhit and Fakheri (2005) for the parallel ow rectangular microchannel heat exchanger showed that more heat is
transferred as the thermal conductivity ratio increased up to
Kr = 40, after which increasing the thermal conductivity ratio
will not enhance the heat transfer. They showed that this is because, the wall will behave as an innitely conducting wall with
negligible temperature gradient, and the heat transfer between
the two uids will be independent of the wall properties,
depending only on the uid conditions and channel geometry.
The difference between the numerical results of Al-bakhit and
Fakheri (2005) and present results is due to the difference in
geometry. However, Fig. 13 may explain this behaviour
(Hasan, 2009). As can be seen the dimensionless mean of wall
temperature Tsm Tsm  Tc;in =Th;in  Tc;in with x* = x/Dh
varied with uid temperature up to Kr = 10. After this value
the Tsm is independent of uid temperature and seen as a
straight line. This means that, the gradient oTs/ox = 0 and
then q00x 0 after Kr = 10.

Study of the axial heat conduction in parallel ow microchannel heat exchanger


1

0.0025

0.8

0.002

average

0.2

Kr = 1

K r = 500

Kr = 5
K r = 10

K r = 1000
0

34

68

102

136

0.001

K r = 100

0.0005
170

x
Figure 13 Longitudinal variation of the wall dimensionless mean
temperature for different thermal conductivity ratios taken from
Hasan (2009).

20

60

100

140

180

220

260

300

Number of channels
Figure 16 Variation of the heat exchanger effectiveness to
average axial heat conduction ratio with number of channels for
different Reynolds numbers.

450

450

400

400

350

350

300

"
q average
(kW m 2 )

"
q average
(kW m 2 )

Re=400

0.0015

Tsm
0.4

250
200
150

Re=100

100

Re=200

50
0

Re=100
Re=200

0.6

129

Re=400

300
250
200

Kr=1

150

Kr=5
Kr=10

100

Kr=40

50
20

60

100

140

180

220

260

300

Kr=100

Number of channels

20

Figure 14 Variation of the average axial heat conduction ux


with number of channels for different Reynolds numbers.

60

100

140

180

220

260

300

Number of channels
Figure 17 Variation of the average axial heat conduction ux
with number of channels for different thermal conductivity ratios.

0.5
0.5

0.45
0.4
0.35

0.4

0.3
0.25

0.3

0.2
0.15

Re=100

0.1

Re=200

0.05

Re=400

20

60

100

140

180

220

260

0.2

Kr=1
Kr=5

0.1

Kr=10

300

Number of channels
Figure 15 Variation of the microchannel heat exchanger effectiveness with number of channels for different Reynolds numbers.

Fig. 14 shows the variation of the average axial heat conduction ux q00average with the number of channels for different
Reynolds numbers for heat exchanger volume 2 108 m3 at
ts = 50 lm and Kr = 1. From this gure it can be noted that,

20

60

100

140

180

220

260

300

Number of channels
Figure 18 Variation of the heat exchanger effectiveness with
number of channels for different thermal conductivity ratios.

for all values of Reynolds number, the q00average increased with


increasing number of channels, due to an increase in the heat

130

M.I. Hasan et al.


0.0014
0.0012

0.0008
0.0006
Kr=1

"
average

0.001

0.0004

Kr=5
Kr=10

0.0002
0

20

60

100

140

180

220

260

300

Number of channels
Figure 19 Variation of the heat exchanger effectiveness to average axial heat conduction ratio with number of channels for different
thermal conductivity ratios.

transfer with increasing number of channels. Also, the gure


shows that the q00average increased with increasing Re, this is
due to an increase in the axial heat conduction with Re.
Fig. 15 illustrates the variation of the microchannel heat exchanger effectiveness e with the number of channels for heat
exchanger volume 2 108 m3 at Kr = 1 and ts = 50 lm.
For all values of Re, the gure claries that, the effectiveness
increased with increasing number of channels. This is due to
the amount of uid that decreased when the channel volume
decreased and then DT increased. On the other hand, the axial
conduction in the solid wall also increased as shown in previous gures but the heat transferred between the two uids is
large enough to overcome the axial heat conduction. Also,
the gure shows that, the effectiveness decreased with increasing Re, this is due to an increase in the velocity of uid with Re
and then DT decrease.
The variation of heat exchanger effectiveness to the average
axial heat conduction ratio e=q00average with the number of channels for different values of Re is indicated in Fig. 16. From this
gure it can be seen that, for all Reynolds number, the e=q00average
increased with an increase in the number of channels due to
increasing effectiveness with the number of channels in spite
of increasing the axial heat conduction. Also this ratio decreased with increasing Re due to decreasing effectiveness with
increasing Re.
Fig. 17 shows the variation of the average axial heat conduction ux q00average with the number of channels for different
thermal conductivity ratios for heat exchanger volume 2 108
m3 at Re = 200 and ts = 50 lm. For all thermal conductivity
ratios, the gure shows that, the q00average increased with
increasing number of channels due to the large amount of heat
transfer that undergoes through the wall in this case up to
Kr = 10, after this value the q00average is equal to zero as
explained before. Also, the gure shows that the q00average at
Kr = 1 is lower than for ve and 10, due to that, there is more
heat transfer in the axial direction at higher Kr.
Fig. 18 illustrates the variation of the heat exchanger effectiveness (e) with the number of channels for different values of
Kr for heat exchanger volume = 2 108 m3at Re = 200 and
ts = 50 lm. From this gure one can see that, for all values
of Kr, the effectiveness increased with increasing Kr due to
more heat that is transferred from hot to cold uid with

increasing Kr. Also the effectiveness increased with increasing


number of channels.
Fig. 19 shows the variation of the heat exchanger effectiveness to average axial heat conduction ratio e=q00average with the
number of channels at different values of Kr for heat exchanger
volume = 2 108 m3 at Re = 200 and ts = 50 lm. From this
gure it can be observed that, for all Kr values, the e=q00average increased with increasing number of channels this is because the
effectiveness increased with the number of channels in spite of
the increase in axial heat conduction. Also, the e=q00average at
Kr = 10 is larger than for one and ve because the effectiveness increases with Kr .
4. Conclusions
From the results obtained the following conclusions can be
drawn:
 The parameters that affect the axial heat conduction in an
isosceles right triangular microchannel heat exchanger are:
thermal conductivity ratio Kr, Reynolds number Re,
hydraulic diameter Dh, channel volume and wall thickness
ts .
 The axial heat conduction increased with increasing Kr up
to Kr = 10 after this value the axial conduction is equal
to zero.
 The axial heat conduction increased with increasing Re, ts
and the number of channels but, decreased with increasing
Dh each separately.
References
Hussien Al-bakhit, Ahmad Fakheri, 2005. Entrance and wall conduction effects in parallel ow heat exchangers. In: Shah, R.K.,
Ishizuka, M., Rudy, T.M., Wadekar, V V. (Eds.), Proceedings of
Fifth Internal Conference on Enhanced, Compact and UltraCompact Heat Exchangers: Science, Engineering and Technology.
Engineering Conferences International, September (2005), Hoboken, NJ, USA.
Al-bakhit, Hussien, Fakheri, Ahmad, 2006. Numerical simulation of
heat transfer in simultaneously developing ows in parallel
rectangular ducts. Applied Thermal Engineering 26, 596603.

Study of the axial heat conduction in parallel ow microchannel heat exchanger


Al-Nimr, M.A., Maqableh, M., Khadrawi, A.F., Ammourah, S.A.,
2009. Fully developed thermal behaviors for parallel ow microchannel heat exchanger. International Communications in Heat
and Mass Transfer 36, 385390.
Ashman, Sean, Kandlikar, Satish G., 2006. A review of manufacturing
processes for microchannel heat exchanger fabrication. In: Proceedings of ICNMM2006, Fourth International Conference on
Nanochannels, Microchannels and Minichannels, June 1921.
Limerick, Ireland.
Hasan, Mushtaq I., 2009. Numerical simulation of counter ow
microchannel heat exchanger with different channel geometries and
working uids. PhD Thesis, Mechanical Engineering Department,
Collage of Engineering, Basrah University, Iraq.
Hasan, Hyder M., 2009. Numerical simulation of parallel ow
microchannel heat exchanger with isosceles right triangular geometry. Thesis, Engineering Collage, University of Basra.
Jian, Pei-Xue, Fan, Ming-Hong, Si, Guang-Shu, Ren, Ze-Pei, 2001.
Thermal-hydraulic performance of small scale microchannel and
porous-media heat exchangers. International Journal of Heat and
Mass Transfer 44, 10391051.
Kroeger, P.G.,1967. Performance deterioration in high effectiveness
heat exchangers due to axial heat conduction effects. Advances in
Cryogenic Engineering, vol. 12. Plenum press, New York, pp. 363
373.

131

Dong Liu, Poh-Seng Lee, 2003. Numerical investigation of uid ow


and heat transfer in microchannel heat sink. Project Report, ME
605 Convection of Heat and Mass, December 09, West Lafayette,
IN, USA.
Harvard Lomax, Pullium, Thomas H., Zingg, David W., 1999.
Fundamentals of computational uid dynamics. Handbook,
August 26.
Titus Petrila, Damian Trif, 2005. Basics of Fluid Mechanics and
Introduction to Computational Fluid Dynamics. Springer.
Stief, T., Langer, O.U., Schubert, K., 1999. Numerical investigation of
optimal heat conductivity in micro heat exchangers. Chemical
Engineering Technology 22, 297303.
Vekatarathanam, G., Narayanan, S., 1999. Performance of counter
ow heat exchanger with heat loss through the wall at the cold end.
Cryogenics 39, 4352.
Versteeg, H.K., Malalasekera, W., 1995. An introduction to Computational Fluid DynamicsThe Finite-Volume Method. Longman.
Wang, J.H., Shyu, R.J.,1991. Thermal-hydraulic characteristics of
micro heat exchangers. ASME Winter Annual Meeting, December
16, Atlanta, GA, USA.
Yin, X., Bau, H.H., 1992. Axial Conduction Effect Performance of
Micro Heat Exchangers, ASME Winter Annual Meeting, November 28December 3, New Orleans, Louisiana, USA.