You are on page 1of 7

Energy Conversion and Management 92 (2015) 244–250

Contents lists available at ScienceDirect

Energy Conversion and Management


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

Enhancing heat transfer in microchannel heat sinks using converging


flow passages
Maziar Dehghan, Mahdi Daneshipour, Mohammad Sadegh Valipour ⇑, Roohollah Rafee,
Seyfolah Saedodin
Faculty of Mechanical Engineering, Semnan University, 35131-19111 Semnan, Iran

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

Article history: Constrained fluid flow and conjugate heat transfer in microchannel heat sinks (MCHS) with converging
Received 14 August 2014 channels are investigated using the finite volume method (FVM) in the laminar regime. The maximum
Accepted 19 December 2014 pressure of the MCHS loop is assumed to be limited due to constructional or operational conditions.
Available online 8 January 2015
Results show that the Poiseuille number increases with increased tapering, while the required pumping
power decreases. Meanwhile, the Nusselt number increases with tapering as well as the convection heat
Keywords: transfer coefficient. The MCHS having the optimum heat transfer performance is found to have a width-
Microchannel
tapered ratio equal to 0.5. For this tapering configuration and at the maximum pressure constraint of
Converging channel
Heat transfer enhancement
3000 Pa, the pumping power reduces by a factor of 4 while the overall heat removal rate is kept fixed
Poiseuille number in comparison with a straight channel.
Pumping power Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction transfer in the solid part leads to an asymmetrical heat transfer. So


that, analytic solutions based on simplified thermal boundary con-
Trend toward reducing the size of parts in devices along with ditions may not be much accurate. Moreover, Lee et al. [9] declared
increasing demand of heat exchange rate in a more concentrated that the real boundary conditions (and the real geometry of
region, was the start point of microchannel heat sink (MCHS) intro- channels) are not exactly same as conditions which are supposed
duction. With heat fluxes greater than 106 W m2, cooling methods in classical solutions. Consequently, discrepancies exist between
with effective temperature control is required in MEMS technology experimental data of MCHS and numerical simulation results based
[1–3]. As a solution for removing high heat fluxes from small areas, on conventional macro-channel assumptions. Morini [10] stated
Tuckerman and Pease [4,5] introduced the use of microchannel that the following phenomena would be important because of the
heat sinks for the first time in early 1980s because of the high size reduction: the axial heat transfer, conjugated heat transfer, sur-
area-to-volume ratios. face roughness, and viscous dissipation. Mirzaei and Dehghan [11]
The Navier–Stokes equations in microchannels have been veri- studied the conjugate 3-D heat transfer of a nanofluid in a rectangu-
fied in early 1990s [6,7]. Gamrat et al. [8] investigated the entrance lar MCHS numerically based on the variable property approach.
effects on the fluid flow and heat transfer in microchannels experi- They showed that because of a high difference between the inlet
mentally and numerically. Their numerical simulations did not and outlet temperatures of the working fluid, the change in thermo-
reveal any significant scale-effect contrary to their experimental physical properties is not negligible. Also, they revealed that the
results. It has been identified that the differences between experi- addition of a small amount of nano-particles with a high thermal
mental data and result of classical relations are due to measurement conductivity results in an overall enhancement in the heat transfer
errors and low accuracy in microchannels construction. As the ability. The analysis of flow and heat transfer in microchannel heat
geometry of microchannels is noncircular (often rectangular), heat sinks is an increasing interest of the modern engineering [12–19].
flux enters from one side, and consequently the 3-dimentional heat Beside investigations on flow and heat transfer in microchan-
nels, the optimization of heat and fluid flow has been an attracting
research area recently. Researchers have focused on optimizing
⇑ Corresponding authors. both the shape and dimensions to improve the heat transfer ability
E-mail addresses: m-dehghan@semnan.ac.ir, dehghan.maziar@gmail.com of microchannels. Only a brief look at the recently published
(M. Dehghan), mahdi.daneshipour@semnan.ac.ir (M. Daneshipour), msvalipour@ articles in this field shows the importance of the subject [13]. Heat
semnan.ac.ir (M.S. Valipour), rafee@semnan.ac.ir (R. Rafee), s_sadodin@semnan.ac.
ir (S. Saedodin).
transfer and flow characteristics in wavy microchannel heat sinks

http://dx.doi.org/10.1016/j.enconman.2014.12.063
0196-8904/Ó 2014 Elsevier Ltd. All rights reserved.
M. Dehghan et al. / Energy Conversion and Management 92 (2015) 244–250 245

Nomenclature

A area (m2) Subscript


Cp heat capacity (J/kg K) app apparent
Dh hydraulic diameter (2WH/(W + H)) (m) ave average
 
f pin pz
friction factor 14 Dzh 1=2 b bulk
qu2av e c cross section
H height (m)
h convection heat transfer coefficient (W/m2 K) in inlet
k conductivity (W/m K) f fluid
Kn Knudsen number (l/Dh) h hydraulic
Nu Nusselt number (hDh/k) s solid
L length (m) W wall
l molecular mean free path (m) z axial location
P pressure (Pa)
q heat flux (W/m2) Superscript
T temperature (K) +, ⁄ dimensionless values
t thickness (m)
u axial velocity component (m/s) Greek letters
~
V velocity vector (m/s) a aspect ratio
W width (m) C perimeter (m)
z axial location (m) l viscosity (kg/m s)
q density (kg/m3)

with rectangular cross-section and various wavy amplitudes ran- micro heat exchangers [27,28]. As it was mentioned that the height
ged from 125 to 500 lm were studied numerically by Mohammed tapering has the least effects on the thermal performance of MCHS,
et al. [20]. In another work, Mohammed et al. [21] investigated only the width-tapered configuration is chosen in the present anal-
effects of different types of flow passage on heat and fluid flow ysis [25,26,29]. To compare the hydraulic and thermal performance
in MCHS. They applied zigzag, curvy, and step channel shapes of different configurations, it is assumed that the maximum pres-
and showed that the zigzag shape has the highest heat transfer sure should be limited to a pre-defined value due to constructional
and pressure drop among other shapes. Türkakar and Okutucu- or operational limitations. This constrained fluid flow in the
Özyurt [3] performed dimensional optimization of silicon micro- presence of heat transfer has not been previously investigated. Both
channel heat sinks by minimizing the total thermal resistance hydrodynamical and thermal characteristics including the pressure
considering multiple heat sources. The effects of three different graph, Poiseuille number, convection heat transfer coefficient, and
channel shapes (hexagonal, circular, and rhombus) on the MCHS Nusselt number are analyzed and discussed in this study.
performance were numerically analyzed by Alfaryjat et al. [22].
They showed that the hexagonal cross-section MCHS has the high- 2. Problem definition and mathematical modeling
est pressure drop and heat transfer coefficient among other shapes
while rhombus cross-section MCHS has the highest value of the top The schematic diagram of the problem is shown in Fig. 1 and
wall temperature and thermal resistance. Dehghan et al. [23], dimensions are presented in Table 1. The following assumptions
based on an extension to the study of Dehghan et al. [24], numer- are invoked in the present study:
ically investigated a combined convection–radiation heat transfer
inside a micro-channel filled with a porous medium in the  The flow is incompressible.
slip-flow regime. They presented the dimensionless temperature  According to the symmetry, only one half of the channel shown
and the Nusselt number. They showed that the presence of a finite in the right-side of Fig. 1 is considered.
temperature jump at the heated walls dramatically decreases the  The microchannel is made of aluminum.
Nusselt number.
Another branch in enhancing the overall performance of the
microchannels is varying the cross-section of the channel in the
flow direction. Hung et al. [25] compared conventional single-
layered and double-layered channels with tapered channels. They
found that the tapered channel has the best performance consider-
ing thermal resistance and temperature distribution uniformity
among other shapes. The optimal tapered channel design of a
MCHS with varying heights and widths was investigated by Hung
and Yan [26] using the objective function of the overall thermal
resistance. They found that the thermal resistance is sensitive to
variations in the channel number, channel-width ratio, or width-
tapered ratio while less sensitive to the height-tapered ratio.
In the present study, the problem of developing conjugate heat
and fluid flow within the range of 75–350 Reynolds numbers in
width-tapered microchannel heat sinks with the inlet aspect ratio
of 5 and the inlet hydraulic diameter of 333 lm is numerically
investigated using FVM. The 3-D heat and fluid flow in the converg- Fig. 1. Schematic diagram of the problem; (a) isometric view of MCHS [11] and (b)
ing channels are practical in chemical reactors, bio-technology, and top view of a channel.
246 M. Dehghan et al. / Energy Conversion and Management 92 (2015) 244–250

Table 1 100
Dimensions of microchannel (lm). Wo=200, Pi=1500

Wi Wo Wc H L t Wo=200, Pi=2000
200 200, 150, 100, 75 400 1000 12000 500
Wo=200, Pi=3000

Mirzaei & Dehghan (2013)


 A maximum pressure value is adopted as the constraint of the

Po
problem. The maximum pressure value of the MCHS loop occurs Philips (1990)
at the entrance of the channel and is assumed to be equal to
1500, 2000, and 3000 Pa.
 The inlet width of the flow passage is fixed for all simulations
(Wi = 0.2 mm). While, the outlet width (Wo) is assumed to be
0.2, 0.15, 0.1, and 0.075 mm.

Two identifying parameters (the outlet width, Wo, and the inlet 10
pressure, Pi) are used in order to distinguish the results of each 0.001 0.01 0.1 1
case-study in Figs. 2–8. z+

Fig. 3. The Po number distribution with respect to z+.


2.1. Governing equations
2.2. Boundary conditions
Continuity, momentum and energy equations for the fluid part
A constant heat flux equal to 100 W cm2 is imposed to the bot-
are as following:
tom surface of the MCHS. Other sides are assumed to be insulated.
As it was discussed in the previous section, left and right walls are
r  ðq~
VÞ ¼ 0
symmetrical. Because of the symmetry, the gradient of the temper-
~
V:rðq~
VÞ ¼ rp þ r  ðlr~ VÞ ð1Þ ature and velocity normal to the symmetric planes are zero. At the
~ rðqcp T f Þ ¼ ð~
V: V:rÞp þ r  ðkf rT f Þ inlet, the temperature of the fluid is 293 K, and it enters with a
uniform velocity. At the exit, the gauge pressure is assumed to be
Energy balance in the solid part is: constant (equal to 50 Pa). Since the channel is long enough, fully
developed condition for velocity at the exit is employed. Since,
r  ðks rT s Þ ¼ 0 ð2Þ the molecular mean free path of water (l) is about 0.25 nm, the
Subscripts f and s represent the ‘‘fluid’’ and ‘‘solid’’ parts, respec- Knudsen number (Kn), defined as the ratio of the molecular mean
tively. ~
V is the velocity vector, p is the pressure, l is the fluid free path to the hydraulic diameter, falls in the no-slip regime
viscosity, q is the fluid density, cp is the specific heat of the fluid, (Kn < 0.001) [11]. So that, conditions of the no-slip and no-jump
k is the thermal conductivity, and T is the temperature. are taken at the interface of the solid–fluid parts as following:

Fig. 2. The procedure used for the case of two predefined pressure limits.
M. Dehghan et al. / Energy Conversion and Management 92 (2015) 244–250 247

100 20
Wo=75, Pi=2000 Wo=75, Pi=3000

Wo=100, Pi=2000
Wo=100, Pi=3000
80
Wo=150, Pi=2000
Wo=150, Pi=3000
Wo=200, Pi=2000 15
60
Wo=200, Pi=3000
Po

Nu
40
10

20

0 5
0 0.05 0.1 0.15 0.2 0.25 0.3 0 0.05 0.1 0.15 0.2
z+ z*
Fig. 4. The Poiseuille number for different width-tapered configurations. Fig. 7. Effects of tapering on the Nusselt number.

3500 4.0E+4
Wo=75, Pi=3000
Wo=75, Pi=3000
3000 Wo=100, Pi=3000

Wo=150, Pi=3000 Wo=100, Pi=3000

2500 Wo=200, Pi=3000 3.0E+4 Wo=150, Pi=3000


h (W/m2)

2000 Wo=200, Pi=3000


P (Pa)

1500
2.0E+4
1000

500

1.0E+4
0 0 0.05 0.1 0.15 0.2
0 0.002 0.004 0.006 0.008 0.01 0.012
z*
z (m)
Fig. 8. Effects of tapering on the convection heat transfer coefficient.
Fig. 5. The area-weighted average pressure distribution along the flow direction.

2.3. Solution method and mesh sizing

The present study is a development to the study of Mirzaei and


Dehghan [11]. Finite volume method (SIMPLEC algorithm of Van
Doormal and Raithby [30]) has been used to solve the governing
equations. Convective terms have been discretized using the sec-
10 ond order upwind method. The artificial transient method in a
non-uniform structured grid has been adopted for solving and
Nu

Wo=200, Pi=1500
decoupling the problem [31–35]. The convergence criterion is the
Wo=200, Pi=2000 maximum dimensionless temperature difference between two
successive time-steps lower than 106. The problem has been
Wo=200, pi=3000
solved with different meshes. Difference between the local heat
Mirzaei and Dehghan (2013) transfer coefficients (h) in 40  75  150 and 20  50  100
Nu_3z; Lee et al. (2005) meshes is less than 3%. However the required computational time
increases dramatically. The code requires up to 20 h for the
1
0.0001 0.001 0.01 0.1 adopted mesh number (x  y  z = 20  50  100) in a PC with
z* 3.1 GHz CPU and 4 GB RAM. For the finer mesh (40  75  150) it
requires at least 50% extra time to converge at the same precision
Fig. 6. The Nusselt number distribution with respect to z⁄. and convergence criterion. Mesh type is hexahedron with uniform
mesh size in the x- and y-directions, but in the z-direction (the flow
~ direction) the mesh size is increased by a factor of 1.01.
V ¼0
As it was mentioned, a maximum pressure value is adopted as
Tf ¼ Ts ð3Þ the constraint of the simulation. The outlet gauge pressure has
@T f @T s been fixed to be 50 Pa. Because only one value for the pressure
 kf ¼ ks
@n @n can be fixed as the boundary condition, the inlet pressure has been
248 M. Dehghan et al. / Energy Conversion and Management 92 (2015) 244–250

adjusted by varying the inlet velocity in a trial–error procedure. 1


Nu3z ¼ þ C4; 1 < a < 10
The inlet velocity has been increased up to a certain value in which C 1 ðz ÞC 2 þ C 3
the area-weighted value of the inlet gauge pressure equals to the
C 1 ¼ 3:122  103 a3 þ 2:435  102 a2 þ 2:143  101 a þ 7:325
pressure limit. This limit has been assumed to be 1500, 2000,
and 3000 Pa. So, the inlet velocity would be different for each case. C 2 ¼ 6:412  101
This procedure, as a flowchart, is shown in Fig. 2. Also, Mirzaei and C 3 ¼ 1:589  104 a2  2:603  103 a þ 2:444  102
Dehghan [11] proposed that the temperature dependency of mate-
C 4 ¼ 7:148  1:328  10=a þ 1:515  10=a2  5:936=a3
rial could have major effects on the simulation result. Conse-
quently, the properties of the fluid phase have been adopted at ð6Þ
the mean operating temperature. To this aim, at first a solution
where a is the channel aspect ratio. To obtain the Nusselt number in
with properties at the inlet temperature has been obtained based
the present study, the way proposed by Mirzaei and Dehghan [11]
on the mentioned constraint. Then, the overall mean temperature
has been adopted. At first, the convection heat transfer coefficient
of the fluid phase has been calculated based on a volume averaging
at a location equal to z should be defined in order to define the
to be used for evaluation of the thermophysical properties of water
Nusselt number (Eq. (8)).
listed in Ref. [36].
qðzÞ
hz ¼ ð7Þ
3. Results and discussion T w ðzÞ  T b ðzÞ

The apparent friction coefficient properly covers inlet effects. It hz Dh


NuðzÞ ¼ ð8Þ
includes the pressure drop due to both the wall shear stress and k
momentum variation due to changes in the velocity profile [11]: Tw(z) is the wall temperature and q(z) is the heat flux at z, which
is calculated by Eq. (9) performing summations on the perimeter of
1 Dh pin  pz the channel at each cross section. Tb(z) is the bulk mean tempera-
f app ðzÞ ¼ ð4Þ
4 z 1=2qu2av e ture given by Eq. (10).
P
2WH C qðx;
P
y; zÞdAðx; y; zÞ
qw ðzÞ ¼ ð9Þ
Dh ¼ ð5Þ C dAðx; y; zÞ
W þH
R
uave is the area-weighted average velocity, Dh is the hydraulic diam- qucp T f dA
eter given by Eq. (5), and fapp(z) is the apparent friction factor at dis- T b ðzÞ ¼ ARc ð10Þ
Ac
qucp dA
tance z. W is the width of channel which varies along the flow
direction for a converging channel. Fig. 3 shows the Poiseuille num- Fig. 6 shows that the Nusselt number has a very high value in
ber (Po = fapp  Re) which is a universal dimensionless number repre- the entrance region, then decreases along the flow direction, and
senting the hydraulic aspect of the problem versus a dimensionless finally tends to a constant value which depends on the geometry
axial location (z+ = z  Dh1  Re1). A good agreement is seen of the channel [11,37]. Also, it confirms that the Nusselt number
between the results of the present study (at different inlet pres- is independent of the axial pressure gradient. To see effects of
sures) and those obtained by Mirzaei and Dehghan [11] and Phillips tapering on the Nusselt number, Fig. 7 is plotted. The Nusselt
[37]. Fig. 4 reveals the variation of Poiseuille number at different number, having a high value at the entrance due to the inlet effect,
width-tapered configurations. It can be seen that the Poiseuille decreases along the channel up to reach a fully developed value for
number increases with increased tapering. This increase dramati- a channel with straight walls (Wo = 200). But, the Nusselt number
cally rises at width-tapered ratio (Wo/Wi) smaller than 0.5 (or increases (after a reduction close to the entrance) in width-tapered
Wo < 100 lm). The scale analysis shows that the velocity gradient channels when the inlet effect diminishes. The converging walls
on the walls has the order of magnitude equal to umax/Dh. The max- impose the fluid to have a velocity component in the x-direction.
imum velocity at each section (umax) increases with increased taper- Therefore, the fluid has more ability to transfer heat from the walls,
ing and simultaneously the hydraulic diameter decreases. On the especially at z⁄ > 0.04 where the straight channel would be fully
other hand, the friction factor (f) is proportional to the velocity gra- developed. It should be noted that the hydraulic diameter
dient, as well as the Po number. Consequently, the Po dramatically decreases in a width-tapered channel along the channel. So, the
increases when the width-tapering is remarkable. Fig. 5 depicts the convection heat transfer coefficient increases faster than the
area-weighted average pressure with respect to the axial location. It hydraulic diameter reduction. The variation of the convection heat
could be seen that the pressure gradient is semi-linear for the transfer coefficient along the channel with tapering is revealed in
straight channel (Wo = 200 lm) and for the channel with a Fig. 8. It could be seen that the enhancement obtained by tapering
relatively low converging walls (Wo = 150 lm). By increasing the occurs somewhere between Wo = 100 and Wo = 75 lm. By increas-
width-tapering, the flow faces higher pressure gradients close to ing the tapering, the heat transfer coefficient would not be
the outlet. It was expected because the flow accelerates along the enhanced. This is in agreement with findings of Hung and Yan
flow direction when the walls are converging at the constraint of [26] predicting that the optimum tapering ratio for the present
a fixed overall pressure difference. Consequently, the higher accel- problem is about 0.5. In other words, the optimum outlet width
erating process needs a relatively higher pressure gradient. It of the present problem should be about 100 lm. For this configu-
should be noted that the pumping power decreases by increasing ration (Wo/Wi = 0.5) the inlet velocity (uin = 0.27 ms1) is half of
the tapering since the overall pressure gradient is fixed and the vol- the inlet velocity of the straight channel (uin = 0.53 ms1) when
umetric flow rate decreases.To compare the accuracy of the thermal the maximum pressure is set to be 3000 Pa. Consequently, the
results of the present study with those of other researchers, Fig. 6 is required pumping power is 0.25% of the straight channel one with
plotted. It shows the Nusselt number (Nu) with respect to a dimen- the same heat removal capacity. Another matter that can be
sionless axial location (z⁄ = z  Dh1  Re1  Pr1). The Nu3z given by remarked here is that the local difference between the area-
Eq. (6) is the Nusselt number based on the assumption that a rect- weighted average temperature of the fluid and solid parts increases
angular channel has one side insulated and three other sides near the entrance, then decreases near the end of channel, with
encountering a heat flux [9]: increasing width-tapered ratio. However, the overall temperature
M. Dehghan et al. / Energy Conversion and Management 92 (2015) 244–250 249

350 represents the temperature difference between the solid and fluid
phases [38–40]. So, the irreversibility of the system and amount of
340 generated entropy would decrease and consequently the thermal
efficiency would enhance as well. Fig. 9 shows the temperature
of the bottom side of the flow passages along the flow direction.
330
It can be seen that the maximum temperatures of the channels fall
within the safe zone even for a tapered channel. For example, the
Tw (K)

320 third generation of IntelÒ core i-series processors can operate at


temperatures as high as 105 °C (378 K). To see the 3D temperature
310 Wo=75, Pi=2000 distribution in the fluid and solid parts of MCHS, Fig. 10 has been
Wo=100, Pi=2000 plotted. Fig. 10(a) shows that the fluid does not effectively absorb
the imposed heat flux and is relatively much colder compared to
300 Wo=150, Pi=2000
the solid part. By tapering and changing the flow pattern, and con-
Wo=200, Pi=2000 sequently lowering the fluid flow, the heat is transferred more
290 effectively to the fluid part. As a result, the overall temperature dif-
0 0.002 0.004 0.006 0.008 0.01 0.012
ference between the fluid and solid parts decreases in Fig. 10(b)
z (m) according to what was predicted by the porous medium approach.
Fig. 9. Variation of the bottom-side temperature of flow passages along the flow
4. Conclusion
direction.

The conjugate heat transfer and constrained fluid flow in width-


tapered microchannel heat sinks have been investigated in the
laminar regime (Re < 350). The maximum pressure of the MCHS
loop has been limited due to a constructional or operational limi-
tation. Three different maximum pressures (1500, 2000, and
3000 Pa) and four different tapering configurations (Wo = 200,
150, 100, and 75 lm) have been analyzed for the problem. High-
lights of this study could be stated as following:

 The Poiseuille number increases with increased tapering. When


the width-tapered ratio (Wo/Wi) is smaller than 0.5, the increase
in the Poiseuille number is high.
 As the high and low pressures of the loop have been assumed to
be under the control, the pumping power decreases with
increased tapering because of the reduction in the flow rate.
 The pressure gradient is semi-linear for straight and semi-
straight ducts. It increases along the flow direction with
increased tapering.
 The Nu number is not sensitive to tapering in the entrance sec-
tion of the channel (z⁄ < 0.02).
 The Nusselt number increases with respect to the tapering
when the entrance effect of the flow diminishes. This happens
due to a velocity component toward the center-plane of the
flow passage.
 The enhancement in the heat transfer coefficient has an opti-
mum value when the width-tapered ratio (Wo/Wi) is about 0.5
in the present problem. For high width-tapered ratios (Wo/Wi)
the pressure gradient devotes to only to a small section near
the exit of the duct and the overall heat transfer performance
would not be enhanced.
 The required pumping power of the optimum width-tapered
configuration is 0.25% of the straight channel one with the same
overall heat removal capacity when the maximum pressure of
the loop is 3000 Pa. It is found that a MCHS with a moderate
width-tapered ratio (of about 0.5 when the inlet aspect ratio
is equal to 5) removes the same amount of heat flux as the
MCHS with straight flow passages, while the required pumping
power is considerably reduced.
Fig. 10. Temperature (K) contours of the MCHS (Pi = 2000 Pa); (a) Wo = 200 lm and
(b) Wo = 100 lm.
References

[1] Kakac S, Yener Y, Avelino M, Okutuko T. Single-phase forced convection in


difference decreases with tapering. It is predictable by the porous minichannel (a state-of-the-art review). In: Microscale Heat Transfer. Springer;
medium approach. As the specific area per volume and the local 2005 [chapter 1].
[2] Bontemts A. Measurements of single-phase pressure drop and heat transfer
convection heat transfer coefficient increase with increasing taper- coefficient in micro and minichannels. In: Microscale heat transfer. Springer;
ing, the local thermal non-equilibrium intensity decreases which 2005 [chapter 2].
250 M. Dehghan et al. / Energy Conversion and Management 92 (2015) 244–250

[3] Türkakar G, Okutucu-Özyurt T. Dimensional optimization of microchannel [23] Dehghan M, Mahmoudi Y, Valipour MS, Saedodin S. Combined conduction–
heat sinks with multiple heat sources. Int J Therm Sci 2012;62:85–92. convection–radiation heat transfer of slip flow inside a micro-channel filled
[4] Tuckerman DB, Pease RFW. IEEE Electron Dev Lett 1981;2(5):126–9. with a porous material. Transp Porous Media in preparation [2014].
[5] Tuckerman DB, Pease RFW. Optimized convective cooling using [24] Dehghan M, Rahmani Y, Ganji DD, Saedodin S, Valipour MS, Rashidi S.
micromachined structure. J Electrochem Soc 1982;129(3):98C. Convection–radiation heat transfer in solar heat exchangers filled with a
[6] Pfahler J, Harley J, Bau H, Zemel J. Liquid transport in micron and submicron porous medium: homotopy perturbation method versus numerical analysis.
channels. Sens Actuat 1990;A21–A23:431–4. Renew Energy 2015;74:448–55.
[7] Makihara M, Sasakura K, Nagayama A. The flow of liquids in micro-capillary [25] Hung TC, Sheu TS, Yan WM. Optimal thermal design of microchannel heat
tubes – consideration to application of the Navier–Stokes equations. J Jpn Soc sinks with different geometric configurations. Int Commun Heat Mass Transfer
Precis Eng 1993;59(3):399–404. 2012;39:1572–7.
[8] Gamrat G, Favre-Marinet M, Asendrych D. Conduction and entrance effects on [26] Hung TC, Yan WM. Optimization of a microchannel heat sink with varying
laminar liquid flow and heat transfer in rectangular microchannels. Int J Heat channel heights and widths. Numer Heat Transfer, Part A 2012;62:722–41.
Mass Transfer 2005;48:2943–54. [27] Hwang JJ, Tseng FG. Chin pan ethanol–CO2 two-phase flow in diverging and
[9] Lee PS, Garimella SV, Liu D. Investigation of heat transfer in rectangular converging microchannels. Int J Multiphase Flow 2005;31:548–70.
microchannels. Int J Heat Mass Transfer 2005;48:1688–704. [28] Lee SS, Yim Y, Kyung H, Lee J. Extensional flow-based assessment of red blood
[10] Morini GL. Scaling effects for liquid flows in micro-channels. Heat Transfer Eng cell deformability using hyperbolic converging microchannel. Biomed
2006;27:64–73. Microdev 2009;11:1021–7.
[11] Mirzaei M, Dehghan M. Investigation of flow and heat transfer of nanofluid in [29] Hung TC, Yan WM. Effects of tapered-channel design on thermal performance
microchannel with variable property approach. Heat Mass Transfer of microchannel heat sink. Int Commun Heat Mass Transfer 2012;39:1342–7.
2013;49:1803–11. [30] Van Doormal JP, Raithby GD. Enhancements of the SIMPLE method for
[12] Morini GL, Yang Y. Guidelines for the determination of single-phase forced predicting incompressible fluid flows. Numer Heat Transfer 1984;7:147–63.
convection coefficients in microchannels. J Heat Transfer 2013;135. http:// [31] Dehghan M, Basirat Tabrizi H. On near-wall behavior of particles in a dilute
dx.doi.org/10.1115/1.4024499. 101004-1. turbulent gas–solid flow using kinetic theory of granular flows. Powder
[13] Wang XD, An B, Xu JL. Optimal geometric structure for nanofluid-cooled Technol 2012;224:273–80.
microchannel heat sink under various constraint conditions. Energy Convers [32] Dehghan M, Mirzaei M, Mohammadzadeh A. Numerical formulation and
Manage 2013;65:528–38. simulation of a non-Newtonian magnetic fluid flow in the boundary layer of a
[14] Valeh-e-Sheyda P, Rahimi M, Karimi E, Asadi M. Application of two-phase flow stretching sheet. J Model Eng 2013;11(34):73–82.
for cooling of hybrid microchannel PV cells: a comparative study. Energy [33] Dehghan M, Basirat Tabrizi H. Turbulence effects on the granular model of
Convers Manage 2013;69:122–30. particle motion in a boundary layer flow. Can J Chem Eng 2014;92:189–95.
[15] Ting TW, Hung YM, Guo N. Effects of streamwise conduction on thermal [34] Dehghan M, Basirat Tabrizi H. Effects of coupling on turbulent gas-particle
performance of nanofluid flow in microchannel heat sinks. Energy Convers boundary layer flows at borderline volume fractions using kinetic theory. J
Manage 2014;78:14–23. Heat Mass Transfer Res 2014;1:1–8.
[16] Yang K, Zuo C. A novel multi-layer manifold microchannel cooling system for [35] Dehghan M, Mirzaei M, Valipour MS, Saedodin S. Flow of a non-Newtonian
concentrating photovoltaic cells. Energy Convers Manage 2015;89:214–21. fluid over a linearly moving sheet at a transient state; new similarity variable
[17] Miner MJ, Phelan PE, Odom BA, Ortiz CA. An experimental investigation of and numerical solution scheme. J Model Eng in press [2015].
pressure drop in expanding microchannel arrays. J Heat Transfer 2014;136. [36] Li Z, Huai X, Tao Y, Chen H. Effects of thermal property variations on the liquid
http://dx.doi.org/10.1115/1.4025557. 031502-1. flow and heat transfer in microchannel heat sinks. Appl Therm Eng
[18] Huang CY, Wu CM, Chen YN, Liou TM. The experimental investigation of axial 2007;27:2803–14.
heat conduction effect on the heat transfer analysis in microchannel flow. Int J [37] Phillips RJ. Microchannel heat sinks, Advances in Thermal Modeling of
Heat Mass Transfer 2014;70:169–73. Electronic Components and Systems. New York: Hemisphere Publishing
[19] Mirzaei M, Saffar-Avval M, Naderan H. Heat transfer investigation of laminar Corporation; 1990.
developing flow of nanofluids in a microchannel based on Eulerian–Lagrangian [38] Dehghan M, Valipour MS, Saedodin S. Perturbation analysis of the local
approach. Can J Chem Eng 2014. http://dx.doi.org/10.1002/cjce.21962. thermal non-equilibrium condition in a fluid saturated porous medium
[20] Mohammed HA, Gunnasegaran P, Shuaib NH. Numerical simulation of heat bounded by an iso-thermal channel. Transp Porous Media 2014;
transfer enhancement in wavy microchannel heat sink. Int Commun Heat 102(2):139–52.
Mass Transfer 2011;38:63–8. [39] Dehghan M, Jamal-Abad MT, Rashidi S. Analytical interpretation of the local
[21] Mohammed HA, Gunnasegaran P, Shuaib NH. Influence of channel shape on thermal non-equilibrium condition of porous media imbedded in tube heat
the thermal and hydraulic performance of microchannel heat sink. Int exchangers. Energy Convers Manage 2014;85:264–71.
Commun Heat Mass Transfer 2011;38:474–80. [40] Dehghan M, Valipour MS, Saedodin S. Temperature-dependent conductivity in
[22] Alfaryjat AA, Mohammed HA, Adam NM, Ariffin MKA, Najafabadi MI. Influence forced convection of heat exchangers filled with porous media: a perturbation
of geometrical parameters of hexagonal, circular, and rhombus microchannel solution. Energy Convers Manage 2015;91:259–66.
heat sinks on the thermohydraulic characteristics. Int Commun Heat Mass
Transfer 2014;52:121–31.

You might also like