Applied Thermal Engineering 31 (2011) 556e565

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Critical investigation of heat transfer enhancement using nanofluids in

microchannels with slip and non-slip flow regimes
A. Akbarinia a, b, *, M. Abdolzadeh c, R. Laur a
a
Institute for Electromagnetic Theory and Microelectronics (ITEM), University of Bremen, 28359 Bremen, Germany
b
Department of Mechanical Engineering, Shahid Bahonar University of Kerman, Kerman, Iran
c
Department of Mechanical Engineering, I. A. U. Kerman Branch, Kerman, Iran

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

Article history: The Reynolds number in nanofluid studies depends on the inlet velocity and the kinematic viscosity of
Received 21 May 2010 nanofluid. The nanofluid kinematic viscosity increases with an increase in nanoparticles volume fraction
Accepted 14 October 2010 while the inlet velocity should be increased to keep the Reynolds number constant. Therefore, it is not
Available online 10 November 2010
clear that either increasing the nanoparticles volume fraction or increasing the inlet velocity has a major
role on heat transfer enhancement in nanofluids flow studies which are done at constant Reynolds
Keywords:
numbers. In this study, forced convection Al2O3-water nanofluid flows in two-dimensional rectangular
Nanofluid
microchannels have been investigated to study heat transfer enhancement due to addition of the
Microfluidic
Heat transfer
nanoparticles to the base fluid especially in microchannels at low Reynolds number. Three different cases
Slip flow are examined to evaluate proportion impact of increasing nanoparticles volume fraction (f) and the inlet
Reynolds number velocity (uin) on heat transfer enhancement. Two-dimensional NaviereStokes and energy equations
Microchannel accompany with the slip velocity and the jump temperature boundary conditions expressions have been
discretized using the Finite Volume Method (FVM). The Brownian motions of nanoparticles have been
considered to determine the thermal conductivity of nanofluids. The calculated results show good
agreement with the previous numerical and analytical data. It is found that at a given Reynolds number,
the major enhancement in the Nusselt number is not due to increasing the nanoparticle concentrations
but it is due to the increasing the inlet velocity to reach a constant Re. Constant Reynolds number studies
of nanofluids are not sufficient approach to evaluate the heat transfer and the skin friction factor due to
the nanofluids usage.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction flow regime, which is characterized by slip flow at wall. Traditionally,

the no-slip condition at wall is enforced in the momentum equations
With the development of science and technology, people began and an analogous no-temperature-jump condition is applied in the
to realize that with the size decreasing the system will own many energy equation. Strictly speaking, no-slip/no-jump boundary
advantages that do not appear in conventional size, including conditions are valid only if the fluid flow adjacent to the surface is in
compact size, disposability, and increased functionality [1]. Micro- thermodynamic equilibrium. This requires an infinitely high
channels, however, are the basic structures in these systems. Based frequency of collisions between the fluid and the solid surface [1,2].
on the Knudsen number (Kn), the flow in microchannels have been In practice, the no-slip/no-jump condition leads to fairly accurate
classified into four flow regimes: continuum flow regime predictions as long as Kn < 0.001. Beyond that, flow in devices shows
(Kn
0.001), slip flow regime (0.001 < Kn
0.1), transition flow significant slip since characteristic length is on the order of the mean
regime (0.1 < Kn
10) and free molecular flow regime (Kn > 10). The free path of the fluid or gas molecules. It means that the collision
flow in most application of these systems, such as Micro Gyroscope, frequency is simply not high enough to ensure equilibrium and
Accelerometer, Flow Sensors, Micro Nozzles, Micro Valves, is in slip a certain degree of tangential velocity slip and temperature jump
must be appearing. Slip velocity at the wall is the most important
feature in micro/nano scale that differs from conventional internal
flow. Therefore, the slip flow characteristics are very important for
* Corresponding author. Institute for Electromagnetic Theory and Microelec- designing and optimizing the micro/nano systems. On the other
tronics (ITEM), University of Bremen, 28359 Bremen, Germany. Tel.: þ49 421
hand, the optimization of microchannels design is affected signifi-
21862517; fax: þ49 421 2184434.
E-mail address: [email protected] (A. Akbarinia). cantly by their reliability and thermal performance. It is found that

1359-4311/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2010.10.017
 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565 557

Nomenclature Greek letters

a thermal diffusivity (¼(k/rCp))
Cp specific heat (kJ kg1 K1) b variable, defined in Eq. (19)
d diameter of nanoparticle or water molecule g specific heat ratio (¼(Cp/Cv))
Dh hydraulic diameter (m) q non-dimensional temperature (¼(T  Tin/Twall  Tin))
f Fanning friction coefficient, ð ¼ sw =ðð1=2Þrnf u2 ÞÞ l mean free path (m)
h heat transfer coefficient (W m2 K1) h variable, defined in Eq. (14)
k thermal conductivity (W m1 K1) m dynamic viscosity (N s m2)
kB Boltzmann constant (¼1.3807  1023 j/K) r density (kg m3)
Kn Knudsen number (l/Dh) sT thermal accommodation coefficient
L length of microchannel (m) sv momentum accommodation coefficient
Nux local Nusselt number, (¼hxDh/knf) sw shear stress at the wall, (¼mnf(vu/vyjwall))
q00 heat flux at the wall y kinematic viscosity (ynf ¼ mnf/rnf)
p pressure (Pa) f nanoparticles volume fraction
P non-dimensional pressure F independent variable
Po Poiseuille number,ðf Re ¼ sw Dh =ðð1=2Þmnf uÞÞ
Penf Peclet number (¼Renf$Prnf) Subscripts
Prnf Prandtl number ð ¼ mnf ðCp Þnf =knf Þ b balk
Renf Reynolds number ð ¼ ðrnf uin Dh =mnf ÞÞ en entrance length
T temperature f base fluid
u velocity (m s-1) fd fully developed
U non-dimensional velocity i, j array indices
W width of microchannel (m) in inlet condition
x, y coordinates N closed to the wall
X, Y non-dimensional coordinates nf nanofluid
X* reciprocal Graetz number (x/(Dh$Pe)) s solid nanoparticles
wall at the wall

the temperature jump effects are very important and should be conductivity has also been investigated numerically by Refs.
included in the modeling of the slip flow heat transfer problems, and [15e20]. They have evaluated the impacts of parameters such as
neglecting these effects leads to a significant overestimate of heat microchannel geometry, Reynolds number and nanoparticles
transfer coefficient [3,4]. volume fraction on the thermal performance of nanofluid flow in
In most of the previous research, low thermal conductivity base microchannels. They have reported higher heat transfer perfor-
fluids such as air or water have been employed to study the fluid flow mance of nanofluids flow compared to the base fluids in micro-
in microchannel. The heat transfer of fluids flow is limited based on channels at constant Reynolds number. However, all of them have
thermal properties of fluids. However, demand for increasing the heat used the non-slip flow regimes in their numerical research works.
transfer in microchannel needs to improve thermal properties of base To the best knowledge of the authors, the majority of previous
fluids. One of the presented solutions for this problem is adding solid numerical and experimental studies on nanofluids flow and heat
nanoparticles with high thermal conductivity such as Al2O3, Cu or CuO transfer have been done with constant Reynolds number. The Rey-
to the base fluid which are called nanofluids. During the past two nolds number in nanofluids studies depend on the inlet velocity and
decades, scientists and researchers have been attempting to develop the kinematic viscosity of nanofluid. The nanofluid kinematic
nanofluids, which can offer better heating and cooling performance viscosity increases with an increase in nanoparticles volume fraction
for variety of thermal systems compared to traditional heat transfer while the inlet velocity should be increased to keep the Reynolds
fluids. It is found by numerous numerical and experimental studies number constant. Therefore, it is not clear that either increasing the
that the nanofluids have high thermal conductivities and abilities to nanoparticles volume fraction or increasing the inlet velocity has
increase the heat transfer [5e8]. The major attempts have been done major role on heat transfer enhancement in nanofluids studies
to investigate the effective thermal conductivity of nanofluids and the which are done at constant Reynolds numbers.
effective dynamic viscosity for predicting the thermal conductivity The objective of the present paper is to study the proportion
and dynamic viscosity of nanofluids [9].Most of the papers have impact of nanoparticle volume fraction (f) and the inlet velocity
reported an enhancement in the heat transfer due to increasing the (uin) on heat transfer of nanofluids flow at constant Reynolds
nanoparticle concentration which are done at constant Reynolds numbers especially in microchannels with low Reynolds numbers
number. But some researchers have reported an uncertainty in heat and answer to this question “ Does the nanofluid increase heat
transfer enhancement of nanofluids flow [10,11]. Akbarinia [10] has transfer?”. To aim to this purpose, laminar forced convection slip
investigated heat transfer in the nanofluids flow with constant mass and non-slip Al2O3-water nanofluid flow in a rectangular micro-
flow rate and reported a reduction in heat transfer with increasing channel have been investigated, employing the slip velocity and the
nanoparticles volume fraction. jump temperature boundary conditions at the fluidewall interface.
Recently, the nanofluids utilization in microchannel has been Three different cases are examined to evaluate proportion impact of
received a great interest amongst researchers. Koo and Kleinstreuer increasing nanoparticles volume fraction (f) and the inlet velocity
[12], Jang and Choi [13] and Chein and Huang [14] have studied (uin) on heat transfer enhancement. Simultaneous effects of the
experimentally improving thermal performance of the nanofluids inlet velocity and the nanoparticles volume fraction on heat
flow in microchannel with different nanofluids such as CuO-water transfer augmentation are studied. The axial velocity, Poiseuille
or Al2O3-water. The thermal performance of microchannels using number, temperature and Nusselt number profiles are presented
nanofluids base on different models for effective thermal and discussed. Although the study is done in a microchannel with
558 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565

Table 1
ThermophysicalProperties of nanoparticles and base fluid at 27  C.

Properties Water Nanoparticle (Al2O3)

Density r(kg/m3) 998.2 3890
Heat capacitance Cp (J/kg K) 4240 880
Thermal conductivity k (W/mK) 0.608 35

The continuity, momentum and fluid energy equations are non-

dimensionalized using following dimensionless parameters:

xi u p T  Tin
Fig. 1. Schematic diagram of two-dimension rectangular microchannel.
Xi ¼ ;U ¼ i ;P ¼ ; and q ¼ (4)
Dh i uin rnf u2in Twall  Tin
slip flow regimes, but results are also presented for the case no-slip
flow (Kn ¼ 0) which also covers the macro-scale. where Dhz2W.
Therefore, the non-dimensionalized governing equation
2. Analysis becomes as follow:
Continuity equation:
2.1. Governing equations
vUj
¼ 0 (5)
The schematic of the channel and coordinate system are shown vXj
in Fig. 1. The flow is considered along the x-axis and the channel
length is chosen 20 times the width in order to achieve hydrody- Momentum equations:
!
namically developed flow at the outlet. v
 1 v 2 Ui vP
The Al2O3-water nanofluid flow and the heat transfer in rect- U i Uj ¼  (6)
vXj Renf vXj2 vXi
angular microchannels have been considered. The nanofluid flow is
laminar, steady state and incompressible with constant properties Fluid energy equation:
while dissipation, pressure work and body forces are neglected. The
used properties of base fluid (water) and solid nanoparticles (Al2O3) v 1 v2 q
are presented in Table 1. The rarefaction effects set the slip velocity ðUi qÞ ¼ (7)
vXi Penf vXi
and the jump temperature at the fluidewall interface. Therefore,
the steady state governing equations describing the fluid flow and where Penf ¼ Renf Prnf ; Renf ¼ ðrnf uin Dh =mnf Þ and Prnf ¼ ðmnf
the heat transfer in the microchannel in the Cartesian coordinate ðCp Þnf =knf Þ, are the Peclet number, the Reynolds number and the
and in the tensors form are as follows: Prandtl number, respectively.
Continuity equation:
2.2. Nanofluid properties
vuj
¼ 0 (1)
vxj The physical properties of nanofluid containing water-Al2O3 can
be calculated as follow:
Momentum equations:
Density:
!!
v
 v vui vp rnf ¼ ð1  fÞrf þ frs (8)
rnf u u ¼ mnf  (2)
vxj i j vxj vxj vxi
Heat capacitance:
Fluid energy equation:



 
rCp nf
¼ ð1  fÞ rCp f þf rCp s (9)

v v vT
rCp nf ðui TÞ ¼ knf (3) The presented expression for dynamic viscosity by Maiga et al. [21]
vxi vxi vxi
which was determined based on available experimental results for
Where i ¼ 1, 2 and ui ¼ (ux, uy) are velocity vectors. water-Al2O3 is given as:

Fig. 2. Grid independent test a) in x direction b) in y direction.

 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565 559

Table 2
Calculated values for the first case (I) at a constant Reynolds number.

Re f uin mnf/rnf Cp knf Pr Pe

10 0 0.0502 1.0048e-006 4242 0.608 7 70
10 5% 0.0734 1.4679e-006 4203 0.687 10.3 103

where lf is mean free path of water molecular (lf ¼ 0.17 nm), kB is

Boltzmann constant (kB ¼ 1.3807  1023 j/k) and h has been
calculated by the following equation:

h ¼ A$10TC ; A ¼ 2:414  105 ; B ¼ 247:8; C ¼ 140

B
(14)

2.3. Slip velocity and jump temperature

The slip velocity at fluid-wall interface exists in either gas or

liquid flows. This effect is more significant in micro and nono-
scales, because in macro-scale the hydraulic diameter is very large
compare to the molecular mean free path (Knz0) and the
assumption of non-slip velocity (i.e. slip velocity is zero at the fluid-
Fig. 3. Compare the fully developed Poiseuille number with the previous numerical wall interface) is valid. In cases of micro and nano-scales, this
results by Renksizbulut et al. [29] and also with analytical results. assumption is no longer correct for both liquid and gas flows. This
fact is confirmed by Tretheway and Meinhar [23] which measured
  the velocity profiles of water flowing through 30300 μm channels
mnf ¼ 123f2 þ 7:3f þ 1 mf (10) experimentally. They found that when a hydrophobic microchannel
(uncoated glass) surface is coated with a 2.3 nm thick monolayer of
The thermal conductivity of water-Al2O3 nanofluid has been hydrophobic octadecyltrichlorosilane, an apparent velocity slip is
determined from Chon et al. [22] correlation which considers the measured just above the solid surface. This velocity is approxi-
Brownian motion and mean diameter of the nanoparticles as mately 10 % of the free stream velocity and yield slip length of
follow: approximately 1 mm. For this slip length, slip flow is negligible for
 0:3690 !0:7476 length scales greater than 1mm but must be considered at micro
knf d ks and nano scales. The results of Ngoma and Erchiqui [24], Yang and
¼ 1 þ 64:7f0:7460 f Pr0:9955
f Re1:2321
f
kf ds kf Kwok [25] and Louis Barrat and Bocquet [26] also confirm the use of
slip velocity for liquid flows in micro and nano-scales. In the
(11)
present work the rang of the Knudsen number is selected between
where Prf and Ref in Eq. (11) are defined as: 0 to 0.1 to consider slip velocity at nano-scales (Knz0.1) as well as
micro-scales (0<Kn <0.1) and non-slip velocity at macro-scale
h (Knz0) in the liquid flows. The second order non-dimensionalized
Prf ¼ (12)
rf af velocity slip condition is expressed as [27]:
   
rkB T 2  sV vU  3 ðg  1Þ Kn2 Re vq 
Ref ¼ U ¼ Kn  þ (15)
3ph2 lf
(13) sV vY wall 2p g Ec vX wall
Here Ec is Eckert number ðEc ¼ ðU 2 =Cp DTÞÞ. The second term can
be negligible if ðv=vYÞ[ðv=vXÞ at the wall and also due to fact that
it is second order in the Knudsen number. Similar arguments can be
applied to the jump temperature boundary condition, and the
resulting from Taylor series leads in dimensionless form [27]:
   "   
2  sT 2g 1 vq 
q  qwall ¼ Kn 
sT g þ 1 Pr vY wall
! #
2 
Kn2 v q
þ þ. ð16Þ
2! vY 2 wall

Values of the thermal and momentum accommodation coefficients

(sV,sT) are near unity for most engineering applications and they
are taken as unity in the present study.

Table 3
Calculated values for the second case (II) at f ¼ 0 and the corresponding inlet
velocity values in the first case (I).

uin f Re Pr Pe
0.0502 0 10 7 70
Fig. 4. Compare the local Nusselt number with the previous numerical results by Yu
0.0734 0 14.6 7 102.2
and Ameel [30] for Kn¼0.04 and b¼10 with infinity aspect ratio.
560 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565

Table 4  At the channel inlet: the nanofluid velocity and the tempera-
Calculated values for the third case (III) at a constant inlet velocity and different f. ture profiles are assumed uniform and constant (X ¼ 0,
f uin mnf/rnf Cp knf Re Pr Pe 0
Y
0.5: U ¼ 1, q ¼ 0).
0 0.0502 1.0048e-006 4242 0.608 10 7 70  At the fluidesolid interface: The velocity in y direction is zero
5% 0.0502 1.4679e-006 4203 0.687 6.9 10.3 71.1 while the slip velocity and the jump temperature of the
nanofluid flow adjacent to the wall is proportional to normal
velocity and the temperature gradients at the fluidewall
After simplifying, the non-dimensional forms of the slip interface (Eqs. (17) and (18)). A constant temperature (qwall ¼ 1)
flow velocity and jump temperature boundary conditions are at the wall is also applied.
given by:  At the channel outlet: the diffusion flux in the direction normal
 to the exit is assumed to be zero for the velocity and the
vU 
UN  Uwall ¼ Kn  (17) temperature while a zero pressure is assigned at the flow exit.
vY wall ðX ¼ 10; 0
Y
0:5 ðvU=vXÞ ¼ ðvq=vXÞ ¼ 0; P ¼ 0Þ

Kn vq 
qN  qwall ¼ (18) 3. Numerical methods and validations
b vY wall
where The sets of coupled non-linear differential equations were
  discretized using the Finite Volume Method (FVM). The power
gþ1
b ¼ Pr 2g ;
law scheme was used for the convective and diffusive terms
(19) while the SIMPLER procedure was introduced to couple the
qwall ¼ 1 and Uwall ¼ 0 velocityepressure as described by Patankar [28]. The slip
velocity and the jump temperature boundary conditions
expressions are discretized and employed in the discretization
equations.
2.4. Boundary conditions A structured non uniform grid distribution has been used for
the computational domain. It is finer near the microchannel
These set of non-linear elliptical governing equations have been entrance and near the wall where the velocity and the tempera-
solved subject to following boundary conditions: ture gradients are high. Channel lengths are set to a value greater

Fig. 5. Variation of the non-dimensional velocity (a,b) and dimensional velocity profile (c,d) with nanoparticles concentration for Re ¼ 10 and Kn ¼ 0.01 at fully developed region a)
and c), at centerline along the channel b) and d).
 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565 561

Fig. 6. The local Poiseuille number profiles for first case (I) and second case (II) at Re ¼ 10 and different f for a) slip flow regime (Kn ¼ 0.01) and b) non-slip flow regimes (Kn ¼ 0).

than the estimated entrance lengths to ensure that fully developed By submitting the term ðvU=vYÞjwall from Eq. (17) into Eq. (21),
conditions are achieved at the exit. Several different grid distri- the local Poiseuille number for the slip flow regimes can be
butions have been tested to ensure that the calculated results are expressed as:
grid independent. As it is shown in Fig. 2 increasing the quantity of
nodes more than 60 nodes in y direction and 160 nodes in x 2
f ReX ¼ U (22)
direction does not change the velocity significantly. Therefore, the Kn N
selected grid for the present calculations consisted of 160  60
The local Nusselt number (Nux) can be obtained by:
nodes in the x and y directions respectively. The solution is
assumed converged when jððFnþ1  Fn Þ=Fnþ1 Þj
106 is satisfied hx Dh
for all independent variables. Nux ¼ (23)
k
After solving the governing equations for velocity, pressure and
temperature other useful quantities such as Nusselt number and where:

Poiseuille number can be determined. q00 00 vT 
The local Poiseuille number (fRex) can be obtained by: hx ¼ and q ¼ k  (24)
ðTwall  TB Þ vy wall

vu
vy wall After normalizing and combination of Eqs. (23) and (24), the
f Rex ¼ 2 Dh (20)
u local Nusselt number for the case of uniform wall temperature can
be calculated from:
After using dimensionless parameters, it becomes as follows: 
vq 

vU  vY wall
f ReX ¼ 2  (21) NuX ¼ (25)
vY wall qB  qwall
562 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565

Fig. 7. a) the non-dimensional temperature at the outlet (the left side) and b) the non-dimensional bulk temperature profiles(the right side) for first case (I), second case (II) and
third case (III).
 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565 563

Fig. 8. The local Nusselt number profiles for first case (I), second case (II) and third case (III) for a) slip flow regime, Kn¼0.01, (the left side) and b) non-slip flow regimes, Kn¼0, (the
right side).

where the bulk temperature can be calculated from: b

Knðq  qwall Þ
NuX ¼ (27)
Z h
L=D qB  qwall
qðX; YÞUðX; YÞdYdX In order to demonstrate the validity and also precision of the model
assumptions and the numerical analysis, fully developed values of
qB ðXÞ ¼ 0
(26)
Z h
W=D the Poiseuille numbers are compared with available numerical and
UðX; YÞdYdX analytical solutions at Re ¼ 100 for different Kn. The calculated
Poiseuille numbers in Fig. 3 show a good agreement with numerical
0
results by Renksizbulut et al. [29] and analytical values, which vary
By substituting the term ðvq=vYÞjwall from Eq. (18) into Eq. (25), the with Kn according tof Refd ¼ ð24=ð1 þ 12KnÞÞ. The local Nusselt
local Nusselt number for the slip flow regimes can be expressed as: numbers are presented in Fig. 4 and are compared also with the
564 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565

results by Yu and Ameel [30] for Kn ¼ 0.04, b ¼ 10 and infinity the first case (I) decreases the temperature and bulk temperature at
aspect ratio. As seen from this figure a significant agreement any cross section at a given Re. Is the reduction due to either
between the calculated and numerical results are observed. increasing the nanoparticles volume fraction or increasing the inlet
velocity?
4. Results and discussion Figures of second case (II) show that increasing inlet velocity
also reduces the temperature and the bulk temperature at any cross
Numerical simulations have been done for three different section like the first case (I). But increasing the nanoparticles
cases to evaluate the impact of the Al2O3 nanoparticle volume volume fraction without increasing the inlet velocity does not have
fraction and the inlet velocity on heat transfer enhancement in any significant effect on the temperature which is illustrated in
the non-slip and slip nanofluid flow regimes. The first case (I) is Fig. 7 third row (third case (III)).
done with a constant Reynolds number where the inlet velocity
increases with increasing f in order to keep Re constant. The
4.4. Nusselt number
second case (II) is done for the pure fluid (f ¼ 0) where only the
inlet velocity increases according to the used inlet velocity in the
Variations of the local Nusselt number in three cases are shown
first case (I) for different f. The third case (III) is done with
in Fig. 8. As it is reported by previous researcher, increasing the
a constant inlet velocity (uin ¼ 0.0502 m/s according to the used
nanoparticles volume fraction increases the local Nusselt number
inlet velocity for f ¼ 0 in the first case (I)) where only the
in the non-slip and slip flow regimes for the first case (I) at a given
nanoparticles volume fraction increases without keeping Re
Re. For the second case (II), increasing the inlet velocity also
constant.
increases the local Nusselt number as well as increasing f for the
Hence the major fluid flows in the microchannels are laminar
first case (I).
flow with low Reynolds numbers, for this study low Reynolds
With considering the third case (III) in Figs. 7 and 8 and Table 3 it
numbers (7
Re
15) are chosen. The non-slip and slip flow
is obvious that adding solid volume nanoparticles to the base fluid
simulations for these three cases have been also presented. The
increases the thermal conductivity and Prandtl number while the
nanoparticles volume fraction varies from 0 to 5%, because most
Reynolds number reduces. As a result, the Peclet number does not
published papers have also done in this range. The used values for
change significantly when the solid nanoparticles volume fraction
the inlet velocity and other calculated values of two f(f ¼ 0 and
increases. Consequently, increasing f without increasing the inlet
f ¼ 5%) for the first case (I), the second case (II) and the third case
velocity does not have any significant effect on the local Nusselt
(III) are presented in Table 2e4, respectively.
number at the third case (III). This behavior also have been
observed by Santra et al. [31] at low Reynolds number (Re<100) in
4.1. Velocity field
which the Copper (with very high thermal conductivity Ks ¼ 386)
nanoparticles volume fraction does not have any notable effect on
The velocity profiles for different nanoparticles volume fraction
the average Nusselt number (Fig. 7 in [31]) in isothermally heated
(f) at Re ¼ 10 and Kn ¼ 0.01 are shown in Fig. 5. The non-dimen-
parallel plates.
sional momentum equation (Eq. (6)) only depends on the Reynolds
number. Therefore, at a given Re the solid volume fraction does not
have any effect on the non-dimensional velocity which is illustrated 5. Conclusions
in Fig. 5a and b. However at a given Re, the inlet velocity should be
increased with increasing the solid volume fraction to keep Re The Reynolds number in nanofluids studies depends on the
constant. Consequently, the velocity in any cross section increase inlet velocity and kinematic viscosity of nanofluid. The nanofluid
which is presented in Fig. 5c and d. As it is shown in Table 2, the kinematic viscosity increases with an increasing in nanoparticles
kinematic viscosity (ynf ¼ mnf/rnf) of nanofluid increases with volume fraction while the inlet velocity should be increased to
increasing the nanoparticle volume fraction in a given Re, therefore keep the Reynolds number constant. Therefore, it is not clear
the inlet velocity should be increased. In the previous works which that either increasing the nanoparticles volume fraction or
were done with constant Reynolds numbers, it is not clear that the increasing the inlet velocity has a major role on heat transfer
reported effects of nanofluid on heat transfer and friction factor are enhancement in nanofluids studies which are done at constant
due to either the adding the solid nanoparticles to the base fluid or Reynolds numbers.
increasing the inlet velocity. Forced convection Al2O3-water nanofluid flows in two-dimen-
sional rectangular microchannels have been investigated to study
4.2. Skin friction factor heat transfer enhancement due to application of the nanoparticles
to the base fluid. Three different cases are examined to evaluate the
The local Poiseuille number at different f, Kn and Re is pre- proportion impact of increasing the nanoparticles volume fraction
sented in Fig. 6 for the first case (I) and second case (P). The (f) and the inlet velocity (uin) on heat transfer enhancement. Two-
nanoparticle volume fraction does not have any effect on the non- dimensional NaviereStokes and energy equations accompany with
dimensional velocity at a given Re (see Fig. 5a and b). Consequently, the slip velocity and the jump temperature boundary conditions
it does not have any effect on the Poiseuille number in the micro or expressions have been discretized using the finite volume tech-
macro (Kn ¼ 0) scale for the first case (I) as depicted in Fig. 6a and b, nique. The Brownian motions of nanoparticles have been consid-
respectively. The Fig. 6c and d show that for the second case (P) ered to determine the thermal conductivity of nanofluids. The
because of small variation in the Reynolds numbers, fRe does not following conclusions are obtained:
change significantly when the nanoparticles volume fraction
increases.  Results clearly show that at a given Reynolds number (Re), the
non-dimensional velocity is not affected by the volume fraction
4.3. Temperature filed of nanoparticles but the inlet velocity should be increased to
keep Re constant.
The non-dimensional temperature and non-dimensional bulk  The nanoparticles concentration has not any notable effect on
temperature for three cases are presented in Fig. 7. Increasing f for the Poiseuille number (fRe) at any Re.
 A. Akbarinia et al. / Applied Thermal Engineering 31 (2011) 556e565 565

 At a given Reynolds number, the temperature decreasing is not [9] S.M.S. Murshed, K.C. Leong, C. Yang, Thermophysical and electrokinetic
properties of nanofluids e a critical review, Appl. Therm. Eng. 28 (2008)
due to increasing the nanoparticles concentration, but it is due
2109e2125.
to increasing the inlet velocity. [10] A. Akbarinia, Impacts of nanofluid flow on skin friction factor and Nusselt
 It is shown that at a given Re, the major enhancement on the number in curved tubes with constant mass flow, Int. J. Heat Fluid Flow 29 (1)
Nusselt number is not due to increasing the nanoparticle (2008) 229e241.
[11] C.T. Nguyen, F. Desgranges, N. Galanis, G. Roy, T. Maré, S. Boucher, H. Angue
concentrations but it is due to increasing the inlet velocity to Mintsa, Viscosity data for Al2O3ewater nanofluiddhysteresis: is heat transfer
reach a constant Re. Increasing the nanoparticle concentration enhancement using nanofluids reliable? Int. J. Therm. Sci. 47 (2008) 103e111.
does not change the Peclet number significantly, therefore it [12] J. Koo, C. Kleinstreuer, Laminar nanofluid flow in microheat-sinks, Int. J. Heat
Mass Transf. 48 (13) (2005) 2652e2661.
does not have any significant effect on the heat transfer at [13] S.P. Jang, S.U.S. Choi, Cooling performance of a microchannel heat sink with
a constant inlet velocity. nanofluids, Appl. Therm. Eng. 26 (17e18) (2006) 2457e2463.
 Increasing the nanoparticles volume fraction increases density [14] R. Chein, G. Huang, Analysis of microchannel heat sink performance using
nanofluids, Appl. Therm. Eng. 25 (17e18) (2005) 3104e3114.
and dynamics viscosity which could increase pumping power. [15] H. Abbassi, C. Aghanajafi, Evaluation of heat transfer augmentation in
On the other hand, increasing inlet velocity increases pumping a nanofluid-cooled microchannel heat sink, J. Fusion Energy 25 (3e4) (2006)
power at a given Re. It has been shown that the same heat 187e196.
[16] T.H. Tsai, R. Chein, Performance analysis of nanofluid-cooled microchannel
transfer enhancement could be reached only with the same
heat sinks, Int. J. Heat Fluid Flow 28 (5) (2007) 1013e1026.
increasing in inlet velocity without employing nanofluids. [17] J. Li, C. Kleinstreuer, Thermal performance of nanofluid flow in microchannels,
 Constant Reynolds number studies of nanofluids are not Int. J. Heat Fluid Flow 29 (4) (2008) 1221e1232.
[18] R. Chein, J. Chuang, Experimental microchannel heat sink performance studies
sufficient to evaluate the heat transfer and the skin friction
using nanofluids, Int. J. Thermal Sci. 46 (1) (2007) 57e66.
factor due to the nanofluids usage. [19] J.Y. Jung, H.S. Oh, H.Y. Kwak, Forced convective heat transfer of nanofluids in
microchannels, Int. J. Heat Mass Transf. 52 (1e2) (2009) 466e472.
Further numerical and experimental investigations are needed [20] C.J. Ho, L.C. Wei, Z.W. Li, An experimental investigation of forced convective
cooling performance of a microchannel heat sink with Al2O3-water nanofluid,
to develop an appropriate method such as constant pumping Appl. Therm. Eng. 30 (2e3) (2010) 96e103.
powers for investigating heat transfer due to the nanofluids [21] S.E.B. Maiga, C.T. Nguyen, N. Galanis, G. Roy, Heat transfer behaviors of
application which could clearly answer the question “Is nanofluids nanofluids in a uniformly heated tube, Superlatt. Microstruct. 35 (2004)
543e557.
application useful and economic?” [22] C.H. Chon, K.D. Kihm, S.P. Lee, S.U.S. Choi, Empirical correlation finding the
role of temperature and particle size for nanofluid (Al2O3) thermal conduc-
References tivity enhancement, Appl. Phys. Lett. 87 (2005) 1e3.
[23] Derek C. Tretheway, Carl D. Meinhart, Apparent fluid slip at hydrophobic
microchannel walls’, Phys. Fluids 14 (3) (2002) L9eL12.
[1] M. Gad-el-Hak, The fluid mechanics of microdevices: the Freeman scholar
[24] Guyh Dituba Ngoma, Fouad Erchiqui, Heat flux and slip effects on liquid flow
lecture, J. Fluids Eng. 121 (1999) 5e33.
in a microchannel, Int. J. Therm Sci 46 (2007) 1076e1083.
[2] A. Beskok, G.E. Karniadakis, Simulation of heat and momentum transfer in
[25] Jun Yang, Daneil Y. Kwok, Effect of liquid slip in elctrokinitic parallel plate
complex micro-geometries, J. Thermophys. Heat Transf. 8 (4) (1994) 355e370.
microchannel flow, J. Colloid Interface Sci. 260 (2003) 225e233.
[3] F.E. Larrodé, C. Housiadas, Y. Drossinos, Slip-flow heat transfer in circular
[26] Jean louis Barrat, Lyderic Bocquet, Large slip effect at a nonwetting fluid-solid
tubes, Int. J. Heat Mass Transf. 43 (2000) 2669e2680.
interface, Phys. Rev. Lett. 82 (23) (1999) 4671e4674.
[4] W. Sun, S. Kakac, A.G. Yazicioglu, A numerical study of single-phase convective
[27] Mohamed Gad-el-Hak, The MEMS Handbook, MEMS: Introduction and
heat transfer in microtubes for slip flow, Int. J. Therm. Sci. 46 (2007)
Fundamentals, second ed. Taylor & Francis Group, 2006.
1084e1094.
[28] S.V. Patankar, Numerical Heat Transfer and Fluid Flow. Hemisphere, Wash-
[5] B.C. Pak, Y.I. Cho, Hydrodynamic and heat transfer study of dispersed fluids
ington, 1980.
with submicron metallic oxide particles, Exp. Heat Transf. 11 (1998) 151e170.
[29] M. Renksizbulut, H. Niazmand, G. Tercan, Slip-flow and heat transfer in
[6] Y.M. Xuan, Q. Li, Investigation on convective heat transfer and flow features of
rectangular microchannels with constant wall temperature, Int. J. Therm. Sci.
nanofluids, J. Heat Transfer 125 (2003) 151e155.
45 (9) (2006) 870e881.
[7] A. Akbarinia, A. Behzadmehr, Numerical study of laminar mixed convection of
[30] S. Yu, T.A. Ameel, Slip flow heat transfer in rectangular microchannels, Int.
a nanofluid in horizontal curved tubes, Appl. Therm. Eng. 27 (8e9) (2007)
J. Heat Mass Transf. 44 (2001) 4225e4234.
1327e1337.
[31] A.K. Santra, S. Sen, N. Chakraborty, Study of heat transfer due to laminar flow
[8] A. Akbarinia, R. Laur, Investigation the diameter of solid particles affects on
of copper-water nanofluid through two isothermally heated parallel plates,
a laminar nanofluid flow in a curved tube using a two phase approach, Int.
Int. J. Therm. Sci. 48 (2) (2009) 391e400.
J. Heat Fluid Flow 30 (4) (2009) 706e714.