P044088085

M. Dhananjaya et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 4( Version 8), April 2014, pp.80-85
www.ijera.com 80 | P a g e
Onset of Convection in a Nanofluid Saturated Porous Layer with
Temperature Dependent Viscosity
I. S. Shivakumara*, M. Dhananjaya**
* (Department of Mathematics, Bangalore University, Bangalore - 560 001, India
Email: shivakumarais@gmail.com)
** (Department of Mathematics, Bangalore Institute of Technology, Bangalore - 560 004, India
Email: darahulm@gmail.com)
ABSTRACT
The effect of nanofluid viscosity varying exponentially with temperature on the onset of convection in a layer of
nanofluid saturated Darcy porous medium is investigated. The nanoparticle flux is zero condition on the
boundaries is invoked to account for physically realistic situation. The resulting eigenvalue problem is solved
numerically using the Galerkin method. It is observed that the instability sets in only as stationary convection
and the occurrence of oscillatory convection is ruled out. The effect of viscosity parameter on the characteristics
of stability is found to be significant and dual in nature. The onset of convection is hastened and the size of
convection cells is enlarged with an increase in the value of modified diffusivity ratio, concentration Darcy-
Rayleigh number, the modified particle density increment parameter and the Lewis number.
Keywords: nanofluid, porous medium, variable viscosity, Galerkin method, thermal convection
I. Introduction
Nanofluids are colloidal suspensions of
nanometer-sized particles in base fluids that show
great potential applications in many fields. The
developments in this field are reviewed by several
authors and well documented in the literature ( Kakac
and Pramuanjaroenkij [1], Yu and Xie [2],
Goharshadi et al. [3] ). Nanofluids are found to have
high thermal conductivity at very low nanoparticles
concentration and considerable enhancement of
convective heat transfer. As a result, buoyancy
driven convection in a nanofluid layer has been
studied extensively (Kim et al. [4] and references
therein). Its counterpart in a nanofluid saturated
horizontal layer of porous medium has also attracted
equal importance in the literature because of its
importance in many fields of modern science,
engineering and technology, chemical and nuclear
industries and bio-mechanics. The problem was first
considered by Nield and Kuznetsov [5] and following
this formalism several studies have been undertaken
subsequently to investigate various additional effects
on the problem by the same authors and others. The
details can be found in the monograph of Nield and
Bejan [6].
It is imperative to note that thermal
convective instability in nanofluids is affected
significantly by their properties and specifically by
nanofluid viscosity and thermal conductivity.
Copious literature available on thermal convection in
nanofluids is mainly concerned with the heat transfer
properties such as heat transfer enhancement, thermal
conductivity measurement, improvement and
enhancement, estimation of thermal conductivity and
so on [7-11]. Nonetheless, the viscosity is an
important property to be considered for all thermal
applications involving nanofluids as it describes the
internal resistance of a fluid to the flow. Moreover,
the pumping power and convective heat transfer
coefficient are influenced by viscosity. Hence,
viscosity is as important as thermal conductivity in
engineering systems involving nanofluid flow.
Experiments have been conducted extensively in the
past to measure the effect of temperature on
nanofluid viscosity and it is observed that the
viscosity decreases with increasing temperature. In
particular, it has been shown that viscosity of
CuO/EG-water nanofluid, CuO/PG-water nanofluid,
Al2O3/EG-water nanofluid and SiO2/EG-water
nanofluid decreases exponentially with the increasing
temperature in different temperature ranges [12-14].
Mahbubul et al. [15] have thoroughly compiled and
reviewed different characteristics of viscosity of
nanofluids.
Under the circumstances, considering
variation in viscosity of the nanofluid with
temperature becomes equally important in order to
predict the correct contribution of nanoparticles on
thermal convective instability in a nanofluid saturated
porous layer. Besides, all the studies on the onset of
convection in nanofluid porous layer are based on
Buongiorno [7] model and it is assumed that the
value of the nanoparticle fraction at the boundaries
can be prescribed in the same way as the temperature.
RESEARCH ARTICLE OPEN ACCESS

But the assumption of boundary conditions
controlling the nanoparticle volume fraction on the
boundaries may be difficult in practice and they are
to be amended. Recently, Nield and Kuznetsov [16]
have commented on the authenticity of these
boundary conditions and argued that vanishing of
nanoparticle flux condition on the boundaries is more
appropriate to account for physically realistic
situation.
The intent of the present study is, therefore,
to investigate the onset of convection in a nanofluid
saturated layer of porous medium by considering
nanofluid viscosity depending exponentially on
temperature and imposing physically realistic
boundary conditions on nanoparticle volume fraction,
that is nanoparticle flux is zero on the boundaries.
The eigenvalue problem is solved numerically using
higher order Galerkin method to obtain more accurate
results and also to account for the influence of all the
parameters on the onset of convection. A detailed
parametric study is under taken and observed that
oscillatory convection is not a preferred mode of
instability.
II. Mathematical formulation
We consider a nanofluid saturated horizontal
porous layer of thickness d as shown in Fig. 1. The
lower and upper surfaces are held at constant but
different temperatures 0T and 1T ( 0T ), respectively
and the model considered incorporates the effects of
both Brownian motion and thermophoresis. A
Cartesian coordinate system  , ,x y z is chosen such
that the origin is at the bottom of the porous layer and
the z -axis vertically upward in the presence of
gravitational field.
The viscosity of the nanofluid  is assumed
to vary exponentially with temperature in the form
 0 exp aT T      (1)
where, 0 ,  are positive constants
and 0 1( ) / 2aT T T  is the average temperature.
The governing equations are:
0q 

(2)
f q
p g q
t K
 



   


 
(3)
     
 
2
m f
T
Bp
a
T
c c q T k T
t
D
c D T T T
T
 
  

   

 
      
 

(4)
  2 21 T
B
a
D
q D T
t T

     


 

(5)
    1 1fp aT T         (6)
where,  , ,q u v w

the velocity vector, p the
pressure,  the overall density of the nanofluid, p
the density of nanoparticles, f the density of base
fluid, g

the gravitational acceleration,  the
nanoparticle volume fraction, K the permeability of
the porous medium,  the porosity of the porous
medium, T the temperature of the nanofluid,  the
coefficient of thermal expansion, c the specific heat,
k the thermal conductivity, BD the Brownian
diffusion coefficient and TD the thermophoretic
diffusion coefficient.
Fig. 1: Physical configuration
The basic state is quiescent and in the
dimensionless form it is found to be
1 2bT z  , 0b AN z   (7)
where,  0 1 0/A T B aN D T T D T  is the modified
diffusivity ratio and 0 is the reference value of
nanoparticle volume fraction. The pressure is of no
consequence here as it will be eliminated
subsequently.
To study the stability of the system, the
basic state is perturbed in the form q q
 
,
( )p p z p  , bT T T  and b    , where
q

, p , Tand  are assumed to be small
compared to their equilibrium counterparts.
Following the standard linear stability analysis
procedure as outlined in [16], the dimensionless
stability equations can then be shown to be
 
 
 
2 2
2 2
2 2
( )
1/ 2 t n
D a W
Pr
f z D D a W
R a R a

 
    
 
  
(8)
 2 2 2
2 BN
M W D a D
Le
      (9)

 
 
2 2
2 2
2
1
A
A
N
N W D a
Le
D a
Le
     
  
(10)
where, D d dz , )( r ii   is the growth
rate, 2 2
a l m  is the overall horizontal wave
number,W is the amplitude of the perturbed vertical
velocity,  is the amplitude of perturbed
temperature,  is the amplitude of perturbed
nanoparticle volume fraction and
 ( ) exp 1/ 2f z z     (11)
In the above equations,
 0 1 0/t fR dK T T g    is the thermal Darcy -
Rayleigh number,   0 0/n p fR Kd g      is the
concentration Darcy - Rayleigh number, / BLe D 
is the Lewis number,    0 /B p f
N c c   is the
modified particle density increment, T   is the
viscosity parameter, 2 2
0 / fPr d K    is the
Darcy-Prandtl number and ( / )M   is the heat
capacity ratio.
The isothermal boundaries are impermeable and the
nanoparticle flux is zero on the boundaries. The
appropriate boundary conditions are:
0, 0,
0A
W
D N D
  
   
at 0,1z  . (12)
III. Numerical solution
Equations (8) - (10) constitute an eigenvalue
problem and solved numerically by the Galerkin
method. The variables are written in a series of basis
functions as
       
   
1 1
1
, ,
N N
i i i i
i i
N
i i
i
W z AW z z B z
z C z
 

   
  
 

(13)
where, , andi i iA B C are unknown coefficients. The
basis functions are represented by the power series
satisfying the respective boundary conditions
     
   
 
2
1
1 , 1,2,3,4...
,
( / ) for 2,3,4...
i
i i
A
i
i A
W z z z z i
z N z z
z N z i i
   
  
  
(14)
Multiplying Eq.(8) by  jW z , Eq.(9) by
 j z and Eq.(10) by  j z ; performing the
integration by parts with respect to z between
0z  and1 , and using the boundary conditions, we
obtain the following system of algebraic equations:
.
i ji i ji i ji i ji
i ji i ji i ji i ji
i ji i ji i ji i ji
AC B D C E A F
AG B H C I B J
A K B L C M C N



  
  
  
(15)
The coefficients ji jiC N involve the inner products
of the basis functions and are given by
2 2
( ) ( )
( )
ji j i j i
j i
C f z W D W a f z W W
f z W DW
  
 
2
(1/ 2)ji j iD a Rt W   , 2
ji j iE a Rn W 
 21
Pr
ji j i j iF DW DW a W W   ,
2ji j iG W  , 2
ji j i j iH D D a       ,
2 B
ji j i
N
I D
Le
    , ji j iJ M   ,
ji A j iK N W   ,
 2
2
A
ji j i j i
N
L D D a
Le
       ,
 21
ji j i j iM D D a
Le
       ,
ji j iN    .
The system of equations given by Eq. (15) is
a generalized eigenvalue problem which can be
written in the form
1 2X X   (16)
where, 1 and 2 are real matrices of order
N N and X is the eigenvector. By using the
subroutine GVLRG of the IMSL library, the complex
eigenvalue is determined when the other
parameters are specified. Then one of the parameters,
say tR , is varied until the real part of vanishes. The
zero crossing of real part of  is achieved by
Newton’s method for fixed point determination. The
corresponding value of tR and a are the critical
conditions for neutral stability. Then the critical
Rayleigh number with respect to the wave number is
calculated using the golden section search method.

The imaginary part of  indicates whether the
instability onsets into steady convection or into
growing oscillations.
IV. Results and discussion
The effect of exponential variation in
nanofluid viscosity with temperature on the onset of
convection in a nanofluid saturated Darcy porous
medium is investigated by considering physically
more realistic boundary conditions for the
nanoparticle fraction. The resulting eigenvalue
problem is solved numerically by employing
Galerkin method. The parametric values vary with
the base fluid and nanoparticles chosen. The ratio of
density of the nanoparticles to that of a base fluid for
Cu (copper) and Ag (silver) is 8.96 and 10.5,
respectively. The ratio of heat capacity based on the
volume fraction of nanoparticles to that of a base
fluid is 0.83 for Cu and 0.59 for Ag. For such
nanofluids, nR turns out to be of the order 1 10
and BN of the order 3
10 1
10
. The value of
modified diffusivity ratio is not more than 10 and Le
is taken in the order of 1 10.
Table 1: Comparison of tcR for different orders of
approximations in the Galerkin expansion
for 2,AN  2nR  , 0.059BN  and 1Le 
N
tcR
0  3  5  8 
1 32.0000 50.2214 91.8729 267.3621
2 30.1340 30.1562 33.8914 53.1522
3 29.7876 28.2585 23.6465 19.2704
4 29.7004 28.1578 22.3523 11.0813
5 29.6998 28.0979 22.2734 9.1550
6 29.6998 28.0914 22.2237 8.9022
7 28.0912 22.2104 8.8728
8 28.0912 22.2092 8.8448
9 22.2092 8.8343
10 8.8343
It is observed that oscillatory convection is
not a preferred mode of instability for all the
parametric values chosen. To have a check on the
accuracy of the numerical procedure adopted, it is
instructive to look at the results for various levels of
the Galerkin approximation to know the process of
convergence and also the accuracy of the results.
Table 1 shows the numerically computed values of
critical Darcy-Rayleigh number tcR for various
values of viscosity parameter  and for different
orders of approximations in the Galerkin expansion
when 2AN  , 2nR  , 1Le  and 0.059BN  .
From the tabulated values, it is clear that the
convergence of the results crucially depends on the
value of  . More number of terms in the Galerkin
expansion is required to get the convergent results
with increasing  . When 0 AN   , the well-
known value of the critical Darcy-Rayleigh number
39.4784 for ordinary fluids is recovered.
Figures 2 (a, b, c, d) show the neutral
stability curves in the ( , )tR a -plane for different
values of AN (Fig. 2a), nR (Fig. 2b), Le (Fig. 2c)
and  (Fig. 2d). The neutral curves are similar to
those observed in the Darcy-Bénard problem for
ordinary viscous fluids and exhibit single minimum
with respect to the wave number. It is observed that
an increase in the value of AN , nR , Le and  is to
decrease the region of stability and hence their effect
is to reinforce instability on the system. In Fig. 2(a),
the result for 0AN  corresponds to that of regular
fluid.
2 4 6 8
0
10
20
30
40
50
60
2 4 6 8
0
10
20
30
40
50
60
7
5
3
1
NA = 0
a
Rt
(a)
3
7
5
Rn = 1
a
Rt
(b)
2 4 6 8
5
10
15
20
25
30
35
40
2 4 6 8
0
10
20
30
40
50
60
5
3
Le = 1
a
Rt
(c)
8
6
4
2
a
Rt
 = 0
(d)
Fig. 2: Neutral curves for different values of
(a) AN with 2,nR  0.059,BN  1Le  and 4  (b)
nR with 2,AN  0.059,BN  1,Le  4  (c) Le
with 2,nR  2,AN  0.059,BN  4  (d) 
with 1Le  , 1nR  , 4,AN  0.059BN 
The influence of various physical
parameters on the onset of convection is exhibited in
Figs. 3-6. The variations of critical Darcy-Rayleigh
number tcR and the corresponding critical wave
number ca as a function of  for different values

of AN are shown in Figs. 3(a) and (b), respectively
for 1Le  , 0.059BN  and 2nR  . From the figures
it is evident that increase in the value of AN is to
decrease tcR and increase ca .
0 2 4 6 8
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
0 2 4 6 8
0
8
16
24
32
40

ac
NA = 0, 1, 2, 3, 4
(b)
20.83
25.42
30.11
34.98
39.63
(Rt c)max
0.1
1.38
1.1
0.38
0
4
3
2
1
NA
4
3
1
2
NA = 0

Rtc
0.11
(a)
Fig. 3: Variation of (a) tcR , (b) ca with  for different
values of AN when 1Le  , 0.059BN  and 2nR 
0 2 4 6 8
0
6
12
18
24
30
36
0 2 4 6 8
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
( Rt c )max
29.98
20.96
25.50
34.87
0.38
1.5
1.0
0.1
4
3
2
1
Rn
4
3
2
Rn = 1

Rtc
(a)
Rn = 1, 2, 3, 4

ac
(b)
Fig. 4: Variation of (a) tcR (b) ca with  for different
values of nR when 1Le  , 2AN  and 0.059BN  .
Hence, the effect of increasing AN is to
reinforce instability on the system and also to enlarge
the size of convection cells. The result for 0AN 
corresponds to regular fluids and it is observed that
higher heating is required to instil instability in this
case indicating the presence of nanoparticles is to
augment heat transfer and to hasten the onset of
convection due to diffusion of nanoparticles by
thermophoresis and Brownian motion. Although
thermophoresis and Brownian motion are responsible
for the motion of nanoparticles in the base fluid, it is
observed that thermophoresis is more dominating
effect to initiate the diffusion of nanoparticles and
hence increase in the value of AN is to hasten the
onset of convection. The results obtained for different
values of concentration Darcy-Rayleigh number
nR for 1Le  , 2AN  and 0.059BN  exhibit
similar behaviour on the stability characteristics of
the system and the same is evident from Figs. 4(a)
and (b). That is, increase in the nanoparticle
concentration in the base fluid is to hasten the onset
of convection.
0 1 2 3 4
3.0
3.1
3.2
3.3
3.4
3.5
3.6
3.7
0 1 2 3 4
27
28
29
30
NB = 0.001, 0.01, 0.05, 0.1

ac
(b)
( Rt c
)max
29.92
30.08
30.22
30.25

0.71
0.8
0.65
0.6
0.1
0.05
0.01
0.001
NB

Rtc
NB = 0.001, 0.01, 0.05, 0.1
(a)
values of BN when 2AN  , 2nR  and 1Le 
0 2 4 6 8
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
0 2 4 6 8
5
10
15
20
25
30
35
Le = 1, 2, 3, 4
ac

(b)
( Rt c
)max








4
2
3
1
Le
4
3
2
Le = 1
Rtc

(a)
values of Le when 0.059BN  , 2AN  and 1nR 
The modified particle density increment
parameter BN does influence the stability
characteristics of the system with higher number of
terms in the Galerkin expansion. Its effect is found to
be missing if single term is considered in the
Galerkin expansion. This suggests the limitation of
single term Galerkin method in studying the problem.
The effect of increasing BN is to hasten the onset of
convection and also to increase the size of convection
cells and this is evident from Figs. 5(a) and (b),
respectively. The results shown in these figures are
for 2AN  , 2nR  and 1Le  . Figures 6(a) and (b)
represent the variation of tcR and ca as a function of
 for different values of ,Le respectively
when 0.059BN  , 2AN  and 1nR  . From the
figures it is noted that the effect of increasing Le is
to reinforce instability on the system and also to
enlarge the size of convection cells.

A closer inspection of the above figures also
reveals that the viscosity parameter  has a strong
influence on the onset of convection in a nanofluid
saturated porous layer. For different values of
AN (Fig. 3a), nR (Fig. 4a), BN (Fig. 5a) and Le (Fig.
6a), two regions are distinguished. With increasing  ,
tcR increases initially up to a certain peak value,
depending on the parametric values, and then
decreases rapidly. The values of  at which
tcR attains its maximum value max( )tcR are computed
numerically and shown in the respective figures for
the parametric values considered. It is seen that
max( )tcR decreases while  increases with
increasing AN , nR , BN and Le . At the maximum
value of tcR , a sublayer starts to form and once the
viscosity parameter is further increased the thickness
of the sublayer as well as tcR are reduced.
Thus  displays a dual role on the onset of
convection.
V. Conclusions
The influence of nanofluid viscosity varying
exponentially with temperature on the onset of
convection is investigated. The eigenvalue problem is
solved numerically using the Galerkin method by
considering vanishing of nanoparticle flux boundary
conditions at the boundaries. The onset of convection
is found to occur only via stationary mode and the
characteristics of stability of the system are strongly
dependent on the viscosity parameter  . The effect of
increasing concentration Darcy - Rayleigh
number ,nR modified diffusivity ratio AN , modified
particle-density increment BN and the Lewis
number Leis to hasten the onset of convection and
also to enlarge the size of convection cells. The
viscosity parameter  exhibits a dual effect on the
onset of convection depending on the values of AN ,
nR , BN and Le. It shows stabilizing effect on the
system initially but displays a reverse trend once
 exceeds a threshold value depending on the other
parametric values.
VI. Acknowledgement
One of the authors (M.D) wishes to thank
the Head of the Department and the Management of
his college for their encouragement and support.
References
[1] S. Kakac and A. Pramuanjaroenkij, Review
of convective heat transfer enhancement
with nanofluids. Int. J. Heat Mass Transf.
52, 2009, 3187-3196.
[2] W. Yu and H. Xie, A Review on
Nanofluids: Preparation, Stability
Mechanisms, and Applications. J.
Nanomaterials. 2012, 1-17.
[3] E. K. Goharshadi, H. Ahmadzadeh, S.
Samiee and M. Hadadian, Nanofluids for
heat transfer enhancement-a review, Phys.
Chem. Res. 1, 2013, 1-33.
[4] J. Kim, Y. T. Kang and C. K. Choi, Analysis
of convective instability and heat transfer
characteristics of nanofluids, Phys.
Fluids16, 2004, 2395-2401.
[5] D. A. Nield and A. V. Kuznetsov, Thermal
instability in a porous medium layer
saturated by a nanofluid, Int. J. Heat Mass
Trans.52, 2009, 5796-5801.
[6] D. A. Nield and A. Bejan, Convection in
porous media (Springer. 4th ed., 2013).
[7] J. Buongiorno, Convective transport in
nanofluids, ASME J. Heat Transf.128, 2006,
240-250.
[8] P. Vadasz, Heat conduction in nanofluid
suspensions, ASME J. Heat Transf.128,
2006, 465-477.
[9] D. Y. Tzou, Thermal instability of
nanofluids in natural convection, Int. J. Heat
Mass Transf.51, 2008, 2967-2979.
[10] K. Khanafer and K. Vafai, A critical
synthesis of thermophysical characteristics
of nanofluids, Int. J. Heat Mass Transf.54,
2011, 4410-4428.
[11] K. B. Anoop, T. Sundararajan and Sarit K.
Das, Effect of particle size on the convective
heat transfer in nanofluid in the developing
region, Int. J. Heat Mass Transf. 52, 2009,
2189-2195.
[12] P. Namburu, D. Kulkarni, D. Misra and D.
Das, Viscosity of copper oxide nanoparticles
dispersed in ethylene glycol and water
mixture, Exp. Therm.Fluid Sci. 32 (2), 2007,
397-402.
[13] D. P. Kulkarni, D. K. Das and R. S. Vajjha,
Application of nanofluids in heating
buildings and reducing pollution, Appl.
Energy 86 (12), 2009, 2566-2573.
[14] M. T. Naik, G. R. Janardhana, K. V. K.
Reddy and B. S. Reddy, Experimental
investigation into rheological property of
copper oxide nanoparticles suspended in
propylene glycol- water based fluids, ARPN
J. Eng. Appl. Sci.5(6), 2010, 29-34.
[15] I. M. Mahbubul, R. Saidur and M. A.
Amalina, Latest developments on the
viscosity of nanofluids, Int. J. Heat Mass
Trans. 55, 2012, 874-885.
[16] D. A. Nield and A.V. Kuznetsov, Thermal
instability in a porous medium layer
saturated by a nanofluid: A revised model,
Int. J. Heat Mass Trans.68, 2014, 211-21.

P044088085

More Related Content

What's hot (15)

Viewers also liked (20)

Similar to P044088085 (20)

Recently uploaded (20)

P044088085