Finite Element Analysis of Convective Heat And Mass Transfer Flow of a Viscous Fluid through a Porous
Medium with Heat Generating Source and Quadratic Density Temperature Variation
Y. MADHUSUDHANA REDDY 1 and D. R. V. PRASADA RAO 2
1 Associate Professor, Department of Mathematics, Sri Venkateswra Institute of Technology, Anantapur, INDIA.
2 Professor, Department of Mathematics, S. K. University, Anantapur, INDIA.
(Received on: April 15, 2012) ABSTRACT
In this analysis we investigate the effect of quadratic density temperature variation on convective heat and mass transfer flow of a viscous fluid through a porous medium in vertical channel with heat generating source. By using galarkin finite element analysis the equations governing flow heat and mass transfer are solved with three noded line segments. The effect of various fluid sources on velocity, temperature concentration is analyzed. The rate of heat and mass transfer on the boundaries are evaluated numerically.
Keywords: Vertical Channel, Nondarcy , Finite element analysis, Porous Medium, Heat and Mass Transfer.
1. INTRODUCTION
Flow and heat transfer in porous medium has been attracting the attention of an increasingly large number of investigators in recent years. The need for fundamental studies in porous media heat transfer stems from the fact that a better under standing of a host of thermal engineering applications in which porous
materials present is required. The
accumulated impact of these studies is two
fold, first to improve the performance of
existing porous media-related thermal
systems, second is to generate new ideas and
explore new avenues with respect to the use
of porous media in heat transfer
applications. Some examples of thermal
engineering disciplines which stand to
benefit from a better understanding of heat
and fluid flow process through porous materials are geothermal systems, thermal insulations, grain storage, solid matrix heat exchangers, oil extractions and the manufacturing of numerous products in the chemical industry. Darcy’s equation give satisfactory results for closely packed porous medium but does not explain the flow through sparsely distributed porous medium in later situation, Brinkman 5 proposed an alternate model by adding a term which accounts for the viscous shear in addition to the Darcy’s equation. The first theoretical investigation of natural convection in porous enclosure using Brinkman model was made by Chan et al 6 . Later on a series of investigations were carried out using the Brinkman model by a few authors notably Poulikakos and Bejan 14 .
Convective flows driven by temperature and concentration differences have been studied extensively in the fast and various extensions of the problem have been reported in the literature with both temperature and concentration interacting simultaneously, the convection has become quite complex. Combined heat and mass transfer along a vertical plate in natural convection flows have been studied in great detail. Bejan and Khair 4 have investigated the vertical free convection boundary layer flow embedded in a porous medium resulting from the combined heat and mass transfer. Similarity solutions for buoyancy induced heat and mass transfer for vertical surfaces in porous media were presented.
Lai and Kulacki 11 have used an integral method to solve the problem of Bejan and Khair 4 . Nelson and Wood 13 has numerically and analytically investigated the combined heat and mass transfer for both boundary
conditions. Mamou 12 et al have analyzed the problem of thermosolutal convection in a rectan cell filled with a brinkman porous medium saturated by a binary fluid.
Sugunama 18 and Ravindranath reddy 15 have analyzed the free convective heat transfer in vertical channel taking dissipative terms.
Sulochana 19 has analyzed the convective heat and mass transfer through a porous medium confined in a vertical channel with viscous and Darcy dissipations.
In all the above investigations, the variation of density is taken in the linear form
( ) ∆ T
−
=
∆ ρ ρβ (1.1)
Where β is the co-efficient of thermal expansion and is 2,07 x 10 4 (OC) -1+ . This is valid for temperature variation near 20 0 c.
But this analysis is not applicable to the study of the flow of water at 4 0 c, the density of water is a maximum at atmospheric pressure and the above relations (1.1) does not hold good. The modified form of (1.1) applicable to water at 4 0 c is given by
( ) ∆ T 2
−
=
∆ ρ ργ (1.2)
Where γ = 8 x 10 -6 (OC) -2 . Taking this fact
into account, Goren showed in this case,
similarity solutions for the fee convection
flow of water at 4 0 c past a semi-infinite
vertical plate exist. An approximate solution
for velocity and temperature has been
obtained by using Karman – Pohlhausen
method together with the method of finding
similarity solution. Following a quadratic
density temperature variation applebaum 3
have discussed the laminar free convection
flow through coaxial circular cylinders with and without heat sources. Taking non-linear density temperature variation Sarojamma 16a has analyzed the hydro-magnetic free convection flow in a cylindrical geometry.
Vajravelu 19a have solved the problem of free convection between vertical walls by taking the non-linear density temperature variation, viz.
( ) 1 ( ) 2
0 g T − T e − T − T e
−
=
∆ ρ ρβ ρβ (1.3)
Where β 0 and β 1 are the constants.
Bhargawa and Agarwal 2a have investigated the fully developed laminar free convection flow in the presence of constant heat sources in a circular pipe taking the same density temperature relationship (1.3). It is found that the flow and heat transfer both depend upo a new parameter
∆ T
=
0 1
β β γ
in addition to the heat source parameter and free convection parameter k.
Keeping the applications in view Jaffer 9 and Venkataramana 20 have studied the effect of Quadratic density temperature variation on convective heat transfer in vertical channel. Recently Alivene 1 and Rao 16 have studied the effect of quadratic density temperature variation on convective heat and mass transfer through a porous medium in vertical channel under varied conditions.
In this paper we investigate the convective heat and mass transfer through a porous medium confined in a vertical channel in the presence of heat generating sources with quadratic density temperature
variations. The equations governing the velocity, energy, and diffusion are non- linear coupled. By employing Galerkin finite element analysis with quadratic interpolation functions the equations are solved. The velocity, temperature, and concentration distributions are analyzed for different variations of the parameters , viz. Gm, Gc, D -1 , S, Sc, and α. The rate of heat transfer and mass transfer are evaluated numerically for different variations of the parameters.
2. FORMULATION OF THE PROBLEM We consider the convective heat and mass transfer flow of a viscous, incompressible, non-conducting, fluid through a Darcy, isotropic, homogenous porous medium. The x-axis is directed along the vertical surface and the y-axis is transverse to this. Both the vertical surface and the fluid are maintained initially at the same temperature and concentration.
Instantaneously they are raised to a temperature Tw and concentration Cw which remain unchanged. The walls are situated at y = ± L. Assuming the Concentration at a low level the soret and Doufour effects are neglected. Incorporating viscous heating effects, wall mass flux and buoyancy, under the Boussinesque approximation, the boundary layer equations may be presented as follows:
v 0
∂ = y
∂ (2.1)
( ) ( )
2 2
2 e * e
p
u u
v g T T g C C u
y y K
υ β β υ
∂ = ∂ + − + − −
∂ ∂
(2.2)
2
2 ( e )
T T
v Q T T
y λ y
∂ = ∂ − −
∂ ∂ (2.3)
2
m 2
C C
v D
y y
∂ = ∂
∂ ∂ (2.4)
where u is the axial velocity in the porous region, T & C are the temperature and concentrations of the fluid, k is the permeability of porous medium, g is the acceleration due to gravity, K p is the permeability of porous medium , Q is the hear source, υ is the kinematic viscosity , ρ is the dencity of the fluid, α is the thermal diffusivity , D m is the coefficient of mass diffusitivity, C s is the concentration susceptibility, C p is the specific heat, β is the coefficient of thermal expansion, β * is the coefficient of volume expansion.
The boundary conditions are
y = 0; u = 0, T = T 1 , C = C 1 , (2.5) y = 1, u = 0 ,
T = T 2, C = C 2.
We introduce the non-dimentional variables as
(x 1 , y 1 ) = (x, y)/L
2
1 2
T T T T θ = −
− 1 2 2
C C C C φ = −
−
The conservation equations are now transformed into the following system of coupled, non-linear ordinary differential equations of f, θ, φ.
1 2
yy y m c
u + Su − D u − = G θ + G φ
(2.6)
Pr 0
yy S y
θ + θ αθ − =
(2.7)
= 0
+ y
yy SSc φ
φ (2.8)
The corresponding boundary conditions are u = 0, θ = 1, φ = 1 at y = L (2.9) u = 0, θ = 0, φ = 0 at y = -L.
where
Gm = 2 3 ( 1 2 )
g β T T
υ −
(Local temperature Grashof number) Gc = 2 3 ( 1 2 )
*
g β C C
υ − (Local mass
Grashof number) D -1 = K p /L 2 (Darcy’s number) α = QL 2 /λ (Heat source parameter)
v L 0
S = υ (Suction Reynolds number)
P r γ
= λ
(Prandtl number) Sc Dm
= ν
(Schmidt number)
3. FINITE ELEMENT ANALYSIS OF THE PROBLEM
To solve these differential equations with the corresponding boundary conditions, we assume
If u i , θ i , and φ i are the approximations of u,
θ, and φ we define the errors (residual)
i i i
i E E E
E 1 , 2 , 3 , 4 as
2 1
( )
i i
i i i i
u
d du du
E S Gm Gc D u
dy dy dy θ φ −
= + − − −
(3.1) Pr
i i
i d d d i
E S
dy dy dy
θ = θ + θ αθ −
(3.2)
i i
i d d d
E SSc
dy dy dy
φ = φ + φ
(3.3)
where
u i =
∑
= 3
1 k
k
f k ψ
θ i =
∑
= 3
1 k
k k ψ θ
(3.4)
φ i =
∑
= 3
1 k
k k ψ φ
these errors are orthogonal to the weight function over the domain of e i . under Galerkin finite element technique we choose the approximation functions as the weight function. Multiply both sides of the equations (3.1) – (3.3) by the weight function i.e., each of the approximation
function
i
ψ j
and integrate over the typical two nodded linear element ( η e , η e + 1 ) we obtain
1
0
e
e
i i
u j
E dy
η η
+ ψ
∫ =
(j = 1, 2, 3, 4) (3.5)
1
0
e
e
i i
E j dy
η θ η
+ ψ
∫ =
(j = 1, 2, 3, 4) (3.6)
1
0
e
e
i i
E j dy
η φ η
+ ψ
∫ =
(j = 1, 2, 3, 4) (3.7)
1
2 1
( ) 0
e
e
i i
i i i i
j
d du du
S Gm Gc D u dy
dy dy dy
η η
θ φ ψ
+
−
+ − − − =
∫
(3.8)
1
Pr 0
e
e
i i
i i
j
d d d
S dy
dy dy dy
η η
θ θ αθ ψ
+
+ − =
∫
(3.9)
1
0
e
e
i i
i j
d d d
SSc dy
dy dy dy
η η
φ φ ψ
+
+ =
∫
(3.10) following the Galerkin weighted residual method and integration by parts method to the equations (3.8 – 3.10), we obtain
1 1 1 1 1
1
1, 2,
e e e e e
e e e e e
i
j k k i i i i
j k j k j k j j j
d du du
dy S dy Gm dy Gc dy D u dy Q Q
dy dy dy
η η η η η
η η η η η
ψ ψ θ ψ φ ψ ψ
+ + + + +
+ − − − − = +
∫ ∫ ∫ ∫ ∫
(3.11)
Where
( ) ( )
1,
k
j j e e
Q du
ψ η dy η
− =
,
( ) ( )
2, 1 1
k
j j e e
Q du
ψ η + dy η +
=
1 1 1
1, 2,
Pr
e e e
e e e
i
j k k i i
j k j j j
d d d
dy S dy dy R R
dy dy dy
η η η
η η η
ψ θ θ ψ αθ ψ
+ + +
+ − = +
∫ ∫ ∫
(3.12)
where
( ) ( ) ( ) ( )
1,
k k
j j e e j e e
d d
R dy dy
θ φ
ψ η η ψ η η
− = +
,
( ) ( ) ( ) ( )
2, 1 1 1 1
k k
j j e e j e e
d d
R dy dy
θ φ
ψ η + η + ψ η + η +
= +
1 1
1, 2,
e e
e e
i
j k k i
j j j
d d d
dy ScS dy S S
dy dy dy
η η
η η
ψ φ φ ψ
+ +
+ = +
∫ ∫
(3.13)
where
( ) ( ) ( ) ( )
1,
k k
j j e e j e e
d d
S dy dy
φ θ
ψ η η ψ η η
− = +
,
( ) ( ) ( ) ( )
2, 1 1 1 1
k k
j j e e j e e
d d
S dy dy
φ θ
ψ η + η + ψ η + η +
= +
making use of equations (3.4), we can write above equations as
1 1 1 1
1 3 3 3 3
3
1
1 , 2 ,
1 1 1 1
1
e e e e
e
e e e e
e i
j k k i i i i
k k j k k j k k j k k j
k j j
k k k k
k
d d d
d y S u d y G m d y G c d y D u d y
u Q Q
d y d y d y
η η η η
η
η η η η
η
ψ ψ
+ψ ψ ψ θ
+ψ ψ φ
+ψ ψ
+ψ ψ
+ −
= = = =
=