15.11.2012 Views

Instability of Rayleigh –Benard -Marangoni convection in a ... - IJAMM

Instability of Rayleigh –Benard -Marangoni convection in a ... - IJAMM

Instability of Rayleigh –Benard -Marangoni convection in a ... - IJAMM

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

BENARD-MARANGONI CONVECTION IN A POROUS LAYER<br />

PERMEATED BY A NON-LINEAR MAGNETIC FLUID<br />

ABSTRACT<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah 1 and Z. Z.Rashed 2<br />

1 Department <strong>of</strong> Mathematics, Faculty <strong>of</strong> Science, Umm Al-Qura University,<br />

Makkah, Saudi Arabia<br />

2 Department <strong>of</strong> Mathematics, Faculty <strong>of</strong> Science, Northern Border University,<br />

Arar, Saudi Arabia<br />

Email: aamohammad@uqu.edu.sa<br />

Received 12 June 2011; accepted 27 September 2011<br />

The onset <strong>of</strong> Benard-<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a horizontal porous layer permeated by a<br />

magnetohydrodynamic fluid with a non-l<strong>in</strong>ear magnetic permeability has been exam<strong>in</strong>ed. The<br />

porous layer is assumed to be governed by the Br<strong>in</strong>kman model, bounded from below by a<br />

rigid surface and from above by a non deformable free surface. The critical effective<br />

<strong>Marangoni</strong> number and the critical <strong>Rayleigh</strong> number are obta<strong>in</strong>ed for different values <strong>of</strong> the<br />

effective Darcy number, Biot number, Chandrasekhar number and nonl<strong>in</strong>ear magnetic<br />

parameter for the cases <strong>of</strong> stationary <strong>convection</strong> and overstability. The related eigenvalue<br />

problem is solved us<strong>in</strong>g the first order Chebyshev polynomial method.<br />

Keywords: <strong>Marangoni</strong> <strong>convection</strong>, Porous layer, Br<strong>in</strong>kman model, Thermal <strong>in</strong>stability,<br />

Magnetic field<br />

1 INTRODUCTION<br />

Thermal <strong>in</strong>stability <strong>in</strong> a horizontal layer <strong>of</strong> a fluid heated from below has attracted<br />

considerable attention due to its relevance to a variety <strong>of</strong> applications such as geophysics,<br />

oceanography and atmospheric sciences as well as eng<strong>in</strong>eer<strong>in</strong>g. Systematic <strong>in</strong>vestigation <strong>of</strong><br />

this topic began <strong>in</strong> the early last century with the experiments <strong>of</strong> (Benard1900;1901).<br />

(<strong>Rayleigh</strong> 1916) provided a theoretical explanation <strong>of</strong> Benard's experimental results by<br />

consider<strong>in</strong>g <strong>in</strong>stability due to the action <strong>of</strong> buoyancy forces and he showed that the numerical<br />

value <strong>of</strong> the non-dimensional <strong>Rayleigh</strong> number, R, decides the stability or otherwise <strong>of</strong> a layer<br />

<strong>of</strong> fluid heated from below. <strong>Rayleigh</strong>-Benard <strong>convection</strong> driven by buoyancy has been a<br />

subject <strong>of</strong> considerable <strong>in</strong>terest s<strong>in</strong>ce the pioneer<strong>in</strong>g work <strong>of</strong> (<strong>Rayleigh</strong> 1916). Many <strong>of</strong> these<br />

studies are mentioned <strong>in</strong> (Chandrasekhar 1981). Numerous other papers exist on this topic;<br />

see for example (Abdullah and L<strong>in</strong>dsay 1990; 1991; Lombardo and Mulone 2005; Guray and<br />

Tarman 2007, Banjar and Abdullah 2011, Dubey et al. 2011).<br />

Apart from the buoyancy forces, convective <strong>in</strong>stability can also occur due to temperature<br />

dependent surface tension forces (known as <strong>Marangoni</strong> <strong>convection</strong>) which was


14<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

firstly studied by (Pearson 1958) who assumed <strong>in</strong>f<strong>in</strong>itesimally small amplitude analysis with<br />

non-deformable free surface at the top and no slip boundary at the bottom and showed that the<br />

variations <strong>in</strong> the surface tension due to temperature gradients could <strong>in</strong>duce motion with<strong>in</strong> the<br />

fluid when the numerical value <strong>of</strong> the non-dimensional <strong>Marangoni</strong> number, M, exceeds a<br />

critical value. <strong>Marangoni</strong> <strong>in</strong>stability has been studied widely for the cases <strong>of</strong> deformable and<br />

non-deformable free surface with various factors such as variable viscosity, magnetic field,<br />

rotation etc.(Takashima and Namikawa 1971; Takashima 1981a; 1981b; Friedrich and<br />

Rudraiah 1984; Hashim and Wilson 1999; Hashim and Arif<strong>in</strong> 2003; Awang and Hashim<br />

2008)).<br />

The theories <strong>of</strong> (<strong>Rayleigh</strong> 1916; Pearson 1958) were discussed by (Nield 1964) who<br />

comb<strong>in</strong>ed both buoyancy and surface tension mechanisms <strong>in</strong>to a s<strong>in</strong>gle analysis (known as<br />

Benard-<strong>Marangoni</strong>) and showed that the two agencies caus<strong>in</strong>g <strong>in</strong>stability re<strong>in</strong>force one<br />

another and are tightly coupled. Some <strong>of</strong> the major features <strong>of</strong> the Benard-<strong>Marangoni</strong><br />

<strong>convection</strong> problem are elucidated by ( Nield 1966; Garcia-Ybarra et al. 1987; Maekawa and<br />

Tanasawa 1989; Perez-Garcia and Carneiro 1991; Wilson 1993; Char and Chiang 1994; Jou<br />

et al. 1997; Chang and Chiang 1998; Douiebe et al. 2001) and others.<br />

<strong>Rayleigh</strong>-Benard <strong>convection</strong> <strong>in</strong> porous layers has been studied extensively s<strong>in</strong>ce the<br />

pioneer<strong>in</strong>g work <strong>of</strong> (Horton & Rogers 1945; Lapwood 1948). The copious literature cover<strong>in</strong>g<br />

different developments <strong>in</strong> this field are well documented <strong>in</strong> (Kaviany 1995; Ingham 1998;<br />

Vafai 2005; Nield and Bejan 2006). The study <strong>of</strong> <strong>Marangoni</strong> and Benard-<strong>Marangoni</strong><br />

<strong>convection</strong>s <strong>in</strong> porous media has drawn little attention compared to the study <strong>of</strong> <strong>Rayleigh</strong>-<br />

Bénard <strong>convection</strong> <strong>in</strong> porous media <strong>in</strong> spite <strong>of</strong> their importance <strong>in</strong> material science<br />

process<strong>in</strong>g, solidification <strong>of</strong> alloy, etc.<br />

<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a porous media has been studied us<strong>in</strong>g the Br<strong>in</strong>kman model by<br />

(Hennenberg et al. 1997; Rudraiah and Prasad 1998; Fadzillah et al. 2008). The onset <strong>of</strong><br />

coupled Darcy-Benard-<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a liquid saturated porous layer has been<br />

<strong>in</strong>vestigated by (Shivakumara et al. 2009) by employ<strong>in</strong>g the Br<strong>in</strong>kman-Forchheimer-<br />

Lapwood-extended Darcy flow model.<br />

In the literature, so far no research have been conduct<strong>in</strong>g regard<strong>in</strong>g the effect <strong>of</strong> magnetic<br />

field on <strong>Marangoni</strong> or Benard-<strong>Marangoni</strong> <strong>convection</strong>s <strong>in</strong> a porous layer. The presence <strong>of</strong><br />

magnetic field <strong>in</strong> an electrically conduct<strong>in</strong>g fluid usually has the effect <strong>of</strong> <strong>in</strong>hibit<strong>in</strong>g the<br />

development <strong>of</strong> <strong>in</strong>stabilities. (Thompson 1951; Chandrasekhar 1981) and others have<br />

exam<strong>in</strong>ed Benard <strong>convection</strong> <strong>in</strong> the context <strong>of</strong> magnetohydrodynamic fluid with a l<strong>in</strong>ear<br />

constitutive relationship between the magnetic field H and the magnetic <strong>in</strong>duction B.<br />

However a non-l<strong>in</strong>ear constitutive relationship between H and B may be appropriate for<br />

certa<strong>in</strong> classes <strong>of</strong> materials. The relevance <strong>of</strong> this criterion to the configuration <strong>of</strong> a neutron<br />

star is discussed by (Roberts 1981; Muzikar and Pethick 1981).(Cowley and Rosensweig<br />

1967; Gailitis 1977) and others use non-l<strong>in</strong>ear magnetization laws to describe the properties<br />

<strong>of</strong> ferr<strong>of</strong>luids. (Abdullah and L<strong>in</strong>dsay 1990; 1991; Abdullah 2000; Jan and Abdullah 2000;<br />

Al-aidrous and Abdullah 2005) have exam<strong>in</strong>ed Benard <strong>convection</strong> problems us<strong>in</strong>g the<br />

nonl<strong>in</strong>ear relationship suggested by (Roberts 1981).<br />

The aim <strong>of</strong> this work is to study the effect <strong>of</strong> magnetic field on the stability <strong>of</strong> Benard-<br />

<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a horizontal porous layer us<strong>in</strong>g a l<strong>in</strong>ear and non-l<strong>in</strong>ear relationship<br />

between H and B. The non-l<strong>in</strong>ear relationship used by (Abdullah and L<strong>in</strong>dsay 1990;1991) has


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

been adopted. The flow <strong>in</strong> the porous medium is controlled by the Br<strong>in</strong>kman model and the<br />

porous layer is bounded from below by a rigid surface and from above by a non deformable<br />

free surface. The first order Chebyshev polynomial method is used to obta<strong>in</strong> the numerical<br />

solutions <strong>of</strong> the correspond<strong>in</strong>g eigenvalue problem (Bukhari 1997).<br />

2 MATHEMATICAL FORMULATION<br />

Consider a horizontal porous layer permeated by an <strong>in</strong>compressible electrically conduct<strong>in</strong>g<br />

magnetohydrodynamic viscous fluid. The porous layer is governed by Br<strong>in</strong>kman model and is<br />

subjected to a constant gravitational acceleration <strong>in</strong> the negative x 3 direction, and to a<br />

constant magnetic field H <strong>in</strong> the positive x 3 direction. The lower boundary, x3 � 0 , is<br />

assumed to be rigid while the upper boundary, x3 � d , is assumed to be free and subjected to<br />

temperature–dependant surface tension forces. The temperatures at x3 � 0 and x3 � d are<br />

T � and T� �Tu<br />

respectively. (see Figure.1)<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

Figure 1: Schematic diagram <strong>of</strong> the problem<br />

The Bouss<strong>in</strong>esq approximation is applied to the layer such that<br />

�1 � �T �T<br />

��<br />

,<br />

� (1)<br />

� �0<br />

�<br />

�<br />

where � is the density <strong>of</strong> the fluid, � � is the density <strong>of</strong> the fluid at T � , T is the fluid<br />

temperature, and � is the coefficient <strong>of</strong> volume expansion. The relation between B i and i H<br />

is assumed to be non-l<strong>in</strong>ear(see Abdullah & L<strong>in</strong>dsay 1990) such that<br />

Hi<br />

�<br />

�<br />

�� (2)<br />

�<br />

�Bi<br />

where � �� , B�<br />

� � � is the <strong>in</strong>ternal energy function. In fact<br />

��<br />

�B<br />

��<br />

B<br />

H<br />

i<br />

i � �� � ��<br />

� ���<br />

Bi<br />

(3)<br />

�B<br />

�B<br />

�B<br />

B<br />

i<br />

15


16<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

� 1<br />

where �<br />

�B<br />

B<br />

�<br />

� is the magnetic susceptibility. We shall assume that the magnetic <strong>in</strong>duction,<br />

B i , is constant <strong>of</strong> the form Bi � �0, 0,<br />

B�.<br />

Thus the govern<strong>in</strong>g field equations are<br />

��<br />

�Vi<br />

�<br />

� �P,<br />

i � V<br />

� �t<br />

K<br />

V<br />

i,<br />

i<br />

� 0,<br />

�� � B �<br />

�T<br />

�� c�<br />

� �� c �<br />

B<br />

�<br />

i,<br />

i<br />

�B<br />

i<br />

�t<br />

m<br />

�t<br />

� 0,<br />

� V<br />

i,<br />

j<br />

�<br />

B<br />

j<br />

�<br />

�<br />

�V<br />

j<br />

i , k<br />

p<br />

B<br />

i<br />

f<br />

i,<br />

j<br />

� �<br />

B<br />

i<br />

k<br />

�<br />

�<br />

�<br />

,<br />

�<br />

� � e<br />

� �<br />

2<br />

ijk<br />

V<br />

e<br />

i<br />

� �<br />

2<br />

V ��T<br />

� k�<br />

T ,<br />

krs<br />

�<br />

�1 ��<br />

�T �T<br />

��<br />

�B�� ,<br />

s,<br />

rj<br />

where V is the velocity, P is the hydrostatic pressure, k is the thermal conductivity, � is the<br />

i<br />

dynamic viscosity, � is the porosity, �eff � � � is the effective viscosity, K is the<br />

permeability, p<br />

c is the specific heat at constant pressure, � � c p � f<br />

�<br />

g<br />

i<br />

� and � � c�m<br />

(4)<br />

� are the heat<br />

capacity per unit volume <strong>of</strong> the fluid and porous medium respectively and �� is the electrical<br />

conductivity. The fluid is conf<strong>in</strong>ed between the planes 3 0 � x and x3 � d and on these<br />

planes we need to specify the follow<strong>in</strong>g boundary conditions<br />

� ��<br />

At x3 � d : V3 � 0,<br />

k� T �n�<br />

qT � 0,<br />

B3, 3 � 0,<br />

2��<br />

Dnt<br />

� �T<br />

�t,<br />

(5)<br />

� �T<br />

where q is the heat transfer coefficient, � is the k<strong>in</strong>ematic viscosity, n , t denote the normal<br />

and tangential unit vectors at the upper free surface respectively, and � is the surface tension<br />

which assumed to be a l<strong>in</strong>ear function <strong>of</strong> temperature such that � � � � ��<br />

�T � T�<br />

�, where<br />

� � be<strong>in</strong>g the surface tension at temperature T � and � is the change <strong>of</strong> surface tension with<br />

temperature.<br />

At x3 � 0 : V3 � 0,<br />

V3,<br />

3 � 0,<br />

T � T�<br />

, B3<br />

� 0.<br />

(6)<br />

Equations (4) have a steady solution <strong>in</strong> which<br />

�B�, B �0, 0,<br />

B�<br />

B is constant<br />

V � 0, ��P<br />

� � g � 0,<br />

T � T�<br />

� � x3<br />

, � � � � ,<br />

where � is the adverse temperature gradient. We shall suppose that the steady solution be<br />

perturbed by the follow<strong>in</strong>g l<strong>in</strong>ear perturbation quantities<br />

V � 0 �V,<br />

P � P<br />

�x �� p,<br />

B � �0, 0,<br />

B��<br />

�b , b , b �<br />

� � � � b � , T � T � � x ��<br />

.<br />

3<br />

, B<br />

3<br />

�<br />

3<br />

1<br />

2<br />

3<br />

,


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

Apply<strong>in</strong>g these perturbations, equations (4) become<br />

1 �Vi<br />

� p � � � 2<br />

� �� Vi<br />

Vi<br />

g i3<br />

t �<br />

�<br />

�<br />

� � � � � �� �<br />

� � � ��<br />

� K �<br />

� � B�b<br />

i,<br />

3 � � b3,<br />

3 �i3<br />

�,<br />

Vi<br />

, i � 0,<br />

��<br />

2<br />

Gm<br />

�V3<br />

� � �m�<br />

� ,<br />

�t<br />

bi<br />

, i � 0,<br />

�bi<br />

� BVi<br />

, 3 �<br />

�t<br />

where<br />

Gm<br />

�<br />

2 2<br />

�� b � � �� b � � b ��.<br />

i<br />

�� c�<br />

�<br />

m<br />

���c p �f i3<br />

3,<br />

i3<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

3<br />

� is the ratio <strong>of</strong> heat capacities,<br />

���c p � f<br />

(7)<br />

17<br />

k<br />

�m � is the thermal<br />

� �<br />

diffusivity, � �<br />

� �<br />

�<br />

B �,<br />

B<br />

is the electrical resistivity and � � is the non-l<strong>in</strong>ear magnetic<br />

�<br />

parameter. The boundary conditions (5) and(6) become<br />

��<br />

V3 � 0,<br />

k � q�<br />

� 0,<br />

�x<br />

3<br />

�b<br />

�x<br />

3 �<br />

3<br />

2<br />

� � V �<br />

2<br />

0,<br />

� �<br />

� 3 �<br />

� ��<br />

�h�<br />

� 0 , (8)<br />

� � 2<br />

x �<br />

� � 3 �<br />

V3 � 0,<br />

V3,<br />

3 � 0,<br />

� � 0,<br />

b3<br />

� 0.<br />

(9)<br />

We now non-dimensionalize equations (7) us<strong>in</strong>g the dimensionless variables<br />

2<br />

d �<br />

�<br />

ˆ , ,<br />

m ˆ μ<br />

x ,<br />

m ˆ , ˆ , � ˆ<br />

i � d xi<br />

t � t Vi<br />

� Vi<br />

p � p bi<br />

� Bbi<br />

� T�<br />

�Tu<br />

�<br />

� d K<br />

m<br />

Thus equations (7) become<br />

Da ˆ<br />

eff �V<br />

i<br />

ˆ ˆ<br />

2<br />

� �� p �V<br />

ˆ ˆ<br />

i � Daeff<br />

� Vi<br />

� R�<br />

�i3<br />

Pr<br />

�t<br />

� Q Pm<br />

i,<br />

3<br />

Vˆ<br />

i,<br />

i � 0,<br />

� ˆ � ˆ 2<br />

G<br />

ˆ<br />

m � H �V3<br />

� � � ,<br />

�t<br />

bˆ<br />

i,<br />

i � 0,<br />

�bˆ<br />

i<br />

�Vˆ<br />

i,<br />

3 � Pm<br />

�t<br />

�bˆ � � bˆ<br />

� �<br />

3,<br />

3<br />

i3<br />

2 � ˆ 2<br />

� b � ˆ ˆ<br />

i � � � b3�i3<br />

� b3,<br />

i3<br />

��<br />

,<br />

,<br />

(10)


18<br />

where<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

�� ��<br />

� � � gdKT�<br />

� Tu<br />

B K<br />

K<br />

Pr � , Pm<br />

� , R �<br />

, Q � � , Daeff<br />

� ,<br />

� �<br />

� �<br />

� � �<br />

2<br />

m<br />

m<br />

m<br />

�<br />

d �<br />

�T � T �<br />

� � u<br />

H � �<br />

T�<br />

� Tu<br />

�1<br />

� �<br />

��<br />

1<br />

Here P r Pm<br />

, R,<br />

Q , Daeff<br />

when heat<strong>in</strong>g from above,<br />

when heat<strong>in</strong>g<br />

from below.<br />

, are non-dimensional numbers denote respectively the viscous<br />

Prandtl number, magnetic Prandtl number, <strong>Rayleigh</strong> number <strong>of</strong> porous layer, Chandrasekhar<br />

number and effective Darcy number. The non-dimensionalization <strong>of</strong> boundary conditions (8)<br />

and (9) yield<br />

Vˆ<br />

3 � 0,<br />

� ˆ �<br />

� Bi ˆ � � 0,<br />

�xˆ<br />

3<br />

2<br />

�<br />

Vˆ<br />

2<br />

M ˆ<br />

3 � eff �h�<br />

� 0,<br />

xˆ<br />

2<br />

� 3<br />

bˆ<br />

3,<br />

3 � 0,<br />

Vˆ 3 � 0,<br />

Vˆ<br />

3 3 � 0,<br />

ˆ � 0,<br />

bˆ<br />

, � 3 � 0 . (12)<br />

where Bi and M eff are the Biot and effective <strong>Marangoni</strong> numbers which are given by<br />

Bi �<br />

qd<br />

k<br />

,<br />

M<br />

eff<br />

� �� d<br />

�<br />

� ��<br />

�<br />

m<br />

2<br />

.<br />

Now we look for a normal mode solution <strong>of</strong> equations (10) <strong>of</strong> the from<br />

� t i<br />

� � � � �a1 x1<br />

�a<br />

2 x<br />

t x ��<br />

x e e<br />

2 �<br />

� ,<br />

i<br />

3<br />

� � , 1 2 a , a are the wave numbers <strong>of</strong> the harmonic disturbance and � is the<br />

growth rate. Thus equations (10) become<br />

where �w ,� , b�<br />

Da<br />

G<br />

m<br />

eff<br />

P<br />

r<br />

�<br />

� b � Dw � P<br />

2 2<br />

2 2<br />

2 2<br />

�D � a �w ��<br />

Q Db � ��D<br />

� a �w � Da �D � a �<br />

�� � H � w<br />

m<br />

� a<br />

2 2<br />

� �D � a ��<br />

,<br />

2 2 2<br />

��D � a ��<br />

� a � b.<br />

2<br />

R�<br />

� Q D<br />

2<br />

w,<br />

eff<br />

2<br />

w<br />

2<br />

(11)<br />

(13)


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

� 2 2<br />

where D � , a � a1<br />

� a2<br />

is the wave number. The boundary conditions (11) and (12)<br />

�x3<br />

become<br />

w � 0<br />

D�<br />

� Bi�<br />

� 0<br />

D<br />

2<br />

w � a<br />

Db � 0,<br />

2<br />

M<br />

eff<br />

� � 0<br />

w � 0,<br />

Dw � 0,<br />

� � 0,<br />

b � 0 . (15)<br />

3RESULTS AND DISCUSION<br />

The govern<strong>in</strong>g equations (13) together with the boundary conditions (14) and (15) are solved<br />

when the fluid layer is heated from below (i.e. H� � �1<br />

). The first order Chebyshev spectral<br />

method is used to obta<strong>in</strong> the numerical results. Here we shall consider the follow<strong>in</strong>g cases.<br />

3.1 The effect <strong>of</strong> buoyancy forces only<br />

In the follow<strong>in</strong>g analysis we shall study the special case <strong>of</strong> buoyancy effect only (<strong>Rayleigh</strong>-<br />

Benard <strong>in</strong>stability). In this case we shall suppose that M eff � 0 <strong>in</strong> condition (14)3(i.e. no<br />

surface tension). We may elim<strong>in</strong>ate b and � from (13)1 us<strong>in</strong>g equations (13)2,3 to obta<strong>in</strong><br />

�<br />

Da �<br />

L�<br />

� Daeff<br />

L ��<br />

P �<br />

�<br />

�<br />

r �<br />

� Q<br />

�L ��<br />

G �<br />

2 2<br />

2<br />

�L ��G<br />

� �L � � a �D w � L �L ��<br />

G �� L � � � a �w<br />

� 0<br />

m<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

m<br />

� � 2 � 2�<br />

� 2 �<br />

�<br />

� L � � � a w R a L a w<br />

P �<br />

� �<br />

�<br />

� � � �<br />

m<br />

P �<br />

�<br />

�<br />

� � m �<br />

m<br />

�<br />

�<br />

�<br />

where we have assumed that the porous medium layer is heated from below ( i.e. H � � �1).<br />

This equation can be expanded to yield<br />

Gm<br />

Daeff<br />

�<br />

3<br />

2 � Gm<br />

Daeff<br />

Daeff<br />

Gm<br />

Daeff<br />

� 2 � 2<br />

Gm<br />

Daeff<br />

Gm<br />

� �<br />

�<br />

Lw � � ��<br />

�L<br />

w � a<br />

�Lw�<br />

Pr<br />

P<br />

�<br />

� �<br />

�<br />

m<br />

Pr<br />

Pr<br />

Pm<br />

p � �<br />

�<br />

�<br />

m<br />

pr<br />

p �<br />

��<br />

�<br />

� �<br />

m � ��<br />

��<br />

Daeff<br />

Daeff<br />

� 3 � 2<br />

Daeff<br />

2<br />

1 � 2<br />

� � ��<br />

Gm<br />

Daeff<br />

�L<br />

w � a<br />

a Daeff<br />

GmQ<br />

Gm<br />

�<br />

�<br />

� �<br />

L w<br />

Pr<br />

p �<br />

�<br />

�<br />

� � � � � �<br />

m<br />

Pr<br />

p �<br />

��<br />

�<br />

� �<br />

m �<br />

�<br />

�<br />

� 2<br />

Ra<br />

� �<br />

2 2<br />

2<br />

4<br />

4 2<br />

�� a G Q � a G Q � � a G �Lw� � � � a G Q �w�<br />

� Da L w � �� a Da � Q � 1�<br />

m<br />

�<br />

Pm<br />

2 2<br />

2 2 2 4<br />

4<br />

�a Q � � a G Q � � a �L w � �Ra � � a Q�Lw<br />

� Ra � w � 0<br />

m<br />

m<br />

m<br />

�<br />

�<br />

pm<br />

m<br />

�<br />

�<br />

��<br />

eff<br />

�<br />

�<br />

�<br />

eff<br />

3<br />

L w<br />

19<br />

(14)<br />

(16)<br />

(17)


20<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

which is an eighth order ord<strong>in</strong>ary differential equation to be satisfied by w . Suppose that both<br />

w � A s<strong>in</strong><br />

s�<br />

x where A is a constant and s is an <strong>in</strong>teger. Thus<br />

boundaries are free and that � �<br />

2<br />

4<br />

6<br />

w � D w � D w � D w � 0,<br />

on x3<br />

� 0,<br />

1.<br />

2<br />

2<br />

3<br />

Let L � D � a , then Lw � ��<br />

w , where � � s � � a , thus equation (17) becomes<br />

3 Gm<br />

Daeff<br />

2 �Gm<br />

Daeff<br />

�<br />

� � �<br />

Pr<br />

Pm<br />

� Pr<br />

� �<br />

� � ��Daeff<br />

�<br />

�Gm<br />

�<br />

� �<br />

2 2<br />

� s � GmQ<br />

� �<br />

Pm<br />

�<br />

1<br />

Pr<br />

�<br />

�<br />

�<br />

�<br />

2 �� � �a<br />

�<br />

2 2 Daeff<br />

2<br />

�� � � a �� � � G �� � � a �<br />

2<br />

2<br />

Ra � �<br />

2<br />

Ra �<br />

2<br />

2 2 2 2<br />

�� � � a �� � � �� � � a � �Da<br />

� � � � Q�<br />

s � � � 0 .<br />

�pm<br />

��<br />

Daeff<br />

�<br />

� �<br />

P �<br />

�<br />

m �<br />

Pm<br />

Now we shall consider two cases <strong>of</strong> <strong>in</strong>stabilities.<br />

3.1.1 Stationary Convection<br />

m<br />

��<br />

1<br />

Pr<br />

2<br />

eff<br />

2<br />

� Gm<br />

�<br />

� Gm<br />

�<br />

��<br />

�<br />

� Pm<br />

�<br />

� ��<br />

To f<strong>in</strong>d the critical <strong>Rayleigh</strong> number for the onset <strong>of</strong> stationary <strong>convection</strong> we set � � 0 <strong>in</strong><br />

equation (18) to obta<strong>in</strong><br />

� 2 2 2<br />

R � �Daeff � � � � Q s � �.<br />

(19)<br />

2<br />

a<br />

S<strong>in</strong>ce this equation does not conta<strong>in</strong> the non-l<strong>in</strong>ear parameter � then we deduce that the nonl<strong>in</strong>ear<br />

magnetic permeability has no effect <strong>in</strong> the case <strong>of</strong> stationary <strong>convection</strong>. In fact <strong>in</strong><br />

absence <strong>of</strong> magnetic field ( Q � 0 ) and � 0 , s � 1 we obta<strong>in</strong><br />

2 2 �a � �<br />

2<br />

2<br />

� �<br />

R � � ,<br />

2 2<br />

a a<br />

Da eff<br />

2<br />

and the critical <strong>Rayleigh</strong> number Rcrit<br />

by (Lapwood 1948). Clearly<br />

� 4�<br />

� 39.<br />

4784 ,which is the same result obta<strong>in</strong>ed<br />

dR<br />

dDaeff<br />

3<br />

6 � 2<br />

a �<br />

4<br />

�<br />

� 2<br />

�1<br />

�<br />

dR � �<br />

� � , � a<br />

� 1 � �<br />

2 2<br />

a<br />

� �<br />

2<br />

� �<br />

� 2 �<br />

�<br />

dQ a � � �<br />

which <strong>in</strong>dicate that both the effective Darcy number and the magnetic field have a stabiliz<strong>in</strong>g<br />

effect on the system.<br />

(18)


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

3.1.2 Overstability Convection<br />

To obta<strong>in</strong> the critical <strong>Rayleigh</strong> number for the case <strong>of</strong> overstability we suppose that<br />

2<br />

a<br />

n1<br />

� ,<br />

2<br />

�<br />

�<br />

�<br />

2<br />

� � � �1 � n1<br />

�,<br />

� i�1<br />

,<br />

2<br />

R<br />

� R1<br />

,<br />

2<br />

Q � Q1<br />

,<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

�<br />

Daeff<br />

1<br />

� Da<br />

2 eff<br />

�<br />

,<br />

2<br />

s � 1,<br />

where � is complex and � 1 � 0 . Substitute <strong>in</strong>to equation (18), we obta<strong>in</strong><br />

� i�<br />

�<br />

�1 � n �<br />

�<br />

� i�1<br />

�<br />

�<br />

�<br />

��1 � n ��<br />

� n �<br />

�1 � n �Da �G<br />

� � ��1 � n ��<br />

�n<br />

�<br />

2 eff 1<br />

�1 � n � � G ��1 � n ��<br />

�n<br />

�<br />

�1 � n �<br />

�<br />

P<br />

3<br />

1<br />

m<br />

1<br />

G<br />

1<br />

1<br />

Da<br />

P P<br />

Da<br />

Da<br />

G<br />

P<br />

P<br />

Q<br />

2 �G<br />

��<br />

1 �<br />

�<br />

1<br />

P<br />

� G<br />

Da<br />

� G<br />

�<br />

� �<br />

� P<br />

m 1 ��1 � n1<br />

��<br />

�n1<br />

�<br />

�1 � n �<br />

n1R1<br />

p<br />

�1 � n �<br />

2<br />

��1 � n ��<br />

�n<br />

� Da �1 � n � � �1 � n �<br />

1<br />

m<br />

�<br />

r<br />

eff 1<br />

1<br />

eff 1<br />

m<br />

m<br />

m<br />

1<br />

1<br />

�<br />

�<br />

�<br />

�<br />

r<br />

eff 1<br />

�<br />

�<br />

�<br />

�<br />

�<br />

�<br />

m<br />

m<br />

eff 1<br />

m<br />

m<br />

P<br />

r<br />

eff 1<br />

1 �<br />

P �<br />

r �<br />

1<br />

1<br />

m<br />

m<br />

�<br />

�<br />

�<br />

�<br />

1<br />

1<br />

�<br />

1<br />

1<br />

1<br />

m<br />

1<br />

�<br />

�<br />

�<br />

� Q<br />

The real and imag<strong>in</strong>ary parts <strong>of</strong> equation (20) are<br />

n<br />

1<br />

R<br />

1<br />

� Da<br />

1<br />

� Pm<br />

�1 � n1<br />

� � n1<br />

�<br />

�1 � n1<br />

� G G<br />

m m<br />

�<br />

�1 � n � � n � P<br />

1<br />

eff 1<br />

1<br />

3 2<br />

�1 � n � � �1 � n � � �1 � n �<br />

m<br />

1<br />

r<br />

1<br />

1<br />

Q<br />

1<br />

1<br />

1<br />

�<br />

n<br />

1<br />

R<br />

1<br />

�1 � n �<br />

1<br />

�<br />

� � 0,<br />

�<br />

2 �<br />

��<br />

�<br />

Daeff<br />

1 � 1<br />

�<br />

�<br />

� Pr<br />

� 2<br />

� Gm<br />

�<br />

��1<br />

� n1<br />

� �<br />

�<br />

(21)<br />

Daeff<br />

1<br />

�1 � n1<br />

��,<br />

P<br />

21<br />

(20)


22<br />

n<br />

�<br />

�<br />

1<br />

R<br />

P<br />

1<br />

m<br />

2 �1 � n � �1 � n � � n �<br />

�1 � n � � n �<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

�1 � n � � G Q �1 � n � � n �<br />

�1 � n � �1 � n ��G �1 � n � � n ��<br />

��<br />

2<br />

1<br />

�<br />

Da<br />

1<br />

1<br />

eff 1<br />

m eff 1<br />

�1 � n � .<br />

1<br />

1<br />

1<br />

� 1<br />

�<br />

� Pm<br />

G<br />

P<br />

m<br />

Da<br />

r<br />

1<br />

P P<br />

m<br />

1<br />

1<br />

�<br />

�<br />

�<br />

m<br />

1<br />

�<br />

�Da<br />

�<br />

m<br />

1<br />

eff 1<br />

From equation (21) and (22) we obta<strong>in</strong><br />

�<br />

p<br />

�<br />

2<br />

1<br />

2<br />

m<br />

�<br />

�<br />

�<br />

�<br />

�<br />

�1 � n �<br />

�<br />

�<br />

�<br />

�1 � n � � n �<br />

�<br />

�<br />

�G<br />

�<br />

1<br />

�<br />

�<br />

�G<br />

�<br />

�1 � � n � n ���1 � n � Da �G<br />

� � � G � .<br />

1<br />

1<br />

1<br />

Da<br />

eff 1<br />

1<br />

1<br />

m<br />

�<br />

1<br />

1<br />

P<br />

r<br />

eff 1<br />

�<br />

�<br />

�<br />

�<br />

�<br />

� � G<br />

�<br />

2<br />

From equation (23), � 1 � 0 provided<br />

P<br />

m<br />

and<br />

G<br />

m<br />

�<br />

�<br />

�<br />

�1<br />

�<br />

�<br />

1<br />

� n<br />

1<br />

� ��� 1�<br />

n<br />

1<br />

�1 � n � � n ��1 � n �<br />

�<br />

m<br />

m<br />

1<br />

1<br />

m<br />

�<br />

� ��<br />

�Q1�<br />

��<br />

�<br />

�<br />

�<br />

��<br />

��<br />

�<br />

�<br />

1 �<br />

P �<br />

r �<br />

1<br />

P<br />

m<br />

r<br />

1<br />

1<br />

P<br />

�<br />

�<br />

�<br />

�<br />

m<br />

��<br />

��<br />

���<br />

� G<br />

m<br />

� � n<br />

�<br />

� �<br />

1G<br />

�<br />

m<br />

�1 � n �<br />

1<br />

�<br />

�<br />

�<br />

�<br />

(22)<br />

(23)<br />

, (24)<br />

�<br />

�<br />

� � �<br />

� 1 1 1<br />

1<br />

Q1<br />

�Daeff<br />

1�1<br />

� n1<br />

�� �Gm<br />

� �<br />

� � Gm<br />

� . (25)<br />

� 1 �<br />

� � ��<br />

�<br />

� Pr<br />

� �<br />

�<br />

� � Gm<br />

�<br />

� 1 n1<br />

� n1Gm<br />

� Pm<br />

�<br />

In fact for overstability to be possible the conditions (24) and (25) must be satisfied. Note<br />

that when the relation between the magnetic field and the magnetic <strong>in</strong>duction is l<strong>in</strong>ear ( i.e.<br />

� � 0 ) and us<strong>in</strong>g Darcy's model, the conditions (24) and (25) reduce to<br />

2 2<br />

P � �<br />

PmGm � 1 ,and<br />

m � � a 2 2<br />

Q1 � � � �� � a ��<br />

Gm<br />

�,<br />

�1 � PmGm<br />

�<br />

Da<br />

where � � . These are the same conditions obta<strong>in</strong>ed by (Bhadauria and Sherani 2008).<br />

� P<br />

r


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

The govern<strong>in</strong>g equations (13) together with the boundary conditions (14) and (15) are<br />

solved numerically <strong>in</strong> absence <strong>of</strong> surface tension when the upper boundary is free and the<br />

lower boundary is rigid. We suppose first that the effect <strong>of</strong> magnetic field is ignored. For the<br />

stationary <strong>convection</strong> case the relation between the Biot number, Bi , and the critical <strong>Rayleigh</strong><br />

number, R , is displayed <strong>in</strong> figure (2) for different values <strong>of</strong> the effective Darcy number,<br />

Da . As Bi <strong>in</strong>creases R <strong>in</strong>creases which <strong>in</strong>dicates that the Biot number delays the onset <strong>of</strong><br />

eff<br />

<strong>convection</strong>. Moreover as Da eff decreases R decreases.<br />

In the presence <strong>of</strong> magnetic field, the relation between Q and the critical R for both<br />

stationary <strong>convection</strong> and overstability cases is displayed <strong>in</strong> figures (3) when � 0.<br />

001.<br />

In<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

Da eff<br />

both stationary and overstability cases as Q <strong>in</strong>creases the critical R <strong>in</strong>creases which <strong>in</strong>dicates<br />

that the magnetic field has a stabiliz<strong>in</strong>g effect. For the overstability case we noticed that <strong>in</strong> the<br />

case <strong>of</strong> nonl<strong>in</strong>ear magnetic permeability ( � � 0 ), the critical R <strong>in</strong>creases as � <strong>in</strong>creases<br />

which <strong>in</strong>dicates that the nonl<strong>in</strong>earity has a stabiliz<strong>in</strong>g effect. However this nonl<strong>in</strong>earity has no<br />

effect <strong>in</strong> the case <strong>of</strong> stationary <strong>convection</strong> which is the preferred mechanism. Moreover as<br />

Da decreases R decreases for stationary <strong>convection</strong> and overstability cases and for the<br />

eff<br />

l<strong>in</strong>ear and non-l<strong>in</strong>ear cases. The numerical results related to figure (3) are listed <strong>in</strong> tables (1)<br />

and (2).<br />

The critical <strong>Rayleigh</strong>, R , is obta<strong>in</strong>ed when Da � � (fluid layer ) and M � 0 for<br />

different values <strong>of</strong> Q . The numerical results <strong>of</strong> this case are listed <strong>in</strong> table (3) which are <strong>in</strong><br />

excellent agreement with those <strong>of</strong> (Biswal and Rao 1999). Note that the relation between Q f<br />

(<strong>in</strong> fluid layer) and Q ( <strong>in</strong> porous layer) is Q � Q f Daeff<br />

.<br />

3.1.3 Special Cases<br />

Here we shall discuss the same problem (the effect <strong>of</strong> buoyancy forces) us<strong>in</strong>g Darcy's model<br />

when<br />

(i) The boundaries are mixed.<br />

(ii) Both boundaries are free.<br />

( i ) Mixed boundaries<br />

In this case we shall assume that the upper boundary is free and the lower boundary is rigid.<br />

Figure (4) shows the relation between Q and critical R for the stationary <strong>convection</strong> and<br />

�6 overstability cases when Da � 10 , � � 0.<br />

3 . Clearly as Q <strong>in</strong>creases R <strong>in</strong>creases for both<br />

cases and for l<strong>in</strong>ear and non-l<strong>in</strong>ear cases. The numerical results related to this figure are listed<br />

<strong>in</strong> table (4). We note that <strong>in</strong> absence <strong>of</strong> magnetic field �Q � 0�the<br />

critical <strong>Rayleigh</strong> number<br />

R � 27.<br />

098 which is <strong>in</strong> excellent agreement with the same result mentioned <strong>in</strong> (Nield and<br />

Bejan 2006).<br />

eff<br />

23


24<br />

(ii) Both boundaries are free<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

In this case we shall assume that both boundaries are free. The critical <strong>Rayleigh</strong> numbers are<br />

obta<strong>in</strong>ed for different values <strong>of</strong> Q for the stationary <strong>convection</strong> case. The numerical results<br />

are listed <strong>in</strong> table (5). These results are <strong>in</strong> excellent agreement with those <strong>of</strong> (Bhadauria &<br />

Sherani 2008).<br />

3.2 The effect <strong>of</strong> surface tension force only<br />

In this case the effect <strong>of</strong> surface tension is studied <strong>in</strong> absence <strong>of</strong> buoyancy forces. First we<br />

suppose that the effect <strong>of</strong> magnetic field is ignored. For the stationary <strong>convection</strong> case the<br />

relation between the Biot number, Bi, and the critical effective <strong>Marangoni</strong> number, M eff , is<br />

displayed <strong>in</strong> figure (5) for different values <strong>of</strong> the effective Darcy number, Da eff . Clearly as<br />

Bi <strong>in</strong>creases M eff <strong>in</strong>creases which <strong>in</strong>dicates that the effect <strong>of</strong> the Biot number is to delay the<br />

onset <strong>of</strong> <strong>Marangoni</strong> <strong>convection</strong>. Moreover as Da eff decreases M eff <strong>in</strong>creases. Similar<br />

relations are presented <strong>in</strong> figure (6) between Bi and the critical effective Darcy-<strong>Marangoni</strong><br />

number<br />

M ( � Daeff<br />

M eff ) for different values <strong>of</strong> eff<br />

Deff<br />

�5<br />

Daeff<br />

� 10 then D � 2.<br />

0532602 and 4.<br />

2875<br />

M eff<br />

result obta<strong>in</strong>ed by (Hennenberg et al. 1997).<br />

Da . We note that if Bi � 0 and<br />

a � which is <strong>in</strong> excellent agreement with the<br />

In the presence <strong>of</strong> magnetic field, the relation between, Q , and the critical M eff for both<br />

stationary <strong>convection</strong> and overstability cases is displayed <strong>in</strong> figure (7) when � 0.<br />

001.<br />

In<br />

Da eff<br />

both stationary <strong>convection</strong> and overstability cases the magnetic field has a stabiliz<strong>in</strong>g effect.<br />

For the overstability case when � 0<br />

M <strong>in</strong>creases as � <strong>in</strong>creases which<br />

� the critical eff<br />

<strong>in</strong>dicates that the nonl<strong>in</strong>ear magnetic permeability has a stabiliz<strong>in</strong>g effect. However this<br />

nonl<strong>in</strong>earity has no effect <strong>in</strong> the case <strong>of</strong> stationary <strong>convection</strong>. In fact we noticed that as<br />

M <strong>in</strong>creases for stationary <strong>convection</strong> and overstability cases and for<br />

Da eff decreases eff<br />

l<strong>in</strong>ear and non-l<strong>in</strong>ear cases. The numerical results related to this figure are listed <strong>in</strong> tables (6)<br />

and (7). Similar relations between M and Q are displayed <strong>in</strong> figure (8).<br />

Deff<br />

Figure (9) shows the effect <strong>of</strong> Bi on M eff for different values <strong>of</strong> Q for the stationary<br />

<strong>convection</strong> case. Clearly as Bi <strong>in</strong>creases M eff <strong>in</strong>creases. In table (8) the critical values <strong>of</strong><br />

M eff are shown for different values <strong>of</strong> Da eff , Bi , � and Q when the rigid boundary is<br />

thermally <strong>in</strong>sulated ( D � 0<br />

excellent agreement with those <strong>of</strong> (Shivakumara et al. 2009).<br />

� ). In absence <strong>of</strong> magnetic field � 0�<br />

Q � the results are <strong>in</strong><br />

The critical <strong>Marangoni</strong> number, M , is obta<strong>in</strong>ed when Da � � (fluid layer) and<br />

R � 0 for different values <strong>of</strong> Q . As Q <strong>in</strong>creases, M <strong>in</strong>creases which <strong>in</strong>dicates that the<br />

magnetic field has a stabiliz<strong>in</strong>g effect <strong>in</strong> this case also. The numerical results obta<strong>in</strong>ed<br />

co<strong>in</strong>cide with those <strong>of</strong> (Biswal and Rao 1999).<br />

eff


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

3.3 The effect <strong>of</strong> both surface tension and buoyancy forces<br />

In this case the effect <strong>of</strong> both surface tension and buoyancy forces is studied <strong>in</strong> the presence<br />

<strong>of</strong> magnetic field. The relation between Q and the critical M eff for both stationary<br />

<strong>convection</strong> and overstability cases is displayed <strong>in</strong> figure (10) when � 0.<br />

01 for different<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

Da eff<br />

values <strong>of</strong> R . In both stationary <strong>convection</strong> and overstability cases as Q <strong>in</strong>creases the critical<br />

M eff <strong>in</strong>creases for l<strong>in</strong>ear and non-l<strong>in</strong>ear cases which <strong>in</strong>dicates that the magnetic field has a<br />

stabiliz<strong>in</strong>g effect. Moreover as � <strong>in</strong>creases M eff <strong>in</strong>creases which <strong>in</strong>dicates that the nonl<strong>in</strong>ear<br />

magnetic permeability has a stabiliz<strong>in</strong>g effect <strong>in</strong> the overstability case. We also note from the<br />

figure that as R <strong>in</strong>creases M eff decreases. The numerical results related to this figure are<br />

listed <strong>in</strong> tables (9a) when � � 0 and table (9b) when � � 0.<br />

5,<br />

1.<br />

Similar relations between<br />

M D and Q are displayed <strong>in</strong> figure (11).<br />

eff<br />

Figures (12) and (13) show the relation between R and M eff for different values <strong>of</strong> � ,<br />

when Q � 50,<br />

Da � 0.<br />

01and<br />

Bi � 0 . Clearly as the buoyancy forces <strong>in</strong>creases, the surface<br />

tension effect decreases and vice versa for the stationary <strong>convection</strong> and over stability cases<br />

and for l<strong>in</strong>ear and non-l<strong>in</strong>ear cases.<br />

4 CONCLUSION<br />

The onset <strong>of</strong> Benard-<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a horizontal porous layer permeated by a<br />

magnetohydrodynamic fluid with a non-l<strong>in</strong>ear magnetic permeability has been exam<strong>in</strong>ed. The<br />

porous layer is assumed to be governed by the Br<strong>in</strong>kman model, bounded from below by a<br />

rigid surface and from above by a non deformable free surface. This problem is exam<strong>in</strong>ed<br />

under the effect <strong>of</strong> surface tension only, the effect <strong>of</strong> buoyancy effect only and the effect <strong>of</strong><br />

both surface tension and buoyancy force.<br />

The critical effective <strong>Marangoni</strong> number and the critical <strong>Rayleigh</strong> number are obta<strong>in</strong>ed for<br />

different values <strong>of</strong> the effective Darcy number, Biot number and Chandrasekhar number. The<br />

non-l<strong>in</strong>ear magnetic permeability has no effect on the development <strong>of</strong> <strong>in</strong>stabilities through the<br />

mechanism <strong>of</strong> stationary <strong>convection</strong> and from the viewpo<strong>in</strong>t <strong>of</strong> terrestrial applications, this is<br />

frequently the preferred process and so we should not expect the non-l<strong>in</strong>ear magnetic<br />

permeability to manifest itself under terrestrial circumstances. However <strong>in</strong> non-terrestrial<br />

applications overstability is the preferred mechanism and <strong>in</strong> this situation, the presence <strong>of</strong> the<br />

non-l<strong>in</strong>ear magnetic permeability <strong>in</strong>fluences the onset <strong>of</strong> overstable <strong>convection</strong>.<br />

Numerical results are obta<strong>in</strong>ed for the critical effective <strong>Marangoni</strong> and <strong>Rayleigh</strong> numbers<br />

under the effects <strong>of</strong> magnetic field <strong>in</strong> the cases <strong>of</strong> stationary <strong>convection</strong> and overstability and<br />

some special cases are produced. It is shown that the results <strong>of</strong> Darcy model case can be<br />

recovered <strong>in</strong> the limit as Darcy number Daeff � 0 , while the results <strong>of</strong> fluid layer can be<br />

recovered <strong>in</strong> the limit as Darcy number Da � � .<br />

eff<br />

25


26<br />

R<br />

R<br />

150<br />

140<br />

130<br />

120<br />

110<br />

100<br />

90<br />

80<br />

70<br />

60<br />

50<br />

40<br />

30<br />

20<br />

10<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

0<br />

-1 0 1 2 3 4 5 6 7 8 9 10 11 12<br />

Figure 2: Variation <strong>of</strong> R with Bi for different values <strong>of</strong> Da eff<br />

1400<br />

1200<br />

1000<br />

800<br />

600<br />

400<br />

200<br />

0<br />

0 10 20 30 40 50 60 70 80 90 100 110<br />

Figure 3: Variation <strong>of</strong> R with Q when � 0.<br />

001.<br />

Q<br />

Bi<br />

Stationary Convection<br />

Da eff<br />

Daeff � 0.<br />

1<br />

Daeff � 0.<br />

01<br />

Daeff � 0.<br />

001<br />

�<br />

� 1<br />

� � 0. 5<br />

� � 0


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

Table 1: The critical values <strong>of</strong> the <strong>Rayleigh</strong> number R for different values <strong>of</strong> Q and Da eff<br />

for the cases <strong>of</strong> stationary <strong>convection</strong> and overstability. Here M eff � 0 , � � 0 , Bi � 0 ,<br />

Gm � 1.<br />

Q<br />

0<br />

1<br />

5<br />

10<br />

20<br />

30<br />

40<br />

50<br />

60<br />

70<br />

80<br />

90<br />

100<br />

stationary <strong>convection</strong><br />

Daeff � 0.<br />

01 Daeff � 0.<br />

001<br />

R a R a<br />

35.364 2.304 28.466 2.342<br />

53.225 2.684 45.345 2.808<br />

115.015 3.505 103.569 3.817<br />

184.723 4.091 169.423 4.557<br />

315.033 4.83 292.937 5.512<br />

439.561 5.331 411.334 6.173<br />

560.877 5.716 526.936 6.687<br />

680.060 6.031 640.706 7.111<br />

797.690 6.298 753.158 7.475<br />

914.118 6.532 864.599 7.794<br />

1029.582 6.739 975.236 8.079<br />

1144.246 6.926 1085.21 8.338<br />

1258.234 7.096 1194.63 8.574<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

overstability<br />

Daeff � 0.<br />

01 Daeff � 0.<br />

001<br />

R a R a<br />

- - - -<br />

- - - -<br />

105.016 3.711 88.819 4.153<br />

140.113 4.319 120.14 4.862<br />

195.204 4.963 171.52 5.684<br />

242.738 5.386 216.87 6.243<br />

286.314 5.709 259.06 6.681<br />

327.317 5.974 299.20 7.046<br />

366.466 6.200 337.87 7.363<br />

404.193 6.397 375.40 7.644<br />

440.783 6.573 412.02 7.899<br />

476.434 6.732 447.89 8.133<br />

511.294 6.878 483.12 8.350<br />

Table 2: The critical values <strong>of</strong> the <strong>Rayleigh</strong> number R for different values <strong>of</strong> Q , Da eff and<br />

� for the case <strong>of</strong> overstability. Here M eff � 0 , Gm � 1,<br />

0<br />

Q<br />

5<br />

10<br />

20<br />

30<br />

40<br />

50<br />

60<br />

70<br />

80<br />

90<br />

100<br />

Daeff � 0.<br />

01<br />

R<br />

114.951<br />

160.569<br />

232.639<br />

295.701<br />

354.201<br />

409.760<br />

463.203<br />

515.028<br />

565.559<br />

615.022<br />

663.584<br />

� � 0.<br />

5<br />

a<br />

3.518<br />

4.196<br />

4.861<br />

5.285<br />

5.607<br />

5.87<br />

6.094<br />

6.290<br />

6.465<br />

6.624<br />

6.769<br />

Daeff � 0.<br />

001<br />

R<br />

98.482<br />

138.89<br />

205.89<br />

266.00<br />

322.59<br />

376.91<br />

429.58<br />

481.01<br />

531.43<br />

581.01<br />

629.88<br />

a<br />

3.972<br />

4.748<br />

5.581<br />

6.139<br />

6.575<br />

6.938<br />

7.254<br />

7.535<br />

7.790<br />

8.025<br />

8.242<br />

Bi � , � 1,<br />

P � 0.<br />

25 .<br />

Daeff � 0.<br />

01<br />

R<br />

-<br />

179.81<br />

270.35<br />

349.96<br />

424.29<br />

495.3<br />

563.93<br />

630.75<br />

696.13<br />

760.32<br />

823.51<br />

a<br />

-<br />

4.752<br />

4.752<br />

5.181<br />

5.501<br />

5.761<br />

5.982<br />

6.174<br />

6.346<br />

6.502<br />

6.644<br />

Pr m<br />

� � 1<br />

Daeff � 0.<br />

001<br />

R<br />

-<br />

156.811<br />

239.888<br />

314.964<br />

386.098<br />

454.726<br />

521.568<br />

587.043<br />

651.426<br />

714.905<br />

777.618<br />

a<br />

-<br />

4.612<br />

5.483<br />

6.042<br />

6.474<br />

6.833<br />

7.145<br />

7.422<br />

7.673<br />

7.904<br />

8.118<br />

27


28<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

Table 3: The critical values <strong>of</strong> <strong>Rayleigh</strong> number R for different values <strong>of</strong> Q for the case <strong>of</strong><br />

stationary <strong>convection</strong>. Here Bi � 0 ,<br />

R<br />

1400<br />

1200<br />

1000<br />

800<br />

600<br />

400<br />

200<br />

0<br />

Q f<br />

0<br />

4<br />

10 �<br />

10 �<br />

10<br />

3<br />

2<br />

10<br />

6<br />

Daeff � 10 , Gm � 1,<br />

M � 0 (fluid layer).<br />

(Biswal and Rao 1999)<br />

R a<br />

668.998 2.086<br />

669.000<br />

669.020<br />

874.862<br />

2424.90<br />

2.086<br />

2.086<br />

2.288<br />

3.128<br />

Q<br />

0<br />

10<br />

10<br />

10<br />

2<br />

3<br />

7<br />

8<br />

10<br />

Present<br />

R a<br />

668.998 2.0856<br />

669.000<br />

669.020<br />

874.862<br />

2424.903<br />

stationary <strong>convection</strong><br />

2.0856<br />

2.0856<br />

2.2884<br />

3.1281<br />

� � 1<br />

� � 0.<br />

5<br />

� � 0<br />

0 10 20 30 40 50 60 70 80 90 100 110<br />

Q<br />

�6 Figure 4: Variation <strong>of</strong> R with Q when Da � 10 , � � 0.<br />

3 (Darcy model)


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

Table 4: The critical values <strong>of</strong> the <strong>Rayleigh</strong> number R for different values <strong>of</strong> Q and � for<br />

�6<br />

the cases <strong>of</strong> stationary <strong>convection</strong> and overstability. Here M � 0 , Da � 10 , Bi � 0 ,<br />

� � 0. 3 (Darcy's Model).<br />

stationary <strong>convection</strong><br />

overstability<br />

Q � � 0 � � 0.<br />

5 � � 1<br />

R a R a R a R a<br />

0 27.098 2.326 - - - - - -<br />

1 43.651 2.810 - - - - - -<br />

5 100.903 3.857 86.541 4.204 95.837 4.014 - -<br />

10 165.800 4.630 117.759 4.950 136.112 4.834 153.493 4.697<br />

20 287.701 5.639 169.050 5.810 202.934 5.708 236.180 5.615<br />

30 404.671 6.343 214.577 6.393 263.166 6.291 311.134 6.204<br />

40 518.949 6.896 257.150 6.848 320.111 6.745 382.389 6.659<br />

50 631.464 7.356 297.838 7.228 374.975 7.123 451.340 7.036<br />

60 742.709 7.753 337.186 7.556 428.369 7.451 518.674 7.363<br />

70 852.982 8.103 375.520 7.848 480.649 7.741 584.790 7.653<br />

80 962.481 8.417 413.047 8.111 532.048 8.004 649.943 7.915<br />

90 1071.344 8.704 449.915 8.352 582.724 8.245 714.311 8.155<br />

100 1179.673 8.967 486.229 8.575 632.795 8.468 778.019 8.378<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

29<br />

Gm � 1,<br />

Table 5: The critical values <strong>of</strong> the <strong>Rayleigh</strong> number R for different values <strong>of</strong> Q for the case<br />

�6<br />

<strong>of</strong> stationary <strong>convection</strong>. Here M � 0 , Da � 10 , � �<br />

model).<br />

Q<br />

0<br />

1<br />

5<br />

10<br />

25<br />

50<br />

100<br />

150<br />

200<br />

500<br />

(Bhadauria & Sherani 2008)<br />

R a<br />

39.4784 3.142<br />

57.5243 3.736<br />

117.4382 4.917<br />

183.9028 5.721<br />

367.12992 7.094<br />

654.1856 8.395<br />

1205.0762 9.959<br />

1742.7393 11.013<br />

2273.5117 11.829<br />

5396.3447 14.863<br />

Bi , Gm � 1,<br />

� � 0. 3,<br />

(Darcy's<br />

Present<br />

R a<br />

39.4784 3.1416<br />

57.5242 3.7360<br />

117.4382 4.9169<br />

183.9027 5.7213<br />

367.1298 7.0940<br />

654.1854 8.3954<br />

1205.0760 9.9593<br />

1742.7391 11.0127<br />

2273.5114 11.8290<br />

5396.3642 14.8631


30<br />

M eff<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

Figure 5: Variation <strong>of</strong> M eff with Bi for different values <strong>of</strong> Da eff for the case <strong>of</strong> stationary<br />

<strong>convection</strong>.<br />

M<br />

Deff<br />

8000<br />

7000<br />

6000<br />

5000<br />

4000<br />

3000<br />

2000<br />

1000<br />

0<br />

-1 0 1 2 3 4 5 6 7 8 9 10 11 12<br />

60<br />

50<br />

40<br />

30<br />

20<br />

10<br />

Daeff � 0.<br />

001<br />

Daeff � 0.<br />

01<br />

Daeff � 0.<br />

1<br />

0<br />

-1 0 1 2 3 4 Bi5<br />

6 7 8 9 10 11<br />

Figure 6: Variation <strong>of</strong> M D with Bi for different values <strong>of</strong> Da eff<br />

eff for the case <strong>of</strong> stationary<br />

<strong>convection</strong>.<br />

Bi<br />

Daeff � 0.<br />

1<br />

Daeff � 0.<br />

01<br />

Daeff � 0.<br />

001


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

Figure 7: Variation <strong>of</strong> M eff with Q when 0.<br />

001<br />

and overstability.<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

Da eff � for the cases <strong>of</strong> stationary <strong>convection</strong><br />

Table 6: The critical values <strong>of</strong> the effective <strong>Marangoni</strong> number M for different values <strong>of</strong> eff<br />

Q<br />

and Da eff for the cases <strong>of</strong> stationary <strong>convection</strong> and overstability. Here R � 0 , � � 0 , Gm � 1<br />

Q<br />

0<br />

1<br />

5<br />

10<br />

20<br />

30<br />

40<br />

50<br />

60<br />

70<br />

80<br />

90<br />

100<br />

M eff<br />

140000<br />

120000<br />

100000<br />

80000<br />

60000<br />

40000<br />

20000<br />

stationary <strong>convection</strong><br />

Daeff � 0.<br />

01<br />

Daeff � 0.<br />

001<br />

M eff<br />

0<br />

372.210<br />

555.645<br />

1192.774<br />

1916.229<br />

3276.515<br />

4581.733<br />

5855.973<br />

7109.293<br />

8347.082<br />

9572.645<br />

10788.193<br />

11995.302<br />

13195.157<br />

0 10 20 30 40 50 60 70 80 90 100 110<br />

a<br />

2.423<br />

2.831<br />

3.701<br />

4.320<br />

5.110<br />

5.659<br />

6.088<br />

6.446<br />

6.754<br />

7.026<br />

7.270<br />

7.492<br />

7.696<br />

M eff<br />

2510.291<br />

4066.627<br />

9576.203<br />

15941.503<br />

28058.077<br />

39792.908<br />

51325.706<br />

62732.865<br />

74054.876<br />

85315.803<br />

96531.023<br />

107710.865<br />

118862.548<br />

Q<br />

a<br />

2.978<br />

3.616<br />

5.044<br />

6.100<br />

7.444<br />

8.363<br />

9.085<br />

9.695<br />

10.232<br />

10.720<br />

11.170<br />

11.590<br />

11.984<br />

stationary <strong>convection</strong><br />

overstability<br />

Daeff � 0.<br />

01 Daeff � 0.<br />

001<br />

M eff<br />

-<br />

-<br />

-<br />

1786.695<br />

2704.793<br />

3550.948<br />

4361.087<br />

5148.130<br />

5918.691<br />

6676.712<br />

7424.767<br />

8164.650<br />

8897.668<br />

a<br />

-<br />

-<br />

-<br />

4.547<br />

5.274<br />

5.754<br />

6.131<br />

6.448<br />

6.724<br />

6.971<br />

7.194<br />

7.399<br />

7.589<br />

�<br />

� 1<br />

� � 0. 5<br />

M eff<br />

� � 0<br />

-<br />

-<br />

9528.054<br />

14200.385<br />

22553.054<br />

30367.372<br />

37898.299<br />

45244.811<br />

52458.225<br />

59569.405<br />

66598.709<br />

73560.439<br />

80465.111<br />

31<br />

a<br />

-<br />

-<br />

5.234<br />

6.022<br />

6.912<br />

7.574<br />

8.126<br />

8.607<br />

9.040<br />

9.435<br />

9.799<br />

10.141<br />

10.461


32<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

Table 7: The critical values <strong>of</strong> effective <strong>Marangoni</strong> number M for different values <strong>of</strong> eff<br />

Q ,<br />

Da eff and � for the case <strong>of</strong> overstability. Here R � 0 , Gm � 1,<br />

Bi � 0 , Pr � 1 , Pm<br />

� 0.<br />

25 .<br />

Q<br />

20<br />

30<br />

40<br />

50<br />

60<br />

70<br />

80<br />

90<br />

100<br />

M Deff<br />

Daeff � 0.<br />

01<br />

M eff<br />

3145.795<br />

4195.587<br />

5205.536<br />

6190.848<br />

7158.764<br />

8113.486<br />

9057.731<br />

9993.375<br />

10921.786<br />

140<br />

120<br />

100<br />

80<br />

60<br />

40<br />

20<br />

0<br />

a<br />

5.141<br />

5.659<br />

6.055<br />

6.386<br />

6.672<br />

6.928<br />

7.159<br />

7.371<br />

7.568<br />

� � 0.<br />

5<br />

Daeff � 0.<br />

001<br />

M eff<br />

26132.837<br />

35676.511<br />

44942.342<br />

54027.250<br />

62981.200<br />

71834.278<br />

80606.382<br />

89311.527<br />

97960.038<br />

a<br />

7.123<br />

7.882<br />

8.501<br />

9.037<br />

9.517<br />

9.948<br />

10.346<br />

10.716<br />

11.062<br />

Daeff � 0.<br />

01<br />

M eff<br />

-<br />

-<br />

-<br />

-<br />

8295.793<br />

9429.100<br />

10550.448<br />

11662.067<br />

12765.521<br />

Q<br />

Figure 8: Variation <strong>of</strong> M Deff<br />

with Q when 0.<br />

001<br />

and overstability.<br />

a<br />

-<br />

-<br />

-<br />

-<br />

6.491<br />

6.748<br />

6.978<br />

7.187<br />

7.380<br />

stationary <strong>convection</strong><br />

� � 1<br />

Daeff � 0.<br />

001<br />

M eff<br />

-<br />

-<br />

-<br />

61996.042<br />

72501.979<br />

82903.700<br />

93221.716<br />

103470.448<br />

113660.512<br />

�<br />

� 1<br />

� � 0. 5<br />

� � 0<br />

0 10 20 30 40 50 60 70 80 90 100 110<br />

a<br />

-<br />

-<br />

-<br />

9.196<br />

9.690<br />

10.134<br />

10.535<br />

10.906<br />

11.252<br />

Da eff � for the cases <strong>of</strong> stationary <strong>convection</strong>


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

Figure 9: Variation <strong>of</strong> M eff with Q for different values <strong>of</strong> Bi for the case <strong>of</strong> stationary<br />

<strong>convection</strong>.<br />

Table 8: The critical values <strong>of</strong> the effective <strong>Marangoni</strong> number M for different values <strong>of</strong><br />

eff<br />

Da , Q , � and Bi when the rigid boundary is thermally <strong>in</strong>sulated ( D� � 0).<br />

Here R � 0 .<br />

eff<br />

Stationary<br />

<strong>convection</strong><br />

(Shivakumara<br />

et al. 2009)<br />

Stationary<br />

<strong>convection</strong><br />

overstability<br />

M eff<br />

�<br />

0<br />

0<br />

0<br />

0.5<br />

110000<br />

100000<br />

90000<br />

80000<br />

70000<br />

60000<br />

50000<br />

40000<br />

30000<br />

20000<br />

10000<br />

Q<br />

0<br />

0<br />

50<br />

50<br />

Bi � 0.<br />

5<br />

Bi � 0<br />

0<br />

0 10 20 30 40 50 60 70 80 90 100 110<br />

Da eff<br />

0.1<br />

0.01<br />

0.1<br />

0.01<br />

0.1<br />

0.01<br />

0.1<br />

0.01<br />

0.1<br />

0.01<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

M<br />

eff<br />

71.974<br />

281.212<br />

71.975<br />

281.212<br />

934.985<br />

7105.323<br />

837.180<br />

5150.602<br />

-<br />

6195.369<br />

Bi � 0<br />

Q<br />

a<br />

0.000<br />

0.000<br />

0.0009<br />

0.001<br />

3.6778<br />

6.3872<br />

4.2934<br />

6.4645<br />

-<br />

6.4142<br />

M<br />

Bi � 1<br />

eff<br />

138.862<br />

479.214<br />

138.862<br />

479.213<br />

1159.569<br />

8092.848<br />

955.449<br />

5636.163<br />

1110.327<br />

6760.099<br />

Daeff � 0.<br />

001<br />

Daeff � 0.<br />

01<br />

a<br />

1.856<br />

2.330<br />

1.8555<br />

2.3297<br />

4.6088<br />

7.9840<br />

4.3950<br />

6.6876<br />

4.2907<br />

6.7381<br />

M<br />

eff<br />

318.279<br />

978.212<br />

318.279<br />

978.207<br />

1899.077<br />

Bi � 5<br />

11159.449<br />

1418.54<br />

7521.53<br />

1638.310<br />

8887.256<br />

a<br />

33<br />

2.467<br />

3.467<br />

2.4672<br />

3.4677<br />

5.9456<br />

11.2008<br />

4.5593<br />

7.2806<br />

4.5294<br />

7.5529


34<br />

M eff<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

Figure 10: Variation <strong>of</strong> M eff with Q when 0.<br />

01<br />

<strong>convection</strong> and overstability.<br />

M Deff<br />

12000<br />

11000<br />

10000<br />

9000<br />

8000<br />

7000<br />

6000<br />

5000<br />

4000<br />

3000<br />

2000<br />

1000<br />

Figure 11: Variation <strong>of</strong><br />

and overstability.<br />

0<br />

0 10 20 30 40 50 60 70 80 90 100 110<br />

120<br />

110<br />

100<br />

90<br />

80<br />

70<br />

60<br />

50<br />

40<br />

30<br />

20<br />

10<br />

0<br />

Q<br />

M with Q when 0.<br />

01<br />

Deff<br />

stationary <strong>convection</strong><br />

Q<br />

R � 30<br />

Da eff � for the cases <strong>of</strong> stationary<br />

stationary <strong>convection</strong><br />

R � 10<br />

R � 30<br />

R � 10<br />

� � 1<br />

�<br />

� 1<br />

0 10 20 30 40 50 60 70 80 90 100 110<br />

� � 0. 5<br />

� � 0<br />

� � 0. 5<br />

� � 0<br />

Da eff � for the cases <strong>of</strong> stationary <strong>convection</strong>


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

Table 9a: The critical values <strong>of</strong> effective <strong>Marangoni</strong> number M eff for different values <strong>of</strong> Q<br />

and R when � � 0 for the cases <strong>of</strong> stationary <strong>convection</strong> and overstability. Here Daeff � 0.<br />

01 ,<br />

Bi � 0 , Gm � 1<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

R<br />

10<br />

30<br />

Q<br />

1<br />

5<br />

10<br />

20<br />

30<br />

40<br />

50<br />

60<br />

70<br />

80<br />

90<br />

100<br />

1<br />

5<br />

10<br />

20<br />

30<br />

40<br />

50<br />

60<br />

70<br />

80<br />

90<br />

100<br />

stationary <strong>convection</strong><br />

� � 0<br />

overstability<br />

M eff a M eff a<br />

472.997 2.710 - -<br />

1122.869 3.580 - -<br />

1853.524 4.209 1709.164 4.472<br />

3221.299 5.015 2626.743 5.209<br />

4530.832 5.574 3472.709 5.695<br />

5808.045 6.012 4282.908 6.077<br />

7063.604 6.377 5070.153 6.397<br />

8303.170 6.690 5840.994 6.676<br />

9530.192 6.967 6599.341 6.925<br />

10746.973 7.214 7347.747 7.150<br />

11955.143 7.439 8087.994 7.356<br />

13155.920 7.645 8821.382 7.548<br />

280.728<br />

967.507<br />

1717.467<br />

3104.218<br />

4424.188<br />

5708.393<br />

6969.108<br />

8212.704<br />

9443.001<br />

10662.517<br />

11873.022<br />

13075.823<br />

2.593<br />

3.393<br />

4.013<br />

4.834<br />

5.410<br />

5.863<br />

6.239<br />

6.562<br />

6.847<br />

7.101<br />

7.332<br />

7.543<br />

-<br />

-<br />

1539.181<br />

2459.244<br />

3306.751<br />

4118.330<br />

4906.881<br />

5678.987<br />

6438.544<br />

7188.113<br />

7929.477<br />

8663.937<br />

-<br />

-<br />

4.343<br />

5.092<br />

5.588<br />

5.976<br />

6.302<br />

6.585<br />

6.837<br />

7.065<br />

7.275<br />

7.469<br />

35


36<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

Table 9b: The critical values <strong>of</strong> effective <strong>Marangoni</strong> number M eff for different values <strong>of</strong><br />

Q , R and � for the case <strong>of</strong> overstability. Here Daeff � 0.<br />

01,<br />

Bi � 0 , Gm � 1.<br />

R<br />

10<br />

30<br />

Q<br />

20<br />

30<br />

40<br />

50<br />

60<br />

70<br />

80<br />

90<br />

100<br />

20<br />

30<br />

40<br />

50<br />

60<br />

70<br />

80<br />

90<br />

100<br />

� � 0.<br />

5<br />

M eff a<br />

3073.969 5.079<br />

4125.166 5.602<br />

5136.297 6.002<br />

6122.658 6.336<br />

7091.526 6.625<br />

8047.122 6.883<br />

8992.175 7.116<br />

9928.571 7.330<br />

10857.685 7.528<br />

2920.928<br />

3976.764<br />

4991.435<br />

5980.727<br />

6952.126<br />

7909.962<br />

8857.029<br />

9795.259<br />

10726.058<br />

4.964<br />

5.495<br />

5.902<br />

6.241<br />

6.535<br />

6.796<br />

7.032<br />

7.249<br />

7.450<br />

M eff<br />

� � 1<br />

-<br />

-<br />

-<br />

-<br />

8235.576<br />

9370.016<br />

10492.372<br />

11604.892<br />

12709.178<br />

-<br />

-<br />

-<br />

6958.300<br />

8111.313<br />

9248.440<br />

10373.148<br />

11487.757<br />

12593.924<br />

a<br />

-<br />

-<br />

-<br />

-<br />

6.447<br />

6.706<br />

6.938<br />

7.149<br />

7.344<br />

-<br />

-<br />

-<br />

6.062<br />

6.362<br />

6.625<br />

6.860<br />

7.075<br />

7.272


8000<br />

7000<br />

6000<br />

5000<br />

M<br />

4000<br />

3000<br />

2000<br />

1000<br />

0<br />

Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

eff<br />

0 50 100 150 200 250 300 350<br />

Figure 12: Variation <strong>of</strong> M efff with R when Q � 50,<br />

0.<br />

01<br />

R<br />

800<br />

700<br />

600<br />

500<br />

400<br />

300<br />

200<br />

100<br />

0<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

Bi � .<br />

Da eff � , 0<br />

� � 0. 5<br />

0 50 100 150 200 250 300 350<br />

Figure 13: Variation <strong>of</strong> R with M efff when Q � 50,<br />

0.<br />

01<br />

M<br />

R<br />

stationary <strong>convection</strong><br />

eff<br />

stationary <strong>convection</strong><br />

� � 1<br />

� � 0. 5<br />

� � 0<br />

� � 1<br />

� � 0<br />

Overstability<br />

Bi � .<br />

Da eff � , 0<br />

Overstability<br />

37


38<br />

REFERENCES<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

Abdullah A (2000). Thermosolutal<strong>convection</strong> <strong>in</strong> a non-l<strong>in</strong>ear magnetic fluid. International<br />

Journal <strong>of</strong> Thermal Sciences 39, pp. 273-284.<br />

Abdullah A and L<strong>in</strong>dsay K (1990). Benard <strong>convection</strong> <strong>in</strong> a non-l<strong>in</strong>ear magnetic fluid. Acta<br />

Mechanica 85, pp. 27-42.<br />

Abdullah A and L<strong>in</strong>dsay K (1991). Benard <strong>convection</strong> <strong>in</strong> a non-l<strong>in</strong>ear magnetic fluid under<br />

the <strong>in</strong>fluence <strong>of</strong> a non-vertical magnetic field. Cont<strong>in</strong>uum Mechanics and Thermodynamics 3,<br />

pp. 13-25.<br />

Al-aidrous A and Abdullah A (2005). <strong>Marangoni</strong> <strong>in</strong>stability <strong>in</strong> a non-l<strong>in</strong>ear magnetic fluid.<br />

Wseas Transactions on Mathematics 4, pp. 415-422.<br />

Awang S and Hashim I (2008). Onset <strong>of</strong> <strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> variable-viscosity fluid<br />

layer subject to uniform heat flux from below. International Journal <strong>of</strong> Heat and Mass<br />

Transfer 35, pp. 948–956.<br />

Banjar H and Abdullah A (2011). Thermal <strong>in</strong>stability <strong>in</strong> superposed porous and fluid layers <strong>in</strong><br />

the presence <strong>of</strong> Coriolis forces. International Journal <strong>of</strong> Applied Mathematics and Mechanics.<br />

7(9), pp. 13-27.<br />

Benard H (1900). Les tourbillons cellulaires dans une nappe liquide. Revue générale des<br />

Sciences Pures et Appliquées 11, pp. 1261–1271and 1309- 1328.<br />

Benard H (1901). Les tourbillons cellulaires dans une nappe liquid transportant de lachaleur<br />

par <strong>convection</strong> enrégime permanent. Annales de Chimie et de Physique 23, pp. 62-144.<br />

Bhadauria S and Sherani A (2008). Onset <strong>of</strong> Darcy-Convection <strong>in</strong> a Magnetic-Fluid-<br />

Saturated Porous Medium Subject to Temperature Modulation <strong>of</strong> the Boundaries. Transport <strong>in</strong><br />

Porous Media 73, pp. 349–368.<br />

Biswal P and Rao A (1999). The onset <strong>of</strong> steady Bénard-<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a two-layer<br />

system <strong>of</strong> conduct<strong>in</strong>g fluid <strong>in</strong> the presence <strong>of</strong> a uniform magnetic field. Journal <strong>of</strong><br />

Eng<strong>in</strong>eer<strong>in</strong>g Mathematics 35, pp. 385–404.<br />

Bukhari A F (1997). A spectral eigenvalue method for multi layered cont<strong>in</strong>uua. Ph. D. thesis,<br />

University <strong>of</strong> Glasgow.<br />

Chandrasekhar S (1981). Hydrodynamic and Hydromagnetic Stability.New York. Dover.<br />

Chang F and Chiang K (1998). Oscillatory <strong>in</strong>stability analysis <strong>of</strong> Benard-<strong>Marangoni</strong><br />

<strong>convection</strong> <strong>in</strong> a rotat<strong>in</strong>g fluid under a uniform magnetic field. International Journal <strong>of</strong> Heat<br />

and Mass Transfer 41/17, pp. 2667-2675.<br />

Char M and Chiang K (1994). Stability analysis <strong>of</strong> Benard–<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> fluids<br />

with <strong>in</strong>ternal heat generation. Journal <strong>of</strong> Physics D Applied Physics27, pp. 748–755.


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

Cowley M D and Rosensweig R E (1967). The <strong>in</strong>terfacial <strong>Instability</strong> <strong>of</strong> a magnetic<br />

fluid.Journal <strong>of</strong> Fluid Mechanics 30, pp. 671 -688.<br />

Douiebe A, Hannaoui G, Benaboud A. and Khmou A (2001). Effects <strong>of</strong> a.c. electric field and<br />

rotation on Bénard–<strong>Marangoni</strong> <strong>convection</strong>. Flow Turbulence and Combustion 67, pp. 185–<br />

204.<br />

Dubey A, S<strong>in</strong>gh U. and Jha R (2011). Mixed Convection <strong>of</strong> non-newtonian fluids through<br />

porous medium along heated vertical flat plate with magnetic. International Journal <strong>of</strong><br />

Applied Mathematics and Mechanics. 7(19), pp. 19-31.<br />

Fadzillah N, Arif<strong>in</strong> N, Nazar R, Ismail F. and Suleiman M (2008). <strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a<br />

fluid saturated porous layer with a prescribed heat flux at its lower boundary. European<br />

Journal <strong>of</strong> Scientific Research 24, ppPP. 477- 486.<br />

Friedrich R and Rudraiah N (1984). <strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a rotat<strong>in</strong>g fluid layer with nonuniform<br />

temperature gradient. International Journal <strong>of</strong> Heat and Mass Transfer 27, pp. 443-<br />

449.<br />

Gailitis A (1977). Formation <strong>of</strong> the hexagonal pattern on the surface <strong>of</strong> a ferromagnetic fluid<br />

<strong>in</strong> an applied magnetic field. Journal <strong>of</strong> Fluid Mechanics 82, pp. 401-413.<br />

Garcia-Ybarra P, Gastillo J. and Velarde M (1987.). A nonl<strong>in</strong>ear evolution for Benard–<br />

<strong>Marangoni</strong> <strong>convection</strong> with deformable boundary. Physics Letters A 122, pp. 107–110.<br />

Guray E and Tarman H (2007). Thermal <strong>convection</strong> <strong>in</strong> the presence <strong>of</strong> a vertical magnetic<br />

field. Acta Mechanica 194, pp. 33–46.<br />

Hashim I and Arif<strong>in</strong> N (2003). Oscillatory <strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a conduct<strong>in</strong>g fluid layer<br />

with a deformable free surface <strong>in</strong> the presence <strong>of</strong> a vertical magnetic field. Acta Mechanica<br />

164, pp.199–215.<br />

Hashim I and Wilson S (1999). The effect <strong>of</strong> a uniform vertical magnetic field on the l<strong>in</strong>ear<br />

growth rates <strong>of</strong> steady <strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a horizontal layer <strong>of</strong> conduct<strong>in</strong>g fluid.<br />

International Journal <strong>of</strong> Heat and Mass Transfer 42, pp.525- 533.<br />

Hennenberg M, Saghir M, Rednikov A. and Legros J (1997). Porous media and theBenard–<br />

<strong>Marangoni</strong> problem. Transport <strong>in</strong> Porous Media 27, pp.327–35.<br />

Horton C and Rogers F (1945). Convection Currents <strong>in</strong> a Porous Medium. Journal <strong>of</strong> Applied<br />

Physics 16, pp. 367-370.<br />

Ingham D and Pop I (1998) (Eds.). Transport Phenomena <strong>in</strong> Porous Media. Pergamon,<br />

Oxford.<br />

Jan A and Abdullah A (2000). Benard <strong>convection</strong> <strong>in</strong> a horizontal porous layer permeated by a<br />

conduct<strong>in</strong>g fluid <strong>in</strong> the presence <strong>of</strong> a vertical magnetic field. Institute for Scientific Research.<br />

Applied Science Research Center. Umm Al-Qura University.<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

39


40<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

A. A. Abdullah and Z. Z. Rashed<br />

Jou J, Kung K. and Hsu C (1997). Effects <strong>of</strong> Coriolis force and surface tension on Benard-<br />

<strong>Marangoni</strong> convective <strong>in</strong>stability. International Journal <strong>of</strong> Heat and Mass Transfer 40/6, pp.<br />

1447-1458.<br />

Kaviany M (1995). Pr<strong>in</strong>ciples <strong>of</strong> Heat Transfer <strong>in</strong> Porous Media, second ed. Spr<strong>in</strong>ger-Verlag,<br />

New York.<br />

Lapwood E (1948). Convection <strong>of</strong> a fluid <strong>in</strong> a porous media. Proc.Cambridge philos. Soc 44,<br />

PP. 508- 521.<br />

Lombardo S and Mulone G (2005). Necessary and sufficient stability conditions via the<br />

eigenvalues–eigenvectors method: an application to the magnetic Bénard problem. Nonl<strong>in</strong>ear<br />

Analysis 63, pp. 2091– 2101.<br />

Maekawa T and Tanasawa I (1998). Effect <strong>of</strong> magnetic field and buoyancy <strong>of</strong> onset<br />

<strong>Marangoni</strong> <strong>convection</strong>.. International Journal <strong>of</strong> Heat and Mass Transfer 32/7, PP.1377-1380.<br />

Muzikar P and Pethick C J (1981). Flux bunch<strong>in</strong>g <strong>in</strong> type II superconductor. Physical Review<br />

24, pp. 2533-2539.<br />

Nield D (1964). Surface tension and buoyancy effects <strong>in</strong> cellular <strong>convection</strong>. Journal <strong>of</strong> Fluid<br />

Mechanics 19, pp. 341-352.<br />

Nield D (1966). Surface tension and buoyancy effects <strong>in</strong> cellular <strong>convection</strong> <strong>of</strong> an electrically<br />

conduct<strong>in</strong>g liquid <strong>in</strong> a magnetic field. Zeitschrift für Angewandte Mathematik und Physik<br />

17/1, pp. 131-139.<br />

Nield D and Bejan A (2006). Convection <strong>in</strong> Porous Media. Third Edition. Spr<strong>in</strong>ger. New<br />

York.<br />

Pearson J (1958). On <strong>convection</strong> cells <strong>in</strong>duced by surface tension. Journal <strong>of</strong> Fluid Mechanics<br />

4, pp. 489–500.<br />

Perez-Garcia C and Carneiro G (1991.) L<strong>in</strong>ear stability analysis <strong>of</strong> Benard–<strong>Marangoni</strong><br />

<strong>convection</strong> <strong>in</strong> fluids with a deformable free surface. Physics <strong>of</strong> Fluids 3 (2), pp. 292–298<br />

<strong>Rayleigh</strong> L (1916).On <strong>convection</strong> currents <strong>in</strong> a horizontal layer <strong>of</strong> fluid when the temperature<br />

is on the under side. Philosophical Magaz<strong>in</strong>e 32, pp. 529–547.<br />

Roberts P H (1981). Equilibria and stability <strong>of</strong> a fluid type II superconductor. Quarterly<br />

Journal <strong>of</strong> Mechanics and Applied Mathematics 34 (3), pp. 327-343<br />

Rudraiah N and Prasad V (1998).. Effect <strong>of</strong> Br<strong>in</strong>kman boundary layer on the onset <strong>of</strong><br />

<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a fluid-saturated porous layer. Acta Mechanica 127, pp. 235-246.<br />

Shivakumara I, Nanjundappa C. and Chavaraddi B. (2009). Darcy–Benard–<strong>Marangoni</strong><br />

<strong>convection</strong> <strong>in</strong> porous media. International Journal <strong>of</strong> Heat and Mass Transfer 52, pp. 2815–<br />

2823.


Benard-<strong>Marangoni</strong> Convection In A Porous Layer Permeated By A Non-L<strong>in</strong>ear Magnetic Fluid<br />

Takashima M and Namikawa T (1971). Surface-tension-driven <strong>convection</strong> under the<br />

simultaneous action <strong>of</strong> a magnetic field and rotation. Physics Letters 37A, pp. 55–56.<br />

Takashima M (1981a). Surface tension driven <strong>in</strong>stability <strong>in</strong> a horizontal liquid layer with a<br />

deformable free surface. I. Stationary <strong>convection</strong>. Journal <strong>of</strong> the Physical Society <strong>of</strong> Japan 50,<br />

pp. 2745–2750.<br />

Takashima M (1981b). Surface tension driven <strong>in</strong>stability <strong>in</strong> a horizontal liquid layer with a<br />

deformable free surface. II. Overstability. Journal <strong>of</strong> the Physical Society <strong>of</strong> Japan 50,<br />

pp.2751–2756.<br />

Thompson W (1951).. Thermal <strong>convection</strong> <strong>in</strong> a magnetic field. Philosophical Magaz<strong>in</strong>e<br />

Series7, pp. 1417-1432.<br />

Vafai K (2005).Handbook <strong>of</strong> Porous Media. Taylor& Francis.<br />

Wilson S (1993).The effect <strong>of</strong> a uniform magnetic field on the onset <strong>of</strong> steady Benard<br />

<strong>Marangoni</strong> <strong>convection</strong> <strong>in</strong> a layer <strong>of</strong> conduct<strong>in</strong>g fluid. Journal <strong>of</strong> Eng<strong>in</strong>eer<strong>in</strong>g Mathematics 27,<br />

pp. 161-188.<br />

Int. J. <strong>of</strong> Appl. Math and Mech. 8 (12): 13-41, 2012.<br />

41

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!