You are on page 1of 19

~ ) Pergamon

PII :

Int..L Heat Mass Transfer. Vol. 40, No. 12, pp. 2755-2773, 1997 1997 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0017 9310/97 $17.00+0.00

S0017-9310(96)00354-7

Combined heat and mass transfer by natural convection with opposing buoyancy effects in a fluid saturated porous medium
D. ANGIRASA and G. P. PETERSONt Department of Mechanical Engineering, Texas A&M University, College Station, TX 77843-3123, U.S.A. and 1. POP Faculty of Mathematics, University of Cluj, R-3400 Cluj, CP 253, Romania
(Received 15 January 1996 and in final form 16 October 1996)

Ahstraet--A numerical study of combined heat and mass transfer by natural convection adjacent to vertical surfaces situated in fluid-saturated porous media is reported. Special attention is given to opposing buoyancy effects of the same order and unequal thermal and species diffusion coefficients. The numerical results support the validity of the boundary layer analysis for high Rayleigh number aiding flows and for opposing flows when one of the buoyant forces overpowers the other. For other cases, such as low Rayleigh number and opposing buoyant forces of the same order of magnitude, full solutions are needed. The structure of the flow, temperature and concentration fields are governed by complex interactions among the diffusion rates and the buoyancy ratio. Numerical results for velocity, temperature and concentration profiles are presented for an extensiverange of parameters and the complexphysical mechanismsunderlying the flow and transport are clearly explained. Nusselt and Sherwood number data are given and the more subtle aspects of the problem are clarified. 1997 Elsevier Science Ltd.

1. I N T R O D U C T I O N

Natural convection flows due to the combined buoyancy effects of thermal and species diffusion in a fluid saturated porous medium have many applications, such as geothermal fields, soil pollution, fibrous insulation and nuclear-waste disposal. The nature of natural convection flows in porous media due to thermal buoyancy alone is well-documented in literature [l, 2]. Substantial amount of work has also been done for viscous fluid flows with combined buoyancies [35]. Comparatively less work exists on the buoyancy induced natural convective flows resulting from the combined buoyancy effects in porous media. A review of natural convective flows due to the combined buoyant mechanisms in porous media was presented by Nield and Bejan [6]. The current investigation focuses on the transport phenomena adjacent to vertical surfaces, with special attention to the opposing buoyancy effects. Raptis et al. [7] presented a series solution considering wall-shear and suction for boundary layer flows with constant wall temperature and cont Author to whom correspondence should be addressed.

centration of the diffusing species. In a later work [8], constant wall flux conditions were considered, followed by an analysis of unsteady flows [9]. Based on the similarity analysis of Cheng and Minkowycz [10] for Darcian natural convection flows with thermal buoyancy alone, Bejan and Khair [11] studied the vertical natural convective flows due to the combined buoyancy effects of thermal and species diffusion. They also presented an order of magnitude analysis of the boundary layer equations which yields functional relations for the Nusselt and Sherwood numbers in limiting cases. Boundary layer analysis was shown to be invalid when the two buoyant mechanisms oppose each other, and are of the same order. Hasan and Majumdar [12] obtained a similarity solution for the flow around a long horizontal cylinder, with the second buoyant force resulting from a transpirationinduced concentration gradient. Jang and Chang [13] obtained finite-difference solutions for boundary layer type flows adjacent to inclined surfaces. Heat and mass transfer rates were shown to increase with buoyancy ratio and the wall inclination. In a related study, Jang and Chang [14] analyzed the vortex instability of natural convection in a porous medium resulting from the combined heat

2755

2756

D. ANGIRASA et al.

NOMENCLATURE

buoyancy ratio (equation 5) concentration on the diffusing species nondimensional concentration (equation 6) Cp specific heat D diffusion coefficient of the species g gravitational acceleration t7 average heat transfer coefficient (equation 12) hm average mass transfer coefficient (equation 13) ku thermal conductivity of the stagnant medium K permeability L height of the surface Le Lewis number = ~/D Nu average Nusselt number (equation 12) p pressure Ra Rayleigh number = 9flAtKL/vo~ Sh average Sherwood number (equation 13) t temperature T nondimensional temperature (equation 6) u, v velocity components U, V nondimensional velocity components (equation 6)

B c C

convective velocity = ,q[~AtK/v V~ space coordinates x, y X, Y nondimensional space coordinates (equation 6). Greek symbols c~ thermal diffusivity ko/(pCp) f coefficient of thermal fi expansion = -- 1/p(@/&)p,~ volumetric coefficient due to fl* concentration = - 1,:p(@/&+)p.t porosity 0 kinematic viscosity v density of the fluid p ratio of heat capacities of the a stagnant medium and the fluid (pCp)o/'(pC,,), time r nondimensional time (equation 6) ~* stream function (equation 10) vorticity (equation 8). ~)
= =

Subscripts f fluid w wall 0 stagnant medium o~ reference.

and species buoyancy effects adjacent to a horizontal surface by considering boundary layer flows. Transient development of velocity, temperature and concentration boundary layers adjacent to a vertical surface was reported by Jang and Ni [15]. The non-Darcy flow over two-dimensional and axisymmetric bodies of arbitrary shape was studied by Kumari et al. [16]. A similarity solution for a general body shape was developed and the governing ordinary differential equations were solved by a shooting method. Non-Darcy mixed convection from vertical and horizontal surfaces was also reported [17]. Kumari and Nath [18] studied the unsteady laminar natural convection of an electrically conducting fluid over two-dimensional and axisymmetric bodies in a porous medium with an applied magnetic field. The unsteadiness of the flow was caused by time-dependent wall conditions. Unsteady mixed convection adjacent to a vertical surface with combined buoyancies in the presence of magnetic field was numerically studied [19]. Unsteady magneto-hydrodynamic mixed convection with combined buoyancies over a horizontal cylinder and a sphere was also reported [20]. Lai et al. [21] obtained similarity solutions for natural convection from slender bodies of revolution. To obtain similarity for a paraboloid with constant temperature and concentration, and a vertical cylinder

with linear wall temperature and concentration, it was necessary to assume a vertical two layer structure. Lai and Kulacki [22] published similarity solutions for boundary layers adjacent to flat vertical surfaces with constant wall temperature and concentration and constant wall flux conditions with aiding buoyancies. Flow injection at the wall was shown to decrease transport rates because the boundary layer thickness increases. Lai [23] reported similarity solutions for coupled heat and mass transfer by mixed convection in aiding buoyancies from a vertical surface in a saturated porous medium. The heat and mass transfer rates were shown to vary between the two extremes of natural and forced convection depending on the values of buoyancy ratio and Lewis number. The transient development of the concentration boundary layer with impulsive mass diffusion from a heated vertical surface after the thermal boundary layer was stably established was reported by Pop and Herwig [24]. Transient effects were also studied by Jang et al. [25] for non-Darcy flows at higher Grashof numbers in which the convective terms in the boundary layer equations were retained. Telles and Trevison [26] included the hydrodynamic, thermal and chemical dispersion effects in the similarity analysis of boundary layer flows. More recently, Nakayama and Hossain [27] obtained an integral solution for aiding

Combined heat and mass transfer by natural convection flows adjacent to vertical surfaces. Rastogi and Poulikakos [28] considered non-Newtonian fluid saturated porous media and presented similarity solutions for aiding flows with constant wall temperature and concentration, as well as constant wall flux conditions. The foregoing survey of natural convection flows adjacent to surfaces in unconfined saturated porous media with the combined buoyancies of heat and mass diffusion shows that all the theoretical work was based on boundary layer type analysis. This is valid when the two buoyant mechanisms aid each other. Boundary layer analysis is valid also for flows with opposing buoyancies, provided one of them overpowers the other and the Lewis number is of the order of unity. In all other cases, the boundary layer approximations break down. In the current investigation, the results of a comprehensive numerical study which is not based upon the boundary layer approximations are reported. Both aiding and opposing flows are considered for a wide range of the governing parameters, the thermal Rayleigh number, Ra, the buoyancy ratio, B and the Lewis number, Le, with emphasis on opposing buoyant mechanisms. The basic physical processes are explained in detail, and Nusselt and Sherwood number data are provided.

2757

Tj.=O, C j . = O ,

OW : 0 OX

-x ou,= o,

ou,= o

T=J C=1
w=O

T=O C=O

01# aY

:0 X
P out

2. ANALYSIS
The physical configuration and the coordinate system under consideration are illustrated in Fig. 1. A vertical surface of height L is embedded in a fluid saturated porous medium. The temperature of the vertical surface is tw and the surface concentration of the diffusing species is Cw. Far away from the surface these values are invariant and are represented by to~ and c~, respectively. It is assumed here that tw > t~ and Cw > Go. The buoyancy-driven Darcy flow and transport, adjacent to the vertical surface due to the combined effects of the thermal and species diffusion can be described by the following two-dimensional conservation equations in nondimensional form

Ti.=O. C j . = O ,

OW : 0 OX

Fig. 1. Geometry and numerical boundary conditions. where fl is the thermal volumetric expansion coefficient, -l/p(Op/0t)p.c and fl* is the volumetric coefficient due to concentration, -1/p(Op/Oc)p.t. The magnitude of the buoyancy ratio, B, indicates the relative strengths of the two buoyant forces and the algebraic sign provides information on the relative direction of the two forces. While the thermal buoyancy always acts vertically upward, the species buoyancy may act in either direction depending on the relative molecular weights, with a heavier species contributing a negative buoyant force that acts vertically downward, resulting in a negative sign on the buoyancy ratio. For the case where the buoyancy ratio is equal to zero, i.e. B = 0, the flow is driven by thermal buoyancy alone. Two other parameters which appear in the nondimensional conservation equations are the Rayleigh number, Ra =9flAtKL/vc~ and the Lewis number, Le = e/D. Here, K is the permeability, 9 is the gravitational acceleration, ct is the thermal diffusivity given by e = ko/(pC)f and D is the species diffusion coefficient. The following nondimensional variables were defined in the development of equations (1)-(4).

OU OU OY

8V

=o

(1)

8V OT OC OX- oY+B~

(2)

or oc

ov vOr oc oc

(o2T l lO2c

o2v)

a~z , + U ~ +

~=Ra\Ox

2 +OY2]

(3)

&, + U~+

V o y - RaLe \8X 2 +

8~;

o2c)

(4)

where 4, is the porosity of the medium and a is the ratio of heat capacities of the stagnant medium to the fluid. The buoyancy ratio, B, is defined as

x : zx'

Y Y=Z' T-

z*

- crVc '
and C

v u=vu ' v=V ' c-- c~ Cw-- c~ (6)

/~*(Cw-C~) fl(tw -- to,)

/3*At

flat

(5)

t-- t~ tw -- t~

2758

D. ANGIRASA et al. Table 1. Ra = 1000, B = 0, Le = 1 and At* = 0.001


Nu Sh

where Vc is a convective velocity defined as V~ = g f l k t K / v . The boundary conditions in nondimensional form are expressed as Y=0
Y+oc V=O,T=I,C=I U=O,T=O,C=O.

and (7)

a = 0.8 and ~ = 0 . 7 5 G = 1.0 and ~ = 1.0

27.81636951 27.81636182

27.81636985 27.81636182

A nondimensional vorticity, defined as


~o

= 8~-

8U

83

8V

(s)

allows equation (2) to be written as

(o = ~

aT

+ Bs-y

8C

of equation (9). The average Nusselt and Sherwood numbers are calculated through numerical integration of equations (12) and (13) using an open-ended formulation of Simpson's rule [30]. The initial conditions used to march the discretized equations, equations (3) and (4), are z*=0
T=O,C=O,U=O,

(9)
V= 0 for all X and Y. (I 4)

and a nondimensional stream function ~, to be defined such that U=~


80

and

V-

8X

,?~,

(10)

Combining equations (8) and (10), the stream function equation can be written as

82@. 02@
8x ~ + 7-ys = o. (ll)

For Darcy flows, the concept of vorticity is rather redundant. The velocity gradient difference on the right-hand side of equation (8) is known as vorticity in viscous flows. Hence, the definition of vorticity is used here for convenient identification. The average Nusselt number is obtained as
hL I aT

The boundary conditions for equations (3), (4) and (11) are shown in Fig. 1. The width of the computational domain was first conservatively estimated from the similarity solution, and then modified by trial and error. Its value lies between 0.15 and 1.0 depending on the values of Ra, B and Le. Both the absolute and relative error criteria with a value less than 10 5 were used to check for steadystate solution. Constant grid spacings were employed in each direction. The number of grid points was varied to make the grid-dependent error less than one percent in field variables. This required 101 grid points in each direction. The calculations were time consuming; typically 40000 time-steps were required to reach a steady-state solution. The time-step was chosen to be 0.001 after checking that the value had no impact on the final steady-state solution.
4. RESULTS AND DISCUSSION

where/7 is the average heat transfer coefficient. Similarly, the average Sherwood number can be expressed

as
D - 8yjv=odX

(13)

where hm is the average mass transfer coefficient.


3. NUMERICAL METHODS

The energy and species conservation equations, equations (3) and (4) are solved first using the alternating direction implicit scheme (ADI) of Peaceman and Rachford described by Roache [29]. The convective terms are discretized with first-order upwind differencing and the diffusion terms with central differencing. The vorticity is then evaluated using equation (9) and a central difference formulation. The stream function equation, equation (11), is then solved iteratively within each time-step using a successive order relaxation (SOR) method. It should be noted that the wall vorticities are not required to obtain the vorticity field. The vorticity is merely evaluated with the field solutions of T and C by means

Extensive calculations have been performed to obtain steady-state flow, temperature, and concentration fields, and the average Nusselt and Sherwood numbers for a wide range of the parameters, buoyancy ratio, B, Lewis number, Le and Rayleigh number, Ra. The literature survey discussed above revealed that similarity (boundary layer) solutions could not be obtained for certain conditions, specifically, when - 1 < B < 0, for low values of Rayleigh number and for Lewis number less than unity when B < 0 [11]. These ranges of parameters are given spacial consideration here because the limitations of boundary layer analysis have been removed. Since the energy and species conservation equations, equations (3) and (4), respectively, are marched in time to the asymptotic steady-state solutions, the influence of ~r and @ on the steady-state solutions was first ascertained. Table 1 presents the computed values for the Nu and Sh for two pairs of values of a and @. As shown, very little difference is observed between the two sets of Nu and Sh. Although the transient results are influenced by r, and qS, steady-state solutions are not. Hence. the values of ~r and q5 each was

Combined heat and mass transfer by natural convection

2759

14 = 12 10 8 6 4 ~ 2 0
-X

Present

JE ~ "
t.v

/ t

Ra= 100 Le= 1

I I I

I I I

-3

-2

-1
B

Fig. 2. Comparison with similarity solution for Ra = 100 and Le = 1.

set to unity without introducing any error in the results to follow. The work of Mahajan and Angirasa [5] and Rahman and Lampinen [31] in viscous natural convection flows with combined buoyancies enunciated the limitations of the boundary layer analysis. Figures 2 and 3 compare the numerical solutions obtained here for the Nusselt number in porous media with those of the similarity analysis obtained by Bejan and Khair [11] for L e = 1 and R a = 100 and 1000, respectively. As illustrated, at the lower Ra of 100 the difference between the two is rather large, especially in the region - 2 < B < 1 (Fig. 2), where the similarity solution underpredicts the transport rates for all values of B because the diffusion transport in the vertical direction has been neglected. At R a = 1000, there is an excellent agreement between the two sets of results over a wide range of B for which boundary layer analysis is accurate (Fig. 3). These results substantiate the validity and accuracy of the numerical model and support the conclusions of Bejan and Khair [11] that similarity solutions fail in the range of - 1 < B < 0 because

in this range, there is no discernible vertical layer structure. While this topic will be discussed further in the results to follow, it is apparent from the results presented in Figs. 2 and 3, that the present numerical model provides accurate results for all B and for a wide range of R a and Le. In Figs. 2 and 3, the thermal and species diffusion coefficients are identical (i.e. L e = 1) and hence, the Nusselt and Sherwood number are equal. F o r aiding flows, N u increases with B as there is an increase in the effective Rayleigh number. The value of N u is at a minimum when B = - l, where there is no flow and the transport takes place entirely by diffusion. The case of B = 0 indicates that convection is due to thermal buoyancy alone. The value of N u obtained in the current computation compared very well with that obtained using the similarity solution of Cheng and Minkowycz [10]. As B is decreased below zero, the buoyancy due to species diffusion is negative and acts vertically downward thereby opposing the vertically upward thermal buoyancy. The transport rates then are lower for B < 0 when compared to those with

2760

D. ANGIRASA et al.

70
Present

6050-

Bejan and Khair [11]

40tOO

o
Z

30-

20-

10_

Ra=lO00 Le=l

J i l l

L I I l

I I I I

I I I I

I l l L

I I I I

-6

-4

-2

B
Fig. 3. Comparison with similarity solution for Ra = 1000 and Le = I.

aiding buoyancies and the same numerical value of B. As the value of B continues to decrease, the magnitude of the species buoyancy increases and overpowers the upward thermal buoyancy. As a result, the transport rates begin to increase again because it is the magnitude of the velocity, not its direction that augments the heat (or mass) transfer rate. The Nusselt (or Sherwood) number for Le = 1 is symmetric about - 1 when shown as a function of the buoyancy ratio, except in the range - 1 < B < 0. Here, diffusion dominates and the opposing buoyancies result in a complex flow.
EJfect o f B on the flow, temperature and concentration fields To understand the flow structure and the temperature and concentration fields, the velocity and temperature profiles at X = 0,5 are plotted in Figs. 4 and 5, respectively. The temperature and concentration fields are identical since Le = 1. As illustrated in Fig. 4, an upward boundary layer flow exists for B = 4. Boundary layer type flow exists for B = - 4 as well,

but the flow direction is reversed due to the downward species buoyancy which dominates the flow. On the other hand, the profiles for B = - 0 . 5 and - 1.5 do not exhibit boundary layer characteristics. The flow is still upward for B = - 0 . 5 because of the larger thermal buoyancy. These two profiles are symmetric about the horizontal axis at U = 0. In Fig. 5, the temperature (concentration) profiles for these two are identical, and are almost linear, indicating the strong influence of diffusion for this range of B. As B decreases from a value of + 4, the temperature profiles become flatter, thereby resulting in lower Nusselt numbers (Fig. 3). After Nu reaches a minimum value at B = -- 1, these profiles become steeper with further decreases in B. These results clearly indicate that the behavior of Nu(Sh) with B is consistent with the flow and temperature (concentration) fields.
EfJect o f Lewis number The Lewis number determines the relative extent of the temperature and concentration fields from the vertical surface. In a fluid saturated porous medium

Combined heat and mass transfer by natural convection

6 X=0.5

2761

B ----o-_ 4

2 U

--":'-- -0.5 -1.5 -4

-2

Ra = 10 3 Le= 1

-4 0.00
I

0.05

0.10 Y

0.15

0.20

Fig+ 4. Velocity profiles for various values of buoyancy ratio.

with a value ofLe(-_- a/D) larger than unity, the extent of the thermal diffusion from the vertical surface is much larger than that of the species diffusion. For Le < 1, these thicknesses are reversed. For opposing buoyancies (B < 0), the flow, temperature and concentration fields then become quite complex because of the unequal extent of the two diffusion processes. Two values of B, - 0 . 5 and -1.5, were chosen to highlight the complexity of the flow and transport for the case of opposing buoyancies. The first value produces upward flow near the surface and the second, downward flow. Vertical velocity profiles are plotted in Fig. 6 for various values of the Lewis number, Le. The corresponding temperature and concentration profiles are shown in Figs. 7 and 8, respectively. Since B -- --0.5, the thermal buoyancy is larger and hence in the region where the flow has a component resulting from thermal buoyancy, the flow is upward. This was the case for all of the flow region for Le > 1. Comparing Figs. 6 and 8, it is apparent that the peak vertical velocity for Le-- 10(>> ]) occurs well away

from the wall at a location where the concentration layer nearly ends. Farther away from the thickness of this layer, only the upward thermal buoyancy exists. Hence, it can be concluded that near the wall, where there is a thin concentration layer, the flow is decelerated at the downward species diffusion drags the flow. However, the weaker species buoyancy cannot reverse the flow entirely. For Le << 1, the species diffusion spreads farther away than the thermal diffusion. By comparing Figs. 6 and 7, it is clear that away from the wall, where the thermal buoyancy is weak, the flow is entirely downward. The vertical velocity reaches zero again where the concentration layer ends. Figure 7 indicates that the temperature profile becomes flatter with decreasing Lewis number, due to the relatively larger spread of the thermal layer when compared to the species concentration layer. This disparity, however, is not large as the Rayleigh number and the buoyancy ratio remain unchanged. The change in concentration profile with Lewis number is substantial as shown in Fig. 8. At lower values of Le, the species spread farther away from the vertical

2762

D. ANGIRASA et al.

1.0 B --o.-- 4 ---o-- -0.5 -1.5


-4

0.8

0.6
O
I--

Ra = 103 Le= 1

0.4

0.2 X=0.5 0.0 --p0.00 0.02 0.04 0.06 0.08 0.10

Y
Fig. 5. Temperature (concentration) profiles for various values of buoyancy ratio.

Combined heat and mass transfer by natural convection

2763

.8

.......

X=0.5 Le 0.6 ---o--0.1 --o--- 0.2 0.7 2.0 - - 0 - - 5.0 .--o-- 10.0 Ra = 104 B = -0.5

0.4

0.2

0.0

-0.2

-0.4

0.00

0.02

0.04

0.06 Y

0.08

0.10

0.12

Fig. 6. Velocity profiles for various values o f Lewis n u m b e r for B = - 0 . 5 .

2764

D. A N G I R A S A et al.

1.0 X=0.5 0.9 Le 0.8 0.7 0.6 I-.- 0.5 0.4 0.3 0.2 0.1 0.0 I 0.00
I l 1

---o-- 0.1 2.0 " 10

Ra = 10 4 B =-0.5

0.02

0.04 Y

0.06

0.08

Fig. 7. T e m p e r a t u r e profiles for various values of Lewis n u m b e r for B = - 0 . 5 .

Combined heat and mass transfer by natural convection

2765

1.0 X=0.5 Le ---o--0.1 ---o-- 0.2 0.5 0.7 -.-o-- 2.0 -.-o-- 10

0.8

0.6

Ra = 104

0.4

B = -0.5

0.2

0.0 -I-0,00 0.02 0.04 Y


Fig. 8. Concentration profiles for various values of Lewis n u m b e r for B = - 0 . 5 .

0.06

0.08

0.10

2766

D. ANGIRASA et al.

1000 = = Nu Sh

100 t-

co
t,--

O Z

10-

R a = 104 B = -0.5

'l

I I

[ 1

0.1

1
Le

10

Fig. 9. The variation of Nusselt and Sherwood number with Lewis number for B = -0.5.

surface. There is an order of magnitude difference in the thickness of concentration layers between the two values of 0.1 and 10 for Le. At Le = 0.1, the concentration layer is of the same order as the height of the vertical surface. Summarizing the results illustrated in Figs. 6-8, in the complex interaction between the different diffusion rates and opposing buoyancies, there are regions of flow reversal and deceleration. These factors are unaccounted for in the approximations made in the boundary layers and, hence, cannot be obtained by similarity and other boundary layer analyses. The variation of the Nusselt and Sherwood numbers with the Lewis number is presented in Fig. 9. As shown, the Sherwood number increases with increasing Le, because for larger Le, the concentration layer is thinner. The Nusselt number, however, does not vary as much as the Sherwood number because the Rayleigh number remains constant and the flow near the surface is always dominated by the thermal buoyancy.

The flow and transport are different for B = -- 1.5, where the downward species buoyancy overpowers the upward thermal buoyancy. For Le 4= 1, the unequal diffusion of thermal energy and species results in interesting flow patterns. The upward velocity, the temperature and the concentration profiles at X = 0.5 for B = - 1.5 are plotted in Figs. 10-12, respectively. For Le < 1, the flow is entirely downward (Fig. 10) and the peak velocity shifts away from the wall to a location where the thermal buoyancy is weak (Fig. 11). As Le increases to values greater than 1, the flow is downward only near the wall where there is a thin concentration layer. For this case, the peak downward velocity occurs on the wall and the greatest upward vertical velocity shifts to a region away from the wall where the species buoyancy is weak. Here, it should be noted that the magnitude of the maximum upward velocity is lower than that of the downward velocity. The temperature profiles exhibit more complex behavior for B = - 1 . 5 than for B = - 0 . 5 , with the steepest temperature profile occurring at Le = 0.1 for

Combined heat and mass transfer by natural convection

2767

0.4 X=0.5

0.2

0.0

-0.2 Le --o---o-Ra = 10 3 B =-1.5 -0.6 " ', --o----o-0.1 0.2 0.5 1.0 2.0 10

-0.4

-0.8

0.0

0.1

0.2 Y

0.3

0.4

0.5

Fig. 10. Velocity profiles for various values of Lewis number for B = - 1.5.

B=-1.5 (Fig. 11) as opposed to the case of B = - 0 . 5 . Also, for B = - 1 . 5 , the entire flow is downward at Le = 0.1, therefore resulting in the steep temperature profile. As Le increases, the extent of the dominant species buoyancy decreases, causing the velocities to fall as shown in Fig. 10 and as a consequence, making the temperature profiles shallower. When the Lewis number is larger than unity, the magnitude of the velocities increases again and the temperature profiles are steeper. The concentration profiles, on the other hand, spread uniformly farther apart with increasing value of Lewis number as shown in Fig. 12. The Nusselt and Sherwood number variation for B = - 1.5 as presented in Fig. 13 represents the trends of the temperature and concentration profiles discussed above. The highest Nu is attained at the lowest Le, decreases with increasing Le and then increases again. The Sherwood number expectedly increases with increasing Le.

Effect of Rayleigh number The Raleigh number is defined with reference to the temperature difference alone and hence, by itself, it does not represent the magnitude of the species buoyancy. If the buoyancy ratio is held constant, the variation of the Rayleigh number does not change the relative magnitude of the species buoyancy. This implies that higher values of the Rayleigh number must then result in larger velocities and steeper temperature and concentration gradients at the wall (not shown in figures) resulting in higher transport rates. Figure 14 shows the variation of the Nusselt and Sherwood numbers with varying Rayleigh number. The Sherwood number is substantially larger than the Nusselt number because of the larger value of the Lewis number ( = I0). The Nusselt and Sherwood number data for this case are correlated as Nu --- 0.586 Ra 5 and Sh = 2.045 Ra 5. General correlations involving all the parameters, Ra, B and Le are not attempted here because of the complexity of the inter-

2768

D. ANGIRASA

et al.

1.0 X=0.5

0.8

Le ----o--0.1
--0-1.0

0.6
I--

10

0.4

R a = 103 B = -1.5

0.2

0.0 --F0.00 0.04 0.08 Y


Fig. 11. T e m p e r a t u r e profiles f o r v a r i o u s values o f Lewis n u m b e r f o r B = - 1.5.

0.12

0.16

0.20

C o m b i n e d heat a n d mass transfer by n a t u r a l c o n v e c t i o n

2769

X=O.5 Le
0.8 -~l~
'

---o--0.1
0.5

1.0 10 0.6 ..) 0.4


R a = 10 3 B = -1.5

0.2

0.0 0.0 0.1 0.2 Y


Fig. 12. C o n c e n t r a t i o n profiles for v a r i o u s values of Lewis n u m b e r for B = - 1.5.

0.3

0.4

2770

D. ANGIRASA

el a/.

100

O9 O Z

10-

--e--

Nu

Ra = 10 3

Sh

B=-1.5

I I

I I I

i I I

0.1
F i g . 13. T h e v a r i a t i o n of Nusselt and Sherwood

1
Le
number with Lewis number for B = 1.5.

10

Combined heat and mass transfer by natural convection

2771

1000
L e = 10
--.,.

/ ! 1

,.-

u';
L,--

100 -

Nu

10

10 a

104

10 ~

Ra
Fig. 14. The variation of Nusselt and Sherwood number with Rayleigh number.

2772

D. A N G I R A S A et al. application to heat transfer from a dike. Journal ~lGeophysical Research, 1977, 82, 2040-2044. Bejan, A. and Khair, K. R., Heat and mass transfer by natural convection in a porous medium. International Journal o[' Heat and Mass Transi[er, 1985, 28, 909 918. Hasan, M. and Majumdar, A. S., Transpiration-induced buoyancy effect around a horizontal cylinder embedded in a porous medium. International Journal of Energy Researeh, 1985, 9, 151-163. Jang, J. Y. and Chang, W. J., Buoyancy-induced inclined boundary layer flows in a porous medium resulting from combined heat and mass buoyancy effects. International Communications in Heat and Mass Transler, 1988, 15, 17 -30. Jang, J. Y. and Chang, W. J., The flow and vortex instability of horizontal natural convection in a porous medium resulting from combined heat and mass buoyancy effects. International Journal of Heat Mass Tran:~ler, 1988, 31,769 777. Jang, J. Y. and Ni, J. R., Transient free convection with mass transfer from an isothermal vertical flat embedded in a porous medium. International Journal c?fiHeat and Fluid Flow, 1989, 10, 59-65. Kumari, M., Takhar, H. S. and Nath, G., Double diffusive non-Darcy free convection from two-dimensional and axisymmetric bodies of arbitrary shape in a saturated porous medium. Indian Journal o[ Technology, 1988, 26, 324-328. Kumari, M., Takhar, H. S. and Nath, G., Non-Darcy double-diffusive mixed convection from heated vertical and horizontal plates in saturated porous media. WarmeStc~ffubertraq, 1988, 23, 267 273. Kumari, M. and Nath, G., Double diffusive unsteady free convection or two-dimensional and axisymmetric bodies in a porous medium. International Journal of Energy Research, 1989, 13, 379 391. Kumari, M. and Nath, G., Double diffusive unsteady mixed convection flow over a vertical plate embedded in a porous medium. International Journal c?f Energy Research, Vol. 13, pp. 419-430 (1989). Kumari, M. and Nath, G., Unsteady mixed convection with double diffusion over a horizontal cylinder and a sphere within a porous medium. Warme-Stoffubertraq, 1989, 24, 103 109. Lai, F. C., Choi, C. Y. and Kulacki, F. A., Coupled heat and mass transfer by natural convection from slender bodies of revolution in porous media. International Communications in Heat and Mass Tran.~r, 1990, 17, 609 620. Lai, F. C. and Kulacki, F. A., Coupled heat and mass transfer by natural convection from vertical surfaces in porous media. International Journal 0[' Heat and Mass Trans~f~'r, 199l, 34, 1189 1194. Lai, F. C., Coupled heat and mass transfer by mixed convection from a vertical plate in a saturated porous medium. International Communications in Heat and Mass Tran,~/i~r, 1991, 18, 93--106. Pop, I. and Herwig, H., Transient mass transfer from an isothermal vertical fiat plate embedded in a porous medium. International Communications in Heat and Mass Tran.~'/er, 1990, 17, 813 821. Jang, J. Y., Tzeng, D. J. and Shaw, H. J., Transient free convection with mass transfer on a vertical plate embedded in a high-porosity medium. Numerical Heat TransJ~,r Part A, 1991, 20, 1 18. Telles, R. S. and Trevisan, O. V., Dispersion in heat and mass transfer natural convection along vertical boundaries in porous media. International Journal of Heat and Mass Tran,f;(~r, 1993, 36, 1357 -1365. Nakayama, A. and Hossain, M. A., An integral treatment for combined heat and mass transfer by natural convection in a porous medium. International Journal ~[' Heat and Mass Tran.~/er, 1995, 38, 761-765.

a c t i o n a m o n g v a r i o u s f a c t o r s a f f e c t i n g t h e flow a n d transport.

t l.

5. S U M M A R Y A literature survey of external natural convection by combined heat and mass transfer from surfaces e m b e d d e d in a fluid s a t u r a t e d p o r o u s m e d i u m s u g g e s t s t h a t b o u n d a r y l a y e r a n a l y s i s is i n v a l i d for o p p o s i n g b u o y a n c i e s a n d w h e n t h e d i f f u s i o n coefficients are u n e q u a l . In t h e p r e s e n t w o r k , c o m p l e t e s o l u t i o n s h a v e been obtained numerically by solving the Darcy type equations without making approximations of boundary layer character. For high Rayleigh number aiding flows, t h e n u m e r i c a l s o l u t i o n s m a t c h v e r y well w i t h t h e s i m i l a r i t y s o l u t i o n s . H o w e v e r , t h e y differ s u b s t a n t i a l l y f o r o p p o s i n g flows a n d for l o w R a y l e i g h n u m b e r flows. T h e flow a n d t r a n s p o r t f o l l o w c o m p l e x p a t t e r n s depending on the interaction between the diffusion coefficients a n d b u o y a n c y ratio. T h e local d i r e c t i o n o f t h e flow c h a n g e s b e c a u s e o f t h e v a r i a t i o n in t h e e x t e n t o f t h e t h e r m a l a n d c o n c e n t r a t i o n layers, a n d t h e opposing buoyant mechanisms. The Nusselt and S h e r w o o d n u m b e r v a r i a t i o n reflects this c o m p l e x interaction.

12.

13.

14.

15.

16.

17.

18.

REFERENCES 1. Cheng, P., Geothermal heat transfer. In: Handbook ~f Heat Transfi, r Applications (eds W. M. Rohsenow, J. P. Hartnett and E. N. Ganic), 2rid edn. McGraw-Hill, New York, 1985, Chapter 11. 2. Gebhart, B., Jaluria, Y., Mahajan, R. L. and Sammakia, B., Buoyancy-Induced Flows and Transport, Ch. 15. Hemisphere, Washington, DC, 1988. 3. Gebhart, B. and Pera, L., The nature of vertical natural convection flows resulting from the combined buoyancy effects of thermal and mass diffusion. International Journal o f Heat and Mass Tran.~[er, 1971, 14, 2025--2050. 4. Ostrach, S., Natural convection with combined driving forces. PhysieoChemieal Hydrodynamics, 1980, 1, 233 247. 5. Mahajan, R. L. and Angirasa, D., Combined heat and mass transfer by natural convection with opposing buoyancies. Journal o f Heat Transfer, 1993, 115, 60~612. 6. Nield, D. A. and Bejan, B., Convection in Porous Media, Ch. 9. Springer-Verlag, New York, 1992. 7. Raptis, A., Tzivanidis, G. and Kafousias, N., Free convection and mass transfer flow through a porous medium bounded by an infinite vertical limiting surface with constant suction. Letters on Heat Mass Tran,~'fer, 1981, 8, 417M24. 8. Raptis, A., Kafousias, N. and Massalas, C., Free convection and mass transfer flow through a porous medium bounded by an infinite vertical porous plate with constant heat flux. Z A M M , 1982, 62, 489~491. 9. Raptis, A. A., Unsteady free convective flow and mass transfer through a porous medium bounded by an infinite vertical limiting surface with constant suction and time-dependent temperature. International Journal of Energy Research, 1983, 7, 385 389. 10. Cheng, P. and Minkowycz, W. J., Free convection about a vertical flat plate embedded in a porous medium with

19.

20.

21.

22.

23.

24.

25.

26.

27.

Combined heat and mass transfer by natural convection 28. Rastogi, S. K. and Poulikakos, D., Double-diffusion from a vertical surface in a porous region saturated with a non-Newtonian fluid. International Journal of Heat and Mass Transfer, 1995, 38, 935-946. 29. Roache, P. J., Computational Fluid Dynamics, revised edn. Hermosa, Albuquerque, NM, 1982. 30. Press, W. H., Flannery, B. P., Teukolsky, S. A. and

2773

Vetterling, W. T., Numerical Recipes: The Art of Scientific Computing. Cambridge University Press, New York, 1986. 31. Rahman, M. M. and Lampinen, M. J., Numerical study of natural convection from a vertical surface due to combined buoyancies. Numerical Heat TransJer Part A, 1995, 28, 409~429.

You might also like