Proceedings of 3rd BSME-ASME International Conference on Thermal Engineering 20-22 December, 2006, Dhaka, Bangladesh

NATURAL CONVECTION FLOW IN A SQUARE CAVITY WITH TEMPERATURE DEPENDENT HEAT GENERATION Md. Mamun Molla1*, M. M. A. Sarker2
2

Department of Mechanical Engineering, University of Glasgow, Glasgow G12 8QQ, UK Department of Mathematics, Bangladesh University of Engineering & Technology, Dhaka-1000, Bangladesh. * Corresponding author: E-mail: m.molla@mech.gla.ac.uk (M. M. Molla), Tel: +44 (0)141 330 8463

ABSTRACT: Unsteady natural convection laminar flows in a square cavity formed by insulated bottom and top walls, uniformly heated left wall and the cooled right wall has been investigated numerically. The governing equations are transformed into non-dimensional form and the resulting partial differential equations are solved applying upwind finite difference method together with Successive Over-Relaxation (SOR) scheme. The effect of the heat generation and the Rayleigh number on streamlines and isotherms as well as on the rate of heat transfer from the heated wall of the cavity are presented. Key words: Natural convection, Unsteady laminar flow, Heat generation, Square cavity 1. INTRODUCTION Ostrach [1] provided a comprehensive review article and extensive bibliography on natural convection in cavities up to 1988. De Vahl Davis [2] investigated the natural convection of air in a square cavity. Valencia and Frederick [3] investigated the heat transfer in square cavities with partially active vertical walls. Selamet et al. [4] studied the laminar buoyancy driven flows in an enclosure. Natural convection heat transfer in rectangular cavities heated from the bottom had been investigated by Hasnaoui et al. [5]. Sundstrom and Kimura [6] observed the phenomena on laminar free convection in inclined rectangular enclosure. November and Nansteel [7], Ganzarolli and Milanez [8], investigated the natural convection in rectangular enclosures heated form below and cooled from the sides. De Vahl Davis [2], Dixit and Babu [9] have recently investigated natural convection of air in square cavity by using lattice Boltzmann method. A large number of physical phenomena involve natural convection driven by heat generation. The study of heat generation in moving fluids is important from the view point of several physical considerations such as those dealing with chemical reactions and those concerned with dissociating fluids. Possible heat generation effects may alter the temperature distribution and, therefore, the particle deposition rate. This may occur in applications related to nuclear reactor cores, fire and combustion modelling, electronic chips and semiconductor wafers. The literature is replete with examples dealing with the heat transfer in laminar flow of viscous fluids. All of the above authors ignored the effect of temperature dependent heat generation due to molecular interaction within the cavities which is not less important from the angle of practical applications. Vajravelu and Hadjinolaou [10] studied heat transfer characteristics in a laminar boundary layer flow of viscous fluid. Molla et al. [11-13] studied natural convection flow with heat generation along a uniformly heated vertical wavy surface, horizontal cylinder and sphere.

In this paper, unsteady natural convection laminar flow in a square cavity consisting of insulated top and bottom walls, uniformly heated left wall along with cooled right wall is investigated .The basic equations of motion are non-dimensionalized and solved numerically by using a upwind finite-difference scheme together with Successive Over Relaxation (SOR) scheme. Effect of heat generation and Rayleigh number on streamlines and isotherms, as well as on the rate of heat transfer from the heated wall of the cavity are discussed and presented graphically. In the whole investigation, the Prandtl number Pr has been chosen to be 1.0. 2. FORMULATION OF THE PROBLEM
A square cavity of height H filled with viscous incompressible fluid is shown in fig. 1. where top and bottom walls are adiabatic whereas the right wall is maintained at constant cold temperature Tc and the left wall is uniformly heated. The effect of temperature-dependent heat generation [10] in the 3 flow region has also been taken into account. The volumetric rate of heat generation q [W/m ], is

Q0 (T TC ), for T TC q = (1) for T < TC 0, where Q0 is a heat generation constant which may be either positive or negative. This source term represents the heat generation when Q0 > 0 and the heat absorption when Q0 < 0.
u = v = 0, T / y = 0

u=v=0, T=TH y

v 0
u = v = 0, T / y = 0

Fig.1: Physical model and coordinate system. Unsteady two-dimensional laminar free convective flow of viscous incompressible fluid having constant properties is assumed where buoyancy effect is included through Boussinesq approximation. Under these assumption, equations of continuity, momentum and energy are u v =0 + (2) x y

gravity, is the fluid density, is the thermal diffusivity, is the coefficient of thermal expansion, is the viscosity of the fluid, Cp is the specific heat at constant pressure. To make the above equations dimensionless, we introduce the following non-dimensional variables

2T 2T Q0 T T T +u +v = (5) x 2 + y 2 + C (T TC ) t x y p where ( u, v ) are the velocity components along the ( x, y ) axes, g is the acceleration due to

2u 2u u u u 1 p +u +v = + x 2 + y 2 t x y x 2 2 v v v v v 1 p +u +v = + x 2 + y 2 + g (T TC ) t x y y

u=v=0, T=TC

(3) (4)

x y t u ,Y= , = 2 , U= , H H /H H / (6) T TC p v V = , P= , = TH TC /H 2 / H 2 where (=/) is the reference kinematic viscosity and is the non-dimensional temperature. Substitution of dimensionless variables (6) into equations (2)-(5) leads to: (U ) (V ) 2 2 Ra + + + + = , (7) 2 2 X Y Pr X Y X X =

(U ) (V ) 1 2 2 + + = + 2 X Y Pr Y 2 X

+ ,

(8)

where

is the vorticity function and is the stream function defined by : U= , V= (10) Y X In the above equations Ra is the Rayleigh number, Pr is the Prandtl number and is the heat generation parameter defined respectively by the following relations. (TH TC )H 3 Q0 H 2 Ra = , Pr = and = (11) C p The dimensionless initial and boundary conditions are: U =V = = 0 for = 0 for 0 Y 1 at X = 0 U = V = 0, = 1 for 0 Y 1 at X = 1 U = V = 0, = 0 (12) U = V = 0, =0 for 0 X 1 at Y = 0 Y U = V = 0, =0 for 0 X 1 at Y = 1 Y Once we know the numerical values of the temperature function we may obtain the rate of heat transfer in terms of the Nusselt number from the left wall of the cavity using: T Nu = (13) X X =0 An upwind finite-difference method, together with a Successive Over-Relaxation (SOR) iteration technique [14] has been employed to integrate model equations (7) and (8) governing the flow. Table 1: Comparisons of present numerical results with the benchmark solutions (listed in bracket) [2] for the case Pr = 0.71 and 51 51 meshes Ra
103 104 105

= X 2 + Y 2

(9)

max

Umax
3.616 (3.649) 16.230 (16.178) 34.92 ( 34.73)

Vmax
3.665 (3.697) 19.468 (19.617) 68.43 (68.59)

1.1655 5.1283 10.141 (9.612)

3. RESULTS AND DISCUSSION Numerical results are presented to determine effects of internal heat generation and Rayleigh number in presence of heat generation in a cavity. Values of heat generation parameter, between 0.0 & 20.0 and Rayleigh number Ra (= 103, 104, 105, 106) are considered. The corresponding problem without the effect of heat generation has been investigated by De Vahl Davis [2] for different values of the Rayleigh number for Pr = 0.71. With 5151 mesh for the case Pr = 0.71, = 0.0 the comparison of the maximum values of, U and V are presented in Table 1, which shows a good agreement with the benchmark solutions [2].
(i)
-0.40

(ii)
0. 9 1

(iii)
2.80

(iv)

-4.04

-5.14
-4.77 -4.04

4. 30

-5.84

= 0.0
0. 8

= 10.0
8 0. 7 0.71 0.64

= 15.0
2.07
1.92 1.78 1.63 1.48 1.34
1.19
1.04 0.89 0.75

= 20.0
4.17 3.87 3.58 3.28 2.98 2.69
2.39
2.10

(i)
0.92

(ii)

(iii)

(iv)

0. 4 5 0.31 0.16

Fig. 2: Streamlines (top) and isotherms (bottom) for different values of with Pr =1.0, Ra=104 The effect of internal heat generation on the flow field has been considered first. The resulting flow and temperature distributions are depicted in fig. 2. where top row gives the streamlines for increasing values of the heat generation parameter, (= 0.0, 10.0, 15.0, 20.0) while Ra = 104. In these figures we see that, without heat generation there is only one cell, called primary cell. With the effect of heat generation, a secondary cell has been developed in the top-left corner of the cavity. The increasing rate of heat generation leads to increase the flow rate in the secondary cell as well as increase in its size until it occupies almost half of the total cavity. This effect of internal heat generation on the flow field is reasonable since internal heat generation assists buoyancy forces by accelerating the fluid flow. On the other hand, the fluid temperature increases significantly due to effect of heat generation which is shown in isotherms of fig. 2 (bottom). It is clearly seen that, owing to increase of the heat generation, the fluid temperature exceeds the surface temperature that negates the heat transfer from the heated surface which is also illustrated in fig. 4 (a). Fig 3 shows the effect of Rayleigh number on the flow and temperature field where the internal heat generation is kept fixed at = 10.0. For the lower Raleigh number, two convective cells dominate the flow, however, with the increase of Rayleigh number, the primary cell becomes larger and occupies whole domain when Ra = 106. This behaviour is almost opposite to the effect of heat generation from lower to higher for a fixed Raleigh number (cf. fig 2). As we discussed earlier that the internal heat generation accelerates the flow and eventually it has an influence on buoyancy, however, for a fixed when Raleigh number increases, the buoyancy effect accelerates and dominates the flow, although there is a weak internal heat generation influence on the flow filed. The same phenomenon has been seen in the temperature contours in fig 3(bottom). Initially, when the Raleigh number is small the internal fluid tempera-

0. 6 2 0.32

0.64

= 0.0

0. 15 0.08

0.57 0.50 0.43 0.36 0.29 2 0.2

8.23 7.65 6.48 5.31 4.73 4.14

3.56

1.80

= 10.0

= 15.0

0.60

1.51 1.21

2.39

2.97
2.39

9 0.

= 20.0

1 1 . 8 . 22 1

-6.55
1.81

-5.18

1.51 1.21

ture is higher but it decreases with increase of Ra. For Ra = 106 the flow is stratified which is similar to a side heated cavity flow with no internal heat generation.
0.50

0.91

0.9 1

-1.99
-1 1. 1

.5 -19

-21.00

-5.18

(iv) Ra = 106

(i) Ra = 103
5.06

(ii) Ra = 104
2.07

5 (iii) Ra = 10
1.31
1.21 1.12 1.03

1.02

4.70 4.34 3.98 3.62

3.27 2.91

1.92 1.78 1.63 1.48 1.34

1.19

0.94 0.84

2 2. 1 . 5 5 9 1.83 1.47 1.11 0.75

1.04 0.89

0.75 0.66
0.84

0.57
0.38 9 0.29 0.1

0.95 0.88 0.81 0.73 0.66 0.59 0.52 0.44 0.37

0.60

0.39

0. 4 5 0.31 0.16

Fig. 3: Streamlines (top) and isotherms (bottom) for different values of Ra while Pr = 1.0, = 10.0
(a) 30.0 Heat transfer from the left surface
(b) 20.0 Heat transfer from the left surface

-30.0 -60.0 -90.0 0.0 0.2

Fig. 4: Rate of heat transfer from the hot wall (a) varying with Ra = 104 and (b) varying Ra with = 10 and Pr = 1.0. Fig. 4 (a) shows how the presence of heat generation inside the fluid influences the overall predictions of heat transfer. Initially when = 0, the heat transfer on the hot surface is positive, however, as the amount of heat generation increases, the total heat transfer decreases and becomes negative. This is due to the fact that the heat generation mechanism creates a layer of hot fluid near the hot surface and at some level when is larger, the resultant temperature of fluid exceeds the surface temperature. The rate of heat transfer also decreases with the height of the cavity. When there is no heat generation effect, the rate is uniform, however, when = 20.0, the temperature decreases largely near the top of the surface line. Fig 4(b) shows the effect of heat transfer from the hot surface on Rayleigh number when the heat generation parameter is fixed at = 10.0. For smaller Rayleigh numbers, the heat transfer rate is negative (Ra = 103, 104). That means this surface is loosing the heat whereas, for higher Rayleigh number e.g. Ra = 106, heat transfer rate is positive, although it is decreasing with the height of the surface again. For higher values of Ra, there is a strong effect of buoyancy force in the flow field which causes to circulate the internally generated heat throughout the cavity. Hence the hot layer, which has formed near the hot surface, disappears and the heat from the hot surface defuses in the cavity.

1.83

(i) Ra = 10

0.0

2.55

(ii) Ra = 10

(iii)Ra = 105

(iv) Ra = 106

0.30

15 0.

10.0

Ra 3 10 4 10 5 10 6 10

0.0 10.0 15.0 20.0

0.0

-10.0

0.4 Y 0.6

0.8

1.0

-20.0 0.0

0.2

0.4Y 0.6

0.8

1.0

4. Conclusions In the present paper a problem on natural convection laminar flow in a side heated square cavity with internal heat generation has been investigated numerically by employing an upwind finite difference method together with a Successive Over-Relaxation (SOR) technique. The results have been presented for the chosen fluid of Prandtl number Pr = 1.0, heat generation parameter (= 0.0, 10.0, 15.0, 20.0) and the Rayleigh number Ra (= 103, 104, 105, 106). From the present investigation the following conclusion may be drawn: Increase in the values of heat generation parameter leads to develop a secondary cell on the left upper corner of the cavity and increases the flow rate in the secondary cell and increases its size as well until it occupies almost half of the cavity. The temperature of the fluid in the cavity also increases due to the increase of the internal heat generation, consequently rate of heat transfer from the left wall decreases. For increasing Rayleigh number, the flow rate increases in both clock wise and anticlock wise directions. Rate of heat transfer from the heated wall increases due to the increase in Rayleigh number. Reference [1] [2] [3] [4] [5] [6] [7] [8] Ostrach, S.: Natural convection in enclosures. ASME J. Heat Transfer, 110(1988), 1175-1190. De Vahl Davis, G.: Natural convection of air in a square cavity: A bench mark numerical solution. Int. J. Numer. Methods Fluids, 3 (1983), 249-264. Valincia, A., Frederick, R. L.: Heat transfer in square cavities with partially active vertical walls. Int. J. Heat Mass Transfer, 32(1989), 1567-1574. Selamet, E. E., Arpaci, V. S., Borgnakke C.: Simulation of laminar buoyancy driven flows in an enclosure. Numerical Heat Transfer, 22 (1992), 401-420. Hasnaoui, M., Bilgen, E., Vasseure, P.: Natural convection heat transfer in rectangular cavities heated from below. J. Thermophysics. Heat Transfer, 6 (1995), 255-264. Sundstrom, L. G., Kimura, S.: On laminar free convection in inclined rectangular enclosure. J. Fluid Mech., 313 (1996), 343-366. November, M., Nansteel, M. W.: Natural convection in rectangular enclosures heated form below and cooled along one side. Int. J. Heat Mass Transfer, 30 (1986),2433-2440 Ganzarolli, M. M., Milanez, L. F.: Natural convection rectangular enclosures heated from below and symmetrical cooled from the sides. Int. J. Heat mass Transfer, 38 (1995), 1063-1073. Dixit, H. N., Babu, V.: Simulation of high Rayleigh number natural convection in a square cavity using the lattice Boltzmann method. Int. J. Heat Mass Transfer, 49 (2006), 727-749. Vajravelu, K., Hadjinicolaou, A.: Heat transfer in a viscous fluid over a stretching sheet with viscous dissipation and internal heat generation. Int. Comm. Heat Mass Transfer 20 (1993), 417-430. Molla, M. M., Hossain, M. A., Yao, L. S.: Natural convection flow along a vertical wavy surface with heat generation/absorption. Int. J. Thermal Sci. ,43 (2004),157-163. Molla, M. M., Hossain, M. A., Paul, M. C.: Natural convection flow from an isothermal circular cylinder in presence of heat generation. Int. J. Engn. Sci. ,44 (2006) 949-958. Molla, M. M., Hossain, M. A., Taher, M. A.: Magnetohydrodynamic natural convection flow on a sphere with uniform heat flux in presence of heat generation. Acta Mechanica (In press). Roache, P.J.: Computational Fluid Dynamics, Hermosa, Albuquerque. NM, 1972.

[9]

[10]

[11] [12] [13]

[14]