Proceedings of the 12th Annual Paper Meet 1

8-9 February 2008, Dhaka

Paper ID: 192-NM11

MIXED CONVECTION INSIDE A LID-DRIVEN PARALLELOGRAM CAVITY WITH ISOFLUX

HEATING FROM BELOW

G. Saha1, S. Saha2, M. A. Maleque3, M. M. K. Chowdhury3 and M. Q. Islam2

1
Department of Arts and Sciences, Ahsanullah University of Science and Technology, Dhaka-1215
2
Department of Mechanical Engineering, 3Department of Mathematics,
Bangladesh University of Engineering and Technology, Dhaka-1000
Email: [email protected]

ABSTRACT: Two-dimensional steady state mixed convection in a lid-driven parallelogram cavity with isoflux
heating from below has been investigated numerically. Two side walls are maintained at constant motion and
constant ambient temperature while the upper wall is fixed and kept adiabatic. A uniform constant heat flux is
applied at the bottom wall with a view to modeling heated microchip structure. Governing equations expressing in
non-dimensional forms are solved by using Penalty finite element method. Parametric studies are carried out for
different values of Richardson number ranging from 0 to 10. The influence of the governing parameters on the
Nusselt number at the heat source surface is observed for three different side angles of the moving lids. Results
presented in the form of streamline and isotherm plots show a comparative explanation for the variation of heat
transfer characteristics.

Key Words: Lid-driven, mixed convection, Penalty finite element, Richardson number.

1. INTRODUCTION
Mixed convection flow and heat transfer in lid-driven cavities have been receiving a considerable attention
in literature. This attention stems from its importance in vast technological engineering and natural applications.
Analysis of mixed convective flow in a lid-driven cavity finds applications in cooling of electronic devices, food
processing, dynamics of lakes, reservoirs and cooling ponds, and among others. Moallemi and Jang [1] numerically
studied mixed convection flow in a bottom heated square driven cavity and investigated the effect of Prandtl number
on the flow and heat transfer process. They found that the effects of buoyancy are more pronounced for higher
values of Prandtl number. Kuhlmann et al. [2] conducted a numerical and experimental study on steady flow in
rectangular two sided lid-driven cavities. They found that the basic two-dimensional flow is not always unique. For
low Reynolds numbers, it is consisted of two separate co-rotating vortices adjacent to the moving walls. Khanafer
and Chamkha [3] examined numerically mixed convection flow in a lid-driven enclosure filled with a fluid saturated
porous medium and reported on the effects of the Darcy and Richardson numbers on the flow and heat transfer
characteristics. Aydin and Yang [4] numerically studied mixed convection heat transfer in a two-dimensional square
cavity having an aspect ratio of 1. In their configuration the isothermal cold sidewalls of the cavity were moving
downwards with uniform velocity while the top wall was adiabatic. A symmetrical isothermal heat source was
placed at the center of the partially insulated bottom wall. They investigated the effects of Richardson numbers and
the length of the heat source on the fluid flow and heat transfer. Oztop and Dagtekin [5] performed numerical
analysis of mixed convection in a square cavity with moving and differentially heated sidewalls.
In what concerns natural convection heat transfer in parallelogrammic enclosures, there is a limited number
of studies, and they highlight the strong potential of this geometry to be used in efficient heat transfer systems.
Naylor and Oosthuizen [6] numerically studied two-dimensional free convective flow in a parallelogram-shaped
enclosure. They performed the calculations for Rayleigh numbers ranging from 103 to 105 and enclosure aspect
ratios from 0.5 to 3. Seki et al. [7] described the experimental measurements for natural convective heat transfer
across a parallelogrammic enclosure with various tilt angles of upper and lower walls. The characteristics of the heat
transfer were discussed through the evaluation of the heat transfer coefficient and effects of inclination angles on the
flow patterns. Aldridge and Yao [8] described the flow features of natural convection in a parallelogrammic
enclosure. This study was conducted by direct simulation for the transient problem using the finite volume method.
The Rayleigh numbers used for water were 2.55 × 106 and 2.92 × 107 with Prandtl number 6.8, for silicone oil 4.3 ×
104 and 3.2 × 105 with Prandtl number 3300. The buoyancy driven flow fields for various regimes were displayed
and discussed.

Mechanical Engineering Division The Institution of Engineers, Bangladesh

Proceedings of the 12th Annual Paper Meet 2
8-9 February 2008, Dhaka

In this study, we consider the steady and laminar flow in a lid-driven parallelogram cavity of dimensions H
× W. The aspect ratio of the enclosure is considered as 1 (H = W). The physical model and coordinate system
considered in this investigation are shown in Fig. 1. The left inclined side wall is moving at an angle φ in the upward
direction with a uniform velocity Uo whereas the right side wall is moving at the same angle and same velocity in
the downward direction. The moving side walls are maintained at constant temperature Tc, a constant heat flux, q is
applied at the bottom wall and upper wall is kept adiabatic. The Renolds number, Re is taken as 100, the Richardson
number is varied from 0 to 10 and Prandtl number, Pr is taken as 0.71. To the best of the author’s knowledge, no
attention has been paid to the problem of mixed convection of flow and heat transfer in a lid-driven parallelogram
cavity with isoflux heating from below. The objective of the present study is to examine the effects of the governing
parameters on the Nusselt number at the heat source surface.

2. MATHEMATICAL MODEL
The governing equations for laminar steady two-dimensional mixed convection, after invoking the
Boussinesq approximation and neglecting the viscous dissipation, can be expressed in the dimensionless form as
∂U ∂V
+ =0 (1)
∂X ∂Y
∂U ∂U ∂P 1 ⎛ ∂ 2 U ∂ 2 U ⎞
U +V =− + ⎜ + ⎟ (2)
∂X ∂Y ∂X Re ⎜ ∂X 2 ∂Y 2 ⎟
⎝ ⎠
∂V ∂V ∂P 1 ⎛ ∂ 2V ∂ 2V ⎞
U +V =− + ⎜ + ⎟ + Ri θ (3)
∂X ∂Y ∂Y Re ⎜ ∂X 2 ∂Y 2 ⎟
⎝ ⎠
∂θ ∂θ 1 ⎛ ∂ 2θ ∂ 2θ ⎞
U +V = ⎜ + ⎟ (4)
∂X ∂Y Re Pr ⎝ ∂X 2 ∂Y 2 ⎠
where X and Y are the coordinates varying along horizontal and vertical directions, respectively, U and V are the
non-dimensional velocity components in the X and Y directions, respectively, θ is the non-dimensional temperature
and P is the non-dimensional pressure. The non-dimensional governing parameters seen in the above, Ri, Re, Gr, and
Pr are the Richardson number, Reynolds number, Grashof number, and Prandtl number respectively, and they are
defined as follows:
Gr ρU 0 W g β q W4 ν
Ri = , Re = ,Gr = , Pr = (5)
Re 2
µ kν 2
α
The dimensionless parameters in the equations above are defined as follow:
X=
x
, Y=
y
, U=
u
, V=
v
, P=
p
, θ=
( T − TC )
(6)
W W U0 U0 ρ U02
qW / k
where ρ, β, ν, α, g, and cp are the fluid density, coefficient of volumetric expansion, kinematic viscosity, thermal
diffusivity, gravitational acceleration, and specific heat, respectively.
The boundary conditions for the present problem are specified as follows:
∂θ ∂θ
Upper wall: = 0, U = V = 0 and Bottom wall: = −1, U = V = 0
∂Y ∂Y (7)
Left wall: θ = 0, U = cos Φ , V = sin Φ and right wall: θ = 0, U = − cos Φ, V = − sin Φ
where Φ is the inclination angle of the vertical side walls of the enclosure.
The average Nusselt number at the heated wall of the enclosure may be expressed as
1
∫ θ ( X ) dX
1
Nu = (8)
0 S
where θS (X) is the local dimensionless temperature of the heated surface.

3. SOLUTION OF THE PROBLEM

The numerical procedure used to solve the governing equations for the present work is the Galerkin finite
element method. The application of this technique is well documented by Zienkiewicz and Taylor [9]. A non-
uniform triangular mesh arrangement is implemented in the present investigation especially near the heated wall to
capture the rapid changes in the dependent variables. The continuity equation (1) will be used as a constraint due to
mass conservation and this constraint may be used to obtain the pressure distribution [9]. In order to solve Eqs. (2) -
(4), we use the Penalty finite element method [10] where the pressure P is eliminated by a penalty parameter. The

Mechanical Engineering Division The Institution of Engineers, Bangladesh

Proceedings of the 12th Annual Paper Meet 3
8-9 February 2008, Dhaka

three noded triangular elements are used in this paper for the development of the finite element equations.
The convergence of solutions is assumed when the relative error for each variable between consecutive iterations is
recorded below the convergence criterion δ such that
Γ n +1 − Γ n
<δ (9)
Γ n +1
where n is the Newton iteration index and Γ = U, V, P and θ. The convergence criterion is set to 10-6.

To test and assess grid independence of the present solution scheme, many numerical runs are performed for
Richardson number, Ri = 10 and Φ = 30° as shown in Table 1. These experiments reveal that a non-uniform spaced
grid of 5966 elements for the solution domain is adequate to describe correctly the flow and heat transfer processes
inside the enclosure. In order to validate the numerical model, the results are compared with those reported by Guo
and Sharif [11], for square straight enclosure with Ri = 10 and L/W = 0.2, where L is the length of heat source. As
shown in Fig. 2, the agreement is found to be excellent which validates the present computations indirectly.
Therefore, a conclusion can be drawn that the present configuration can be handled very well by this numerical
procedure.

4. RESULTS AND DISCUSSION

The buoyancy and shear-driven convection in a parallelogram cavity with a centrally heated source at the
bottom wall and moving cold side walls is investigated. Also the flow and temperature fields in terms of computed
streamlines and isotherms for various values of inclination angle of the side walls, Φ = 30°, 45° and 60° are shown
in Figs. 3-5. The different values of the mixed convection parameter, Ri = 0, 1 and 10 are considered here as
representative cases for a constant value of Reynolds number, Re = 100 and Prandtl number, Pr is taken as 0.71. The
value Re = 100 indicates that the flow is laminar where viscous forces are dominant, and is characterized by smooth,
constant fluid motion. The value Pr = 0.71 indicates that the flowing fluid is air.
Because of the asymmetry of the problem geometry and the symmetry of the boundary
conditions, the flow and temperature fields are asymmetrical about the midlength of the cavity. Due to
shear force induced by the moving lids, the flow directions in the left and right halves of the cavity are
same except for the sense of rotating strength. In each case, the main flow rises along the inclined cold
side wall from the left-end portion of the bottom wall and gets blocked at the top adiabatic wall, then the
flow descends downwards along the vertical center axis of the cavity and turns horizontally to the corner
end after hitting the bottom wall. Similarly, another clockwise rotating cell forms on the right half of the
cavity indicating buoyancy assisted flow within the cavity. Thus the moving fluid forms two asymmetrical
rolls with clockwise rotations inside the cavity. For low Richardson number, the presence of the
stagnation point is noticed along the symmetric vertical axis. As expected, the streamline plots do not
reveal any significant difference for increasing Ri. But noticeable difference is observed in the isotherm
plots for different inclination angle of the sides of the cavity.
For Φ = 30°, as it can be seen from Fig. 3, the streamlines and isotherms at Ri = 0, 1, and 10 are
nearly the same because of the moving side walls with negligible buoyancy. The geometry of the flow
field and temperature field are not affected considerably by changing Ri and also the left circulating cell
assisted by the buoyancy effect grows in size and covers more than half of the cavity. The effect of
Richardson number, Ri on the fluid flow and heat transfer for other inclination angle of the side walls
considered here, is found to be in a similar behavior to that of Φ = 30°, which can be seen from Figs. 4-5.
For Φ = 45° and 60°, two recirculating cells of same directions of motion are developed. An increase of Φ
leads the left circulating cell to expand and the right circulating cell squeeze. Here the symmetry is
completely destroyed due to the action of the buoyancy flow and the forced flow. Also tilting the vertical
sidewalls of the cavity to an extend yields the right side shipment of the cores.
The average Nusselt number at the hot surface, which is a measure of the overall heat transfer rate, is
plotted in Fig. 6. The effect of Richardson number on the overall heat transfer process is clearly depicted in this
figure. From Fig. 6, we can see that the inclination angles of the side walls plays an important role on heat transfer
for different values of Richardson number. The average Nusselt number distribution shows that the heat transfer
increases with the increase of Richardson number. For Ri ≤ 1, heat transfer rate becomes linear but for Ri >1, the
heat transfer rate increases sharply. It can also be seen from Fig. 6 that, maximum heat transfer occurs in the case of
inclination angles of the side walls, Φ = 30°. This is due to the fact that the cold side walls acting as heat sinks
within the enclosure come very close to the heat source at low inclination angles.

Mechanical Engineering Division The Institution of Engineers, Bangladesh

Proceedings of the 12th Annual Paper Meet 4
8-9 February 2008, Dhaka

5. CONCLUSION
In this investigation, convection heat transfer due to both buoyancy induced flow and forced flow inside a
parallelogram cavity is numerically investigated. Effect of buoyancy parameter for mixed convection in a bottom
heated corrugated cavity is studied. This result shows that the heat transfer increases with the increases of Ri for any
inclination angles of the sidewalls of the parallelogram cavity. Moreover, the maximum heat transfer is obtained
when the inclination angle of the side walls of the cavity is Φ = 30°.

REFERENCES
[1] Moallemi, M. K. and Jang, K. S., 1992. Prandtl number effects on laminar mixed convection heat transfer
in a lid-driven cavity, Int. J. Heat Mass transfer, 35, 1881-1892.
[2] Kuhlmann, H. C., Wanschura and M., Rath, H. J., 1997. Flow in two sided lid-driven cavities: non-
uniqueness, instabilities, and cellular structures, J. Fluid Mech., 336, 267-299.
[3] Khanafer, K. and Chamkha, A. J., 1999. Mixed convection flow in a lid-driven enclosure filled with a fluid
saturated porous medium, Int. J. Heat Mass Transfer, 42, 2465-2481.
[4] Aydin, O. and Yang, W. J., 2000. Mixed convection in cavities with a locally heated lower wall and moving
sidewalls, Numer. Heat Transfer, Part A., 37, 695-710.
[5] Oztop, H. F. and Dagtekin, I., 2004. Mixed convection in a two-sided lid-driven differentially heated square
cavity, Int. J. Heat Mass Transfer, 47, 1761-1769.
[6] Naylor, D. and Oosthuizen, P. H., 1994. A numerical study of free convective heat transfer in a
parallelogram-shaped enclosure, Int. J. Numer. Meth. Heat Fluid Flow, 4, 553–559.
[7] Seki, N., Fukusako, S. and Yamaguchi, A., 1983. An experimental study of free convective heat transfer in
parallelogrammic enclosure, ASME J. Heat Transfer, 105, 433-439.
[8] Aldridge, K. D. and Yao, H., 2001. Flow features of natural convection in a parallelogrammic enclosure,
Int. Comm. Heat Mass Transfer, 28 (7), 923-931.
[9] Zienkiewicz, O. C. and Taylor, R. L., 2000. The finite element method, 5th edn. Oxford: Butterworth-
Heinemann.
[10] Natarajan, E., Basak, T., and Roy, S., 2008. Natural convection flows in a trapezoidal enclosure with
uniform and non-uniform heating of bottom wall, Int. J. of Heat and Mass Transfer, 51, 747-756.
[11] Guo, G. and Sharif, M. A. R., 2004. Mixed convection in rectangular cavities at various aspect ratios with
moving isothermal sidewalls and constant flux heat source on the bottom wall, Int. J. of Thermal Sciences,
43, 465-475.

TABLES

Table 1: Comparison of the results for various grid dimensions (Ri = 10 and Ф = 30°)
Elements 4560 5924 5966 7162
Nu 4.661 4.698 4.701 4.702

Mechanical Engineering Division The Institution of Engineers, Bangladesh

Proceedings of the 12th Annual Paper Meet 5
8-9 February 2008, Dhaka

FIGURES

Fig. 1: Schematic diagram of the physical domain

Guo and Sharif [11] Present Work

Fig. 2: Comparison of isotherm plots of the square enclosure with Guo and Sharif [11] at Ri = 10 and L/W = 0.2
Streamlines
Isotherms

Ri = 0 Ri = 1 Ri = 10

Fig. 3: Variation of streamlines and isotherms plots for Ф = 30°

Mechanical Engineering Division The Institution of Engineers, Bangladesh

Proceedings of the 12th Annual Paper Meet 6
8-9 February 2008, Dhaka

Streamlines
Isotherms

Ri = 0 Ri = 1 Ri = 10
Fig. 4: Variation of streamlines and isotherms plots for Ф = 45°
Streamlines
Isotherms

Ri = 0 Ri = 1 Ri = 10
Fig. 5: Variation of streamlines and isotherms plots for Ф = 60°

Fig. 6: Variation of average Nusselt number with the Richardson number for different inclination angles of the side
walls

Mechanical Engineering Division The Institution of Engineers, Bangladesh