Contents

List of tables .................................................................................................................................... 2

List of Figures ................................................................................................................................. 3

CFD Assignment ............................................................................................................................. 5

Natural Convection in Cube........................................................................................................ 5

Introduction to Problem Statement ............................................................................................. 5

Methodology ............................................................................................................................... 5

Computational Fluid Dynamics .................................................................................................. 6

Geometry..................................................................................................................................... 7

Meshing..................................................................................................................................... 10

Mesh Independence Test........................................................................................................... 12

Case Setup ................................................................................................................................. 13

Setup Validation........................................................................................................................ 15

Analysis..................................................................................................................................... 17

Results ....................................................................................................................................... 18

Post Processing ......................................................................................................................... 19

 List of tables

Table 1 .......................................................................................................................................... 13

Table 2 .......................................................................................................................................... 16

Table 3 .......................................................................................................................................... 17

Table 4 .......................................................................................................................................... 18
 List of Figures

Figure 1 Cube Geometry................................................................................................................. 8

Figure 2 Adiabatic Walls generations ............................................................................................. 8

Figure 3 Hot Face Definition .......................................................................................................... 9

Figure 4 Cold Face Definition ........................................................................................................ 9

Figure 5 Mesh Setting ................................................................................................................... 10

Figure 6 Surface Mesh on Cube.................................................................................................... 11

Figure 7 Volume Mesh ................................................................................................................. 12

Figure 8 ......................................................................................................................................... 14

Figure 9 ......................................................................................................................................... 14

Figure 10 ....................................................................................................................................... 15

Figure 11 ....................................................................................................................................... 15

Figure 12 Temperature Contours .................................................................................................. 19

Figure 13 ....................................................................................................................................... 20
Figure 14 ....................................................................................................................................... 21

Figure 15 ....................................................................................................................................... 22

Figure 16 ....................................................................................................................................... 23

Figure 17 ....................................................................................................................................... 24
 CFD Assignment

Natural Convection in Cube

Introduction to Problem Statement

Flows of Newtonian fluids driven by buoyancy in square, rectangular or cubic enclosures are found

in a wide range of engineering applications such as nuclear reactors, solar devices, food products,

and polymer processing industries. For differentially heated three-dimensional enclosures with

adiabatic side walls, the heat transfer characteristics are influenced by Prandtl number (Pr) and

Rayleigh number (Ra) based on the height of the cube. The problem aims at simulating the heat

transfer problem. The heat transfer takes place through the convection. The heat transfer problem

is described by a dimensionless parameter called Rayleigh Number. For a given Rayleigh number,

the height of the cube needs to be evaluated and the description of the whole process of numerical

simulation is to be discussed.

Methodology

The first step in the analysis is to find the height of the cube. The given Rayleigh number is

10*5. The formula for the Rayleigh number is given as:

g is the gravitational acceleration taken as 9.8 m/s2

ß is the thermal expansion coefficient of air taken as 0.00369/k

α is the thermal diffusion constant of air

υ is the viscosity of air taken as 1.81*10^5

Delta T is the change in temp between the left and right side of the cube. The change in temp is

taken to be 100 C.

Putting all these values in the above formula the height of the cube appears to be 0.22m.

Computational Fluid Dynamics

Computational fluid dynamics (CFD) deals with study of fluid flows or fluid dynamics and

problems associated with it by using numerical techniques and data structures, it is a branch of

fluid mechanics. Advancement from manual to computers become essential to perform the

calculations required to understand the behavior of free stream flow when the fluid interact with

the surfaces defined through boundary conditions.

The field of CFD became frequently used tool for computing solutions for fluid flows with or

without solid interaction because of rise in computational powers. (Carmelietcd, 2010) The

physical properties which are used for the examination of fluid flows are velocity, pressure,

temperature, density, viscosity and so on. The Navier-Stokes equations are used as mathematical

model for the physical case which describes both the fluid flow and heat transfer
 Define Choose
Choose model Iteration and Post
import CAD model Generate Grid boundry type variable
for solver save results Processing
and condition (monitor)

The figure above shows the general procedure of the CFD analysis, the very first step is to create

a water tight geometry and then generate a grid which covers the fluid region using a fine good

quality mesh and then define the boundary conditions and then define the equations to be solved

and then the iterations are run and after the iterations are done and the solution is converged the

post processing is done to check the flow picture and to check the output results obtained.

Geometry

As specified in the problem, the geometry of the cube is made in the commercially available

software for performing CFD named as Star CCM plus. The geometry in star CCM plus looks

like following:
 Figure 1 Cube Geometry

The left face of the cube is the hot face, whereas the right face is cold on. The upper lower and

front and back face are adiabatic according to the specified condition. The cube is made hollow

and the fluid inside the cube is air.

Figure 2 Adiabatic Walls generations

After the geometry, the next step is to define the adiabatic wall. Figure 2 represents the adiabatic

walls created and represented in pink color. While creating the geometry it is made sure that the

cube is made such a way that there is no gap left so that the air shall not leak inside the cube.

This process is known as cleaning the geometry or building the topology of the cube. The water

tight geometry of the cube is the most important part of the geometry creation.

Figure 3 Hot Face Definition

Figure 4 Cold Face Definition

Meshing

Once the geometry is made, the next step is to do the pre-processing setup. This is done by making

a mesh/ grid over the entire domain. The concept of this grid is that the equations will be solved at

each node. The finer the grid, the better results would be there. But a fine grid requires more time

and resources to compute the results. So, there is always a tradeoff between the size of the mesh

and the accuracy. The following mesh settings are made in Star CCM Plus

Figure 5 Mesh Setting

The polyhedral elements were selected for the meshing and 5 prism layers are created. The mesh

is created such a way that the important areas are covered using the finer mesh and the relatively

less important areas are covered with coarse mesh. The finer mesh enables us to cover the variation

in the flow of air inside the cube.

The mesh on the cube looks like following:

 Figure 6 Surface Mesh on Cube

The volume mesh is created in a way that the mesh at the surface is finer and then the mesh

grows continuously as it moves in the volume region. The elements are growth in such a way tha

the boundary layer is covered properly. The Volume mesh looks like following
 Figure 7 Volume Mesh

The mesh has a very good quality and hence we are ready to go for the analysis part.

Mesh Independence Test

The mesh independence test is necessary for any simulation based analysis. The mesh

independence test proves that the results are independent of the size of the mesh. The mesh

independence test is performed by forming different sizes mesh and running analysis over them.

The size of the mesh is changed, it is increased and decreased from the original value and analysis

is run over it (Cristi Angheluta, 2019). Finally a stage is reached where the changes in the mesh

size does not affect the results produced. The factor that is analyzed is the Nusselt number. The

analysis for different mesh is summarized in the following table:

 Table 1

Grid Name Size Nusselt Number

Mesh 1 200*70 4.52

Mesh 2 250*50 4.51

Mesh 3 300*60 4.509

Mesh 4 250*100 4.5091

The result from all the cases is the same. Hence the mesh independence is carried out and is

validated. Any mesh size can be used because all the mesh produce the same results. The same

value for the Nusselt number verifies that the result is independent of the mesh size. The analysis

for the mesh independence was carried out on the Prandtl Number of 0.71. the value of n for the

analysis is 1.

Case Setup

The case is setup as follows:

The steady state analysis is done, with the couple implicit technique used.
 Figure 8

One of the main property which was observed is heat transfer, we observed heat being

transferred during iterations.

Figure 9

The boundary conditions were set and the two walls were given the specific temperatures and the

other four walls were given the adiabatic boundary condition.

 Figure 10

The monitors were set as shown in the figure and the stopping criteria was the equation of

continuity up to 1 e -06 for convergence criteria. This tells that the further no more iterations are

required and the result obtained is the final result.

Figure 11

Setup Validation

Whenever a real world problem is analyzed though a numerical simulation, the first step is to

validate the numerical setup. The validation of this numerical setup ensures the accurate result.

The model is the processing machine for the case (htt1). If the model is not correctly selected, the

results would never be the correct even the simulation is analyzed properly. The setup is validated
by solving a problem of the same nature on the present setup. The results for that problem already

exist. These results are experimental or published results. To validate the given problem of the

heat transfer through the natural convection, the already existing data is taken from (Xie JF, 2017)

Xie JF, Cao BY (2017) Natural convection of power-law fluids under wall vibrations: A lattice

Boltzmann study, Numerical Heat Transfer A 72, 600-627.

The result from the above mentioned paper and our present case are presented in the table below:

Table 2

Parameter Present Study Research Paper % Error

Nusselt Number 4.5 4.52 0.4

Umax/V* 67 68.3 1.9

Vmax/V* 35 34.7 2

The percentage error in all of the above parameters is less than 5 percent, hence the setup is

validated and we process with the analysis.

Analysis

The validated setup is used for numerically simulating the heat transfer problem for the Newtonian

fluids. The Lattice Boltzmann method is used. The validation for this method is presented in the

reference paper. The analysis was run on the following boundary conditions:

Table 3

Cube Face Boundary Condition Temp ( C )

Top face Adiabatic wall -

Bottom face Adiabatic wall -

Front face Adiabatic wall -

Back face Adiabatic wall -

Left Side wall Velocity Inlet 100

Right Side Wall Velocity Outlet 0

The non dimensionless parameters were kept as follows:

 Parameter Value

Rayleigh Number 10^5

Prandtl Number 0.7

n 1

Height 0.22 m

Results

The results were calculated when the solution has properly converged. The following results

were obtained

Table 4

Parameter Value

Nusselt Number 4.5

 Umax/V* 67

Vmax/V* 35

Post Processing

Figure 12 Temperature Contours

Here was a case in which the velocity was static, and we can see that the heat is being transferred

from higher intensity surface to the lower intensity surface. The variation in the temperature can

be seen from the figure above. (Snoussi, 2013) The reddish region shows the high temperature

region, and the bluish region shows the lower temperature region. The region between shows the

temperature lying in the range as shown in the figure above.

 Figure 13

As the velocity is varied the temperature variation is also changed. The velocity affects the

change in transfer and the convection inside the cube. The increase in the velocity increases the

kinetic energy of the particles and hence resulting in increase in the change in the temperature

variation in the surface. The greenish and the bluish region in the figure above shows the mixing

of the high and low temperature regions. The band between the reddish and bluish region has

increased as compared to the case discussed before which has proved that the increase in the

velocity has helped in increased in the absorption of the heat.

 Figure 14

Here we can observe a smooth band of temperature variation, the better temperature variation is

due to the mixing of fluid inside the cube due to the kinetic energy induced by the temperature

difference. The picture shows that there is a motion induced inside the cube, and the temperature

has varied smoothly and hence distributed inside the region of the cube.
 Figure 15

So the final case of the temperature variation is the figure above. The smooth transfer of hear can

be seen easily.
 Figure 16

The pressure contours in the above figure are the values at the surfaces which will be around

zero due to no velocity due to no slip condition. This tells that the flow is static at the surface of

the cube so the pressure is not built that high inside the cube and the velocity in side the cube is

also not induced that high to create much of a pressure at the surface of the cube.
 Figure 17

Here we can see that the velocity at the surface of the cube is approximately zero because of no

slip condition. The contours in the above figure shows that the velocity which is induced will be

inside the cube and won’t effect much at the surface of the cube.

Conclusion
Finally, we can conclude that if two sides of the cube are set at different temperatures keeping all

other sides as adiabatic, i.e. no heat transfer, the heat will transfer from the high temperature side

to the low temperature sides increasing the temperature of the other sides and the region in

between the surface. There is also flow induced into the space in the cube due to the change in

the temperature level and the velocity inside the cube increases and the flow starts to circulate.

The change in the temperature variation also increases with the increase in the velocity of air.
 References

(n.d.). Retrieved from https://docs.oracle.com/en/cloud/saas/applications-

common/22d/farhd/validate-the-setup-for-classifying-service-requests-created-from-
inbound-emails.html

Carmelietcd, T. D. (2010). CFD analysis of convective heat transfer at the surfaces of a cube
immersed in a turbulent boundary layer.

Cristi Angheluta, F. M. (2019). Mesh Independence .

Snoussi, L. (2013). CFD studies of natural convection in a Cube.

Xie JF, C. B. (2017). Natural convection of power-law fluids under wall vibrations .