FDM 1 PDF

1.
Problem Description
Solve the unsteady 1-D heat conduction equation using the finite difference method
Specify the material and thermal conductivity
Solve the heat conduction equation
Plot the temperature distribution with respect to time
Fig. 1. Problem Description
2. Governing Equation
The governing equation of this problem is 1D conduction equation in transient state.
The conduction equation
is partial differential equation of t of time and x,y,z of length as below.

2 T 2 T 2 T 1
+
+
+ =
x 2 y 2 z 2
In this case, since = 0, and only x-direction will be considered, the governing equation reduces to
2 T 1
=
x 2
3. Solution Method
The problem was solved with MATLAB code and the result also verified with commercial CFD code, FLUENT
ver. 6.3.
I assumed achievement of linear temperature distribution occur when average of temperature at each
node is equal to 335.15K.
Average of temperature along the bar should be same with average of temperature at
both ends, which is 335.65K.
1) MATLAB code
To solve this problem, I used explicit method to calculate temperature distribution of the rod.
By
discretizing 1D conduction equation for explicit method, the finite difference of temperature was
obtained as below.
2 T 1
=
x 2
1 (+1 ) +1
2 + 1
=
()2
+1 =
(+1
2 + 1
)
+
()2
The notation i is for element and n is for time.

this equation is attached at Appendix.
The entire MATLAB code of formulation based on
For formulation, the bar was dissected into 10 pieces with
even length of dx=0.1 m.
2) FLUENT
To solve this problem, 2 dimensional problem solver of FLUENT ver. 6.3.26 was used. The solver is
pressure based, unsteady solver with 1st order implicit formulation.
Energy equation was activated to
solve heat transfer at the system.
4. Boundary Condition
1) Grid Generation for CFD (FLUENT)
Fig. 2. Grid Generated
Fig.2. is the grid generated and it was generated in Gambit.

shape.
The length is 1, height is 0.1 m.
The grid was generated in 2D, square
The number of grid in longitudinal direction is 100, which
means the length between two adjacent nods is 0.01 m.
The number of grid in transverse direction is
10 to make same length with grid in longitudinal direction.
2) Boundary Condition
To make 1D conduction condition, it was assumed that there is no heat flux to the y-direction.
At the
left side of the bar, where x=-0.5m, temperature was set to 298.15K and at the opposite side of the bar,
where x=0.5m, temperature was set to 373.15K.
Time step length was set to 0.1s for both method.
3) Material Properties
I selected three different materials to examine the effect of thermal conductivity to the time elapsed to
achieve linear temperature distribution along longitudinal direction.
own thermal conductivity are tabulated in Table. 1.
The selected materials and their
Thermal properties of the materials was obtained
from material database installed in FLUENT.

Table 1. Thermal Properties of the Selected Materials
Material
Aluminum (Al)
Gold (Au)
Copper (Cu)
Thermal Conductivity
202.4 W/mK
297.73 W/mK
387.6 W/mK
Density
2719 kg/m3
19320 kg/m3
8978 kg/m3
Cp
871 J/kgK
129.81 J/kgK
381 J/kgK
Thermal Diffusivity
8.546e-05 m2/s
1.19e-04 m2/s
1.13e-04 m2/s
5. Result
1) MATLAB with Explicit Solution Method
a) Aluminum
Elapsed Time: 4776 s
Temperature Distribution along x-Direction and Change along Time
b) Gold
Elapsed Time:3430 s
c) Copper
Elapsed Time:3612 s
2) FLUENT with Implicit Solver

a) Aluminum
Temperature Distribution along x-Direction and Change along Time (Visualized)
b) Gold
c) Copper
3) Discussion
In this project, I derived the governing equation of one-dimensional heat transfer phenomenon
of different metallic materials Aluminum, Gold, Copper and solved it with two different method
and tools; finite difference method with explicit solution method and commercial CFD code, FLUENT
with implicit solver.
As shown in 5.1 and 5.2, the elapsed time to achieve proper average
temperature, which is set to 335.15 K (-0.5K error with theoretical value), was almost similar. The
elapsed time for each case was tabulated in Table. 2. There was approximately 100 seconds of
difference in every case, which was considered mainly because of the different length of a section 10
pcs for MATLAB, 100 pcs for FLUENT and difference of solution method.
In overall, the elapsed
time were similar in same material, and quite different along with material property, heat diffusivity .
Material\Solver
MATLAB (Explicit) (10pcs)
FLUENT (Implicit) (100pcs)
Aluminum
4776 s
4863 s
Gold
3430 s
3520 s
Copper
3612 s
3713 s
Appendix
MATLAB Explicit Solution Method Source Code
clear
%Generating Grid
n=input('Insert number of grid... ');
dx=1/n;
x=linspace(0,1,n+1);
%Material Property
a=input('Insert thermal diffusivity of the material... ');
%Setting Time Step and Initializing
dt=input('Insert time step size... ');
j=0; % # of iteration
Temp=298.15*ones(1,n+1);
Temp(n+1)=373.15;
T0=Temp;
%Formulation goes until avg of temperature of each point is equal to=335K
while mean(Temp(j+1,:))<335
j=j+1
for i=2:n
Temp(j+1,i)=(a)*(dt)/(dx)^2*(Temp(j,i+1)-2*Temp(j,i)+Temp(j,i-1))+Temp(j,i);
Temp(j+1,1)=298.15;
Temp(j+1,n+1)=373.15;
end
end
Elaptime=j*dt

FDM 1 PDF

Uploaded by

Copyright:

Available Formats

FDM 1 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FDM 1 PDF

Uploaded by

Copyright:

Available Formats

1.

Specify the material and thermal conductivity

Solve the heat conduction equation

Plot the temperature distribution with respect to time

Fig. 1. Problem Description

The conduction equation

is partial differential equation of t of time and x,y,z of length as below.

node is equal to 335.15K.

both ends, which is 335.65K.

The notation i is for element and n is for time.

The entire MATLAB code of formulation based on

For formulation, the bar was dissected into 10 pieces with

even length of dx=0.1 m.

Energy equation was activated to

solve heat transfer at the system.

Fig. 2. Grid Generated

Fig.2. is the grid generated and it was generated in Gambit.

The length is 1, height is 0.1 m.

The grid was generated in 2D, square

The number of grid in longitudinal direction is 100, which

means the length between two adjacent nods is 0.01 m.

The number of grid in transverse direction is

10 to make same length with grid in longitudinal direction.

Time step length was set to 0.1s for both method.

The selected materials and their

Thermal properties of the materials was obtained

from material database installed in FLUENT.

2) FLUENT with Implicit Solver

In overall, the elapsed

MATLAB (Explicit) (10pcs)

FLUENT (Implicit) (100pcs)

You might also like