Analysis and numerical solution of the temperature gradient

across a 2-D Square Plate

SUMMER 2017

Instructor Students

Contents
Problem Statement......................................................................................................................................3
Sketch of the problem.................................................................................................................................4
Difference Equation Derivation....................................................................................................................4
Results.........................................................................................................................................................5
Analytical Solution...................................................................................................................................5
Numerical Solution..................................................................................................................................7
Plots.......................................................................................................................................................10
Discussion and Conclusion.........................................................................................................................11
Problem Statement
The temperature T is maintained at 0C along three edges of a square plate of length 6 cm, and

the fourth edge is maintained at 100C until steady state conditions prevail. This groups

assignment was to derive the Difference Equation using the finite difference method (with steps

h=k=1) and using it to solve the temperature at all unknown node points. Furthermore, it is

required to use the analytical solution to evaluate the temperature at locations (2,2) and (2,4)

compare it to the value obtain through the numerical solution and finally plot the temperature

contours of the plate.

Difference Equation Derivation

equation:

(1)

The steps to derive the equation are shown below.

Starting from Laplaces Equation:

By changing the notation such that u(x,y)= ui,j and x h = i1 , we can rewrite eq. 5 and

rearrange it in the form:

Therefore, each node points can be approximated by of the summation of its direct neighbors.

Results
Analytical Solution
The general solution of the Laplace Equation for the temperature gradient of a hot plate is the

following:
In our case, the dimensions are a=b=6cm and the fourth edge is a maintained at f(x)=100 C.

When substituting a,b and f(x), equation 6 becomes:

To validate the equation the boundary conditions are checked:

For x=0 the sin term becomes 0:

For x=6 the sin term becomes 0:

For y=0 the sinh term becomes 0:

For y=6 the estimated value was 106

This value does not seem to coincide very well with the boundary condition u(x,6)=100 Thats

because only three terms of the series were used to estimate that value. However, Wolfram Alpha
was used to estimate the value u(x,6) for n=1 to 1000 terms which resulted to be 100 which

checks the last of the boundary conditions.

Once it was proved that the equation is valid for the boundary conditions, equation 7 was used to

estimate the temperature at nodes (2,2) and (2,4). The values obtained with the first three terms

of the series resulted to be very close to the evaluation of the series for n=1 to 1000. Here, we

report the values for up to 3 terms to comply to the assignment:

Numerical Solution

In one of the previous sections, we have shown the derivation of the Difference Equation which

leads to equation 1. Equation 1 was used to solve for the numerical solution.

given that,
 given that,

given that,

given that,

given that,

By similarity, it is possible to express the explicit equation for all the remaining nodes. These

equations can be inserted as formulas in the cells of Microsoft Excel spreadsheet. Each cell

represents a node. By using the iterative method, the temperature values at each node were

computed by the software and are reported in the following tables.

Table 2: Unknown and known values of the nodes

By iterative method,the obtained results were the following:

100 100 100 100 100

0 46.86791 62.92132 66.94133 62.92162 46.86825 0

0 24.55108 37.87703 41.92326 37.87747 24.55159 0

0 13.46037 22.11363 24.99824 22.11407 13.46088 0

0 3.130874 5.346078 6.133763 5.346243 3.131066 0

0 0 0 0 0

Table 3: Tabular form of the temperature gradient with steps h=k=1

Plots
The contour plots shown were created with Microsoft Excel.

Discussion and Conclusion

The results obtained by the two different methods agree with each other: the margin of error is

between 1% and 2% which strongly supports this evidence. For the purpose of theproject, the

more accurate solution is considered to be the one obtained through the numerical iterative

method. With the numerical method, the iteration stops whenever the nth+1 value is almost equal

to the nthvalue. Their difference needs to be smaller than a parameter that comes from the design

of the engineer. The engineer, through appropriate calculations, has the responsibility to establish

a margin of error that is small enough to make the softwares computation satisfactory to its real-

life application. The reason why the numerical method can be more accurate of the analytical

method is because of the mathematical limitation that comes with the analytical expression itself.

The analytical solution is indeed expressed a series that goes from n=1 up to infinity and it is

practically impossible to sum an infinite number of terms to get to the real value. However, it is
possible to sum a number of terms that is high enough to get a solution that approximates well

the real value. Usually, it is not too difficult to figure out what the upper term of the n value

should be. However, even with the computational softwaresavailable today, this method might

take up a lot of time when the expressions are complicated enough for the capacity of the

calculator or computer.

It is important to notice that in the Laplace Equation, for steady state conditions, the temperature

gradient does not depend on the plate thickness nor on the material properties. As it can be seen

in the analytical solution, the temperature gradient will depend only on the plate surface

dimensions (a and b) and on the boundary conditions. This can be explained by the fact that the

properties of the material (such as thermal conductivity or thermal diffusivity) can affect the rate

of the heat transfer through the material which also leads to the distinctions between conductors

(with a high conductivity and then high heat transfer rate) and insulators (with a low conductivity

and very low heat transfer rate). However, in the problem assigned, only the steady state

response of the gradient is taken into consideration. The steady state response is not affected by

the material properties.

In conclusion, the conductivity of a material determines the transient response of the material

which can be seen as the rapidity of a material to cool down or heat up. When steady state

conditions prevail, the net heat transfer is equal to 0 (Qin=Qout), the temperature gradient will

depend only on the boundary conditions and on the dimensions of the plate independently of the

conductivity of the material.