[ramakant]

Dear all

I need this CFD code of the given problem.If anybody CFD code of the problem please try to share with me.
my e-mail id:ramakantgadhewal@gmail.com
Thank you

With Best Regards

Ramakant Garhewal

[mprinkey]

We don't do homework problems? Most of us already have done them--all of them. Several of us *assign* them to students. If you have a specific question about how to write your own code, please ask. But there are no shortcuts to learning this topic--you have to do it yourself.

[arjun]

Originally Posted by mprinkey

We don't do homework problems? Most of us already have done them--all of them. Several of us *assign* them to students. If you have a specific question about how to write your own code, please ask. But there are no shortcuts to learning this topic--you have to do it yourself.

What he said plus if you want hint look into patankar's book or malalasekera's book. They have worked example all you need to do is code it.

Try it is not difficult and plus what you will learn will stay with you a long time. There is no learning like learning by doing.

[LuckyTran]

I echo the sentiments above.

This isn't even a Fluent problem. If done in Fluent you would be solving the heat diffusion equation and not the fin equation. It's also a really simple problem and is really great for learning the FVM technique. I admit I did not appreciate how FVM works until I did it myself. This is a really simple problem! 1D steady fin equation. It can hardly get much easier than this. This is also a very robust problem, it's extremely well-behaved and numerically stable.

It doesn't even require a complicated code.

Just discretize the governing equation in the FVM sense. Integrate the governing equation over a control volume, use the Divergence Theorem to convert the volume integral of a divergence term into a surface integral of fluxes. Then discretize the temperature gradient at the faces using your favorite discretization scheme (backward differencing, forward differencing, central differencing). Do this for all cells~

Notice the pattern! All the interior cells will have the same structure. The boundary cells will have slightly different terms. i.e. you only need to do the left, right, and one interior cell to recognize the pattern. If you are not imaginative, then you might need to do it for all cells.

Setup the linear system of equations (writing it in matrix form can help the pattern fall out). If you use central differencing, you'll get a tri-diagonal system. Then it's just a matter of solving the system of linear equations (inverting the matrix). Tri-diagonal algorithm works wonders. Matrix form just helps to visualize, if you can solve it without going into matrices, then kudos.

This final step, solving the system of linear equations is what needs to be coded. You basically need to write a code that systematically assembles the coefficient matrix and then solve it.

If you have trouble, just do it on the simplest case with only 3 cells (the left boundary, 1 interior cell, and the right boundary cell). Hopefully you can invert a 3x3 matrix (or can solve a linear system of 3 equations with constant coefficients). You just need to codifiy this procedure for however many cells you need.