[shinde.gopal]

Hi Foamers,

I have a question related to the cavity problem (tutorial case).
In the cavity tutorial case, domain size is 1 x 1 x 0.1 with 20 x 20 x 1 as the cell count in the blockMesh dict file.
I have increased the domain size to 100 x 100 x 0.1 with 2000 x 2000 x 1 as the cell count in the blockMesh dict file.
The controlDict, fvSchemes, fvSolution files are the same in both cases.
The tutuorial case gets over in a jiffy and the new cavity case (with increased domain and cells) takes a long time almost an hour to reach 100 itrs (t = 0.5 sec with dt = 0.005).

In the new case, iterations of orthogonal correctors for p are in the range 1000 and in the original case it is 1 or 2.
Please guide / help me in understanding why this is happening ?

Looking forward for your expert help and tips.

Gopal.

[al_th]

Hello,

In the original tutorial you solve a system with 400 cells.
In your modified case, you solve a system with 4 000 000 cells.

Remember that when you compute your solution, you need to compute it for each and every cell of your domain, and increasing drastically the number of cells will of course drastically increase the time you need.

Edit : in the cavity case there is no "non-orthogonal" corrector set up. Did you mean iterations for the pressure solutions ?

[shinde.gopal]

Dear Al_th,

Yes I meant the large number of iterations for pressure solutions.
How they are so large in number ?

Gopal.

[al_th]

Dear Gopal,

I am not an expert of CFD at all but I can try and do some wild guess of what is happening.

I did run two or three iterations of your case (cavity with increased number of cells) and apparently the final residual of the pressure is far higher than what is prescribed in the controlDictFile.

The domain might be too large, thus it may be difficult for the solver to reach a residual low enough.

Maybe you could try to put a higher residual threshold, but I'm not sure how it will affect the accuracy of the results.