[michujo]

Dear all, I finally got my little CFD code to work (FVM, SIMPLE method, steadyflow, 2D incompressible code for the lid-driven cavity problem). However I'm experiencing some convergence issues.

I'm using upwind discretization for the convective terms and Gauss-Seidel + relaxation (SOR) solution method. I am having a hard time making the pressure correction to converge.

In each iteration of the SIMPLE method loop I update the velocity and pressure field like this:

u=ustar+u'
v=vstar+v'
p=pstar+alpha_p*p'

, where u and v are the velocity components, p is the pressure, ustar, vstar and pstar refer to the guessed velocity and pressure fields and alpha_p is a relaxation coefficient I use to avoid instabilities.

For a 20cell x 20cell grid and Reynolds number 100 I am obliged to use an extremely low value for alpha_p (alpha_p = 0.01).

Did anyone found the same problem? Do you think this is a normal value for the relaxation coefficient? If you computed the same problem, which value of alpha_p did you use?

Thank you in advance for your help.

Cheers,
Michujo.


P.S: The code seems to work well for coarse grids, but it seems impossible to me to compute larger grids, it does not converge.

[michujo]

Sorry, I just wanted to bring this up again.

Any hint from anyone?

Thanks.

Michujo.