CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Phoenics (http://www.cfd-online.com/Forums/phoenics/)
-   -   Convergence criteria (http://www.cfd-online.com/Forums/phoenics/51963-convergence-criteria.html)

Sileno October 29, 2005 18:11

Convergence criteria
 
I am using Flair 3.6 to simulate the wind flow through a perforated wall with several cavities in a simplified model. The domain size is 2.15 m(x), 3 m (y), 0.9 m(z) and the cells sizes vary from 1 cm at the cavities entrances to 4 cm at the domain extremities.

I am not achieving the properly convergence of the case. I have used positive values for the velocities and negative ones for the rest of the variables, without automatic convergence control, but the solution diverges and the error is big.

Are there any convergence criteria that I could use in that case?

Thanks


Mike December 6, 2005 08:25

Re: Convergence criteria
 
Set the inertial relaxation on the velocities to (small cell size in main flow direction/(no of cells in main flow direction*wind speed). Use linear relaxation of 0.3 on the turbulence variables. It is best to model the perforated wall in terms of the superficial velocity.


All times are GMT -4. The time now is 13:22.