Convergence problem with fluent
 

Hi every one.
I'm dealing with a Rayleigh-Bénard problem on Fluent. The problem is that the solution refuse to converge even after thousands of iterations here the screen capture of the residuals:
http://i1159.photobucket.com/albums/...Untitled-1.jpg
thank for all :)

The convergence criteria is problem specific. It seems that your solution has already been converged.

Thanks for reply.
Actually I'm dealing with a translational periodic flow with fixed pressure gradient. If you have any tips to judge the accuracy of convergence I will be so grateful.

I really do not remember what my college used to judge convergence when he simulated a similar problem. I think the convergence of mass flow rate might be a good indicator.

Try creating several monitors of properties (pressure/velocity/temperature etc) that are important to your physics. Monitors can be points/surface average monitors etc. If the problem is steady (as it appears) the monitors should be flat.

You can also monitor volume averaged mass imbalance, it should small.

OJ