CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Convergence problems (http://www.cfd-online.com/Forums/fluent/29705-convergence-problems.html)

Emilien May 2, 2002 12:09

Convergence problems
 
I am a beginner with FLUENT

I have problems of convergence. I tried to decrease the under-relaxation factors, and also to modify the grid (e.g. more cells), but I still have convergence problems. The plots stay constant (at about 500 iterations) after having decreased. What can I do to make the solution converge Can I consider the solution as accurate when the plots stay constant?

Thanks for your help

Peter May 2, 2002 14:14

Re: Convergence problems
 
You can say that convergence has been achieved when residuals are fallen several orders or magnitude and they level off. You should check the mass flow in the interior of your domain and check that it has reached a constant value. Don't forget to check that your solution is consistent with the near wall treatment. Check values or ystar or yplus.

chiseung May 2, 2002 21:39

Re: Convergence problems
 
1. Check if boundary conditions are set properly with a physical point of view. There are many divergence cases because of improper b.c. setting.

2. Besides, check if physical models are applied properly.

3. And then, reduce your relaxation factors gradually with considering number of cell.

4. Convergence is totally problem dependent,I believe, so if you want to get a good advice explain your system little bit specificlly. If not, there would be just abstract answers.


Emilien May 3, 2002 08:43

Re: Convergence problems
 
Thanks for your answers.

My problem is about a solid part moving in translation in a fluid (a resin). I used the moving mesh model.

(It is only a first problem I try to model, to get used to FLUENT. Because actually, my problem is a blade moving atop of a fluid (resin). So I will also need to use the multiphase model.)

But for the moment, even only a moving part is problematic. My problem is modelled by a rectangular part in a rectangular resin vat(2D). My B.C. are a pressure inlet of 101325 Pa at the top and an outflow at the right.


All times are GMT -4. The time now is 06:19.