CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   convergence and residuals (

ebm February 2, 2009 18:08

convergence and residuals
I am currently running a model and I have a few questions about the convergence.

My residuals (RMS Vol and Mom) make a nice plot where after bouncing around for 20 iterations or so, they gradually decrease and then level off. The problem is that the value that they approach is only between 1e-4 and 1e-3 RMS. The documentation states that this is poor convergence and my values may be off. I have a fairly fine mesh size and my time scale factor is at 0.1, and they go 100 iterations at the same level.

When I look at the results, I see that my P-Vol-fluid_in domain imbalance is 49.9992 %. Is this large imbalance due to the poor convergence that I see? Or, could it be an error somewhere else that is causing the poor massflow, resulting in an inablility to get good convergence?

I have checked all my GGI connections (I am using frozen rotor with no pitch change), and everything appears to be set up normally. My turbulence model is SST and I am using the Automatic Wall Function.

If my 2 issues do not appear to be related, can anyone think of why my massflow would be off by 50%?

The application is a turbomachinery and I have a velocity profile inlet and static pressure outlet.

Thanks for you help.

ebm February 2, 2009 18:27

Re: convergence and residuals

My velocity at the inlet and outlet satisfy my calculated velocity of Q=AV (with the expected Q, not the incorrect value I got at the outlet). I think that velocity and massflow are calculated independently of each other, so there is something wacky with just my massflow that is throwing it off.

Glenn Horrocks February 2, 2009 18:55

Re: convergence and residuals

All times are GMT -4. The time now is 03:21.