[hotboy]

Hello friends!
When I change the under-relaxation factors ,the result is converged.Some people say that "we must change the under-relaxation factors to its defalt ,and let the result run for a time ".Is that right?If I change the under-relaxation factors ,I must change it to its defalut and run for a time to check if it was converged?Thank you very much!

[edoan]

URF do not have to be default. Small URF mean stable but slow solution. larger URF mean potentially unstable but fast solution. If you want to you can increase the URF as the solution progresses to get faster convergence once the solution has stepped over the initial unstable part.

[Kokemoor]

Convergence is often judged by the residuals, which are reduced by lower relaxation factors. A simulation with lower relaxation factors and lower residuals may not be as well converged as one with higher relaxation factors, even though the residuals are larger. The best solution is to judge convergence on the actual quantity you're trying to solve (e.g. pressure drop, drag force, etc.). If you can't do that, and you don't want to come up with new criteria for the residuals, then running some more iterations with the default relaxation factors will give you a better idea of how well converged it is.

[hotboy]

Originally Posted by edoan

URF do not have to be default. Small URF mean stable but slow solution. larger URF mean potentially unstable but fast solution. If you want to you can increase the URF as the solution progresses to get faster convergence once the solution has stepped over the initial unstable part.

Dear edoan !Thank you very much!

[hotboy]

Originally Posted by Kokemoor

Convergence is often judged by the residuals, which are reduced by lower relaxation factors. A simulation with lower relaxation factors and lower residuals may not be as well converged as one with higher relaxation factors, even though the residuals are larger. The best solution is to judge convergence on the actual quantity you're trying to solve (e.g. pressure drop, drag force, etc.). If you can't do that, and you don't want to come up with new criteria for the residuals, then running some more iterations with the default relaxation factors will give you a better idea of how well converged it is.

Dear Kokemoor!Thank you very much for your reply!