CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Phoenics (http://www.cfd-online.com/Forums/phoenics/)
-   -   Relaxation and convergence (http://www.cfd-online.com/Forums/phoenics/52245-relaxation-convergence.html)

 sammi March 20, 2008 04:32

Relaxation and convergence

Dear Sir,

It's me again. I read some posts said that the speed of convergence depends much on relaxation setting. And which relaxation setting should be used depends on experience.

I know the following questions sounds stupid, but I need help...

1)What is the actual meaning of linear relaxation and false time step? What is the different between them? (I only known a negative value means linear relaxation, but I don't actually understand what the meaning of them is.)

2)What is the effect of tighter and looser the relaxation, when should I make it tightr/looser?

3)I found an equation calculating relaxation: (smallest cell distance in main direction of flow)/(fastest velocity) So, if I ran a model using 16 wind directions, does it means that all 16 directions should have different relaxation? If I ran NE direction wind, how should I determain the cell distance. If the smallest cell distance is at 2.5m while the fastest velocity is at 596m, is the different of height improtant?

Regards, Sammi

 All times are GMT -4. The time now is 19:55.