[wanggl]

Hello,

I am running buoyant driven flow using Openfoam. and find that even using relax_U=0.3 and relax_p=0.7, it still converged very slowly.
Anyone know how to speed up the convergence of buoyant driven flow?

Thanks in advance.

[Taataa]

It very much depends on the problem and how you setup a caselike schemes, solutions and so on. There are some default values though that could be a starting point:
for transients it's always 1 -> ".*" 1;
for non-transients: 0.3 for pressure and 0.7 for the rest.

Again, they are starting values and you should play with them. Ideally, you want set them as close as possible to 1.

[wanggl]

Originally Posted by Taataa

It very much depends on the problem and how you setup a caselike schemes, solutions and so on. There are some default values though that could be a starting point:
for transients it's always 1 -> ".*" 1;
for non-transients: 0.3 for pressure and 0.7 for the rest.

Again, they are starting values and you should play with them. Ideally, you want set them as close as possible to 1.

Thanks for your reply.
I use non-transient method to run the case, and see p=0.7 u=0.3 really the best choice. But the final residual (after 5000 iterations) of Uz is around 1E-3, while it is 5E-7 for TKE.
It is hard to say converged for Uz.

[Taataa]

Relaxation is only one of many parameters, you should consider changing linear solvers, tolerances and more importantly schemes. Since you didn't provide any details about the way you setup the case, it's not possible to provide more info.