[Chander]

I am simulating turbulent flow with conjugate heat transfer and I have specified the following convergence criteria:

RMS residual for mass, momentum and energy equations as 1E-6. AND Conservation target as 0.001%

Now I am facing trouble with getting the required convergence. Please refer the attached residual plots. The residual for mass drops below 1e-6. But residual for momentum stabilizes around 1e-5 and that for energy around 5e-6. Regarding the conservation target, the maximum domain imbalance is around 0.006% for H-energy for the fluid domain.

Can this solution be regarded as converged. The question may sound stupid as clearly the specified convergence criteria has not been met. But I had no theoretical basis for specifying 1e-6 as the convergence criteria in the first place. Had i specified 1e-5 as the convergence criteria for equations with no criteria for conservation target, I would not have even observed this behavior.

[siw]

It might a good idea to show the same plots again but when it is solved an a finer mesh, apply a gloabl scaling factor to reduce element size, to see if the convergence changes.

[stumpy]

Looks like good convergence to me! Check the T-Energy imbalance in the solid just to be safe, but I expect it's fine.

[ghorrocks]

This question has been asked hundreds of times.... So I wrote an FAQ:

http://www.cfd-online.com/Wiki/Ansys...gence_criteria