Normal DIC?

Oh I think you are right

I can't really comment on variants, but if this is only happening for one particular model/family of models, then maybe it doesn't capture the relevant dynamics well without the stability of 1-o upwind.

I will change the preconditioner to

DIC

I have read also somewhere that the blow up could be solved by tightening the relative tolerance for p equation

I will run the simulation now

using that

Okay.. let us know if you manage to fix it. :)

Any comments about the schemes?

Your p residuals are way lower compared to tolerance, so this change won't bring about any change to your strategy. The only currently active criterion will be minIter 2.

Is it okay to keep the inner loop tolerances so high?

could you please elaborate

What should I change?

I typically keep them very low..1e-10 or so at max, while my outer residuals are something comparable to 1e-6

But I would like feedback on this..From you or others.

What do you mean "keep inner loop tolerances so high"?

here is the PIMPLE settings

Yeah these are the outer residuals.. by inner ones I mean the ones you have specified for the algebraic solvers.

Do I need to increase

minIter

?

I would rather decrease residuals, set them to a low number.. maybe try 1e-10. How long does it take before you find out if your solution will crash or not?

Also, you could decrease the relTol by a couple of orders, 0.01 maybe. Be warned that this would obviously increase a few iterations.

Just to make sure I am getting your right

Right

nice