[pda]

Hey,

I am solving the incompressible, turbulent flat plate. The problem itself is of the steady-state type. As for now, I used the transient solver turbFoam. I think I could also use SimpleFoam, which is a steady-state solver.

I have a question/problem related to this. I want to run the case and make it stop if the changes in the state variables (p,U,k,etc) drop below a certain user-specified threshold. My runs with turbFoam stopped if an endtime was reached. Running it with simpleFoam needs a specific number of iterations. Is there a way in SimpleFoam to specify as stopping criterium a tolerance instead of a number of runs?

[Note that I am not speaking about the tolerances of the linear solver.]

A second question, should there be a difference in the (assume converged) solution if the same case is solved with turbFoam or simpleFoam? I don't know about the implementation, but aren't you using a physical time in a steady-state solver instead of the physical time? Hence, in the end, the time derivatives of the transient solver should vanish. I do believe, however that the steady-state solver will go faster to the converged solution. Is this correct?

Any help is greatly appreciated!

PJ

[olesen]

Did you check the message board for previous posts dealing with convergence criteria?
I know that I've already posted a possible solution to this several times. The last one within the past few weeks (again).