[titio]

Hi Foamers,

One of the limitations of OpenFoam is the lack of a suitable convergence control, to avoid doing calculations after the simulation has already converged.

Does anyone know of some code, or form to implement convergence control in OpenFoam, based on pre determined criteria defined by the user?

Regards,

Titio

[camoesas]

Is there really no answer to this important question?

[chegdan]

There is an answer and it exists in simpleFoam. Specifically, look at the headers initConvergenceCheck.H and convergenceCheck.H int eh simpleFoam solver directory. They can be used in conjunction with the word "convergence" followed by your desired level of convergence (eg. 1e-6 in your SIMPLE subdictionary in fvSolution. There are a few threads on the forum about this. Hope this helps.

Dan

[camoesas]

HI Daniel,

Thanks for the answer, as you said I've found some threads on this topic

Camoesas

[chegdan]

Also, The convergence control has changed for 2.0.x version of OpenFOAM to something like:

SIMPLE
{
    nNonOrthogonalCorrectors 2;
    convergence              1.0e-6;//old way of specifying convergence criteria
  
    residualControl//OF 2.0.x
    {
        p               1e-5;
        U               1e-5;
        "(k|epsilon)" 1e-5;
    }
}