[mnajafi]

Hi

1.As I understand, p.correctBoundaryConditions(), updates boundary values after explicit update of p. But p is a volScalarField and it's not defined on the boundaries. So what would be the benefit of this function.

2. In the other hand when I try this I get error.

surfaceScalarField P=fvc::interpolate(p);
P.correctBoundaryConditions();

How should I make the above code work?

Thanks

[mnajafi]

1.I found out that it difference is revealed when I call interpolate as code below:

     forAll(p,pi)
    {
        p[pi]=1;
        i++;
    }
    
    Info<<endl<<"p beofre correct:"<<p<<endl;
    Info<<endl<<fvc::interpolate(p)<<endl;
    p.correctBoundaryConditions();
    Info<<endl<<"p after correct:"<<p<<endl;
    Info<<endl<<fvc::interpolate(p)<<endl;

2. Apparently functions that are called by correctBoundaryConditions() (like evaluate) are not defined for fvsPatchField and so this function can't be applied for a surfaceScalarField.

[gentodin]

If we cant apply correctBoundaryConditions(); on surfaceScalarField, how do we apply it to correct boundary conditions for surfaceScarlarField at every timestep due to the update of certain variables during parallel simulation?

[sreehahaha]

Originally Posted by gentodin

If we cant apply correctBoundaryConditions(); on surfaceScalarField, how do we apply it to correct boundary conditions for surfaceScarlarField at every timestep due to the update of certain variables during parallel simulation?

Hi gentodin,

Did you find answer to your questions about correcting boundary conditions of a surfaceScalarField using correctBoundaryConditions()? Please let me know. I've been struggling with a similar issue.

Thanks,
S

[joshwilliams]

This is fundamentally wrong. In 0/p, you define the internal field and boundary values of p. The boundary values can be found by something like "p.boundaryField()[patchName]".

Originally Posted by mnajafi

But p is a volScalarField and it's not defined on the boundaries.