[mheinz]

Hello together,

I am currently working with the wallDist functionality to compute the distance of each cell to its nearest wall. After that I want to cut the data regarding some criterions. Everything works fine, when I calculate on one processor. However, when performing a parallel run, I need to access the patches that connect the seperated meshes of each processor. That's where I need help.
Here is some of the code, that needs fixing:

cellDist = wallDist(mesh).y();

const fvPatchList& patches = mesh.boundary();
forAll (patches, iPatch)
{
    const fvPatch& singlePatch = patches[iPatch];
    forAll (singlePatch, iFace)
    {
        label iFaceCell = singlePatch.faceCell()[iFace];
        cellDist[iFaceCell] = 0.0;
    }
}

This way, I tried to access all boundary patches (mesh.boundary()), but the "processor patches" are not recognized as such.

I'm thankful for any advice.

Michael

[chriss85]

You can combine lists from separate processors using the following function. This can be used for face positions, areas, field values etc..

template<class Type>
    void combineFields(Field<Type>& field)
    {
        List<Field<Type> > allValues(Pstream::nProcs());

        allValues[Pstream::myProcNo()] = field;

        Pstream::gatherList(allValues);
        Pstream::scatterList(allValues);

        field =
        ListListOps::combine<Field<Type> >
        (
            allValues,
            accessOp<Field<Type> >()
        );
    }