CFD Online Discussion Forums - Looping over all cells in parallel

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- OpenFOAM Programming & Development (https://www.cfd-online.com/Forums/openfoam-programming-development/)

- - Looping over all cells in parallel (https://www.cfd-online.com/Forums/openfoam-programming-development/88280-looping-over-all-cells-parallel.html)

Looping over all cells in parallel

Hi Foamers,

The following code works good on a single processor but not on multiple processors...

forAll( mesh.C(), celli)
{
.
.
.
}

How can I make this work on multiple processors?

Thanks
Senthil

Quote:

Originally Posted by skabilan (Post 307402)

Hi Foamers,

The following code works good on a single processor but not on multiple processors...

forAll( mesh.C(), celli)
{
.
.
.
}

How can I make this work on multiple processors?

This should work fine on every processor. Where is the issue?

Quote:

Originally Posted by olesen (Post 307462)

This should work fine on every processor. Where is the issue?

I guess his problem is what's in the loop. The answer to his problem will probably contain "reduce" or "correctBoundaryConditions" ;)

Thanks for the response...Actually, I figured out what was missing...I had to use a "reduce" call to sum over all the processors and not just one processor. Like

reduce(U_flux, sumOp<scalar>());

Senthil

Hello everyone.

I have a problem with forAll loop in parallel too. There is a piece of my code, and i don't know how to extend it for parallel calculations. It calculates mass flux deltaM through cell faces.

Code:



scalarField& deltaMi = deltaM.internalField();    

deltaMi = 0 ;



    forAll(owner, faceI)

    {

        

        label P = owner[faceI] ;

        label N = neighbour[faceI] ;



        dir = (Ui[P]+ Ui[N]) & mesh.Sf()[faceI] ;



        if (dir >= 0)

        {

            deltaMi[faceI] = (0.5*rho[P]*(deltaT.value())*(U[P] + U[N])) & mesh.Sf()[faceI] ;

        }

        else

        {

              deltaMi[faceI] = (0.5*rho[N]*(deltaT.value())*(U[P] + U[N])) & mesh.Sf()[faceI] ;

         }

    

        

        if (mag(deltaMi[faceI]) < 1e-20) 

            

            deltaMi[faceI] = 0 ;



    } ;



    forAll(mesh.boundary(),patchI)

    {

        fvsPatchScalarField& pfdeltaM = deltaM.boundaryField()[patchI] ;

        const fvPatchVectorField& pfU = U.boundaryField()[patchI] ;

        const fvPatchScalarField& pfp = p.boundaryField()[patchI] ;

        const fvPatchScalarField& pfrho = rho.boundaryField()[patchI] ;

        const unallocLabelList& faceCells = mesh.boundary()[patchI].faceCells();

        const vectorField& pSf = mesh.Sf().boundaryField()[patchI];

        

        forAll(mesh.boundary()[patchI], faceI)

        {



        dir = (pfU[faceI] + U[faceCells[faceI]]) & pSf[faceI] ;



        if (dir >= 0)

        {

            pfdeltaM[faceI] = (rho[faceCells[faceI]]*deltaT.value()*(pfU[faceI])) & pSf[faceI] ;

        }

        else

        {

            pfdeltaM[faceI] = (pfrho[faceI]*deltaT.value()*(pfU[faceI])) & pSf[faceI] ;

        }

        } ;



    };

Any tips ?