[xiuying]

Hi,

I need compute wallshearstress and wrote a code as followings. when I use single processor, it worked well. But if I use openmpi and multiprocessor, it couldn't work and there was not wallshearstress data written. Could you inform me how to deal with the problem? Thanks.

volSymmTensorField R(turbulence->R());

volVectorField wallShearStress
(
IOobject
(
"wallShearStress",
runTime.timeName(),
mesh,
IOobject::NO_READ,
IOobject::AUTO_WRITE
),
mesh,
dimensionedVector("wallShearStress", R.dimensions(), vector::zero)
);

//label patchA = mesh.boundaryMesh().findPatchID("lowerWall");

//patchID=0;
// Info<< "End2\n" << endl;


forAll(wallShearStress.boundaryField(), patchi)
{
wallShearStress.boundaryField()[3] =
(
-mesh.Sf().boundaryField()[3]
/mesh.magSf().boundaryField()[3]
) & R.boundaryField()[3];
}

if(runTime.outputTime())
{

wallShearStress.write();
}

kang