[ec91]

Hi all,

I'm launching simulations with a solver that i developed for a deforming elastic body.

I calculate the displacements with a displacementEqn, then i move the mesh (similarly to the solid mechanics solvers in foam-extend).

Then, i calculate the density of the deformed particles as follows:

forAll(rho, cellI)
{
    rho[cellI] = mass[cellI] / (mesh.V())[cellI];
}

The solver works fine in serial, but when i run it in parallel i see a strange behavior at the processor boundaries. When i look at the density, I see darker lines in which the density is higher. These lines correspond to the processor domain boundaries. I have the same problem for the displacements, even if it's less evident.