[JNSN]

Hi Foamers,

I have experienced a strange behaviour, when runnig a case with a partly rotating mesh in parallel. The Solver is transientSimpleDyMFoam, the mesh is partly rotated with the mixerGgiFvMesh.
When I run in parallel the dynamicFVMesh finds only the cells of the rotating zone of one processor and the simulation blows up. If I compute one iteration in serial, and then do a decomposition, everthing works fine. I think, this is because then in constant/polyMesh there is already a cellZone called "movingCell", and in the processor*/constant/polyMesh the cellZones are empty. The dynamic Mesh gives then the output "void mixerGgiFvMesh::addZonesAndModifiers() : Zones and modifiers already present. Skipping."
The simulation is runnig now, but running first in serial is a quite ugly workaround. Does anyone have a smarter solution?

Thanks,
Jan