[ganeshv]

hi !

I need to perform a mapping from a fine mesh to a coarse mesh every few time steps. The coarse mesh can be a sub-domain of the full fine mesh. This needs to run in parallel. My search for such a utility took me here

http://www.cfd-online.com/Forums/ope...time-step.html

But it's not very conclusive. I thought I could modify the source code of mapFields to do something in this regard. The idea is to read in every processor sub-mesh of the coarse mesh on all the processors. Initiate the desired fields to zero. Map the fields from the fine mesh to the coarse mesh and sum them all up on one processor and write it out. This would essentially become some sort of parallel mapFields utility. I can't get all processors to read in a processor mesh like so

        Time runTimeCoarseMesh
          (
           Time::controlDictName,
           word("."),
           word(".")/fileName(word("processor")+name(procI))
           );

        fvMesh coarseMesh
          (
           IOobject
           (
            "coarseSolution",
            runTimeCoarseMesh.timeName(),
            runTimeCoarseMesh,
            IOobject::MUST_READ
            )
           );