I have been experimenting with
I have been experimenting with the Dynamic mesh refinement in parallel, and have hit a bit of a snag.
If a mesh is decomposed without any pre-existing refinements (all cells at refinement level 0), a case can be executed in parallel without any problems as no cells pairs between processors will be flagged for un-refinement. However, when I attempted to use reconstructPar to reform the data, I noticed that the following files were missing: boundaryProcAddressing cellProcAddressing faceProcAddressing pointProcAddressing Is there a method of forcing these files to be written? Or is a future item for the adaptive mesh refinement functionality? I appreciate any insight that you might provide. |
reconstructPar - reconstruct f
reconstructPar - reconstruct fields
reconstructParMesh - reconstruct mesh It does use a geometric merge so be careful of the results. |
Thank you Mattijs, This will w
Thank you Mattijs, This will work for what I am looking to do.
|
All times are GMT -4. The time now is 17:40. |