|September 16, 2013, 06:35||
decomposePar -region (OF 2.1) without splitting regions in separate parts
Join Date: Oct 2010
Location: Delft, Netherlands
Posts: 97Rep Power: 7
I am running a simulation with a multiregion solver (a modification of chtMultiRegionFoam) and want to run some of my calculations in parallel.
I know I can use decomposePar -region regionName to decompose the different regions separately, but this gives quiet an illogical mesh division because 1 region is almost surrounded by another region but gets split over all processors regardless. (See the image of the full mesh with the 3 regions coloured and a snapshot of what part is set to processor0).
My question is: can anybody tell me how I can make sure that the decomposition is done based on position in the mesh, not based on position in the region? For this test I just used the simple decomposition method.
|November 15, 2013, 06:07||
Join Date: Jun 2013
Posts: 10Rep Power: 4
I'm just facing the same problem in these days...
Have you found the solution?
Thanks in advance,
|Thread||Thread Starter||Forum||Replies||Last Post|
|Multi region meshing & recovering the original patch names||fluidpath||OpenFOAM Native Meshers: snappyHexMesh and Others||4||May 19, 2013 19:13|
|Using starToFoam||clo||OpenFOAM Other Meshers: ICEM, Star, Ansys, Pointwise, GridPro, Ansa, ...||33||September 26, 2012 04:04|
|StarToFoam error||Kart||OpenFOAM Meshing & Mesh Conversion||1||February 4, 2010 05:38|
|Trimmed cell and embedded refinement mesh conversion issues||michele||OpenFOAM Other Meshers: ICEM, Star, Ansys, Pointwise, GridPro, Ansa, ...||2||July 15, 2005 04:15|
|Import gmsh msh to Foam||adorean||Open Source Meshers: Gmsh, Netgen, CGNS, ...||24||April 27, 2005 08:19|