[cosimobianchini]

Hi all,
I would like to map results from a mesh to another one with the same geometry but different boundary partitions (the number and the names of the patches do not correspond). However I'm not interested in mapping boundaryFields but only the internalFields.
I was not able to make mapFields work without the -consistent option.
Is there a way for mapping only values on the internalMesh?
Thanks a lot
Cosimo

[sgr]

Hi Cosimo,

did you come up with an answer or a work around to your post? I have a case, where I refined certain regions using refineMesh. If I use mapFields to map field variables of the coarse Mesh on to the refined one, the internal Fields are not mapped. It is just the boundary conditions that are.

I dearly hope for a reply. Best regards,

Simon

[canopus]

The question in still unanswered!
Is there a way for mapping only values on the

internalField

and not boundary values?

[sgr]

I have not found a way and stopped looking, but am still interested in a solution, of course.

Regards.

[aboth]

Hello,

I think I found a way to do this. I know it is late for you, but for future reference I leave this here:

I use:

mapFields ../../014LuLaw188/baseCase/ -sourceTime 0.05 -mapMethod mapNearest -targetRegion region0

with a mapFieldsDict file with blank patchMap and cuttingPatches entries
E.g.:

cp ~/OpenFOAM/OpenFOAM-2.3.x/tutorials/lagrangian/DPMFoam/Goldschmidt/system/mapFieldsDict ./system/

The important part is the

-targetRegion region0

Apparently region0 is the default name for the internal region if you do not specify names in blockMeshDict. (I do not have info on the default name used by other mesh generators.)

For me the mapping created some nan entries, and some programs cannot deal with this, so I substituted them with 0 by executing the following line in the 0 directory:

find ./ -type f -readable -writable -exec sed -i "s/nan/0/g" {} \;

[Adriana Correia]

Hello everyone,

The solution of Ambrus worked well for me.

In my case, I needed to do a field reinitialization by keeping the internal field and changing the boundary conditions. I had the same geometry with the same mesh refinement but as my files were in binary (to keep the double-precision) I couldn't change the boundaries manually, so the workaround I found was:

1. create the new files (p and U for example) with the desired boundary conditions
2. map the internal field from a previous case according to Ambrus instructions

I hope it may be useful if someone faces the same problem as me.

Best regards.