[avk1985]

I am using dynamicMultiMotionSolverFvMesh with coded solvers to which are in turn based on modified versions of inkjet mesh motion class. I am experiencing errors mentioned below while trying to solve the case in parallel.

I viewed processor* meshes in paraview and confirmed that processor boundary mesh is matching but when I viewed the procBoundary0to1_faces.obj and procBoundary1to0_faces.obj only one of the points of procBoundary1to0_faces.obj seems to be off

I have added snap of overlay of all 4 surfaces i.e. boundaries from processor* and obj files

I wanted to have some inputs on how can I conclude if mesh is OK or not ?

Any thoughts on this ?

Errors:

[0] processorPolyPatch::calcGeometry : Writing my 100 faces to OBJ file "/home/amodk/processor0/procBoundary0to1_faces.obj"
[1] processorPolyPatch::calcGeometry : Writing my 100 faces to OBJ file "/home/amodk/processor1/procBoundary1to0_faces.obj"
[1] processorPolyPatch::calcGeometry : Dumping cell centre lines between corresponding face centres to OBJ file"/home/amodk/processor1/procBoundary1to0_faceCentresConnections.obj"
[0] processorPolyPatch::calcGeometry : Dumping cell centre lines between corresponding face centres to OBJ file"/home/amodk/processor0/procBoundary0to1_faceCentresConnections.obj"
[1]
[1]
[1] --> FOAM FATAL ERROR:
[1] face 0 area does not match neighbour by 183.09336% -- possible face ordering problem.
patchrocBoundary1to0 my area:0.00021953419 neighbour area:9.6884642e-06 matching tolerance:1.3598406e-07
Mesh face:497840 vertices:4((4.6616869e-310 4.6616895e-310 0) (-0.039019907 0.03478563 0.01779234) (-0.039870005 0.035692938 0.01779234) (-0.039870005 0.035692938 0.01))
If you are certain your matching is correct you can increase the 'matchTolerance' setting in the patch dictionary in the boundary file.
Rerun with processor debug flag set for more information.

[avk1985]

I think I resolved anyways thanks

[Mparuz]

Originally Posted by avk1985

I think I resolved anyways thanks

Hi avk1985,
how did you manage the problem? I have the same issue for a 2D case and I suspect it's due to the empty type faces assigned to "front" and "back" patches, but I don't now how to face that.


Thanks

[avk1985]

It was problem with coding.
I had to use following to access point0 co-ordinates and returned tnewPoints.
const pointField& p0 = points0();
tmp<pointField> tnewPoints(new pointField(p0));
pointField& newPoints = tnewPoints.ref();