polyDualMesh and volume calculation...
 

I have simulated the flow of a liquid that escapes from a container. I used the 'interFoam' solver. I generated the thetrahedral mesh with gmsh, then I converted it with the polyDualMesh utility. The so obtained mesh is funny to use : it increases the calculation speed and reduces the courant number. In my simulation, the liquid flows against a wall and a part of it passes above this wall, and finaly reaches an other container. I have to calculate the volume of liquid that passed above the wall. I thus use the 'integrate variables' filter of paraview in order to integrate the gamma-field in a definite region.

My problem is the following : the 'integrate variables' filter returns an absurd (and negative) volume !!! (but it works well on the initial thetrahedral mesh)... Please, do you know a methods (software), that can calculate the correct volume on this kind of mesh (onbtained with polyDualMesh) ???

Any answer is welcomed...
JB

Hi

Did your checkMesh report some strangeness on the poly mesh?

I ask because I've been in the same situation and checkMesh on the tet mesh is just fine but after running polyDualMesh the checkMesh reports with errors that cells have negative volume.

In a post I've read Hrv Jasak states that polyDualMesh only works for delaunay tet meshes. The errors mostly occurs in cells in areas with sharp corners.

read 1 (page 2 especially)
read 2

Hi Linnemann,
Your are right : polyDualMesh seems to work for Delaunay thetrahedral mesh. Actually, I use 'gmsh' to generate this kind of mesh (+ Netgen optimization). The 'checkMesh' utility does not return any errors about the cell volume in my case, but il crashes (seg fault) just after having said : 'Mesh is OK'. I think my problem does not come from the mesh itself, but from paraview. Indeed, (it seems that) OpenFoam works fine on this mesh. But the integration of the gamma field with paraview fails... I am very embarassed, because I did a lot of calculations on this kind of mesh (obtained with polyDualMesh), but I can't calculate the volume of fluid in a specific region... Do you think however all my calculations could be wrong because of bad cells generated by polyDualMesh ?
JB