[MOAB-dev] Problems with remove_adjacencies

Iulian Grindeanu iulian at mcs.anl.gov
Fri May 4 11:40:21 CDT 2012


Hello, ----- Original Message -----
| Hi again,
| Thanks for your help. It's really appreciated you getting back to me
| so quickly. I think the major problem I am having is that the I have
| already created the faces and edges of every TET in the mesh before I
| start changing things. The analysis checks whether to split the mesh
| of not at every load step in my analysis so I require edge and face
| info in different places unfortunately. This is why I go about
| changing edges and faces as they are already there.
| I'll try and figure out what I am doing wrong as clearly I am missing
| something.
One question: Does the vtk file contain only tets? or it contains edges and faces too? If you started from tets only, and if you created edges and faces using moab get_adjacency methods, the edges and faces shared by elements are NOT duplicated. You can obtain the side and orientation of each face/edge within a tet, using"side" methods. Creating the edges and faces with "create_element" method would be an overkill, probably, as you have to carefully set the right connectivity yourself. But you can be more selective, in the sense that you maybe want to create only in specific areas (crack??) those edges and faces. But even if you want them created in a specific area, maybe you should use "moab::INTERSECT"; if you have 2 elements, one on each side of the crack, you can create just the face and edges between them (in the crack plane), with proper get_adjacency call If you already created all faces and edges, you may have to duplicate yourself. Again, the "set_connectivity" will adjust up-vertex adjacency, but it will not create new edges/faces. (a subsequent call to get_adjacency will, though, with "create" flag true) Iulian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/moab-dev/attachments/20120504/a121ceb8/attachment.htm>


More information about the moab-dev mailing list