Very interesting. :-) As George suggests, it should be possible to identify tetrahedral or hexahedral elements in a post-hoc fashion by looking at whether the faces bound such an element appropriately. However, it might be faster to tag this information at a lower level (i.e. inside the C code) as elements are created.
Do you have any ideas about supporting isoparametric elements with non-linear shape functions? Could there perhaps be a way to use subdivision surfaces to compute nodal positions? It seems that other linear element types, like beams and shells, are already quite well-supported by the existing mesh structures, but I'm uncertain about how non-linear elements could be represented / edited conveniently. Current open source meshing tools (like Salome-Meca and Gmsh) build their meshes "top-down" from higher-order surfaces. For export purposes, it would be very useful to tag groups of nodes, faces and elements for later application of material properties, boundary conditions and forces. Finally, it might be useful to see if there's any interest in using FEA within Blender itself for simulation purposes. Having some kind of volumetric mesh element might be the first step in allowing people to explore simulation-oriented FEA. A more formal approach to FEA seems gradually to be making its way into simulation pipelines. One example is Weta's "Digital Tissue" system: http://www.youtube.com/watch?v=7VlthWa5pu8 Jonathan Merritt. PS - Disclaimer: I lecture Finite Element Analysis in Mechanical Engineering at The University of Melbourne, so I have a vested interest. :-) _______________________________________________ Bf-committers mailing list [email protected] http://lists.blender.org/mailman/listinfo/bf-committers
