PythonOCC as pre-processor

Perhaps you’ve taken note of the work we’ve put in such that PythonOCC is capable of FEM meshing. However, so far we were not able to add groups ( boundary conditions ) to the mesh, but yesterday this issue was resolved. Getting rid of a single character was all it took. However trivial the fix was, it has rather interesting consequences. As of now, geometry created with PythonOCC and advanced analyses can be coupled. That paves the way for performance based generative design, optimization of parametric designs. Sure, the analyses shown below is really trivial, but the point is that the mesh, with boundary conditions was produced from pythonocc. Which is meaningful, since this way one is able to produce lots of simulation without user interaction, which is what you do in optimization. In the /samples/SMESH/ you’ll find an example that sets the groups on the mesh. Take note that the groups are set after computing the mesh. If you happen to be so lucky to have Code_Aster installed, than you can solve a simple linear elasticity problem as found in the code_aster folder. The solve_lin_elas.sh script is there to load the code_aster enviroment, you’ll have to to edit the paths in this file and the .export file. The .comm file describes the commands for code_aster to run. If you don’t know code_aster, check out this flyer, which is really informative. Think of code_aster as an open-source equivalent of abaqus, capable of running coupled multiphysics problem, mesh refining and capable of processing in parallel. Serious stuff. Post-processing of the results can be done in Salome, Paraview or GMSH.