Re: [deal.II] problem with CylindricalManifold<3> after removing cells from a triangulation

2021-04-26 Thread Luca Heltai
The issue is with the function that removes the cells. It creates the grid correctly but forgets to set the ids of the edges. It is not with your code. For boundary ids this is usually not an issue, since interpolation and boundary conditions are set using the face ids (not the edges), but for

Re: [deal.II] problem with CylindricalManifold<3> after removing cells from a triangulation

2021-04-26 Thread Simon
Hello Luca, your suggestion fixed my problem, so thank you very much! Just to make sure that I understand it correctly: In my code snippet from above I call the member function "triangulation.set_all_manifold_ids_on_boundary(1,1). I actually thought that this call does the same thing as you

Re: [deal.II] problem with CylindricalManifold<3> after removing cells from a triangulation

2021-04-26 Thread luca.heltai
There may be a problem with the way boundary ids are set. Can you try the following? after creating the grid with removed cells, loop over all cells and all faces, and if at boundary with boundary id == 1, then call cell->face(f)->set_all_manifold_ids(1); notice the “_all_”, i.e., it

[deal.II] problem with CylindricalManifold<3> after removing cells from a triangulation

2021-04-23 Thread Simon
Dear all, I created a mesh using GridGenerator::hyper_cube_with_cylindrical_hole and removed after that half of the total number of the cells in order to make use of symmetry. Then I set a SphericalManifold<2> (2D) respectively CylindricalManifold<3> (3D) in order to get a true circle (2D)