XuliaDS

Thanks for the info and it actually makes a lot of sense... (vtk mesh and field altogether). If I understood correctly, I should do: ref = 4 mesh.PrintVTK(os.path.join(output_dir, "final_mesh.vtk"), ref)...

Hi! thank you very much @sshiraiwa and @v-dobrev. I can confirm that this approach works !! I have added x and the stress values to a VTK and both display...

Hi again, Extra comment / doubt: I have tried saving exactly the same mesh but the output vtk from mfem changes the original data. The image shows: left the original...

Cool. So the answer was: call the function as simple as possible :) Ok, this works for the mesh but then how do you match it with the solution ?...

yep, I think it would be very useful. OK, I am currently using the pyMFEM wrap so I should implement it i C++ and then add the python call, yes...

that sounds great. FYI: one thing that I noticed is that when you visualize the mesh after running mfem, the internal patches are not shown. Original mesh:  The mfem...

Hi ! Edit: H1_FECollection has default Gauss-Lobatto rules... which actually makes sense... But when I construct my GridFunctions, I don't seem to be able to evaluate at different quadrature points......

Last question on this: I have tried without success to run the displacement problem with a H1_collection space but then when outputting the solution, evaluate at internal nodes so when...

Thank you ! Yes, this helps and I can now evaluate at the new points. only that if I want to write to vtk, I need to update the mesh...