Can I import a mesh into pymapdl?

[mkofler96]

Hello,

I would like to automize the mesh-generation process, by using custom tools such as meshio (https://github.com/nschloe/meshio) for example.
Is it somehow possible to import a mesh, that could be for example a .vtk, .inp, or .msh file? So to speak reverse the grid = mapdl.mesh.grid functionality?

[mkofler96]

OK problem solved, I found the function save as archive: from ansys.mapdl.reader import save_as_archive
https://mapdl.docs.pyansys.com/version/stable/examples/extended_examples/ex_01-gmsh_example/ex_01-gmsh_example.html

[germa89]

Hi @mikerife

We should not deprecate PyMAPDL-Reader without providing alternatives to their main features.

[mkofler96]

Hi @mikerife,
I stumbled across another problem, which is that I'd like to import a mesh containing lines, which in Ansys should then be treated as beams. Unfortunately, when I want to export my mesh from meshio to the Ansys .msh format, I get the error

KeyError: "Illegal ANSYS cell type 'line'. (legal: triangle, tetra, quad, hexahedron, pyramid, wedge)"

If I use another mesh format (e.g. .vtk), then I get a similar error when importing the mesh

save_as_archive("my_mesh.vtk", mesh)

results in

RuntimeError: `grid` contains no allowable cell types. Contains types [3] and only [<CellType.TRIANGLE: 5>, <CellType.QUAD: 9>, <CellType.TETRA: 10>, <CellType.VOXEL: 11>, <CellType.HEXAHEDRON: 12>, <CellType.WEDGE: 13>, <CellType.PYRAMID: 14>, <CellType.QUADRATIC_TETRA: 24>, <CellType.QUADRATIC_HEXAHEDRON: 25>, <CellType.QUADRATIC_WEDGE: 26>, <CellType.QUADRATIC_PYRAMID: 27>] are allowed.

Do you know by chance if there is any possibility of importing a mesh consisting of lines into PyMAPDL?

Thank you,
Michael

[mikerife]

Hi @mkofler96 What are you using to generate the mesh? The 'Ansys msh' file type supported by meshio does not include link/link/beam elements. Pretty sure that is meant for volumetric CFD models. I see that almost 10 years ago the Mechnical APDL cdb format was requested to be supported by meshio.

Maybe @germa89 can comment on vtk to cdb. My guess is that vtk may not have a data structure for beam cross-sections and orientation points.

Mike

[mkofler96]

Hi Mike, thanks for the reply. I am using the tool SplinePy in combination with the visualization/export tool gustaf that uses meshio as the export engine. You are right, the Ansys msh file is probably intended for fluid simulations and therefore there are no 1D elements needed. Also the cross-sections and orientations are not saved in the vtk format. However, I managed to import the geometry into Abaqus, where I then need to assign cross-sections and the orientation.

One way would be to export the geometry only (without mesh) to IGES and then import that into Ansys. But also there I struggle with using the igesin functionality. It seems that 1D elements are causing troubles. Howver, I'd still prefer using the mesh over the geometry, as I have more control in the custom tools I mentioned (splinepy and gustaf).

Thanks anyways,
Michael

[germa89]

Maybe @germa89 can comment on vtk to cdb. My guess is that vtk may not have a data structure for beam cross-sections and orientation points.

Yeah... I dont think so either. That is condensed stuff.. I dont think there is a standard for this.

I dont think there is an easy solution here... you will always lose the section information. I'm sorry for this.