This post is about the increase of capabilities of salome meshing plunging following another feature request for MESQUITE, Mesh Quality Improvements Toolkit ([Feature request] integration of MESQUITE, Mesh Quality Improvement Toolkit (LGPL))
I just had contact with another meshing commercial software tool, coreform-cubit, and reading its documentation, most of the hexahedral meshing is done by a ‘simple’ quad meshing of a surface and a sweep of this face, for different blocks. This made me compare salome with its capabilities to cubit, which it does not have anything to be ashame of, in contrary. nevertheless, this issue (Quasi-Stuctured Quad conformal with 1D submeshes) still reminds. In salome today this is somewhat ‘achievable’ only for one body (and not completely sure that it does not give any error, also pretty sure it is only feasible for a body with equal polygonal faces), by creating using a 2D quasi structured quad mesh inside gmsh, and then extruding it.
The feature request would be: to be able to combine this algorithm with 1D meshing algorithms, such as 3D gmsh algorithms were made possible in last release (9.12) to be compatible with 2D/1D algorithms. this would enable the creation of more easily unstructured hexahedral meshes of more complex geometries putting salome to the same level of it. for reference of what I mean about the 2D quad+sweep one can refer to Tutorials – CUBIT (cubit200.zip which contains a presentation in this feature)