Group CGNS boundary sections into markers using the ZoneBC of the file - #2976

Proposed Changes

Meshers such as Pointwise often split one boundary into several CGNS sections. For example, a wing with triangles and quadrilaterals becomes tri_wing and quad_wing. The CGNS reader made one marker per section, so the config file had to list both names and set the same boundary condition twice.

With this PR, the reader uses the ZoneBC of the file:

• Each boundary section gets a marker name from the BC that covers it:
  • If the BC has a family, the marker name is the family name.
  • If the BC has no family, the marker name is the BC name.
• Sections with the same marker name become one marker.
• Sections that no BC covers, and files without a ZoneBC, work as before: one marker per section, named after the section.
• A section that is only partly covered by BCs, or covered by BCs with different names, also keeps its own name. A section cannot be split into two markers, so the reader prints a warning instead of guessing.
• BCs given by elements are used: FaceCenter/EdgeCenter/CellCenter, or ElementRange/ElementList in older files. BCs given by points (Vertex) cannot be matched to a section and are ignored.
• The reader prints which sections it grouped, e.g. Marker back groups the sections quad_back tri_back.
• Only the finite volume reader groups sections. The FEM/DG reader still makes one marker per section.

So the grouping follows the names written by the mesher. The reader does not guess from similar names. For two BCs lower_wall and lower_WT:

Meshers such as Pointwise usually give each BC its own family with the same name. So two BCs in the mesher stay two markers in SU2, unless the user puts them in the same family.

Test case

New regression test cgns_reader_sections (serial and parallel), TestCases/cgns_reader/channel_sections.cfg. The mesh is a small 3D channel with 96 cells, made of hexahedra and prisms. It covers each case:

• With develop, the case stops with: The configuration file doesn't have any definition for marker lower_1.
• With this PR, it runs.

I also checked a real Pointwise mesh of a wing (1.1M cells, 4 ranks):

• develop stops on tri_back.
• This PR gives 7 markers. For example, back has 4209 elements, which is tri_back (1002) + quad_back (3207).

The regression tests that already read CGNS meshes give the same results as with develop: wedge, bend, cylinder_split, MFR_coupling, flatplate_supersonic, flamelet, and the CGNS output/readback tests.

Related Work

• Together with Add the Paraview parallel output format (.pvtu with one piece per rank) #2977, this PR and that one are the last PRs of the input/output improvements that started with Fix integer overflows and parallelize the output writers for large meshes #2924.
• The test mesh is in Mesh with boundaries split in sections for the CGNS reader test TestCases#212.
• Temporary: .github/workflows/regression.yml uses the cgnsReaderMarkers branch of TestCases until Mesh with boundaries split in sections for the CGNS reader test TestCases#212 is merged. I will set it back to develop before merging.

PR Checklist

• I am submitting my contribution to the develop branch.
• My contribution generates no new compiler warnings (try with --warnlevel=3 when using meson).
• My contribution is commented and consistent with SU2 style (https://su2code.github.io/docs_v7/Style-Guide/).
• I used the pre-commit hook to prevent dirty commits and used pre-commit run --all to format old commits.
• I have added a test case that demonstrates my contribution, if necessary.
• I have updated appropriate documentation (Tutorials, Docs Page, config_template.cpp), if necessary. (No new option.)

🤖 Generated with Claude Code

https://claude.ai/code/session_01L7oxzVir4zENrpy3K3W6Fm