You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
21 of the 24 configs under SU2/TestCases/hom_euler/ cannot be run because their mesh files are absent from the TestCases repo.
TestCases/hom_euler/ contains only NACA0012_5thOrder, while SU2/TestCases/hom_euler/ has configs for Ringleb (7 variants), SubsonicChannel (3), InviscidVortex (3), Sphere_4thOrder_Hexa, Sphere_4thOrder_Tet, and NACA0012_3D_Hybrid_4thOrder. Note that hom_navierstokes is unaffected — its meshes are present in the TestCases repo as expected.
To reproduce:
git clone https://github.com/su2code/SU2.git
git clone https://github.com/su2code/TestCases.git
cd SU2/TestCases/hom_euler/Ringleb/Quad100X50_nPoly3
SU2_CFD Ringleb.cfg
gives:
The SU2 mesh file named Quad100X50_nPoly3.su2 was not found.
The same happens for all Ringleb, SubsonicChannel, Sphere and NACA0012_3D_Hybrid configs. Note these configs also use the deprecated REF_LENGTH_MOMENT, which needs replacing with REF_LENGTH before the mesh error is reached. The three InviscidVortex configs fail earlier with ITER must not be used when running multizone and/or unsteady problems, so they were not tested for the mesh issue.
I found this while working through configs that still use the deprecated EXT_ITER option (#2866). I left the hom_euler cases out of that PR because I can't verify them without the meshes.
Are these meshes available somewhere, or should these cases be removed?
Activity
bigfooted commentedon Aug 26, 2026
@vdweide do you still have the meshes?
vdweide commentedon Aug 27, 2026
Looks like I do. Where should I upload them?
bigfooted commentedon Aug 27, 2026
In the Testcases repository, in the hom_euler/ subdirectory, and then use the same directory structure as in the existing SU2 repository.