Known bugs
-
Solver divergence
- the convergence might be impeded, and the solver can even diverge, in cases where strong shocks appear in the flow. Try to coarsen (locally) the mesh.
- in cases where the Mach number and the lift coefficient is high, a strong vortex flow is generated at the trailing edge of the wingtip. This can lead to local unphysical shocks occurring on one or two cells. Coarsening (locally) the mesh usually works.
-
Non deterministic meshes
Since the meshes are unstructured, they can be slightly different each time gmsh is run. This can slightly affects the results. Particularly, the grid from the test "tbox/meshDeformation3.py" can even contain elements with negative jacobian.