Dear all,
I have a problem of a wave traveling through a quarter section of a cylinder. I need to enforce symmetry with y and z planes. So far, this is no problem.
In the next step, I need to do some computations with the surface normals (on facets and points) and integrate patches over surface nodes. If I integrate on my quarter-domain, facets that touch the boundaries are not integrated correctly because they do not see the normals from the other side of the symmetry planes and have only 1/2 the area. If I use periodic boundary conditions, the normals are not "mirrored" correctly, but made periodic...
Is there an easy way to just mirror a mesh and a solution vector on two planes to remove the quarter symmetry for post-processing my calculations?
Alternatively: I am interpolating the facet normals into a CG 1 function space on the surface to get point normals. Is there a way to make this procedure aware of the symmetry planes?