I am supposed to model heat transfer through a cavity that absorbs concentrated solar radiation. The discretization of the cavity is sort of set by a previous ray-tracing model. I end up with something like in the attached image, where two plates meet with a butt joint. The points at the joint have different discretizations on each side so I'm having trouble drawing the volume element in a way that avoids double counting the conduction fluxes.
I have tried searching for ways to handle discontinuous grids, but all I find are results for discontinuous functions (shock waves, etc.). Direct answers or better search terms/keywords are appreciated.
