Closed rbooth200 closed 8 years ago
Fixed the compile bug with mpi. Ran a couple more sod tests with mpi, results seem sensible.
It seems fine to me. David, do you want to add it also to the meshless?
Yes, I can do this (add to the meshless) before we accept the merge. I will do it in a few days after my first lecture.
I've added this to the meshless now since I needed to make a modification the SPH implementation anyway - the condition wasn't yet strict enough to cover all edge cases
… - only if periodic/reflecting boundaries are used.
I've tested it by varying the number of neighbours in a sod shock tube test. It seems to work fine. Not tested with MPI.