I recently use pylith to simulate 2D fault dynamic rupture procession, the result is perfect when used cubic meshing and slip_weaking constitutive, but when i used tri meshing or add a fault parallelling the exist one,the error ‘‘RuntimeError: Error occurred while adjusting topology to create cohesive cells for fault ‘nucleation’.’’ occurred, it’s detail is ‘’PETSC ERROR: Petsc has generated inconsistent data
PETSC ERROR: Point 26246 not found in support of cone point 6549.’’
My fault model made up with three part like this: ------fault(long)------nucleation(very little)---------(long), and the edge nodes of different parts in same cubic/tri unite but no plex.
I am very appreciat if you give me help about this question. Thanks.