So I have hit my first snag in the definition of the FE_Nothing element. I created a derived FiniteElement class with zero degrees of freedom and trivial implementations of all necessary virtual functions (see fe_nothing.h). I then created a simple test case where I divide the mesh into two regions and assign either a FE_Q(1) or a FE_Nothing () from my hp fe collection to each cell. The test case (also attached) fails when I attempt to distribute dofs, with an assertion from "hp_vertex_dof_identities()". It seems the hp dof_handler expects all elements to be of type FE_Q, and therefore to have dofs at the cell vertices. Any thoughts on how to proceed?

Thanks again,
Josh


Attachment: fe_nothing.h
Description: Binary data




Attachment: test-case.cc
Description: Binary data




_______________________________________________
dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii

Reply via email to