Hi, I was wondering if there would be a minimal working example where an AdaptiveNonlinearVariationalSolver with a changing mesh was succesfully run with fenics_adjoint functions like ReducedFunctional?
The fenics_adjoint seems to be requiring some information passing that might be broken when doing something like:
‘’‘p, u, T = w.leaf_node().split()’‘’
with T becoming a dolfin object.
Thank you.