Non-linear solver convergence failure

dolfinx
1

Hello everyone,

A code that worked correctly with a non-linear problem with the following structure has stopped converging.

problem = dolfinx.fem.NonlinearProblem(F, u, bcs=[bc])
solver = dolfinx.nls.NewtonSolver(MPI.COMM_WORLD,problem)
r = solver.solve(u)

When I re-computed the code it no longer recognised dolfinx.nls and I added the following line:

import dolfinx.nls

With that single modification I got the following message:

RuntimeError: Newton solver did not converge because maximum number of iterations reached

Does anyone know if this is due to a change in the internal code of the nls function or how can I solve this problem?

2

You Need to supply a minimal code of what was working prior to the change and after the update, as there is no specific reason for it no longer converging