Skip to content

Bounday aware explicit one step method

René Heß requested to merge bugfix/explicit-onestep-with-constraints into master

Make the explicit one step method aware of dirichlet boundary constraints. Because of the way we apply constraints to residual and jacobians we have to reformulate the linear system to a residual formulation.

It would be good if someone could look at the changes and think about wether that makes sense from a theoretic point of view. Before this change the boundary of the solution was equal to zero on all dirichlet boundary parts after the initial solution. Right now I only have a test case from code generation and I'm leaving for holidays soon. After my holidays I can also provide a proper test.

Merge request reports