Hi Krzysztof, I don't have a direct comment on the result that you've shown here, but I think it would be useful if you mentioned what type and order of FE you are using here.
One thing that you should be aware of is that Paraview does not interpolate data in the same way as deal.II does, so the postprocessor itself may be giving you visually misleading results. In particular, each quad is split into two triangles and the solution field is linear on each patch. If you look carefully you can see this in the images you've provided - they're not exactly symmetric about the vertical axis. This, along with the imposition of the hanging node constraints, may play a role in the different shape of the isocontour you've plotted. But again, without knowing the details of the FE space I'm only making a first guess. That said, it does seem a bit strange that the transfer of the solution from a coarse to fine grid renders those two local maxima on the upper part of the geometry. I don't think anything I've mentioned here could result in this artefact. Out of curiosity, have you tried to see what happens when you interpolate onto a globally refined mesh? Best, J-P On Thursday, January 12, 2017 at 9:14:12 AM UTC+1, Krzysztof Bzowski wrote: > > Hi, > I am working on the project with two different triangulations with a > common coarse grid (one without refinement and second with a > refinement). I need to temporary interpolate solution from a coarse grid > to fine grid so I decided to use interpolate_to_different_mesh. Results > are a little bit strange (see attachment). The white line is an > isocontour equal 0. My code looks like: > > DoFHandler<2> fine_dof_handler(fine_triangulation); > fine_dof_handler.distribute_dofs(fe); > > ConstraintMatrix tmp_hanging_nodes; > DoFTools::make_hanging_node_constraints(fine_dof_handler, > tmp_hanging_nodes); > tmp_hanging_nodes.close(); > > Vector<double> fine_level_set(fine_dof_handler.n_dofs()); > VectorTools::interpolate_to_different_mesh(coarse_dof_handler, > coarse_solution, fine_dof_handler, tmp_hanging_nodes, fine_level_set); > > tmp_hanging_nodes.distribute(fine_level_set); > > Best regards, > Krzysztof > -- The deal.II project is located at http://www.dealii.org/ For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en --- You received this message because you are subscribed to the Google Groups "deal.II User Group" group. To unsubscribe from this group and stop receiving emails from it, send an email to dealii+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
