Dear Bjorne, There are quite a number of examples of Laplace and Poisson problems in deal.II's tutorials <https://www.dealii.org/8.4.1/doxygen/deal.II/Tutorial.html>. The picture that you've shown could well be produced by a derivative of step-6 <https://www.dealii.org/8.4.1/doxygen/deal.II/step_6.html>, but I'm not entirely sure.
As for solving such an ill-conditioned system, deal.II offers a number of options for direct solvers (e.g. UMFPack <https://www.dealii.org/8.4.1/doxygen/deal.II/classSparseDirectUMFPACK.html>, MUMPS <https://www.dealii.org/8.4.1/doxygen/deal.II/classPETScWrappers_1_1SparseDirectMUMPS.html> and other options through Trilinos's direct solver interface <https://www.dealii.org/8.4.1/doxygen/deal.II/classTrilinosWrappers_1_1SolverDirect.html>) and high quality preconditioners (for example, Trilinos's AMG <https://www.dealii.org/8.4.1/doxygen/deal.II/classTrilinosWrappers_1_1PreconditionAMG.html> and AMGMueLu <https://www.dealii.org/8.4.1/doxygen/deal.II/classTrilinosWrappers_1_1PreconditionAMGMueLu.html>, and PETSc's BoomerAMG <https://www.dealii.org/8.4.1/doxygen/deal.II/classPETScWrappers_1_1PreconditionBoomerAMG.html> and Parasails <https://www.dealii.org/8.4.1/doxygen/deal.II/classPETScWrappers_1_1PreconditionParaSails.html>) which might be able to work around this problem. I hope this helps! Regards, Jean-Paul On Thursday, January 12, 2017 at 1:17:20 PM UTC+1, bjorne wrote: > > <http://www.dealii.org/images/wiki/gallery/Single_layer.jpeg> > Hi, > > I'm trying to solve an elliptic problem with discontinuous material > properties. The values can vary by ~ 10^5, so the system is > ill-conditioned. Currently I'm using various other FE solvers that don't > deal well with this conditioning. > > The first picture in the deal.II website (http://www.dealii.org/) is the > included image. It appears also in the gallery ( > https://github.com/dealii/dealii/wiki/Gallery). My solutions looks > similar to this. I was hoping somebody in the community could point me to > any examples that solve similar problems, or the code that produced this. > > Thanks, > > B. > -- 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.
