Hi Chong, MeshWorker does work without much effort in MPI parallel code and is made to help with exactly this (who assembles what is non-trivial if you have hanging nodes and processor boundaries). The only thing you have to watch out for is supplying the right flags that determine the cells and faces that will be assembled. I don't think we have a step-12 or -74 with MPI, but there is the error estimator in step-50: https://www.dealii.org/current/doxygen/deal.II/step_50.html#Theerrorestimator
You probably want the flags MeshWorker::assemble_own_cells | MeshWorker::assemble_ghost_faces_once | MeshWorker::assemble_own_interior_faces_once (or _both for both of them if you want to assemble each face twice). On Fri, Aug 12, 2022 at 11:41 PM chong liu <liuchong...@gmail.com> wrote: > > Dear community, > > I have a problem in the assembly part when solving Step-74 (DG method) by > MPI. Generally speaking, we can first determine whether one cell is locally > owned or not by cell->is_locally_owned(), and then assemble matrices and > residuals by hands (e.g. Step-40 for CG and “Distributed LDG Method” for DG > in code gallery). However, Step-74 using Meshworker and FEInterfaceValues > classes automatically loops and assembles all cells, faces, and interfaces. > My question is how to identify one cell which is locally owned or not in > Step-74. Is there any deal.II code combining discontinuous Galerkin method > (which uses FEInterfaceValues class, e.g. Step-12 and Step-74) with MPI > (either PETSc or Trilinos)? Any help is going to be appreciated. > > Thanks and regards, > > Chong > > -- > 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. > To view this discussion on the web visit > https://groups.google.com/d/msgid/dealii/eed5b3fb-2052-44a3-8840-1be2f1484c7a%40Spark. -- Timo Heister http://www.math.clemson.edu/~heister/ -- 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. To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/CAMRj59HahTQf0%3D0h3aLXo64DkXeoeW3RYv3pHmo8aqb91_-zYQ%40mail.gmail.com.
