Dear James, To me, the most likely point of failure here relates to the application of boundary conditions. It might be that one of the constrained boundaries is not traction free, and so the problem is not free of rigid body modes. Can you please confirm that you apply the same numbering scheme to the boundary in 2d as is used in the tutorial? You can confirm the setup in either case by looping over all boundary faces after the mesh is created, and printing their ID. I think that this is the best starting point to diagnose what the problem could be. If that’s not the problem, then can you please provide an example mesh that fails, and confirm which version of deal.II you’re using.
Best, Jean-Paul > On 13 Apr 2020, at 02:57, James Gorman <[email protected]> wrote: > > To Whom It May Concern, > > I am learning deal.ii to work on problems in solid mechanics. In the interest > of computational efficiency, I plan to run most problems in 2D (plane stress > assumption). Since the interesting problems I am concerned about require a > non-trivial mesh, I learned how to use Gmsh based on the suggestion of > Step-49. > > With that in mind, I ran Step-44 in 2D and 3D (no change to the tutorials) > and got results. I then used gmsh to generate a hypercube mesh in 2D and 3D, > modifying Step-44 using the Step-5 as a guide. While the 3D calculation with > the imported mesh successfully converged, the 2D calculation with imported > mesh did not converge. The 2D calculation failed with the residual from > 10^5-10^16 (I tried multiple initial mesh densities). > > Looking at the error output, it states that when a residual is large the > matrix may be invertible or the wrong solver may be used. Since the only > thing I changed was the origin of the mesh, I don't think either of these > explanations should be the case. > > As a sanity check, I ran a ~5 of the other tutorials with imported meshes. > None of the other tutorials I tried ran into a similar issue when I imported > a mesh from Gmsh. The issue also stuck around whether or not I used a > structured or unstructured mesh, or changed the shape from a square to a > rectangle. > > Has anyone else ran into this issue? If so, how did you fix it? > > Sincerely, > James Gorman > > -- > The deal.II project is located at http://www.dealii.org/ > <http://www.dealii.org/> > For mailing list/forum options, see > https://groups.google.com/d/forum/dealii?hl=en > <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 [email protected] > <mailto:[email protected]>. > To view this discussion on the web visit > https://groups.google.com/d/msgid/dealii/f8bd877d-5d1c-477d-8d84-70903f15efb2%40googlegroups.com > > <https://groups.google.com/d/msgid/dealii/f8bd877d-5d1c-477d-8d84-70903f15efb2%40googlegroups.com?utm_medium=email&utm_source=footer>. -- 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 [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/C46D1292-A298-4747-8149-052212C81B5D%40gmail.com.
