Dear Daniel The code solve a solid mechanic model with a plastic material model. The WorkStream is used for matrix and residual assembly (like step-44). So I just add some new classes (PerTaskData and ScratchData) and the most parts of the code is not changed. The code works similar to the its serial version without any problem when the local adaptive refinement is not active. During a load increment, let the solution to time step Tn be given. The full problem is solved. First, I refine the mesh based on the new solution and interpolate the old solution (at Tn ) onto the new mesh. Then I solve for the solution at Tn+1 again, but on the refined mesh. This process is tested in the previous version of the code. However, in the new code after the first refinement, the solution for Tn+ has a constant residual error during Newton iteration.
Thank you P. On Wednesday, February 1, 2017 at 8:24:51 PM UTC+3:30, Daniel Arndt wrote: > > Pasha, > > I have a code that works fine. Now, I am working to change the code >> structure to use “workstream”. The new code works fine without adaptive >> refinement. However, when the mesh is refined and the triangulation has >> hanging node, a constant residual error remains during Newton iteration. I >> cant find source of the problem >> > You are providing way too few information to expect useful answers. What > are you doing in your code? > For which part in your code do you want to change to WorkStream? > What exactly did you change? How is this related to whether you are doing > adaptive refinement or not? > > Best, > Daniel > -- 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]. For more options, visit https://groups.google.com/d/optout.
