Matrix-free multigrid with anisotropic mesh refinement based on step-75?

[subramanya gautam]

Hi,

I am trying to solve a laser-heating type problem over a rectangular plate. So that the x-y dimensions are much larger than the z dimension. At present I am using a modification of step-37 to solve it, but I don't need too much discretization in the z-direction. One of my thoughts was to use the anisotropic refinement (cut_xy) to only refine the elements in one direction and then solve it using multi-grid. However - the local-smoothing multigrid -does not allow this as it needs to limit_level_differences at the vertices.

In  the step-75 example however, I don't see the limit_level_difference_at_vertices flag explicitly passed to the triangulation. Does that mean I could construct a set of triangulations using anisotropic refinement and use a multigrid method to solve it?

Thanks,

Subramanya.

[Daniel Arndt]

Subramanya,

the anisotropic refinement support is pretty limited and won't work
with parallel::distributed::Triangulation. I would be surprised if it
works with the geometric multigrid framework.

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 dealii+un...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/dealii/e8c122b7-5182-4a08-8c55-30b2b0ced92an%40googlegroups.com.

[Subramanya G]

Thanks! I tried to anisotropically refine and coarsen a parallel::distributed triangulation and saw that it didn't work. 

Subramanya.  

सुब्रह्मण्य .

To view this discussion visit https://groups.google.com/d/msgid/dealii/CAOYDWbKt0FV%3DZBFMVE3EAzu9QLsScksEtGtS2%3D2JjB4w63KMuw%40mail.gmail.com.