[Dune] longest edge alberta

[Martin Nolte]

Hi Marco,

sorry, but Alberta does newest vertex bisection. You can only choose the 
refinement edge on the macro grid. It is always the edge between vertex 0 and 
1 (in local numbering).

You can ask the Alberta GridFactory to rotate elements such that the longest 
edge is the refinement edge. However, this interface extension is specific for 
AlbertaGrid. In a DGF file, you can add the following parameter block:

GRIDPARAMETERS
REFINEMENTEDGE LONGEST
#

If you already have such a block, simply add the line "REFINEMENTEDGE LONGEST" 
to it. If this grid is read into AlbertaGrid, the longest edge will be 
selected as refinement edge.

Unfortunately, there is a catch: In 3d, you might end up with a grid that 
segfaults on refinement due to the recursive bisection used in Alberta. As far 
as I know, this only happens if the two longest edges of a tetrahedron do not 
have a common vertex. The grid factory will give a warning in this case. So, 
if your grid is generated from cubes by Kuhn triangulation, you should be fine.

If you consider this explaination helpful, would it be possible for you to add 
it (after possible improvements) to our FAQ?

Best,

Martin

On 06/26/2013 05:23 PM, Marco Cisternino wrote:
> Hi duners,
> I'm trying to refine Alberta. I'm able to do it but not in the way I'd like it.
> I would like to use only the longest edge as the edge to be divided. Is it
> possible?
> For now it uses bisection with I don't know which choice of the edge.
> I saw an example here http://www.dune-project.org/gallery/p1fem.html
> How can I get that result?
>
> Thanks a lot.
>
> Marco
>
> --
> ------------------------------------
> Marco Cisternino, Ph.D.
> OPTIMAD Engineering s.r.l.
> Via Giacinto Collegno 18
> 10143 Torino - Italy
> www.optimad.it
> marco.cisternino at optimad.it
> +39 011 19719782
> ------------------------------------
>
>
>
> _______________________________________________
> Dune mailing list
> Dune at dune-project.org
> http://lists.dune-project.org/mailman/listinfo/dune
>

-- 
Dr. Martin Nolte <nolte at mathematik.uni-freiburg.de>

Universität Freiburg                                   phone: +49-761-203-5630
Abteilung für angewandte Mathematik                    fax:   +49-761-203-5632
Hermann-Herder-Straße 10
79104 Freiburg, Germany