[Dune] integration and assembly for polygons

Andreas Dedner a.s.dedner at warwick.ac.uk
Tue Oct 6 18:27:42 CEST 2015


Hi Ganesh.
I think this is the wrong place to pose that question because the answer is
going to depend on the underlying fem package you use. Perhaps better
to put it for example to the dune-fem mailing list (and I'll answer it 
from there).
Best
Andreas

On 06/10/15 15:58, Ganesh Diwan wrote:
> Hi Dune
>
> I recall one of my old posts on this topic: 
> http://lists.dune-project.org/pipermail/dune/2015-February/013459.html
>
> In order to "build" a polygon: I first create a mesh the usual way 
> (either Gmsh or in Dune itself) and then partition it using ParMETIS 
> to agglomerate the finite elements. One such agglomerated unit is one 
> of my polygons and so on.
>
> Now, I am trying to understand if it will be possible to treat these 
> polygons (which are basically agglomerated finite elements) as finite 
> elements in their own right in Dune? I can compute a given function f 
> at the quadrature points inside the polygons and do int_Omega f,  but 
> obviously I need more. For eg, if an agglomeration can be treated as a 
> finite element, how do I evaluate the basis function phi_i, its 
> derivatives and then do a subsequent assembly for them? To be precise, 
> the methods which are already in place to do exactly this, can I use 
> them to operate on my polygons? For eg. when reusing the dune-fem 
> code, I would like to use methods such as localFunction.evaluate, 
> rhsLocal.axpy and rhs.communicate but I do not understand how to and 
> what to change in order to achieve this.
>
> Many thanks for your help,
> Ganesh
>
>
> _______________________________________________
> Dune mailing list
> Dune at dune-project.org
> http://lists.dune-project.org/mailman/listinfo/dune

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.dune-project.org/pipermail/dune/attachments/20151006/a9b4696a/attachment.htm>


More information about the Dune mailing list