fenics team mailing list archive

Thread
Date

Re: FEniCS, Sieve and AMR

To: "tomtzigt" <tomtzigt@xxxxxxxxxxxxxxxx>
From: Matthew Knepley <knepley@xxxxxxxxxxx>
Date: Tue, 15 Nov 2005 15:18:42 -0600
Cc: 'Discussion of FEniCS development' <fenics-dev@xxxxxxxxxx>
In-reply-to: <200511152106.jAFL6DXi005912@aphrodite.gwi.net> (tomtzigt@suscom-maine.net's message of "Tue, 15 Nov 2005 16:06:20 -0500")
User-agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4 (Jumbo Shrimp, linux)

"tomtzigt" <tomtzigt@xxxxxxxxxxxxxxxx> writes:

> :o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns="http://www.w3.org/TR/REC-html40";>
>
> What is the roadmap of the functionality integration between FEniCS and Sieve? I am looking at using CUBIT for some industrial strength meshing but need some help understanding the impact of that choice in the context of AMR
> within FEniCS.:p>

  The idea is to represent meshes in a way that allows us to easily construct algebraic structures
over them. Any mesh generator could be used. The downside is really in the flexibility/optimization
tradeoff. Currently, our implementation would be slower/bigger than some custom structured AMR like
Chombo (or at least shuold be :) However, our approach to this will be to generate custom code for
EACH mesh that is output and compiled, much like the FFC philosophy of generating code for EACH
weak form. Does this help?

     Matt
-- 
"Failure has a thousand explanations. Success doesn't need one" -- Sir Alec Guiness

References

FEniCS, Sieve and AMR
From: tomtzigt, 2005-11-15