iMechanica - finite elements - Comments

"Easy to Use" Or "Easy To Develop"

Roozbeh Sanaei — Fri, 14 Sep 2007 15:26:37 -0400

i think research programs must be "easy to develop" in addition to be "easy to use". is it not better to develop procedures and libraries instead of complete platform?. procedures are really useful and developing new programs is more easier with libraries instead of complete softwares. i think always concentrate on one think is very better than working on all the world. I think Diffpack and Deal.ii are very good examples of this.

Re:FEBio and mesh generation

Biswajit Banerjee — Fri, 14 Sep 2007 13:08:38 -0400

Dorival,

You should talk directly to Jeff Weiss about the status of the code (open source or otherwise) and on any collaborations you might be interested in. I am sure he will entertain specific questions. With the development of excellently behaved tetrahedral elements (by Mike Puso and others) many people are moving away from hexahedral mesh generation. I would guess that Jeff is also thinking along similar lines.

A promising recent move away from mesh generation has been initiated by Tom Hughes and others on directly using Bezier-Bernstein and other shape approximation functions (that avoid the mesh generation problem and instead use the infrastructure used to build engineering solid models). I would appreciate it if someone could comment on the status of that work.

Biswajit

Open source?

pedroso — Tue, 11 Sep 2007 07:07:58 -0400

Hi guys,

Is that software open source?

I was wondering whether we can collaborate to the hexahedral mesh generator.

Cheers,

Dorival.