[SciPy-user] FEM

Andrew Reid Andrew.Reid@nist....
Mon Mar 19 10:03:08 CDT 2007


On Mon, Mar 19, 2007 at 03:51:24PM +0100, Håkan Jakobsson wrote:
> Thanks everyone for your replies, they have been most helpful. I've studied
> the pros and cons for python, and I'm fairly sure I'll be using python more
> and more.

  But wait!  It's not over yet!

  I'm part of a team working on an FEM project at NIST which 
uses Python quite heavily.  Our brief is to address Materials
Science users, so we have a lot of custom tools for adapting
our mesh to user-provided microstructural image, and we also
have an extensible interface for adding new types of 
user-provided custom constitutive rules.  The actual FEM 
part is pretty standard.

  Our numerical work and solvers are all in C++, 
and we don't actually use SciPy/NumPy, we just use SWIG
to get to the numerical code.  

  The flexibility of Python has been crucial in working through 
our API and mesh-construction tools -- it has the right level of 
easy scriptability, and the right sort of object-orientation,
for what we need.

  The project page is at <http://www.ctcms.nist.gov/oof>, 
and there's a relatively recent talk (in PDF form) at 
<http://www.cacr.caltech.edu/projects/danse/talks/kickoff/22-Reid/danse_oof.pdf>, 
which very strongly resembles the talk I gave at SciPy2006.

				-- A.
-- 
Dr. Andrew C. E. Reid, Guest Researcher 
Center for Theoretical and Computational Materials Science
National Institute of Standards and Technology, Mail Stop 8910
Gaithersburg MD 20899 USA
andrew.reid@nist.gov


More information about the SciPy-user mailing list