yade-users team mailing list archive

[Bruno Chareyre <question235266@xxxxxxxxxxxxxxxxxxxxx>]

Question #235266 on Yade changed:
https://answers.launchpad.net/yade/+question/235266

    Status: Open => Answered

Bruno Chareyre proposed the following answer:
Finite elements were in the "implementation request list" at the very
beginning of yade. It went out, since we couldn't seriously pretend to
compete with the many well written FEM codes available.
If, instead of a general purpose FEM module,  you plan to develop a very
specific feature that would be named "FEM", it becomes more realistic,
and can be interesting (at this point I don't know exactly what you have
in mind though, would deformable elements interact with each other, and
how?).

For integrators, I'm also sorry to tell you that we didn't plan to write
different ones. It seems that the need to recompute all interactions at
each substep of the multisteps/iterative algorithms make them less
efficient in the end. I have never seen any report of performance
improvement in DEM by using another integrator than the original
explicit 2nd order scheme. The situation may be completely different in
finite elements of course, where the stiffness matrix does not have to
be recomputed at each substep.

Anyway, if you seriously start such a project, it would be nice to keep
contacts via the yade-dev mailing list.

Looking forwards.

Bruno

-- 
You received this question notification because you are a member of
yade-users, which is an answer contact for Yade.