yade-dev team mailing list archive

[Václav Šmilauer <eudoxos@xxxxxxxx>]

> In the top of my list "todo" is the implementing "bonded particle
> schema', like described in
> http://www.springerlink.com/content/l2306412v1004871/

OK, I read that paper, it is not different (in general) from what should
have been Dem6DofGeom which I never implemented, but I had preliminary
implementation of bending&twisting in SpheresContactGeometry (the code
is still in there, commented), with a demonstration of twisting chain of
20 spheres (pendulum-like thing).

Though now I think that rather than implementing Dem6DofGeom, a unified
geometry classes should be made, where the displacement computation wil
be done in the geometry functor rather than in constitutive law or in
the IntrGeom class itself (like Dem3DofGeom does). That would allow to
switch between incremental and absolute formulations by switching Ig2
functors, with no change to the law itself. Also, virtual function call
would be avoided.

I will try to blueprint that one and perhaps prototype it; if someone
can help, he/she will be the most welcome.

Cheers, Vaclav