yade-dev team mailing list archive

[Václav Šmilauer <eudoxos@xxxxxxxx>]

> just a curiosity;) I have been reading your paper "Discrete continuum
> models and their implementation" (good material, btw) and I come over
> the definition of normal strain once particles overlap. May I ask you
> if this is a proper definition? Are not we talking about rigid bodies,
> are we? I can see that the overlap accounts for a sort of deformation,
> but can we really define a normal or shear strain dealing with rigid
> bodies? 
The main reason is that the concrete model (which is not mentioned in
that paper deliberately) was derived from continuous models, so the
convention of using displacement normalized by nodal distance (=strain
on the fictious element) was naturally more convenient. Besides that, I
was thinking about interactions as beam-like elements between
"nodes" (sphere's centers): there is a separate implementation of the
concrete model in FEM for checking correctness of DEM, giving
lattice-like structure with such beams; and it was convenient to use the
same equations for both.

Cheers, v.