yade-users team mailing list archive

[Chareyre <question171223@xxxxxxxxxxxxxxxxxxxxx>]

Question #171223 on Yade changed:
https://answers.launchpad.net/yade/+question/171223

Chareyre proposed the following answer:

> Conclusions. Number of facets doesn't matter. Yade doesn't work out
> sphere/facet interactions properly. Maybe I can use a sphere with huge
> radius instead of the facet?

As suggested before, you could use boxes, but actually your conclusion
does not hold, the problem is in your script.
Bodies motion should always be defined in terms of prescribed velocity
(or using motion Engines for the most complicated cases).
Here, you are changing the position of the shaking table, which breaks
the current implementation of verlet distance in the collider. The
result wuld be the same with facet, box, or a big sphere.

Setting verletDist=0 fixes the problem, but still it is not good:
viscous laws for instance would suffer from this pos= in the script.
Prescibe velocity, then it will work like a charm, and you can keep
using verlet>0 (default), which will speed up the simulations.

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