yade-users team mailing list archive

[Bruno Chareyre <question220785@xxxxxxxxxxxxxxxxxxxxx>]

Question #220785 on Yade changed:
https://answers.launchpad.net/yade/+question/220785

    Status: Open => Answered

Bruno Chareyre proposed the following answer:
The problem is apparently linked to fact that the bounding boxes are exactly touching each other in the initial state (b1.zmax = b2.zmin).
Moving the small sphere a bit further solves the problem, like this for instance:

center_sph = Vector3(center_bigSphere[0],
center_bigSphere[1]+1./5.*DbigSphere, center_bigSphere[2]-DbigSphere-
radius_sph)

closer works to:
center_sph = Vector3(center_bigSphere[0], center_bigSphere[1]+1./5.*DbigSphere, center_bigSphere[2]-DbigSphere/2.-radius_sph*1.01)

Remapping on the other side also fix the problem as it is like starting
with fresh (not touching) bounding boxes.

Why it happens only for a special size of the period is not clear, and
the fact that it is the size of the period along x which changes the
behaviour is really weird (since the boxes are touching  along z)...

It is bug clearly, I opened a report. Thanks for script.
It may be only a ">" to be replaced by a ">=" somewhere...

As long as it is not fixed, I guess you have to avoid bodies whose size
is exactly the size of the period if their bounding boxes are exactly
touching another box at step 0 (you were very unlucky to hit this
situation...).

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