yade-dev team mailing list archive

[Václav Šmilauer <eudoxos@xxxxxxxx>]

> I try to use clumps for sometimes and it seems to me that there is a  
> small "bug" (maybe I'm wrong...).
> When clumps are used, they are part of the bodyContainer (like common  
> bodies). But when interaction forces are computed, it seems that the  
> resulting moment is computed at the body center and not at clump  
> centroid.
> I repeat that I may be wrong but in the "contact law engines" I read :
[...]
> addTorque(id1,-c1x.Cross(f), rootBody);

See NewtonsDampedLaw.cpp:61-73 for explanation. Yes, forces are first
put to the bodies, but then recomputed then WRT clump center.

> PS: I've got a python example where one can see strange behavior of  
> clumps. I can send it to anyone if necessary.

Please do send the script! (I am still not quite confident about clumps
in this respect, but I was also suspecting that
InteratingSphere2InteractionBox4SpheresContactGeometry is broken somehow
for shear... Not sure, we will see).

Regards, Vaclav