yade-dev team mailing list archive

[Václav Šmilauer <eudoxos@xxxxxxxx>]

> You can repeat the test, is simple. The fact that the shear force
> changes sign from one case to the other, is not correct. The shear
> force is dependent on the relative shear velocity that in turn depends
> on the ids at contact as we define them. To me this happens because we
> do not define a local system. 
> 
> Please correct me if you see I am wrong or please explain me why the
> shear force changes sign (if it was really a global system as was
> already pointed out, it should not happen).

Hi Chiara, if I understand right, shearForce get negated if you swap
ids. That is consistent. shearForce is in global coordinates, but to one
body (id1*), it applies positive (action), while to the other body
(id2*) negated (reaction). 

Does this explain what you describe?

* This is not directly related to which body we assign the positive
force; we already agreed it should be id1 everwhere
(https://bugs.launchpad.net/yade/+bug/493102 and our discussion
http://www.mail-archive.com/yade-users@xxxxxxxxxxxxxxxxxxx/msg01285.html); while it is not consistent in the code ATM, one way or another, you will always end up with the inverse force on the other particle.

Cheers, Vaclav