yade-users team mailing list archive

[Bruno Chareyre <question632092@xxxxxxxxxxxxxxxxxxxxx>]

Question #632092 on Yade changed:
https://answers.launchpad.net/yade/+question/632092

    Status: Open => Answered

Bruno Chareyre proposed the following answer:
Hello,
I think what you find is mainly a flaw of this whole "initially overlapping particles" concept, yet it also suggests a small change in the code.

What you get is logical from an algorithmic point of view (interaction broken, then reset with unp=0), yet not logical from a physical point of view.
You could say: "ok, let's just keep the zero force until there is no overlap". It can be implemented easily and it would certainly improve the situation, I can do it if you like.
However, if the contact is created again after the reset, it will be fresh and the equilibrium overlap will be zero. Which means that some volume is created overall.
Conversely, the interaction will not be reset if the bond breakage is followed by a large rolling motion (while continuously overlapping).
I don't see the rational behind the fact that moving apart is resetting the equilibrium distance while rotating is not. It should be the opposite in my view: if the surface of the grain is locally flat translating it back and forth keeps the flatness unchanged in the contact region; rotating is changing it.

Anyway, again, if there is an agreement on what the physical assumptions should be implementing them is easy.
My suggestion is to maintain fn=0 until the overlap disappears.

Bruno

-- 
You received this question notification because your team yade-users is
an answer contact for Yade.