yade-users team mailing list archive

[Bruno Chareyre <question234129@xxxxxxxxxxxxxxxxxxxxx>]

Question #234129 on Yade changed:
https://answers.launchpad.net/yade/+question/234129

    Status: Open => Answered

Bruno Chareyre proposed the following answer:
1. I will only put Jan's comment a bit differently. The "du" that should
be used in "Ftt=Ft-kt*du" is the future displacement  between t and
t+dt. Since t+dt is not computed yet, du is not know. So you are using
du of the previous step, but it is inconsistent.

2. I didn't know but searching alphaKr in the code shows this:
Real AlphaKr = 2.0*sdec1->alphaKr*sdec2->alphaKr/(sdec1->alphaKr+sdec2->alphaKr)

division by zero = no good... A small bug to fix. 
It is not really a problem for you. If you do nothing (keep all defaults) there will be no moment (momentRotationLaw=False) and it will not crash (default alpha's are not zero).

3. There are a few known sources of energy changes. For instance creating a contact leads to unbalanced energy change. How big is the error compared to total energy? Does it decrease with the timestep?
I know energy is conserved very accurately in dense quasi-static systems. Is it what you are interested in, or do you need also energy conservation in dynamic problems?

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