yade-users team mailing list archive

[Bruno Chareyre <question267815@xxxxxxxxxxxxxxxxxxxxx>]

Question #267815 on Yade changed:
https://answers.launchpad.net/yade/+question/267815

Bruno Chareyre proposed the following answer:
Hi, the main source of energy sink is usually due to the fact that there is no accurate creation-deletion of contacts. When a new contact is created during a timestep it is only reflected at the next timestep.
It makes more accurate integration schemes useless usually. Even if they are nominally of order 3, 4 or more, the dominant error is due to this failure to precisely acknowledge the contact duration, then everything is of order 2 ultimately (unless adaptative timestepping is used, a nightmare with thousands of particles). Worst that that, some higher order and/or implicit methods have the property to not conserve energy even for simple systems.
If your initial impact is the fastest of all collisions in the problem it explain why the energy change is most noticable at this time.
As always with numerical methods the result will be more accurate if the timestep is decreased.
Bruno

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