| Thread Previous • Date Previous • Date Next • Thread Next |
Question #248996 on Yade changed:
https://answers.launchpad.net/yade/+question/248996
Status: Open => Answered
Anton Gladky proposed the following answer:
Hi Andy,
2014-05-20 20:11 GMT+02:00 Andy Torres <question248996@xxxxxxxxxxxxxxxxxxxxx>:
> I have been working in a toy model, trying to figure out how all the viscoelastic forces works, and I think there is a bug here, because there are forces acting even when the particles are not touching each other. This happens if the particles separate after a collision and then meet again.
If penetration depth < 0, then the interaction should be removed, so
no forces at all: line 70-71, 45 [1]
[1]
https://github.com/yade/trunk/blob/master/pkg/dem/ViscoelasticPM.cpp#L45
> The simulation I made is just two aligned spheres ( one right above the other, so only normal force is involved) one fixed and the other free to fall under gravity force and bounce on the first fixed sphere.
> In the first bounce, the behavior of the system is the right one, the Penetration Depth (PD) starts at 0, the normal Force (Fn) has a large initial value because of the velocity of the falling particle at time of contact. The interaction doesn't disappear after the bounce, but the code has an "if" to set the force to zero and prevent an attractive force due to a viscous component:
> (line 104-106 of /pkg/dem/ViscoelasticPM.cpp)
> const Real normForceReal = phys.kn * geom.penetrationDepth + phys.cn * normalVelocity;
> if (normForceReal < 0) {
> phys.normalForce = Vector3r::Zero();
Please provide minimal working example, so we can test it. Better with
the proposal, how can it be fixed.
Regards
Anton
--
You received this question notification because you are a member of
yade-users, which is an answer contact for Yade.
| Thread Previous • Date Previous • Date Next • Thread Next |
