yade-users team mailing list archive

[Ruidong LI <question707278@xxxxxxxxxxxxxxxxxxxxx>]

Question #707278 on Yade changed:
https://answers.launchpad.net/yade/+question/707278

Ruidong LI posted a new comment:
Thanks for your reply, Jan!

> probably it should be opposite, Vector3(x_cyl-x, y_cyl-y, 0) to point inside the cylinder (not outside)?
Currently the membrane is stretched rather then compressed (tested with maxVelocity=100,iterPeriod=1)
> That's true. We have to reverse the direction vector.

>And maybe different parameters kP, kI, kD then default 1.0?
> Yes. But the difficulty is that it's hard to determine proper proportional parameters for kP, kI,  and kD.

> I don't know, then the behavior is also strange, the membrane is pulled inward once there is node-sphere interaction...
> Yes. I also encounter this problem. I try to reduce the maximum velocity of servo control but still failed, only delaying the time of membrane destruction.

Actually, the PFacet is demonstrated to be effective in modelling
flexible membranes successfully [1]. The problem is that there are some
critical parameters such as velocity are not mentioned in this paper,
which increases the difficulty for descendants to replicate the work.

[1] https://doi.org/10.1061/(ASCE)GM.1943-5622.000134

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