Nails not falling correctly

   1182   4   2
User Avatar
Member
8 posts
Joined: Oct. 2016
Online
Hello, everyone,

I'm really sorry but I'm going insane with this simulation. Does anyone of you have an idea why the nails don't fall correctly on the ground?

The geometry of the nails should be correct, I checked the volume representation and tried all other conceivable parameter changes. Another scale didn't help either. It just doesn't work.

Maybe it has something to do with the center of mass? But I've already set it manually - without success.

Thank you for your help and best regards
mark

Attachments:
nails.hiplc (655.9 KB)
flip.mp4 (1.3 MB)

User Avatar
Member
133 posts
Joined: March 2016
Offline
Hello - Just having a little play with this (only got apprentice on my laptop)
It looks like the collision geo on the nails is not very good and causing issues, changing it to 'capsule' in the bullet data of the RBD packed object, and also upping the rotational stiffness on the solver, allows it to getting closer to what i think you want. Attached the tweaked file - I also used the sop RBD solver (unlocked it to be able to go in and change the RBD packed obj collision to 'capsule')

Attachments:
nails.hipnc (2.1 MB)

Love Houdini
User Avatar
Member
238 posts
Joined: March 2013
Offline
The RBD Solver collision tab is not what you need to be looking at, I wish sidefx would remove it, catches a lot of people out.
You want the Bullet Data tab, set your Geometry to "capsule" as mentioned above.

But always turn on Display Collision geo before you sim anything, to see what your colliders actually look like. If you
enable it on your scene, you'll see the issue instantly.

L
Edited by lewis_T - Nov. 7, 2022 00:29:00
I'm not lying, I'm writing fiction with my mouth.
User Avatar
Member
8 posts
Joined: Oct. 2016
Online
Hello,
Thank you for your feedback and the file. And sorry for the late reply. I'll turn on email notifications for the forum. Just missed that. I have intentionally used the RBD solver. So far I've had the impression that the bullet solver simulates a bit more sloppy and is more suitable for short action clips where you lose track of the movements. In addition, the RBD solver is more flexible in terms of collision geometries.
I looked at Hatchery's simulation. The problem is that the nails no longer behave asymmetrically.
With the RBD solver, a large number of nails fall correctly - only a few move incorrectly. And I couldn't figure out why.
Here is my final result with the RBD solver.

Attachments:
Nails_Editor_1.mp4 (1.7 MB)
Nails_1.mp4 (1.8 MB)

User Avatar
Member
475 posts
Joined: July 2005
Offline
I get better results after choosing division method by size (value ~ 0.01) for the piece node (volume tab).
Edited by Aizatulin - Dec. 5, 2022 12:47:00

Attachments:
nails_piece_divi_by_size_001.hipnc (616.0 KB)

  • Quick Links