Obi Official Forum
Is this a bug? - Printable Version

+- Obi Official Forum (https://obi.virtualmethodstudio.com/forum)
+-- Forum: Obi Users Category (https://obi.virtualmethodstudio.com/forum/forum-1.html)
+--- Forum: Obi Softbody (https://obi.virtualmethodstudio.com/forum/forum-12.html)
+--- Thread: Is this a bug? (/thread-2495.html)



Is this a bug? - StudioTatsu - 18-09-2020

New Project. Clean install. No changes.

Maybe I need to tweak the settings, to prevent this.

What is happening?



Note: I tried it multiple times, and the same thing happens.


RE: Is this a bug? - StudioTatsu - 18-09-2020

Just as a sanity check, I updated Unity to 2020.1.5f1 and it still happens



This might be a bug, but at the same time - it could be a cool game mechanic.  Lengua


RE: Is this a bug? - josemendez - 18-09-2020

(18-09-2020, 04:17 AM)StudioTatsu Wrote: Just as a sanity check, I updated Unity to 2020.1.5f1 and it still happens



This might be a bug, but at the same time - it could be a cool game mechanic.  Lengua

Not a “bug” as in an error in the code, rather just constraint fighting as a consequence of tunneling. This is similar to what happens when you use a dynamic attachment (pin) inside a collider: the collision constraint tries to get the particle outside, and the pin constraint takes it back inside. Neither constraint ever "wins", and the result is some dance moves Guiño. This is described in the manual for the very specific case of pins+collisions, but can actually happen to other constraint types.

In this scene , the barrels are surface softbodies (no particles filling their volume) so when they interpenetrate each other or a ball, they become interlocked and try to get outside of each other with little success because shape matching and collisions begin undoing each other's work: that’s why they jump around.

Any measure taken to prevent tunneling would eventually fix this: using more substeps, larger particles, etc. Using volume softbodies instead would also make this scene more robust if slightly more expensive, we might just do that in the future.


RE: Is this a bug? - StudioTatsu - 18-09-2020

(18-09-2020, 06:59 AM)josemendez Wrote: Not a “bug” as in an error in the code, rather just constraint fighting as a consequence of tunneling. This is similar to what happens when you use a dynamic attachment (pin) inside a collider: the collision constraint tries to get the particle outside, and the pin constraint takes it back inside. Neither constraint ever "wins", and the result is some dance moves Guiño. This is described in the manual for the very specific case of pins+collisions, but can actually happen to other constraint types.

In this scene , the barrels are surface softbodies (no particles filling their volume) so when they interpenetrate each other or a ball, they become interlocked and try to get outside of each other with little success because shape matching and collisions begin undoing each other's work: that’s why they jump around.

Any measure taken to prevent tunneling would eventually fix this: using more substeps, larger particles, etc. Using volume softbodies instead would also make this scene more robust if slightly more expensive, we might just do that in the future.

Thanks for the explanation, I was thinking it might be tunneling, just wanted to be sure. 
This definitely entertained my 4-year-old son, he can't stop laughing "it's dancing"  Sonrisa