Physics Self Mesh


https://forum.reallusion.com/Topic430286.aspx 		Print Topic | Close Window

By Lord Ashes - 6 Years Ago
I am creating a Skeetball object. Part of the object is a board with holes in it. Since I want the Skeetball object to be physics enabled, I constructed the holes in the board at the mesh level instead of using opacity textures.
When I activate Physics on the object and then select Self Mesh, I see that the mesh lines properly wrap around the holes (i.e. recognize the mesh holes). However, when I run the project (Realtime or Frame) the physics enabled ball does not go through the hole. The physics act like the board does not have any holes. I have tried making the physics enabled ball much smaller than the hole (and ensured that it uses a Sphere mesh collider) but the results are the same. The wholes in the board act as if they are solid.

Any tricks to making this work correctly? Is there some settings I should ensure that I have set?
By GOETZIWOOD STUDIOS - 6 Years Ago
Could you share your scene ? Or a simplified scene showing the issue ?
By Lord Ashes - 5 Years Ago
Below is a link to a project which demonstrates the problem. Please note that all contents in the project a (C)opyright by Lord Ashes with permission to use the contents only for diagnostic purposes to help resolve the issue. All other use of the contents is forbidden.

https://drive.google.com/open?id=1bhhJKNX4ftlZbouPXySJDQFhr8UJsYZR

When I apply a force to the ball, I generally get one of three behaviors:

1. The ball gets stuck in the ring area but does not "fall down" to rest on the ring.
2. The ball ends up resting on one of the rings but does not go through the hole.
3. The ring area is solid and the ball cannot enter it.

I also get different results depending if I am in Realtime or Frame mode.     
By 4u2ges - 5 Years Ago
Couple of notes.

1. For every physics enabled prop set the Collision Margin to minimum: 0.01  (This is a main problem)
2. Simplify the Backboard geometry. Take it to Blender. Do Limited Dissolve and then Triangulate faces.

Then vary initial force for the ball until you get it into the hole :)
By GOETZIWOOD STUDIOS - 5 Years Ago
@Lord Ashes
Yep, collision settings was the main issue and yes, when doing physics simulations the rule is to better have an homogeneous meshe(s) tesselation. Physics engines hate when they have to deal with both big and small polygons at the same time.
Also, always run your physics simulation "By Fame", never in "Realtime". Only once you have baked your physics simulation you can switch back to "Realtime".
By Lord Ashes - 5 Years Ago
Thanks for your suggestions. I will give that a try.

I am familiar with the Triangulate (but seems I didn't apply it) but what, if I may ask, does the Limit Dissolve do?
By 4u2ges - 5 Years Ago
I am familiar with the Triangulate (but seems I didn't apply it) but what, if I may ask, does the Limit Dissolve do?         


Limited Dissolve arbitrary simplifies the mesh topology by getting rid of vertices and edges on planar faces. Most of the time it creates ngons.
That is why Triangulate Faces should be applied afterwards.
The reason I suggested it, is because overly complicated mesh (with holes) does not work well with Physics SelfMesh.
It might create some sort of abnormal partitions inside the hole preventing objects from going through.