Hello everybody,
i am facing a problem with my liquid simulation.
All i am doing is having one flip-fluid object colliding with an animated geometry.
The problem is that as my geometry deforms itself ( scale ) the liquid intersects it
instead of colliding with it.
I have tried both static and rbd solver to my geo.
I ve attached a still bellow showing wireframes of both the liquid and the intersected geometry.
(Yellow is the liquid)
Any ideas would be great?
Thnx guys!!
yannis
Fluid Intersection Problem ??
2792 4 0- yanioaioan
- Member
- 31 posts
- Joined: Dec. 2011
- Offline
- ragupasta
- Member
- 349 posts
- Joined: Aug. 2006
- Offline
Also check the collision tab on your static/rbd, and make sure the collision geometry has no holes in it. Turn off display geometry, and under collisions turn on “show collision guide geometry”. If there is any holes, then increase the divisions to close them up.
Also remember to make sure your static/rbd has “use deforming geometry” toggle enabled. As scaling is happening the volume collision rep needs recalculating per frame.
I would assume that the fluid is moving quickly, if this is the case then the speed may be negating the integer values the dopsim is calculating. Try upping the substeps on the dynamics node at object level to anywhere around 3<–>5.
Calculating more substeps might catch the collisions before they pass through the mesh.
Also remember to make sure your static/rbd has “use deforming geometry” toggle enabled. As scaling is happening the volume collision rep needs recalculating per frame.
I would assume that the fluid is moving quickly, if this is the case then the speed may be negating the integer values the dopsim is calculating. Try upping the substeps on the dynamics node at object level to anywhere around 3<–>5.
Calculating more substeps might catch the collisions before they pass through the mesh.
- yanioaioan
- Member
- 31 posts
- Joined: Dec. 2011
- Offline
- yanioaioan
- Member
- 31 posts
- Joined: Dec. 2011
- Offline
- ragupasta
- Member
- 349 posts
- Joined: Aug. 2006
- Offline
Increasing the substeps does increase the time it takes to simulate (if you increase it to 2, it will calculate twice per frame instead of once. 3 times will be 3 calculations per frame, ect, ect). You need to find a substep as low as you can before the collisions fail. Reduce the collision guide geometry samples to as low as you can before it starts to create holes, and increase the particle seperation parameter to reduce fluid particles.
It's a careful balancing act between these values, so it's try and tweak values, then try again and tweak more values.
It's a careful balancing act between these values, so it's try and tweak values, then try again and tweak more values.
-
- Quick Links