flip How to fix flip particle trembling at collision volume junction

[woohhh]

Hi, I'd like to set up a simulation where there's a small container, and half of it will be filled with liquid. The initial state of the liquid should be static, like a cup of water left on the table for a long time. Then I'll introduce a rigid object to collide with the liquid. Sounds simple, right? But what's puzzling me is how to achieve the static state of the water before the impact of the rigid object. I created the container and converted the lower half of its volume into flip particles, then applied gravity. What happened next was that the behavior of the particles at the collision volume junction wasn't as normal as expected. They kept jittering, and as the simulation continued, the jittering became even more intense. Whether it's what I expected or what happens in the real world, the water should become more and more static as time goes by. However, with gravity turned on in the DOPNET, I can't make them completely calm.

I found that if I adjust the size of the flip solver volume limits to replace the role of the container and turn on the flip object's closed boundary option, the particles at the boundary junction don't jitter as much. As time passes, the performance of all fluid particles becomes increasingly calm. This is exactly the effect I'm looking for. However, this method only works with cube-shaped containers. If I use a spherical or tubular container, it doesn't solve the problem.

So, how can we truly eliminate the jittering of flip particles at the collision volume junction?  Below is my .hip file. Thank you all very much.

 

TANKtest.hip

Edited October 22, 2023 by woohhh