Hello, I have a question on flip simulations, with a quick movement of the glass, the liquid particles just evaporate.


How to overcome this problem? I've already tried adding substeps wherever possible.

can you share you hip file?

Try to use more sub-steps in the solver and use a better velocity extrapolation option, you can find all of this in the solver, but also is very very important that your collision geo is having correct velocities for a correct collision detection; Or slow down your coup sim to then re-time the sim again to compensate.
Unfortunately this kind of sims are a nightmare for a FLIP pipeline, I think you should try with a pop fluid pipeline if possible.

Cheers!

Alejandro

I think I found the solution for FLIP, use Transformation not deforming.
Still sorting out, please wait for me to test.

Greetings, I deal with the same problem in a more basic level, I'm trying to fill a glass with water and I'm losing volume as you can see below
How many substeps is a good starting point?

billyslol09
Greetings, I deal with the same problem in a more basic level, I'm trying to fill a glass with water and I'm losing volume as you can see below
How many substeps is a good starting point?

Hello.If your collider is not moving at high speed, then you do not need to increase substep.
You need to decrease the Grid scale.

ysysimon

billyslol09
Greetings, I deal with the same problem in a more basic level, I'm trying to fill a glass with water and I'm losing volume as you can see below
How many substeps is a good starting point?

Hello.If your collider is not moving at high speed, then you do not need to increase substep.
You need to decrease the Grid scale.

But if my geo collider moves definitely fast, I use pop fluid in basic pop solver, because flip doesn`t fit to this type of sim I think.

akvahouse

ysysimon

billyslol09
Greetings, I deal with the same problem in a more basic level, I'm trying to fill a glass with water and I'm losing volume as you can see below
How many substeps is a good starting point?

Hello.If your collider is not moving at high speed, then you do not need to increase substep.
You need to decrease the Grid scale.

But if my geo collider moves definitely fast, I use pop fluid in basic pop solver, because flip doesn`t fit to this type of sim I think.

I got a feeling that POP Fluid feels unnatural for some reason. Also you don't have all the attribute and parameters you have in flip

akvahouse

ysysimon

billyslol09
Greetings, I deal with the same problem in a more basic level, I'm trying to fill a glass with water and I'm losing volume as you can see below
How many substeps is a good starting point?

Hello.If your collider is not moving at high speed, then you do not need to increase substep.
You need to decrease the Grid scale.

But if my geo collider moves definitely fast, I use pop fluid in basic pop solver, because flip doesn`t fit to this type of sim I think.

It can fit.The trick is to use transform instead of deformation, if your colliding body is rigid body motion instead of deformation

ysysimon
better velocity extrapolation

Do you mean this transform?

Because I have an alembic from which I take a glass, and so far I haven’t succeeded with the flip.

akvahouse
transform

I ’m not in the industry, but you should find a way to tell the solver how fast the object moves,such as importing point velocity to the RBD object.If you use the Deformation, the solver simply recalculates the new collision body every frame(Imagine that the collision object is just generated every frame. it's not move at all). It will be difficult or impossible to calculate the correct collision result.

I ’m just an amateur and don’t know much about alembic,If you find solution plz tell me.
Maybe you can extract transform information from alembic file,simply check option “Use OBj Transform” has no effect, because there is no Transform on your obj at all.

ysysimon

akvahouse

ysysimon

billyslol09
Greetings, I deal with the same problem in a more basic level, I'm trying to fill a glass with water and I'm losing volume as you can see below
How many substeps is a good starting point?

Hello.If your collider is not moving at high speed, then you do not need to increase substep.
You need to decrease the Grid scale.

But if my geo collider moves definitely fast, I use pop fluid in basic pop solver, because flip doesn`t fit to this type of sim I think.

It can fit.The trick is to use transform instead of deformation, if your colliding body is rigid body motion instead of deformation

Can you share your scene?