like always they are a strong technical explination but for the moment i don't understand why…
why when the emitter is into the fluid, the amount of fluid reduce drastically?
how i can fill a volume gradully ?
https://vimeo.com/93061136 [vimeo.com]

Are the particles count the same in both liquid sim ?

(au fait tu as bossé à Chez Eddy ? )

this is my problem… everything is identical between the both simulations. the only one parameter who change is the emitter height.
(oui pour une seule et unqiue prod )

just a wild guess. but maybe it has to do with the reseeding. the fluid on the right has more velocity, therefore more separation on collision. and then the reseeding emits more points.

This is because of the way the default flip setup handles sourcing. In essence it does not “push” fluid into the emitter, it just copies fluid into the volume of the emitter, and more specifically the part of the emitter that do not already have fluid.

In your case this obviously does not work as you would intend, although in a general case it is a good setup.

Attached is a modified scene. I just added some velocity and divergence to the source, so it “pushes” or “pumps” the fluid out as it is sourced.

For some reason the divergence seems to have a proportionally larger effect the higher the water level rises, I'm not 100% sure why. Probably something to do with the greater pressure the source is under, and how the divergence enforcing outwards flow ends up causing a larger force.

yeah!
thanx a lot it work verywheel

eetu is spot on. You need to create some initial velocity on your emitter in order to push them outward to leave room for new particles. Think of this as initial “pressure”. If you have a hose with water, you can add initial velocity to your particles as that initial pressure to emit the fluid.


In order to have the fluid actually push out, you need to enable “Apply Particle Separation” and then adjust the separation rate. With these settings you “can” get the particles to push outward but this incurs a big overhead and is not the way you want to set up your FLIP simulations in general but for specialized sims, it is there to try to push the particles apart that are packed denser than the given simulation particle separation setting.

FLIP simulations have an interesting behaviour where if you have two particles that are closer together than the set particle separation, you can have the two particles tag together and not separate through out the simulation adding not much at all but overhead. This is why you may need to add the particle separation expensive step.

FLIP simulations are tricky to set up and keep going. Without certain safe-guards in place, it is quite possible and quite likely to have your particles too dense for the simulation. It is for this reason that many safe guards have been added to carefully add the particles in to the simulation at a fixed density set by the simulation particle separation.

You can dig and find out the actual nodes set up to do this as there are no black boxes in FLIP. Then defeat everything and do what you want. But it's buyer beware at this point and you need to know what you are getting in to.

—-

Fluid Source SOP creates the particles
The particles are actually being created by the original object's Fluid Source SOP. Look at the Particles tab and you will see that the particle density is referencing the simulations particle separation.
You can break the nice ordered particle density by increasing the density of the points right here. Note that increasing the fluid density parameter in the sim doesn't add any new particles. It is here where you can break the designed relationship.

Sink Logic in SOP Solver called “sink_particles”
If you dive inside the Source Volume DOP in your DOP network with the FLIP simulation, look for the SOP Solver called “sink_particles”. Unlocking the Source Volume DOP asset and then bypass this SOP Solver DOP will remove the logic to cull particles that are birthed in to existing density. That's exactly what this micro-solver is doing. It fetches the density from the sim and uses it as a group to cull any net new particles you are trying to add.

Again this is to ensure that you aren't birthing particles that go beyond the particle separation.


—-
By constructing these safe-guards and ensuring the fluid is emitted in at the particle separation rate, not too dense or too sparse, you can turn off the expensive Particle Separation step and let the pressure projection on the fields do their work and project the particles forward through the sim.

This will wreak havoc on any of the predefined OceanFX tank simulations. When would you even consider these steps? I don't know off hand but perhaps a specialized simulation with contained emitters to create a non-realistic fluid sim… Nothing large scale for sure as you need to enable the Particle separation logic which does incur overhead and can lead to other issues.