More detail to small scale FLIP

[venator11387]

I've been learning some FLIP fluids from a small tutorial from 3D Artist. Due to the sake of time and speed, I deviated from the lesson by lowering the scale of the geometry from 1 to .4, so as a result there are changes in detail. I attached two links, one of a Flipbook of my scene and the other is a video of the tutorial FLIP example.

Are there any ways to add more detail to FLIP fluids on a smaller scale, like a sort of upres for it? What kind of attributes could I use to create those nice tendrils, less-rounded shapes, and breakup like in the tutorial example?

My scene:

Tutorial Scene:

http://www.youtube.com/watch?v=qCLWFRwf7ak

Edited December 23, 2012 by venator11387

[Fabiano Berlim]

To get more detail you need to raise the resolution. You can also change your Grid Scale to 1, but normally the default is very good. In the Volume Motion tab you can change Smoothing, if set to one it will only use the volume velocity and zero particle velocity. That can change a lot the behavior. But none of that will give what you want.

The key thing is resolution and patience.

Run the simulation always checking your memory and try to get the most of it. Don't ever explode your memory, it will start to send data to the swap, slowing down the process in Linux and probably crashing the system in Windows.

When doing your final simulation, try to turn off the cache simulation. Even more important, turn off the cache to disk. Because if your temp is not big enough, the simulation will stop in the middle.

To get good large scale great looking simulations you will need to wait for the computer. I'm talking about 2 to 20 hours. There's nothing you can do. Don't let anyone touch the computer and go watch a movie or something.

[venator11387]

Don't let anyone touch the computer and go watch a movie or something

In most cases I use that time to catch up on sleep Yeah I figured I would have to raise the resolution more by lowering the particle separation. The problem is that when I did that the particles would collapse or compress into themselves. Instead of the particles maintaining the volume level of filling the pool, it would "sink" into the geometry or would go through the geometry even after I put the uniform division and thickness VERY high. In another test, where the FLIP particles are falling into a bowl, instead of pushing outward and filling the bowl they would once again collapse (and in some cases go through the geo again). I'm just wondering what I can do to counteract this kind of weird behavior?