Physx fluids - Viscosity and Stiffness

[Bullit]

I have problem filling a cup, it is to slow filling it with water, initially it seems to go okay, but then it slows down while the same number of particles are being emitted. I nailed it to the stiffness, it seems value of 1 the particles still are compressed. So the liquid don't grow how it should- water doesn't compress.
But if i increase the stiffness the simulation doesn't behave so well. Any ideas?

[xsi_fanatic]

I have problem filling a cup, it is to slow filling it with water, initially it seems to go okay, but then it slows down while the same number of particles are being emitted. I nailed it to the stiffness, it seems value of 1 the particles still are compressed. So the liquid don't grow how it should- water doesn't compress.
But if i increase the stiffness the simulation doesn't behave so well. Any ideas?

Personally I don't like using the Lagoa fluid. It's inaccurate. You're better off using Realflow or SPH Fluid plugin for Softimage. But getting it is another quest on its own. I don't even know where to begin lookin.

[Bullit]

I am using Physx implementation from Mr.Core , member here in the forum,, isn't it based on SPH?

[Tekano]

it is sph but on the gpu by nvidia , i think stiffness is like a friction force so could help but not really for water. surely it is not actual radius or collision radious you want to be increasing to see some effect?

[Bullit]

Thanks

I mean when the stiffness is at 1 appears more watery than higher values, but at this value the fluid is compressible so takes a lot of time to fill a cup.

Radious and colision radius are not in ppg, do i found it inside compounds? and how they can help me when the issue is compressibility?

[Bullit]

Btw another thing, what is the best "scale" to work with? should i make a cup with 4 units diameter ( i think this means 40cm) or bigger or tinier?

[druitre]

First off, always assume 1 SI unit is 10 cm, for everything simulation in Softimage. Then twiddle and tweak with PhysX parameters untill you get something recognizable, mark settings, twiddle untill it's something else that's recognizable etc. In other words, calibrate. Porridge, water, honey, mud etc. Try and find relation between parameters this way. As everything is relative to each other there's no real objective 'best' settings, just settings that work well under some conditions.

Parameter with huge influence in my experience is particle size (just the basic one in your emitter), that decides the volume occupied by the particle and thus the compressability. Mass, on the other hand, seems to be ignored completely.

Thingthat bothers me the most is that PhysX in Ice does not correctly calculate deforming obstacles, so is useless for many situations. Too bad, I absolutely love what mr.Core has done.

[Bullit]

I am now in another issue, it seems that liquid inside bottle doesn't work if the bottle is moving. I can only get results if it is static.

[druitre]

Yeah, that's what I mean by 'incorrectly calculating dynamic obstacles': it's doing a concave mesh, not an actual geometry calc. Don't know if that is a PhysX restriction or rather in the Ice implementation.

[Bullit]

Thanks druitre. MrCore advised me to try his Plutonia fluid system instead of Physx. At first look it seems to work.

[druitre]

Plutonia? I'll have to check that, what is it?

[Bullit]

http://softimage.ru/forums/index.php?showtopic=9685

I fyou need any help just PM me.