How to avoid particles getting inside of objects?

[victorwol]

So I'm running quite a few tests with different sizes of cells, only my third day learning this, but I can't seems to find a way to keep particles from getting inside of the objects, and I think this is crucial for any success, I do not see anything what let me adjust it, and the problem with this is like if I splash and object, some particles are left in there hanging from the object....

Thanks

[Ivaylo Katev]

can you show an image?

[victorwol]

yes.... not only that, I'll post the movie, there are several problems I can see.

water keeps dripping from inside of the emerging object like if it was faucet.

there is like a current where you can see water is clearly going to the left at high speed

you can see particles stuck in between the walls of the object.

the right side borders of the simulation are way to visible, and they change the height with the waves but they get stuck at a certain height. I know my box is small, I need to make it wider, but why they would stay stuck and why I have that current?

also I can clearly see the slopes of different heights of the water, that is kind of unusable as it is.

When I calculated this in low resolution, just a few cells, it looked completely different, so how can use design an effect and then when you are happy with it, up the resolution if the effect is going to be totally different. For example, on the low resolution, I didn't see any slopes on the surface of the water, on the high resolution one I can clearly see steps.

Thanks for any help you can provide. I do really appreciate it.

I will provide the scene too in a few minutes.

Thanks

[victorwol]

experimentII.zip

Here is the scene without the cache, with the cache is like 20 Gigs, so I don't think you want to download such thing? but if you want or can, I'll just put it on my Dropbox.


Thanks

[victorwol]

I have being trying different settings and I always get particles inside of the objects and they drip and never stop, I have even rendered a scene where I have made the object that comes out of the water transparent, and you can clearly see how there are particles inside that never stop emitting water down, also the way it behaves is very weird, instead of clam down over time, things get more agitated when an object is close to the surface, but attraction is set to zero, so why a sphere being close to the surface will disturb the surface that way?

[Ivaylo Katev]

now it's perfectly clear what you mean and i can say what is wrong
our liquid solver is grid based and this means the geometry is represented as voxels. the simulator just "sees" the object like it is made of lego blocks. this leads to very annoying problems, like the one you have.
there is a solution, so you have to do change certain settings
1. the voxelization mode must be turned to inscribed, by default it is circumscribed to avoid leaks through thin geometry
2. turn the wetting off.
3. turn on the "clear inside" option. both the options you can find in the right click menu of the torus.
normally i would recommend to use the flip solver (it is particle based), that is accessible in the nightly builds, but unfortunately it can't achieve perfectly flat surface, like the grid based, and the instruments that will correct this are still not implemented.

the second issue is a flow that arises from nothing. this is a known issue that is related to the smooth conservation type. try to change it to symmetric, if the symmetric produce not as smooth surface as you need, you can try to increase the spf and to change the boundary conditions. both of them are related to this issue. i know that you already need jammed boundaries in order to avoid the edge problem, but i still believe that you can resolve it just by fine ocean level tuning.
in general, almost no user is able to achieve the pursued result from the very beginning, and your case is especially heavy. most of the users rendering in ocean mode are making ships, oil platforms, and other like this, and their task is much much easier.

[victorwol]

Well.... I wonder if I got the wrong software for the job.... sounds like I should probably gone with RealFlow and not with this one. Although more complicated to use and interface with MAX, it seems it have the right tools for the job.

The issues I see with the solutions you propose are, if I turn wetting off, the object will arise from the water and have no water dripping from it, will look unnatural, I have tested it already, looks like one of those hydrophobe surfaces that reject water.

I will give it a quick try to what you said, but I'm afraid turning off wetting will kill the effect i'm looking for. And I'm sort of surprised to see that there is no individual per object friction setting to adjust the level of wetting, is either ON or OFF... and leather wouldnot have the same wetting effect than steel....

[Ivaylo Katev]

really particle flow deals better in situations like yours, however you will get another problems with rf, related to the perfect flat infinite surface. we started as fire simulator, not intended for liquids, that's why we are developing a particle based solver especially for liquids.

[victorwol]

I've done what you suggest, I can only find the Voxelization mode on the object right click, I think that is the place right? And turned off wetting. I still have water dripping down the inside of the object like before. Nothing changed.

FYI, my interaction mode was set to inscribed already on the examples I have rendered and uploaded. May be this just can't do it?

I was given access to the nightly versions, but I can't seems to find the Flip solver, the software still looks the same.

[victorwol]

Never mind... I just found where the FLIP thing is, I though it was a totally different thing without the voxels grid.

[victorwol]

this is totally confusing me now, I installed the latest build from today, and all with default settings, if I turn ON liquids and set the initial setup to 50%, the things goes nuts and looks like it is boiling, even with the old solver...



here is a link to the MAX file

https://www.dropbox.com/s/ze3akmms2b...itial.max?dl=0

[victorwol]

if I load an old scene it works as expected, but I create one from scratch and I get all this crazy thing going on.... newbie problem or bug?

By the way, please as communication per text we can see our faces, let me tell you , I'm frustrated to not be able to get what I want, but delighted with the help and fast responses. Great support, don't let my frustration make you think I there is a bad attitude from me please. Not my language either...

[victorwol]

Well... after a lot of testing, I do really need the wetting to get the effect I need. without the wetting the object comes out of the water practically not doing anything to it, there is almost no splash, is like the object had no drag at all, and look totally unrealistic. With the wetting ON, I get a much better effect, the problem is that the object keeps dripping forever like if there was a source of water inside, it never ends, there has to be something wrong in the way that is calculated. I get it that the system is subdividing the space, and some particles might stay inside of the object, but why just don't let them fall? or kill them? how can it become an infinite source of liquid coming down? That must be an error or an overlook into the fluid engine.

I tested the FLIP, it is unusable, all I get looks like boiling water, and when I turn ON ocean, if FLIP is selected, the render is just garbage.

So close to get what I want... so close.... but I can't without the wetting, it just looks unreal....

[Svetlin.Nikolov]

Hello,

Do you really mean the physical size of your container to be 17x17x9 cm? I'd suggest you increase your unit scale, so your simulations will be much more stable.

[victorwol]

No, it is about 4mts by 4 mts, and have tried all kind of sizes, particles still get stuck inside of the object and that ruins the effect, since objects keeps dripping forever. And not using the wetting effect makes it looks totally fake. And, have even tried a 100 mts by 100 mts, and then liquid moves so slow that feels like chocolate.