Creating particles inside the PhoenixFD liquid

[JuhaHo]

I think You already have that scene. It crashes at frame 1462. It also crashes when I restart Max, open the scene and load the last simulated frame 1462.

[Svetlin.Nikolov]

Hey, just an update on the simulation speed issue - I found that the Grid Texture in the scene might be one of the reasons it's so slow even without having any complicated geometry. Can you check if disconnecting the Grid Texture from the "Keitos" material's diffuse slot (just for the duration of the simulation) makes it better?

[JuhaHo]

Thanks,

I disabled the Grid texture, and it seems to be a little faster 1m 52s per frame in foam only resimulation. I'm using Foam amount of 0,25 and Birth treshold 100. I also have an animated birth volume to reduce the total foam amount.

[Svetlin.Nikolov]

Hmmm, I think the animated birth volume shouldn't be a problem speed-wise - the animated objects that collide with the fluid are the ones to watch out for...

[JuhaHo]

The birth volume is located at the liquid surface level. I wanted to reduce the height of the birth volume, so I animated it according the liquid level.

[JuhaHo]

I managed to get the Foam Resimulation finished. Now trying to reduce the size of the tyFlow cache files. At the moment I'm getting 1,2 GB/frame files and running out of HDD space on my simulation workstation. What would be the variable to start with?

[georgi.zhekov]

Hey,

If it comes to the tyflow cache settings - maybe the tyflow people can help.

[JuhaHo]

I think it's all about getting the foam right in resimulation. At the moment the foam could be better as well as the resimulation speed. It took weeks to get it through once only to see, it could be better. TyFlow is baking what it sees.

I have used foam amount of 0,25 with birth treshold of 100. The foam mattress is too thick, there's too much foam and it's located partly in a wrong place. Anyway, I try to get at least one tyFlow caching through and after that I'll iterate the Phoenix foam settings once more.

[georgi.zhekov]

Then how about you run the sim with a very low foam count, just to check how it behaves and if you like the look and then increase it? You can increase the threshold or reduce the foam amount to somethin like 0.01.
The more foam you have the slower the sim will be.

[JuhaHo]

I finally got some results. I used foam amount of 0.01 as suggested and treshold 50. In tyFlow input particles was foam only and 100%. There's still maybe too much particles and some of them are out of container. I'll investigate that next.

Edit: I also attached a rendering of the foam component. Maybe I should make the Birth value even smaller, since the tyFlow particles are escaping out of the container through the collision object.

[JuhaHo]

I have another one.

In this system the particles and liquid are coming from separate ducting and will be mixed in the container. I don't know what I'm not doing right. When enabling the Fluid force operator, the particles are losing the velocity and dropping down as soon as they enter to Phoenix simulator.

[JuhaHo]

Here's a new approach. Main issue seems to be, that particles are flying out of the container.

[JuhaHo]

I simply scaled the scene up by 10. It was instantly better. However the container is still leaking particles.

[georgi.zhekov]

Hey,

If you wish that there are no particles outside of the grid, maybe you can create a dummy geometry with the size of the grid and delete all the particles that get outside of it using a Surface test.

[JuhaHo]

Hey,

If you wish that there are no particles outside of the grid, maybe you can create a dummy geometry with the size of the grid and delete all the particles that get outside of it using a Surface test.

Yes, but before that I try to minimize the amount of those particles leaving the grid. When I'm done I will transfer the settings of this test scene to the actual project scene. Here's my latest test rendering.