realistic beer foam

[omere]

how can i get access to the nightly builds?

[Svetlin.Nikolov]

You should write to support at chaosgroup dot com - they will verify your license and give you access.

[omere]

thank you very much

[omere]

testing with a nightly from last week.
i get the following results:
1. the foam is building up like soap bubles in a bathtub. how can i make the foam build up but to keep more of a flat surface.
2. the foam have this step fr some reason

[Ivaylo Katev]

can't say without the scene. the first problem is perhaps related to the b2b value

[omere]

here is a link to the archived scene:
http://we.tl/iRgzLAYoy2

[Ivaylo Katev]

what can i say until now, your scene is relatively hard to simulate, the small scale and the small bubbles requires high spf. in our sample scene the scale is not real , it's bigger exactly to avoid the extensively high spf. i tested with a control scene, there is no such a problem, the head is flat. i will attach my scene if you want to use it to check the settings

[omere]

this scene was created with the official release?
the simulator still have the liquid section.

[Ivaylo Katev]

yes, the scene is actually the sample scene, but reworked to work with the flip solver. the b2b is ten times smaller, the grid resolution is smaller, because the flip solver can handle thin geometry, and no need to simulate huge liquid particles, here are no waves, splashes and other cool liquid features, the foam is the important part. the scene is for nightly builds and uses the flip solver, this is achievable with the combined simulator object setting the advection method to flip. during the test i saw one foam issue and today will try to fix it, but except this it works fine. the main difference - your scene has correct physical scale, in contrast our scene has bigger scale than the real one. we really recommend the real scale of the simulations, but the small scales become extremely slow to simulate, and usually you have to do may simulations until the pursued result is achieved.
the second difference is the small bubble size and the lack of big bubbles. i know the real foam has very small bubbles, but if you observe the size, the bubble count is huge. anyway, if the spf is high enough, the result must be the same as with bigger scale, but you have to wait for it. i suppose the small bubble size plays a role for the shape in your case, because in the real world the smaller bubbles are also tending to produce non flat foam surface. in fact, many creamy substances are actually foams with very low bubble size. in general the "hardness" of the foam is controlled by the b2b value, that is also higher in your scene, this must be considered too, you can try to simulate with 200 like our scene.

[omere]

Thank you. i will try this and will tell you how it goes.

[omere]

can you setup the same scene but with the LiquidSim object?
i don't have your object in the nightly. only the fireSmoke, Foam and LiquidSim

[Ivaylo Katev]

the liquidsim object has the same parameters as the general purpose simulator, just the ones related to the fire simulations are removed.
you can transfer settings between simulators using the simulation presets button in the main panel.
btw, the foam issue that i mentioned above was fixed, it was related to forming a thin layer directly over the liquid with different properties that produces visible "belt" of the foam, if you meet this issue use the latest nightlies

[omere]

when i'm comparing between the two simulator i see some differences. the liquidsim doesn't have the liquid section

[Ivaylo Katev]

yes, it would be strange the liquid simulator to have liquid section, the parameters inside are controlling the liquid dynamics and were moved to the dynamics section. juts use the preset system to transfer the parameters

[omere]

did you encountered this error?