Liquid: a few questions about flow and shape

[simmsimaging]

I did it both ways. The main point was that I got a much different flow of liquid out of the temperature when setting the liquid to "smoke". It flows much faster and smoother. I'm unable to take advantage of it though, because if the liquid channel and the effects channel don't match you get that solid rendering.

I'm trying to figure out how to get that nice smooth flow of liquid and still be able to render it.

[simmsimaging]

Here is what I am gathering from all the testing on a pour shot so far:

- The docs and online stuff I could find indicated that it didn't really matter which channel you choose for the Liquid source, but the choice of using Fuel, Temp, or Smoke as the Liquid source makes a big difference in the result/dynamics of the fluid.
- Temperature seems to provide the most "full" liquid. Fuel the most sparse and the most blobby (seemingly regardless of the settings used).
- If you try to render using one of the other (non-liquid source) channels for the Effects channel you can get a result, but it can be pretty unusable (too chunky, or in the case of Temperature it is basically a solid block that fills the Phoenix grid entirely).

If you look at the preview it really looks like the Liquid would almost need to use *all* the channels to get a nice full and smooth result. In the preview you can see the Temperature and Smoke particles interact and work around each other, so using one or the other as the effects channel makes for a lumpy/oddly shaped liquid mesh.

If I try to just use Fuel, and set Fuel as the liquid source and effects channel, I cannot seem to get a nice smooth and controllable flow of liquid. I've tried fixed SPF at all different values and nothing really looks very good.

I'm basically at a loss of where/how to push things along to get more than a simple gush of fluid from a simple emitter. I'm sure it's mostly me - help please!
/b

[Paul Oblomov]

I think first of all - phx not a particle based simulator and fluid sim part is still in beta stage Don't expect fully controlable hi-end results. Maybe it would be best to just show us the result you are trying to achieve ?

P.S. I've tried to make very complex splash with phx - well, with realflow it was faster to get what I want/in lots of cases more logical (it is not a fire/smoke simulator at the end ), but still, implicit surfaces rullez the planet. Phx is faster and can easily burn ur water So, I think just post your realflow scene, with something you like and we will try to make the same result with phx. Or at least we try

[simmsimaging]

Hey Paul -
I know it's not a particle simulator, but what else do you call them? I figure particles gets the idea across anyway

I don't have any real expectations yet for Phoenix (high or low). At this point I am trying to figure out what the expectations *should* be For sure it looks to be faster than Real Flow, but only if you can get something workable out of it...

I don't want to really spend a bunch of time setting up a Realflow file to create a target. It's not really apples-to-apples. I just want to be able to create a realistic shot of a pour into a glass. As soon as I tried to create one (rather than just use a pre-built scene like the beer.max) I ran into a lot of trouble. It's not a super specific or crazy complex effect I want (not yet anyway )

/b

[Ivaylo Katev]

The docs and online stuff I could find indicated that it didn't really matter which channel you choose for the Liquid source, but the choice of using Fuel, Temp, or Smoke as the Liquid source makes a big difference in the result/dynamics of the fluid.

well, this sounds really scaring for me , and of course i made a fast test - fortunately the result with different liquid channel was exactly the same, until the last pixel. i suppose you setup was wrong, perhaps you forgot to disable the temperature in the source and this changes the result. do not forget that not only the liquid is simulated, the air is simulated too, and its temperature does matter.
in general, pour a glass is not too hard i think
- load the "liquid" preset (not the liquid-flowing!)
- create a source
- play with the SPF value and the "surface balance" parameter (not available in 1.2) to achieve stable settled liquid.
- add a gizmo to render smooth liquid-glass interface

[simmsimaging]

I will make a video for you. I don't know what I could be doing wrong, but I get wildly different results with different sources for the fuel.

Can you provide some guidance as to how to play with the surface balance parameter that doesn't involve just shooting in the dark and not really knowing what is supposed to happen?

- I have no idea what it means to add a gizmo to render smooth liquid-glass, but that certainly sounds useful. Could you elaborate on that a bit please?

b

[simmsimaging]

Here is a video showing what I mean. First I set Liquid to use Fuel and sim a bit, then Temperature. You can see that none of the 3 channels (fuel, temp, smoke) look the same in in the two versions.

What am I doing wrong that gives this result? Advection is on classic and the SPF is fixed at 5/5

/b

http://screencast.com/t/9QuiXEI34qq

[simmsimaging]

Also: I just tried loading the liquid template - is that what you meant by preset? If so, then it does not seem to work at all. I checked "render" and "simulation" settings, loaded the preset, but nothing changed in my Phoenix settings. Is there another step, or does it only work on fresh/new Phoenix simulators or something?

b

[simmsimaging]

As per Paul's request - I made a super simple scene in Realflow. This is not an example of what I think would be "ideal" for a pour, it's just to see what the (more or less) default setup would create. The flow of liquid just seems to behave far more like liquid. Phoenix just seems way to turbulent and.... gassy? still. I am sure it's partly my settings etc, but I don't think that is all of it.

Thoughts?

b

http://screencast.com/t/acVhqvO7M

[Paul Oblomov]

Does this look like something you're looking for ?
Gizmo used and very fast settings, so the phx remains faster, than rf

[simmsimaging]

....sort of?...

Thanks for doing the test on this. It's an okay beginning but try going a bit beyond something like this to make it look more like a real liquid, and it all falls apart for me. You also appear to have some of the weird dynamic stuff going on towards the end of your simulation - what's going on there?

I'd like to see it with a more realistic input flow, which is why I was bouncing mine off a cylinder, and have it first hit the glass side and go in from there so it would be closer to a "real" beer pour shot. I find the liquid behaviour as it pools up looks a bit off too. It's too... controlled?... like it doesn't have the right bounce/splash quality or it settles down a bit too fast. Something anyway.

Still waiting for some answers from Chaos on a few things so for now the testing is basically stalled out for me.

Thanks again Paul
/b

[Paul Oblomov]

And a closeup of this relatively coarse grid Some grading added.

edit

Well, more testing - later.

[simmsimaging]

Here is a video showing what I mean. First I set Liquid to use Fuel and sim a bit, then Temperature. You can see that none of the 3 channels (fuel, temp, smoke) look the same in in the two versions.

What am I doing wrong that gives this result? Advection is on classic and the SPF is fixed at 5/5

/b

http://screencast.com/t/9QuiXEI34qq

Have you had a chance to look at this?

[Ivaylo Katev]

as i supposed somewhere in the thread, you did not change the source settings, only switched between temperature and fuel.

[simmsimaging]

? I clearly changed the source from fuel to temperature. What else would need to be changed? The directions I have been able to parse all suggest you can use any of the 3-4 channels, set to output of 1. I get different results with each.