FollowPath Settings wrong or bug?

[ben_hamburg]

Hey there, I have some complex tube following water to simulate, so I wanted to try out the new followPath helper.
But either there is a bug (just updated to latest nightly) or I don´t understand the settings:

At an influence of 0.0001 the fluid stll seems to almost completely follow the path and loose all the chaotic details.

Here without the followPath (Gravity is zero, spf about 10)



And here with the follow path at influence 0.0001:



Any ideas or do you need more info?

[Svetlin.Nikolov]

Hey,

What is the scale of your simulator (the x+y+z size from the Grid rollout)? Since the follow speed is 350cm/sec, even with a low influence, this might be too strong. Does it take the liquid the same time to stretch out as much as in both screenshots?

Cheers!

[ben_hamburg]

Yeah, it takes around 7 frames for both sims to stretch out that far. The liquid source speed is 350 cm as well, so I tried to match that, when reducing it I think it just became slower, but equally uniform. I´ll try that again.
Size of the grid is roughly 2000cmX250cmx250cm, adjusted by the scene scale parameter, which is at 10.

[George Barzinski]

Hi,
would you please attach the scene here to investigate the problem further?

[ben_hamburg]

Hey,

I´ve attached the file. I´ve done some more R&D, so its not exactly as it were, but there is still a bunch of problems with the scene.
I tried following a curved spline instead of a straight one, but the results are still quite different then from the example scene on vimeo:
https://vimeo.com/130441394

I noticed that the source is different (two sources coming at the spline at an angle, instead of one source pointing straight at the spline) than what I have and that might already contribute to the problem.
And it also seems like the spline is rotating around itself like a helix along the path, but that might just be the path follow doing its thing.

Anyway, thats what I want, and the scene is what I got, so if you could share the scene I might be able to recreate my setup with it and see if I get something different...

[George Barzinski]

Thanks, mate. Ok, I understand now. I'll check out your scene now.

Rotation is due to Rotation speed parameter.

[George Barzinski]

Well, I managed to do something similar to the video. It's really hard to explain how I've made it - it's a trial and error and previous experience. So, my guideline to you is something like that:

- Chose the right emitter shape, emit mode and discharge value (discharge is based on Follow speed, because you'll need more liquid for faster Follow Speed). These are the first very crucial steps.
- Adjust the proper Follow Speed.
- Adjust Steps per Frame and fluid's surface tension. This would give you some consistence and smoothness of the fluid or vice versa.
- Play with Rotation speed until you're happy with the spiral shape of the fluid.
- At this point the result should be something like smooth gliding spiral
- Add a Turbulence for randomness and tearing the fluid apart.
- If the fluid is going far away from the spline, then increase the Pull Speed. At this scene I set the Influence to 1 and it work fine.

FollowPathAndHelix_max2016.zip

[ben_hamburg]

Ok, thanks for looking into it, I´ll give it a try and let you know if I get stuck again tweaking it...

[ben_hamburg]

I just had time to try it out, thanks, works like a charm now.
I haven´t quite figured out, why my scene was´n t working,but I think the major difference is probably the amount of turbulence, which was like a tenth in my scene compared to yours and there was also an animated noise map in the discharge slot, so I probably just needed more going on here, so I could see the effect of the follow path force.

Currently I´m struggling a bit with upressing the simulation: Apparently when increasing the cell count, I also have to decrease the injection power, or I´m getting crazy explosions in the first couple of frames.

No idea why this should happen, I´m guessing its a numbers game..like everything with physics in it...