Fire with PF as source

[lander]

Creating some tool for spreading fire.
I made a PF system as source for a spreading fire and when I preview the fire I can see what looks like a gas leak where the PF source is located.

The Phoenix system has two sources
1. uses the blue particles as source for smoke with a lower temperature
2. uses the small purple particles as fire with a higher burn temperature

It looks ok where the particles are but it looks as if PhoenixFD system is using the icon from the PF system as a source with a massive emission?
Note that the entire PF system spreads away from the inital emission where the Fire is absolutely furious.

https://vimeo.com/270328972 Source particles
https://vimeo.com/270328992 Phoenix Preview
https://vimeo.com/270331912 Entire PF system

using Max2016 and 3.10.01 Nightly, Build ID: 2018030828105

[Svetlin.Nikolov]

Hey,

Could we see the settings you are using? Most importantly, the Simulation, Grid, Dynamics rollouts of the sim, and the Source settings.

Thanks!

[lander]

I'll send the file to support@chaosgroup.com

[Svetlin.Nikolov]

Hey,

Got the scene! So when you start it, you get a warning that the simulation will be affected by the Gravity force. This means that fire will rise up faster than usual and it will keep creating higher and higher velocities each frame, causing the flame burst you see.

You just need to exclude the force from the Interaction rollout and it will be okay

Cheers!

[lander]

Nope, that was not the issue. If you look at the vimeo files there are two distinctly different areas of the fire simultation. Where it's most furious is where there should be no fire at all when compared to the particles.

[Svetlin.Nikolov]

Hey,

I am talking about this process here:



There is constantly increasing velocity at the border of the grid that is caught in an infinite feedback process. Do you still get that if you exclude the gravity so it won't feed more and more velocity in the jet?

[lander]

Yes, I have excluded gravity from the simulation and the jet is still there. You don't get that on your end?

[Svetlin.Nikolov]

Hmmm, I think I know why I don't get the same result here - the Buffered conservation is the only one that produces different results on different number of cores. How many core do you simulate on?

Cheers!

[lander]

It's an old Dell Precision 3620, Xeon CPU-E3-1225 v5 @3.3GHz, 64GB RAM.

[lander]

I have now had the time to play and no matter what conservation method I use I still get the same result. But you're saying you don't get the same independent fuel source as I? I have also moved the PF source away from the sim and started the emission from a different place but I still get that well of energy. Or maybe (probably) I'm just doing something wrong.

[Svetlin.Nikolov]

Hey,

I've run the sim on all machines I can access, but it seems like I need to hand it over to the QA guys so they can try their magic on it on Monday - it won't make this jet appear. Asking just in case - since we did a lot of changes to geometry and particle interaction since 3.10, can you try a very latest nightly and see if there is some change in behavior on your end? Also, sorry for asking, but are you absolutely positive you are running the exact same scene as we are?

Cheers!

[lander]

I will double check both file and with latest nightly build. There is no rush really. It's just tests on my side.

[Svetlin.Nikolov]

Hey, so far we haven't been able to reproduce this here...

[lander]

Good thing staying away from the computer. I downloaded the latest nightly and the problem is solved. Not sure when but it looks great now.

[Svetlin.Nikolov]

Great news! Thanks for getting back to us for this!