Creating particles inside the PhoenixFD liquid

[JuhaHo]

I think the crash wasn't PhoenixFD related at all. This was just a follow up to resimulation times. Even with very low amount, the processing time for foam only resimulation seems to be over one week.

[Svetlin.Nikolov]

If possible, please send over the Phoenix log from this week long resimulation - only a few frames will show where the bottleneck it.

[JuhaHo]

If possible, please send over the Phoenix log from this week long resimulation - only a few frames will show where the bottleneck it.

Thanks Svetlin,

I see the log file being about 2 MB. It seems to contain some data, that I cannot post in public forum. Where I should email it?

[Svetlin.Nikolov]

Ah, in this case, please shoot it to support at chaosgroup dot com and the guys or gals will forward it to the dev team

Thank you so much!

[Svetlin.Nikolov]

Ohh, so from the log file it seems like there is something very complex that the simulation interacts with - by any chance, are there any very high poly or animated geometries that intersect the simulator? If yes, then is it possible to replace them with low-poly stubs to be used only for the simulation, while the high poly ones get rendered?

[JuhaHo]

That's confusing,

I use only lowres dummy geometry. Everything else is hidden and exclude hidden geometry is on. I don't much geometry during simulation. Maybe Phoenix sees something I don't. Anyway on my resimulation with very few particles, the frame time is more than one hour at the moment.

[Svetlin.Nikolov]

Yeah, something very heavy in the scene seems to be interacting with the simulation.

[JuhaHo]

So, is there something I can do to see what might be the reason for that?

[Svetlin.Nikolov]

Sure, make a backup of the scene file and start removing all nodes in the scene until it starts simulating faster, and this way you will find what node causes the problem.

[JuhaHo]

OK,

I have now only lowres dummy container geometry and the liquid itself visible. However there's little or no change on simulation speed.

[Svetlin.Nikolov]

Ok, would it be possible to send the scene over to Support and we will figure out what's wrong.

Thanks!

[JuhaHo]

Thanks,

I sent the scene to Your support by email. I hope there's enough information.

I also included there a feature request. Is it possible to add feature where I could set up a desired liquid level where Surface force would be automatically adjusted to create constant flow through the system. It was not easy to iterate the Outgoing velocity parameter to match the liquid leaving the simulation, so there would not be big changes on the surface level during the simulation.

[Svetlin.Nikolov]

Thank you! Will get back to you with more info once we know more...

[JuhaHo]

During the holidays I created a new version of my Phoenix configuration. I use now a birth volume for the foam; it's less than 10 % of the simulation volume. Foam amount is set to 0.5 and birth threshold to the default 50.0. I could now simulate a little less then three quarters of whole simulation. The cache size per frame is now a little less than 1 GB. The simulation had crashed last night and if I move to the last simulated frame Max crashes immediately. Maybe I should reduce the foam amount and birth volume further more.

Any thoughts on that?

[georgi.zhekov]

Can you send over the crashing scene along with the Phoenix log so that we can check what is going on?