Page MenuHome

Mantaflow Liquid - Gravity not respecting Time Scale from domain.
Closed, ResolvedPublicBUG

Description

System Information
Operating system: Linux-5.4.0-29-generic-x86_64-with-debian-bullseye-sid 64 Bits
Graphics card: GeForce RTX 2080 SUPER/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 440.64

Blender Version
Broken: version: 2.90 (sub 0), branch: master, commit date: 2020-05-04 16:18, hash: rB375a50dad9c5
Worked: (newest version of Blender that worked as expected)

Short description of error
It seems to me that gravity is not respecting the time scale setting from the liquid domain. The liquid is freezing but the gravity keeps acting over the fluid at the normal speed...

Exact steps for others to reproduce the error
Just playback the attached file and see what I mean. The time scale is animated from default 1.0 to 0.1.

Event Timeline

Philipp Oeser (lichtwerk) changed the task status from Needs Triage to Confirmed.May 5 2020, 10:58 AM

Not sure how this is supposed to work (observed behavior seems unpredictable to me as well...)

@Sebastián Barschkis (sebbas): mind checking? (will confirm for now...)

The expected behavior is the same as in the SPH Particles example here:


Note that the timestep starts with default value and is animated to a lower value, slowing down the particle simulation AND also takes into account the gravity and forces influences. Note that the velocity color of the particles doesn't change when the timestep is slowing down, the velocity remains the same but the timestep is slower...

Force Fields are not working also...
Note in the attached file that right when the timestep is slowing down the liquid particles starts to go crazy, their velocity color goes red immediately...
The above mentioned SPH particles behavior takes into account the force fields and gravity as expected, keeping the motion of the particles and only slowing down the timestep.

Philipp Oeser (lichtwerk) changed the subtype of this task from "Report" to "Bug".May 6 2020, 10:55 AM

Great work! Thank you SO MUCH! You are an awesome team! But special thanks to sebbas! Thank you! thank you! thank you!

Please, check the blend file attached in the comment that I said about Force Fields... after checking this problem with forcefields I see the same problem still happening...

Now there's an even stranger thing happening... see the attached file, please. Also, after checking the file please test the minimum timestep value of the domain, default is 1 change it to 4(same value as maximum) and the result seems to be slower with time scale 1...

Thanks, @Sebastián Barschkis (sebbas) I just compiled blender now after your additional fix, and everything related to this report is working fine now! The slowdown in time scale is taking gravity and forcefields into account. THANK YOU!
There's something weird already with forcefields but I will open a separate report. The animated "Max Distance" under "Fallof" are affecting particles(SPH) and liquid(MANTAFLOW) even outside of radius scale, there's full influence inside the Max Distance radius but outside of the radius there's still a little of influence, it's not totally 0.

@Everton Schneider (eversimo) Ok, great! Yes, a new report would appropriate for that!

Is there a way to simulate the old behavior ? eg. divide gravity by the time scale? (edit: <= almost but not quite) to allow older blends to be edited in 2.9 without a complete rework.

e.g., this relied on the bug. https://www.youtube.com/watch?v=Rc5-R12XZFk

of course going forward i'll use the new behavior