Mantaflow Initial velocity not working PROPERLY #84001

New Issue

Vitor Druzian · 2020-12-21T00:18:17+01:00

Vitor Druzian commented

2020-12-21 00:18:17 +01:00

System Information
Operating system: Windows 10 Pro
Graphics card: Geforce GTX 1070

Blender Version
Broken: 2.91.0, 0f45cab862, master, 2020-12-20
Worked: 2.82.7

Short description of error
In blender 2.91.0, mantaflow working with particles: The option of "Initial velocity" it's bearly doing nothing, if anything; even with the max setting (100). In 2.82 setting initial velocity to 1 was tremendoustly more agressive than the new version

Exact steps for others to reproduce the error
Plane with "Fluid" as Physics Properties; "Flow Source" as Particle System; "Initial velocity" active (any value)

**System Information** Operating system: Windows 10 Pro Graphics card: Geforce GTX 1070 **Blender Version** Broken: 2.91.0, 0f45cab862b8, master, 2020-12-20 Worked: 2.82.7 **Short description of error** In blender 2.91.0, mantaflow working with particles: The option of "Initial velocity" it's bearly doing nothing, if anything; even with the max setting (100). In 2.82 setting initial velocity to 1 was tremendoustly more agressive than the new version **Exact steps for others to reproduce the error** Plane with "Fluid" as Physics Properties; "Flow Source" as Particle System; "Initial velocity" active (any value)

Vitor Druzian commented

2020-12-21 00:18:17 +01:00

Added subscriber: @Druzian

Robert Guetzkow commented

2020-12-21 11:00:53 +01:00

Added subscriber: @rjg

Robert Guetzkow commented

2020-12-21 11:00:53 +01:00

Changed status from 'Needs Triage' to: 'Needs User Info'

Robert Guetzkow commented

2020-12-21 11:00:53 +01:00

The velocity of the particle system appear to influence the smoke just fine. Does the file attached below work for you?

T84001_smoke_2.blend

If this works properly for you, please provide more detailed instructions and a minimal project that demonstrates when it doesn't work as intended.

The velocity of the particle system appear to influence the smoke just fine. Does the file attached below work for you? [T84001_smoke_2.blend](https://archive.blender.org/developer/F9521373/T84001_smoke_2.blend) If this works properly for you, please provide more detailed instructions and a *minimal* project that demonstrates when it doesn't work as intended.

Vitor Druzian commented

2020-12-21 17:54:58 +01:00

See, in older vesions of Blender, Initial velocity has a lot more influence in smoke than at 2.91.0. Not saying that in newer versions the option has no influence, but it have an almost insiginificant one. Follow the image.

![Mantaflow Initial velocity.jpg](https://archive.blender.org/developer/F9522145/Mantaflow_Initial_velocity.jpg) See, in older vesions of Blender, Initial velocity has a lot more influence in smoke than at 2.91.0. Not saying that in newer versions the option has no influence, but it have an almost insiginificant one. Follow the image.

Vitor Druzian commented

2020-12-23 04:10:11 +01:00

Changed status from 'Needs User Info' to: 'Needs Triage'

Fernando Alcala commented

2020-12-23 06:52:41 +01:00

Added subscriber: @Fernando-Alcala

Fernando Alcala commented

2020-12-23 06:52:41 +01:00

Here I compare between 0 and 100 (max) for the initial velocity. There's very little difference in using the maximum value.
230210001-0013.mp4

Here I compare between 0 and 100 (max) for the initial velocity. There's very little difference in using the maximum value. [230210001-0013.mp4](https://archive.blender.org/developer/F9526250/230210001-0013.mp4)

Robert Guetzkow commented

2020-12-23 10:08:28 +01:00

Added subscriber: @sebbas

Robert Guetzkow commented

2020-12-23 10:08:28 +01:00

If I setup two separate simulations with different particle systems that have the same parameters except for the Initial Velocity, use bake type All and bake them, the resulting simulation is significantly different for me. Perhaps this is more of an issue with a cache not being invalidated (e.g. baking it once with a low initial velocity, modifying the value and replaying it is not updating the simulation)?

I've tested this in both 2.91.0 and 2.92.0.

T84001_smoke_5.blend

@sebbas you might want to take another look at this as well.

If I setup two separate simulations with different particle systems that have the same parameters except for the *Initial Velocity*, use bake type *All* and bake them, the resulting simulation is significantly different for me. Perhaps this is more of an issue with a cache not being invalidated (e.g. baking it once with a low initial velocity, modifying the value and replaying it is not updating the simulation)? I've tested this in both 2.91.0 and 2.92.0. [T84001_smoke_5.blend](https://archive.blender.org/developer/F9526579/T84001_smoke_5.blend) ![initial_velocity.png](https://archive.blender.org/developer/F9526581/initial_velocity.png) @sebbas you might want to take another look at this as well.

Fernando Alcala commented

2020-12-30 00:09:59 +01:00

With respect to particles, initial velocity source does change the simulation, but at an imperceptible level compared to using meshes.
no work0001-0048.mp4

Here I have both participating in the same simulation with the only difference being that one is a particle and one is a mesh. This is the first frame of the sim, source 100.
{F9536494, size=full}

And while I'm here, initial velocity normal doesn't seem to do anything useful.
normals brokeded0001-0024.mp4

With respect to particles, initial velocity *source* does change the simulation, but at an imperceptible level compared to using meshes. [no work0001-0048.mp4](https://archive.blender.org/developer/F9536493/no_work0001-0048.mp4) Here I have both participating in the same simulation with the only difference being that one is a particle and one is a mesh. This is the first frame of the sim, source 100. {[F9536494](https://archive.blender.org/developer/F9536494/p_vs_obj.png), size=full} And while I'm here, initial velocity *normal* doesn't seem to do anything useful. [normals brokeded0001-0024.mp4](https://archive.blender.org/developer/F9536503/normals_brokeded0001-0024.mp4)

Philipp Oeser commented

2021-02-09 12:50:07 +01:00

Added subscriber: @lichtwerk

Philipp Oeser commented

2021-02-09 12:50:07 +01:00

Changed status from 'Needs Triage' to: 'Needs User Info'

Philipp Oeser commented

2021-02-09 12:50:07 +01:00

Having trouble reproducing (at least the normal case).

@Fernando-Alcala: is this file working for you?
#84001.blend

#84001.webm

If so, please share the file thaat is failing for you.

Having trouble reproducing (at least the `normal` case). @Fernando-Alcala: is this file working for you? [#84001.blend](https://archive.blender.org/developer/F9748525/T84001.blend) [#84001.webm](https://archive.blender.org/developer/F9748547/T84001.webm) If so, please share the file thaat is failing for you.

Fernando Alcala commented

2021-02-09 23:04:30 +01:00

The "issue" that I had was that the size of the simulation was too small for the timestep size. Normal 50 in meter sized sims works fine:

But when you have centimeter sized sims, you need to increase the timestepping:

The "issue" that I had was that the size of the simulation was too small for the timestep size. *Normal 50* in meter sized sims works fine: ![no more.png](https://archive.blender.org/developer/F9756581/no_more.png) But when you have centimeter sized sims, you need to increase the timestepping: ![yer more.png](https://archive.blender.org/developer/F9756802/yer_more.png)

Robert Guetzkow commented

2021-02-20 23:38:45 +01:00

@Fernando-Alcala @Druzian Am I understanding you correctly that this is considered resolved then?

Fernando Alcala commented

2021-02-22 08:50:26 +01:00

Changed status from 'Needs User Info' to: 'Archived'

Fernando Alcala closed this issue

2021-02-22 08:50:26 +01:00

Fernando Alcala commented

2021-02-22 08:50:26 +01:00

It's a limitation.

Vitor Druzian commented

2021-02-24 20:12:30 +01:00

Fire Initial velocity issue.blend
Initial Velocity.mp4

I hope, from the bottom of my heart, that this time will be clear what I'm saying! No excuses, I have the blend file and a video showing whats the issue. Something changed in smoke simulations and I think is only when usining the INITIAL VELOCITY option. If this change was intentionaly, wasn't in the realease notes! But i suppose that it is not intentionaly once the fire and smoke simulation looks horrible compared with the last version of Blender!
Here I annex a blend file that was simulated at 2.91 and 2.82 AT THE SAME SETTINGS WITH A PROPER SIZE! If you don't belive me, download it and try it yourself!

the last time,

Vitor Druzian

[Fire Initial velocity issue.blend](https://archive.blender.org/developer/F9835719/Fire_Initial_velocity_issue.blend) [Initial Velocity.mp4](https://archive.blender.org/developer/F9835717/Initial_Velocity.mp4) I hope, from the bottom of my heart, that this time will be clear what I'm saying! No excuses, I have the blend file and a video showing whats the issue. Something changed in smoke simulations and I think is only when usining the INITIAL VELOCITY option. If this change was intentionaly, wasn't in the realease notes! But i suppose that it is not intentionaly once the fire and smoke simulation looks horrible compared with the last version of Blender! Here I annex a blend file that was simulated at 2.91 and 2.82 AT THE SAME SETTINGS WITH A PROPER SIZE! If you don't belive me, download it and try it yourself! the last time, Vitor Druzian

Vitor Druzian commented

2021-02-24 20:15:47 +01:00

@rjg Read my last comment please. I don't think it's invalid, but it's your forum, I will not change the status again

Robert Guetzkow commented

2021-02-24 20:23:04 +01:00

Changed status from 'Archived' to: 'Needs Triage'

Robert Guetzkow reopened this issue

2021-02-24 20:23:04 +01:00

Eitan Traurig commented

2021-02-24 20:24:29 +01:00

Added subscriber: @EitanSomething

Eitan Traurig commented

2021-02-24 20:24:29 +01:00

Changed status from 'Needs Triage' to: 'Needs User Info'

Eitan Traurig commented

2021-02-24 20:24:29 +01:00

@Fernando-Alcala Please do not the status of the report, leave that to the developers and bug triagers.

Eitan Traurig commented

2021-02-24 20:25:22 +01:00

Changed status from 'Needs User Info' to: 'Needs Triage'

Jakub Grodzki commented

2021-03-09 13:19:26 +01:00

Added subscriber: @kubagrodzki

Jakub Grodzki commented

2021-03-09 13:58:01 +01:00

@Druzian Hi! I also noticed some weird behaviour of "Initial Velocity" parameter. Look at my example (please don't mind the shading difference):

0001-0060.mp4
In 2.83 setting IV to 1.0 gave really huge difference, now it's almost unnoticeable. Question is: is it bug or feature? I noticed some other differences in Mantaflow behaviour in 2.9x series compared to 2.83, but we don't know what causing them.

I really don't understand Initial Velocity in 2.92. This is a ddirect quote from Manual:

Source
Factor for the inherited velocity. A value of 1 will emit fluid moving at the same speed as the source.

I'm confused. In my example where IV is 1, smoke is definitely moving slower than the source.....

@Druzian Hi! I also noticed some weird behaviour of "Initial Velocity" parameter. Look at my example (please don't mind the shading difference): [0001-0060.mp4](https://archive.blender.org/developer/F9881088/0001-0060.mp4) In 2.83 setting IV to 1.0 gave really huge difference, now it's almost unnoticeable. Question is: is it bug or feature? I noticed some other differences in Mantaflow behaviour in 2.9x series compared to 2.83, but we don't know what causing them. I really don't understand Initial Velocity in 2.92. This is a ddirect quote from Manual: > Source >Factor for the inherited velocity. A value of 1 will emit fluid moving at the same speed as the source. I'm confused. In my example where IV is 1, smoke is definitely moving slower than the source.....

Ankit Meel commented

2021-06-18 19:59:28 +02:00

Closed as duplicate of #88257

Ankit Meel closed this issue

2021-06-18 19:59:28 +02:00

Ankit Meel commented

2021-06-18 20:39:16 +02:00

Added subscriber: @ankitm

Jay Soren commented

2021-08-14 03:10:35 +02:00

Added subscriber: @Cultmethod

Jay Soren commented

2021-08-14 03:10:35 +02:00

Just wanted to comment here that something funky is definitely happening with Initial Velocity in Mantaflow. The values barely do anything and you have to set them to absurdly high values to get what you want.

David DeLuca commented

2021-09-05 15:43:45 +02:00

Added subscriber: @Dd3d

David DeLuca commented

2021-09-05 15:43:45 +02:00

This comment was removed by @Dd3d

*This comment was removed by @Dd3d*

Sign in to join this conversation.

No Label

Download

What's New

Blender Studio

Manual

Developers Blog

Documentation

Benchmark

Blender Conference

Development Fund

One-time Donations

Mantaflow Initial velocity not working PROPERLY #84001