Maniphest T76812

[Mantaflow] Inflow particles not continuous
Closed, Resolved

Assigned To
Richard Antalik (ISS)
Authored By
Pavel (Pavel_Blend)
May 16 2020, 9:51 PM
Tags
  • BF Blender
  • Nodes & Physics
Subscribers
Ankit Meel (ankitm)
Pavel (Pavel_Blend)
Richard Antalik (ISS)
Sebastián Barschkis (sebbas)

Description

System Information
Operating system: Windows-10-10.0.18362-SP0 64 Bits
Graphics card: AMD Radeon HD 7560D ATI Technologies Inc. 4.5.13399 Core Profile Context 15.201.1151.1008

Blender Version
Broken: version: 2.83 (sub 15), branch: master, commit date: 2020-05-11 19:18, hash: rB8956e9e5f6c0
Worked: (newest version of Blender that worked as expected)

Short description of error
Particle flow is not continuous.

Exact steps for others to reproduce the error
Bake the liquid in this scene:


You will see such a result:

Revisions and Commits

rB Blender

Related Objects

Mentioned Here
rBM6667: Fluid: Added note to CFL settings
rB7bb3d9787ead: Fix T76812: [Mantaflow] Inflow particles not continuous

Event Timeline

Pavel (Pavel_Blend) created this task.May 16 2020, 9:51 PM
Ankit Meel (ankitm) renamed this task from [Mantaflow] Inflow particles bug to [Mantaflow] Inflow particles not continuous .May 17 2020, 10:58 AM
Ankit Meel (ankitm) added a project: Nodes & Physics.
Ankit Meel (ankitm) added a subscriber: Ankit Meel (ankitm).

Could you add which settings could have caused this ?

Pavel (Pavel_Blend) added a comment.May 17 2020, 11:05 AM

It seems to me that this error does not depend on the settings. If the fluid source emits particles for a long time, then this error is visible, if not for long, then this error is less noticeable.

Richard Antalik (ISS) closed this task as Archived.May 18 2020, 3:58 PM
Richard Antalik (ISS) claimed this task.
Richard Antalik (ISS) added a subscriber: Richard Antalik (ISS).
Pavel (Pavel_Blend) added a comment.May 18 2020, 5:22 PM

How to get rid of this problem? Do not use adaptive steps?

Richard Antalik (ISS) added a comment.May 19 2020, 11:50 AM
In T76812#934081, @Pavel (Pavel_Blend) wrote:

How to get rid of this problem? Do not use adaptive steps?

As far as I can tell that would be the only solution

Pavel (Pavel_Blend) added a subscriber: Sebastián Barschkis (sebbas).May 19 2020, 4:49 PM

@Sebastián Barschkis (sebbas) what do you say about this behavior of the simulator? Maybe this is still a fluid engine error?

Sebastián Barschkis (sebbas) added a comment.May 19 2020, 5:33 PM

I would say there are two problems here:

  1. Flow is not continuous: This can be improved by setting a smaller CFL value.
  2. After time, there are no particles in the upper inflow region: Particles are being sampled and moved immediately afterwards (all in one step). This way new sampled particles never fill up the upper inflow area.

The fix for (1) you can try immediately, for (2) I'll try to come up with something.

Sebastián Barschkis (sebbas) reopened this task as Confirmed.May 19 2020, 5:55 PM

Reopening this report because would like to add a fix for issue (2) (see my comment above).

Sebastián Barschkis (sebbas) closed this task as Resolved by committing rB7bb3d9787ead: Fix T76812: [Mantaflow] Inflow particles not continuous.May 19 2020, 9:14 PM
Sebastián Barschkis (sebbas) added a commit: rB7bb3d9787ead: Fix T76812: [Mantaflow] Inflow particles not continuous.
Sebastián Barschkis (sebbas) added a comment.May 19 2020, 11:20 PM

A few closing remarks:

  • With 7bb3d9787ead the issue with missing particle in the upper inflow region should be solved (issue (2) from above)
  • When using a lower CFL value for a more continuous flow, it is also recommended to adjust the Timesteps Maximum value (issue (1) from above). This is also described in the manual now (rBM6667).

Thanks for reporting!