Scaled Rigid Body causes issue (moving origin or mesh breaks collision, double transforms on negative scale) #81814

Open
opened 2020-10-18 20:59:31 +02:00 by Mikko Rantalainen · 13 comments

System Information
Operating system: Ubuntu 18.04 LTS
Graphics card: Intel HD 4000

Blender Version
Broken: 3.0
Broken: 2.90.1
Worked: I don't know if this works with any other version, I created the file using 2.90.1.

Short description of error

See the attached .blend file. This file contains two errors with Rigid body simulation. I'm filing both errors in one bug because these may be related (both happen in one example simulation).

Error 1: object origin distance from another object changes interaction between objects

  1. Open attached example case (.blend file)
  2. Play the animation to verify that "Ball" is attached to the chain.
  3. Stop the animation and jump to first frame.
  4. Select object "Ball" and from menu Object - set Origin - Origin to Center of Mass (Volume)
  5. Play the animation to verify that "Ball" is no longer attached to the chain.

Error 2: Rigid body type "Active" changes orientation of the object

  1. Open attached example case (.blend file)
  2. Play the animation to verify that chain is attached to the object "Cube".
  3. Stop the animation and jump to first frame.
  4. Change "Cube" Physics properties - Rigid body - Type: Active and Settings: Dynamic
  5. Play the animation and notice that Cube is now rotated 180 degress around the X axis. Obviously this results in chain no longer connected to the cube as expected and chain and Ball drop directly down during the animation.

Exact steps for others to reproduce the error
Example .blend file (cannot simplify more because Blender seems to be unstable with this file):
physics-test.blend

Additional information
It seems that I can work around Error 1 by selecting all items and applying "Rotation & Scale". I think this should not be required to get two objects to interact during simulation so I'm reporting the problematic file as-is. I cannot figure out the reason nor workaround for Error 2.

$ apt policy blender
blender:

Installed: 2.90.1--release~98526-3e85bb34d0d-0thomas~bionic0
Candidate: 2.90.1--release~98526-3e85bb34d0d-0thomas~bionic0
Version table:

*** 2.90.1--release~98526-3e85bb34d0d-0thomas~bionic0 500
500 http://ppa.launchpad.net/thomas-schiex/blender/ubuntu bionic/main amd64 Packages

**System Information** Operating system: Ubuntu 18.04 LTS Graphics card: Intel HD 4000 **Blender Version** Broken: 3.0 Broken: 2.90.1 Worked: I don't know if this works with any other version, I created the file using 2.90.1. **Short description of error** See the attached .blend file. This file contains two errors with Rigid body simulation. I'm filing both errors in one bug because these may be related (both happen in one example simulation). Error 1: object origin distance from another object changes interaction between objects 1. Open attached example case (.blend file) 2. Play the animation to verify that "Ball" is attached to the chain. 3. Stop the animation and jump to first frame. 4. Select object "Ball" and from menu Object - set Origin - Origin to Center of Mass (Volume) 5. Play the animation to verify that "Ball" is no longer attached to the chain. Error 2: Rigid body type "Active" changes orientation of the object 1. Open attached example case (.blend file) 2. Play the animation to verify that chain is attached to the object "Cube". 3. Stop the animation and jump to first frame. 4. Change "Cube" Physics properties - Rigid body - Type: Active and Settings: Dynamic 5. Play the animation and notice that Cube is now rotated 180 degress around the X axis. Obviously this results in chain no longer connected to the cube as expected and chain and Ball drop directly down during the animation. **Exact steps for others to reproduce the error** Example .blend file (cannot simplify more because Blender seems to be unstable with this file): [physics-test.blend](https://archive.blender.org/developer/F9013340/physics-test.blend) **Additional information** It seems that I can work around Error 1 by selecting all items and applying "Rotation & Scale". I think this should not be required to get two objects to interact during simulation so I'm reporting the problematic file as-is. I cannot figure out the reason nor workaround for Error 2. $ apt policy blender blender: ``` Installed: 2.90.1--release~98526-3e85bb34d0d-0thomas~bionic0 Candidate: 2.90.1--release~98526-3e85bb34d0d-0thomas~bionic0 Version table: ``` *** 2.90.1--release~98526-3e85bb34d0d-0thomas~bionic0 500 500 http://ppa.launchpad.net/thomas-schiex/blender/ubuntu bionic/main amd64 Packages

Added subscriber: @mtrantalainen

Added subscriber: @mtrantalainen

Added subscriber: @iss

Added subscriber: @iss

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

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

I can confirm both issues in 2.90. Issue 1 also exists in 2.91, issue 2 doesn't. But in 2.91 the chain isn't attached to the cube and I can't make it work. I suppose it isn't neccessary to demonstrate first issue so you can just remove it and make first chain link static.

Please check this with build from https://builder.blender.org/download/ and make one report per issue.

I can confirm both issues in 2.90. Issue 1 also exists in 2.91, issue 2 doesn't. But in 2.91 the chain isn't attached to the cube and I can't make it work. I suppose it isn't neccessary to demonstrate first issue so you can just remove it and make first chain link static. Please check this with build from https://builder.blender.org/download/ and make one report per issue.

I re-tested with following version:
Blender 2.91.0 Alpha
October 20, 23:15:53- 7167a57197
Linux 64 bit

It seems that I can make the scene work with 2.91 by applying Scale for objects "Cube" and "Ball". I think this issue has been fixed in 2.91 7167a57197.

It turns out that the object "Cube" having negative scale was the cause for Error 2 in the description. And the ball having scale appears have caused computing the ball Ball connection point being computed to be not interlinked with "Link 6".

Is it intentional that the rigid body physics applies scale twice? It seems that scale is applied once for the object sizing and then another time for the simulation resulting in double application for the physics part. This issue is obviously hidden if I apply scale because then it's just double application of identity.

I re-tested with following version: Blender 2.91.0 Alpha October 20, 23:15:53- 7167a57197e9 Linux 64 bit It seems that I can make the scene work with 2.91 by applying Scale for objects "Cube" and "Ball". I think this issue has been fixed in 2.91 7167a57197e9. It turns out that the object "Cube" having negative scale was the cause for Error 2 in the description. And the ball having scale appears have caused computing the ball Ball connection point being computed to be not interlinked with "Link 6". Is it intentional that the rigid body physics applies scale twice? It seems that scale is applied once for the object sizing and then another time for the simulation resulting in double application for the physics part. This issue is obviously hidden if I apply scale because then it's just double application of identity.

It seems that scale is applied once for the object sizing and then another time for the simulation resulting in double application

I don't think it would be intentional, especially if it produces wrong result. But I am not able to replicate this issue with very simple simulation (scalled cube falling onto plane)

Also I see that I have to only apply scale for cube to make simulation work.

So as far as I can see issue 1 and issue 2 no longer exists, but now we have issue with scale?

> It seems that scale is applied once for the object sizing and then another time for the simulation resulting in double application I don't think it would be intentional, especially if it produces wrong result. But I am not able to replicate this issue with very simple simulation (scalled cube falling onto plane) Also I see that I have to only apply scale for cube to make simulation work. So as far as I can see issue 1 and issue 2 no longer exists, but now we have issue with scale?

Here's an example file which shows an example of insane physics when scale is not (1,1,1): negative-suzanne-physics.blend

Note that I modified face of Suzanne to not be symmetric. After you press "Play", the face will immediately mirror and rotate 180 degrees before dropping due gravity.

This is caused by scale (-1, -1, -1) which is technically okay and is already applied to model when you load the scene. When you press Play, this scale will be applied again before starting the simulation. I think this is a bug and should be fixed.

Is there a sensible use case for applying the scale or rotation again just for the start of the simulation? Also note that this only seems to affect the initial state because if the scale were re-applied on each frame, you'd see Suzanne mirroring and rotating each frame.

Here's an example file which shows an example of insane physics when scale is not (1,1,1): [negative-suzanne-physics.blend](https://archive.blender.org/developer/F9052363/negative-suzanne-physics.blend) Note that I modified face of Suzanne to not be symmetric. After you press "Play", the face will immediately mirror and rotate 180 degrees before dropping due gravity. This is caused by scale (-1, -1, -1) which is technically okay and is already applied to model when you load the scene. When you press Play, this scale will be applied again before starting the simulation. I think this is a bug and should be fixed. Is there a sensible use case for applying the scale or rotation *again* just for the start of the simulation? Also note that this only seems to affect the initial state because if the scale were re-applied on each frame, you'd see Suzanne mirroring and rotating each frame.

Also note that the negative scale is applied between frames 1 and 2!

Also note that the negative scale is applied between frames 1 and 2!

As for negative scale I have found these reports: #56700, #44720 So that is not considered a bug here.

As for negative scale I have found these reports: #56700, #44720 So that is not considered a bug here.
Member

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

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

Added subscriber: @lichtwerk

Added subscriber: @lichtwerk
Member

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

Changed status from 'Needs Triage' to: 'Confirmed'
Member

Sorry for the long silence here.

So I think we have to just agree that scaling (esp. negative -- or non-uniform, which is not used here --) is just over the head of our current RB simulation capabilities.

For the ball case (with positive scaling), this also happens for any mesh or origin manipulation really, I think this can be rephrased to "moving origin or mesh on a scaled RB causes issues".
Will keep it as Known Issue for now, but might eventually be closed (as this might never be fixed with the current system.

Sorry for the long silence here. So I think we have to just agree that scaling (esp. negative -- or non-uniform, which is not used here --) is just over the head of our current RB simulation capabilities. For the ball case (with positive scaling), this also happens for any mesh or origin manipulation really, I think this can be rephrased to "moving origin or mesh on a scaled RB causes issues". Will keep it as Known Issue for now, but might eventually be closed (as this might never be fixed with the current system.
Philipp Oeser changed title from Unstable Rigid body simulation with example file to Scaled Rigid Body causes issue (moving origin or mesh breaks collision, double transforms on negative scale) 2021-09-23 11:21:22 +02:00
Philipp Oeser removed the
Interest
Nodes & Physics
label 2023-02-10 08:46:29 +01:00
Sign in to join this conversation.
No Label
Interest
Alembic
Interest
Animation & Rigging
Interest
Asset Browser
Interest
Asset Browser Project Overview
Interest
Audio
Interest
Automated Testing
Interest
Blender Asset Bundle
Interest
BlendFile
Interest
Collada
Interest
Compatibility
Interest
Compositing
Interest
Core
Interest
Cycles
Interest
Dependency Graph
Interest
Development Management
Interest
EEVEE
Interest
EEVEE & Viewport
Interest
Freestyle
Interest
Geometry Nodes
Interest
Grease Pencil
Interest
ID Management
Interest
Images & Movies
Interest
Import Export
Interest
Line Art
Interest
Masking
Interest
Metal
Interest
Modeling
Interest
Modifiers
Interest
Motion Tracking
Interest
Nodes & Physics
Interest
OpenGL
Interest
Overlay
Interest
Overrides
Interest
Performance
Interest
Physics
Interest
Pipeline, Assets & IO
Interest
Platforms, Builds & Tests
Interest
Python API
Interest
Render & Cycles
Interest
Render Pipeline
Interest
Sculpt, Paint & Texture
Interest
Text Editor
Interest
Translations
Interest
Triaging
Interest
Undo
Interest
USD
Interest
User Interface
Interest
UV Editing
Interest
VFX & Video
Interest
Video Sequencer
Interest
Virtual Reality
Interest
Vulkan
Interest
Wayland
Interest
Workbench
Interest: X11
Legacy
Blender 2.8 Project
Legacy
Milestone 1: Basic, Local Asset Browser
Legacy
OpenGL Error
Meta
Good First Issue
Meta
Papercut
Meta
Retrospective
Meta
Security
Module
Animation & Rigging
Module
Core
Module
Development Management
Module
EEVEE & Viewport
Module
Grease Pencil
Module
Modeling
Module
Nodes & Physics
Module
Pipeline, Assets & IO
Module
Platforms, Builds & Tests
Module
Python API
Module
Render & Cycles
Module
Sculpt, Paint & Texture
Module
Triaging
Module
User Interface
Module
VFX & Video
Platform
FreeBSD
Platform
Linux
Platform
macOS
Platform
Windows
Priority
High
Priority
Low
Priority
Normal
Priority
Unbreak Now!
Status
Archived
Status
Confirmed
Status
Duplicate
Status
Needs Info from Developers
Status
Needs Information from User
Status
Needs Triage
Status
Resolved
Type
Bug
Type
Design
Type
Known Issue
Type
Patch
Type
Report
Type
To Do
No Milestone
No project
No Assignees
3 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: blender/blender#81814
No description provided.