Bug on Rigid Body Simulation when an object is parented to another #69065

New Issue

wenye he · 2019-08-22T22:10:58+02:00

wenye he commented

2019-08-22 22:10:58 +02:00

System Information
Operating system: Windows 10
Graphics card: GTX 1060 6G

Blender Version
2.80

Short description of error
Bug on Rigid Body Simulation when an object is parented to another

Exact steps for others to reproduce the error
Please check the video......
BL280BUG.mp4

More, I have some questions on Rigid Body. I want to ask something, coding and algorithm, to the developers who work on it...
How can I reach them?

Thanks.

**System Information** Operating system: Windows 10 Graphics card: GTX 1060 6G **Blender Version** 2.80 **Short description of error** Bug on Rigid Body Simulation when an object is parented to another **Exact steps for others to reproduce the error** Please check the video...... [BL280BUG.mp4](https://archive.blender.org/developer/F7689093/BL280BUG.mp4) More, I have some questions on Rigid Body. I want to ask something, coding and algorithm, to the developers who work on it... How can I reach them? Thanks.

wenye he commented

2019-08-22 22:10:58 +02:00

Added subscriber: @WenyeHe

blender-admin commented

2019-08-22 22:10:58 +02:00

#72232 was marked as duplicate of this issue

blender-admin commented

2019-08-22 22:10:58 +02:00

#68183 was marked as duplicate of this issue

Abid Maqbool commented

2019-08-23 05:53:21 +02:00

Added subscriber: @AbidMaqbool

Abid Maqbool commented

2019-08-23 05:53:21 +02:00

@WenyeHe meets the devs real-time at https:*blender.chat/channel/blender-coders.. Are you have q about development stuff ask at https:*devtalk.blender.org/..

Abid Maqbool commented

2019-08-23 06:13:15 +02:00

Can confirm in Blender 2.8 and today daily build 2.81..
I think, the expected behavior should be child object will be effected by both Parent (transform) and rigid body.
In this case as parent object is still on it's place so child object can freely act as rigid body

Can confirm in Blender 2.8 and today daily build 2.81.. I think, the expected behavior should be child object will be effected by both Parent (transform) and rigid body. In this case as parent object is still on it's place so child object can freely act as rigid body

Philipp Oeser commented

2019-08-23 10:59:31 +02:00

Added subscribers: @Sergey, @sreich, @lichtwerk

Philipp Oeser commented

2019-08-23 10:59:31 +02:00

In 2.80, I can see object jumping around, but for me, 2.81/master is stable (in that it ignores the sim completely)
#69065.blend

Think this is expected behavior since there is a dependency circle:

Dependency cycle detected:
  SCScene/Transform Component/RIGIDBODY_SIM() depends on
  OBSphere/Transform Component/TRANSFORM_SIMULATION_INIT() via 'Object Transform -> Rigidbody Sim Eval'
  OBSphere/Transform Component/TRANSFORM_EVAL() via 'Transform Eval -> Simulation Init'
  OBSphere/Transform Component/TRANSFORM_PARENT() via 'Eval'
  OBSphere/Transform Component/TRANSFORM_LOCAL() via 'ObLocal -> ObParent'
  OBSphere/Transform Component/TRANSFORM_INIT() via 'Transform Init'
  OBCube/Transform Component/TRANSFORM_FINAL() via 'Parent'
  OBCube/Transform Component/RIGIDBODY_TRANSFORM_COPY() via 'Rigidbody Sync -> Transform Final'
  SCScene/Transform Component/RIGIDBODY_SIM() via 'Rigidbody Sim Eval -> RBO Sync'

2.79 magically behaved "better" in that it seemingly completely ignored the parent transform.

I think, the expected behavior should be child object will be effected by both Parent (transform) and rigid body.

Dont think this is possible with the current system (unless we somewhat translate parent transforms into bullet forces?)

CC @Sergey, @sreich

Afraid we have to close this?

In 2.80, I can see object jumping around, but for me, 2.81/master is stable (in that it ignores the sim completely) [#69065.blend](https://archive.blender.org/developer/F7692351/T69065.blend) Think this is expected behavior since there is a dependency circle: ``` Dependency cycle detected: SCScene/Transform Component/RIGIDBODY_SIM() depends on OBSphere/Transform Component/TRANSFORM_SIMULATION_INIT() via 'Object Transform -> Rigidbody Sim Eval' OBSphere/Transform Component/TRANSFORM_EVAL() via 'Transform Eval -> Simulation Init' OBSphere/Transform Component/TRANSFORM_PARENT() via 'Eval' OBSphere/Transform Component/TRANSFORM_LOCAL() via 'ObLocal -> ObParent' OBSphere/Transform Component/TRANSFORM_INIT() via 'Transform Init' OBCube/Transform Component/TRANSFORM_FINAL() via 'Parent' OBCube/Transform Component/RIGIDBODY_TRANSFORM_COPY() via 'Rigidbody Sync -> Transform Final' SCScene/Transform Component/RIGIDBODY_SIM() via 'Rigidbody Sim Eval -> RBO Sync' ``` 2.79 magically behaved "better" in that it seemingly completely ignored the parent transform. > I think, the expected behavior should be child object will be effected by both Parent (transform) and rigid body. Dont think this is possible with the current system (unless we somewhat translate parent transforms into bullet forces?) CC @Sergey, @sreich Afraid we have to close this?

Firepal commented

2019-12-18 12:19:32 +01:00

Added subscriber: @Firepal

Firepal commented

2019-12-18 12:19:32 +01:00

Lots of issues with parenting and rigid bodies in 2.8x it seems.
Encountered a similarly unpredictable effect with an empty parented to a rigid body object that uses the empty for its Simple Deform modifier.

Result is that the Simple Deform modifier acts as if the empty is moving (even though it's transforming around the object origin and thus with the mesh).
Haven't found a way to fix this other than just applying the modifier.

Lots of issues with parenting and rigid bodies in 2.8x it seems. Encountered a similarly unpredictable effect with an empty parented to a rigid body object that uses the empty for its Simple Deform modifier. Result is that the Simple Deform modifier acts as if the empty is moving (even though it's transforming around the object origin and thus with the mesh). Haven't found a way to fix this other than just applying the modifier.

Philipp Oeser commented

2019-12-18 18:38:01 +01:00

Added subscribers: @justasb, @hgabor47

Philipp Oeser commented

2019-12-18 18:39:57 +01:00

In #68183#831214, @lichtwerk wrote:
Yes, this has problems with the dependency graph [it is producing dependency cycles in 2.8, which is correct, whereas 2.79 somewhat made its way through in a hackish way...]
Sooo, dependency cycles are not treated as bugs [unless they are plain wrong -- which doesnt seem to be the case here]
Note that #58044 (Issue with rigid body and modifiers dependent on transform) is related.
Still need to think a bit about passive rigid bodies though [unsure whether a passive rigid bodies transform really depends on Scenes Rigidbody Sim?]...

> In #68183#831214, @lichtwerk wrote: > Yes, this has problems with the dependency graph [it is producing dependency cycles in 2.8, which is correct, whereas 2.79 somewhat made its way through in a hackish way...] > Sooo, dependency cycles are not treated as bugs [unless they are plain wrong -- which doesnt seem to be the case here] > Note that #58044 (Issue with rigid body and modifiers dependent on transform) is related. > Still need to think a bit about **passive** rigid bodies though [unsure whether a passive rigid bodies transform really depends on Scenes Rigidbody Sim?]...

Jacques Lucke commented

2020-02-12 15:25:46 +01:00

Added subscriber: @JacquesLucke

Jacques Lucke commented

2020-02-12 15:25:46 +01:00

Changed status from 'Needs Developer To Reproduce' to: 'Confirmed'

Jacques Lucke commented

2020-02-12 15:25:46 +01:00

I set this as known issue, because it fixing it might require deeper changes to how Rigid Bodies are implemented in Blender and it is unlikely, that someone is fixing this in the next couple of month.

Pauan commented

2020-10-15 14:32:24 +02:00

Added subscriber: @pauanyu_blender

Philipp Oeser added the

Download

What's New

Blender Studio

Manual

Developers Blog

Documentation

Benchmark

Blender Conference

Development Fund

One-time Donations

Bug on Rigid Body Simulation when an object is parented to another #69065