Is it possible to set up a separate custom physics that shares collision data with the main game Chaos scene through low-level Chaos APIs?

Development Asset Creation
, ,
1

Hi Epic,

i stumbled upon this document [Content removed]

It mentions a feature which I have not been able to locate .

Is it possible to set up a separate custom physics that shares collision data with the main game Chaos scene through low-level Chaos APIs?

  • This is something we are currently implementing but missed the cut for 4.25, so it’s definitely possible but isn’t a trivial amount of work

Looking at ChaosSolversModule.h suggest that this feature has not yet been implemented ? Is there a plan to do so, or is it already available somewhere ?

thanks,

Magnus

2

Any update on this ?

3

Hi Magnus,

Apologies this has taken a while to respond to. Thanks for the link - I’ll reach out to the dev side to find out if this was ever done.

Can I find out the use case here - ie what problem you are trying to solve for this? It could be that we’ve solved it another way.

Best

Geoff Stacey

4

Hi Geoff,

Thanks for getting back to me. I’ll try to outline our use case.

We have an in-house animation system that runs post-physics. We recently introduced a powered ragdoll, which by default is simulated during physics, as it’s built on Chaos constraints.

However, since animations are updated post-physics, we end up tracking our powered ragdoll against the animated pose from the previous frame.

This issue could be mitigated by either predicting the next pose or moving the animation system update to pre-physics, but we’d like to avoid both of these options.

Ideally, we’d like to run the powered ragdoll’s constraint update after our animation system has finished updating.

Best,

Magnus

5

I think I’m following you. Your powered ragdoll sounds similar to our Physical Animation Component in that it is essentially ‘powered’ physics constraints.

In the proposal, you’d run the physics step for everything apart from the powered ragdolls? Then run the character simulation post animation. How would that work for the world update if there are non static objects in contact with the ragdoll since they can’t be accurately solved without knowing what the powered ragdoll is doing?

Normally I’d just suggest you update the animation first, but that sounds like it comes with it’s own issues.

6

Yes, we would run the entire simulation as usual, completely ignoring the powered ragdoll.

This works for us since the powered ragdoll uses one-way interaction and should not influence any other actors in the solver.

Imagine a character that can be pushed by static or dynamic geometry, but never affects any of them, almost like a cloth simulation.

7

That makes sense.

It is something we have an interest in doing, but it hasn’t been done yet. If you are keen to try this yourselves, the Geometry Collection components have done something similar as a potential reference source.

Best

Geoff

8

Al right, thanks. We will give it a go.

9

No worries Magnus, hopefully we’ll get some bandwidth soon for these kind of things!

Is this ticket all good to be closed now?

Best

Geoff

10

Yep .