mikeporetti
Hey Mike, thanks for sending that through.
I can confirm that I'm seeing the same issue for this scene. I've filed a bug report here and we'll see if we can get to the bottom of it.
In the meantime, some observations about your scene 🙂
I see your clavicle is bending a bit. As a general rule it's safe to assume that bones don't deform. We usually opt for one Maya joint per bone object. In your particular setup, this results in self intersections which will mess with your collisions with other objects. (the solver can't do the point in volume test properly because it uses normals)
This value for volume conservation is exceptionally high. It will almost certainly introduce noise into your simulation with these values. Personally I wouldn't go higher than 10 * 106 at the most. Beyond this point diminishing returns kick in any way.
25 for point sensitivity is also pretty high. It means your muscle is going to contract completely at the slightest change in the input curve's length.
So useful to be able to poke around a scene like this and give feedback. Thanks for sharing it! Will hopefully have an idea on that substeps issue soon.
Cheers,
Andy