Closed Cubature1 closed 4 months ago
This is because FleX naturally assumes a 1/60th timestep, what 100% does is multiply 1/60 by one, which equals itself. However, you'll notice that everything is slow. But, at the same time, if you increase gravity to match source, eventually the constraint solver in the physics engine exhibits mass instability and glitches.
This is an issue with no clear solution in sight yet, sorry. I'll keep it open in case I figure new things out
if you increase gravity to match source, eventually the constraint solver in the physics engine exhibits mass instability and glitches.
Is this because the raw numbers are just too big for what it expects? Can you not translate hammer units to metric when you write to the solver, and back again when you read? (I suggested this ages ago)
Superseded by the ticket #53
GPU
RTX 3050
Operating System
Win 11
Is this specific to performance?
Issue Type
None
Incident Description
When the Gelly timescale is set above 100% (As in 1/10th of the default), particles behave improperly. Friction does not exist, and adhesion does not function properly. Seems to fix the other issues on the 1.3.0 version to an extent as well.
Log
No response