Bindings - Lack of predictability

[Fraine7]

I'm having issues understanding and predicting the effects/controls regarding bindings (Both particle and PhysX)

Stiffness doesn't seem to have much of an impact, a value of 1.0 in my mind should make things super-rigid, and anything less than that would allow for particles to have some indpendence but be constrained via the binding length.

I'm finding myself adjusting length to values I can't quite understand, and still not seeing the rigidity that I would expect.

From other forum posts and Tyson's explanation in the tyFlow Facebook group I believe the binding solver steps should be quite high when simulating particles which are supposed to stay rigid, but this slows things down considerably and still requires me to adjust length and number of bindings in a non-intuitive way.

Would anyone be kind enough to explain how I might be able to better predict the binding controls so I can reduce the number of simulations required to achieve certain levels of rigidity?

Thanks

[MeHoo]

Try 1/16 frame instead of more substeps. That worked for me. The stiffness did seem a bit arbitrary though.

[Fraine7]

Does make a big difference but I’d much prefer stiffness value to determine overall flexibility, with 1.0 being rock solid. This would mean binding lengths could be small enough to connect only to neighbouring particles, then if we want to break bindings it would be much more predictable. I’m sure I’ll get used to the current binding methodology with more experience