I'm trying to understand the utility of 'contact breaking threshold' (default value: 0.02) in source code but it's hard
If the distance between two bodies is less than contact breaking threshold, we create a contact point, is that right ?
So, we can create contact points even if bodies doesn't collide. What is the utility because they won't be used by constraint solver when contact processing thershold = 0 ?
I see only one utility to that but I don't found the proof in the source code:
I guess we persist some contact points before bodies really collide in order to have more than one contact point when the bodies will really collide. It allow to improve stability.
Am I right ? If yes: why the broadphase works on AABBox of the shape and not on AABBox extended by contactBreakingTherhold ? Indeed, it will allow to narrowphase to persist more than one contact point before bodies collide.
Sorry for my poor English.
Thank you in advance.