Why constraint_type = all is different from position and rotation combined?

[bbrcn]

Hi

I been playing houdini constraint recently. I find the behaver of bullet hard constraint is not very easy to predict, especially combined with condof, condir, position, rotation and angular motor.

I have an simple file as an example.

One box hard constraint to world space.

Using one wrangle node with s@constraint_typle = "all". it works as expected. Box stay there.

 

When I split constraint with position and rotation. Give them all condof = 3. Ideally it should be act same as the last one stay there (I assume), but box start to fall. I tried constraint force mixing and error reduction parameter it can improve but I'm not sure I should do that since this is super simple setup shouldn't use too much tweak.

 

 

File:hard_constratin_question.hip

 

I'm working on pretty complicate rigid constraint setup I need to fully understand the bullet constraint behaver.  Really appreciate the help. 

 

I also want to ask is rbd constraint the better way for complicated rigid body constraint with lots gear and mechanical parts and vellums. 

 

Thanks

 

Edited January 17, 2022 by bbrcn

[bbrcn]

anyone?

[vicvvsh]

If you set physical parameters back to default values on /obj1/dopnet14/rbdpackedobject1 behavior of constraints is fine. Don't ask me why, I don't know, maybe a bug.

[bbrcn]

12 hours ago, vicvvsh said:

If you set physical parameters back to default values on /obj1/dopnet14/rbdpackedobject1 behavior of constraints is fine. Don't ask me why, I don't know, maybe a bug.

 

Cool, thanks