Contact forces between two btRigidBodies

[apsabelhaus]

Hello Bullet Physics folks!

First, thank you so much for creating this incredibly useful tool. My research lab (http://www.best.berkeley.edu) is putting it to good use to simulate our robots (https://github.com/NASA-Tensegrity-Robo ... it/NTRTsim). You all might know our project already!

I'm doing some work that requires measuring forces between two rigid body objects. I'll be having a robot stand (with its btRigidBody feet) on our world's ground, which is also a btRigidBody, and then I want to measure the distribution of forces between its feet with different foot positions. This will be static: I'll want the forces while these objects have zero velocity (or close to it), and the only external force will be gravity.

What is 'the best way' to get the forces, statically, between two btRigidBodies? From what I understand, m_TotalForce (and related) is zeroed out once the solver has positioned everything. (http://www.bulletphysics.org/Bullet/php ... f=9&t=1601)

Folks from my lab have done estimates before by dropping a rigid body a small distance then measuring the impulse and integrating over time. I've also come across some posts suggesting putting a small spring between objects, and getting the force that way (http://bulletphysics.org/Bullet/phpBB3/ ... php?t=9925).

Any suggestions?
Many thanks!
Drew Sabelhaus

[apsabelhaus]

Hi all - just an update, I decided to use the method suggested here (http://bulletphysics.org/Bullet/phpBB3/ ... php?t=9925) and created a small, high-spring-constant spring between my two objects. If folks want that code, it should be up soon on our repository for the NASA Tensegrity Robotics Toolkit (https://github.com/NASA-Tensegrity-Robo ... it/NTRTsim), though the spring is built on top of our framework and so requires all of NTRT to run.

Thanks again for the great tool!
Drew