Set cone friction model in triball.world

[scpeters]

This illustrates the different paths taken when started with the same linear velocity but different initial angular velocities.

With the default friction model: pyramid_model, the center model which does not spin travels a bit farther than the models with a small bit of initial angular velocity.

triball_spin_pyramid_model.mov

With the cone_model, the center model which has no initial angular velocity travels the shortest distance, which makes intuitive sense to me because it has the least amount of kinetic energy of any of the models. The models with the most angular velocity travel the farthest and exhibit some lateral motion as well.

triball_spin_cone_model.mov

cc @yaswanth1701

[scpeters]

as a bonus, here is a simulation with the cone model and center of gravity height for each triball artificially set to Z=0 (the level of the contact with the ground plane). Notice that there is no lateral motion

triball_spin_cone_model_cgz0.mov

[yaswanth1701]

This illustrates the different paths taken when started with the same linear velocity but different initial angular velocities.

With the default friction model: pyramid_model, the center model which does not spin travels a bit farther than the models with a small bit of initial angular velocity.

triball_spin_pyramid_model.mov
With the cone_model, the center model which has no initial angular velocity travels the shortest distance, which makes intuitive sense to me because it has the least amount of kinetic energy of any of the models. The models with the most angular velocity travel the farthest and exhibit some lateral motion as well.

triball_spin_cone_model.mov
cc @yaswanth1701

Hi Steve, I was working on writing the description for the tricall benchmark. So, for the complex case, we are going to compare the states (position, velocities, and contact forces) of the model with the pyramid against cone friction model right?