Hi all! I’m trying to simulate a sequence of collisions: each sphere transferring its velocity to the next sphere and stopping in the point of collision. The last Sphere is supposed to touch the frame and bounce back to its original position. Still didn’t get the spheres to act like i wanted… :frowning: Any help? Thanks! (33.0 KB)

Hi and welcome,
Here’s a slight improvement -
I got rid of the anchors, and for the initial push I put a button, so you press it once to launch the first one then it turns off. (31.1 KB)
I must admit though, Kangaroo2 is still a bit limited for dynamics - damping can never really be reduced to zero, and there aren’t easy ways to control things like mass or elastic restitution. I go into the details of this more here, but the summary of it is that Kangaroo2 is still primarily for finding exact equilibrium solutions, and while the end results can be accurate, the dynamics which get them there are only an approximation of real physical dynamics.

1 Like

Hi Daniel, thanks very much for your input and for the reference: i understand where you’re going with damping.
To keep the sphere-movement going, is there a way to make the box-boundaries “elastic”?

Thanks again!
G. (35.6 KB)

I am looking at how I can allow momentum preservation in a more consistent way throughout the solver for all goals, but in the meantime, here’s a modified version of the curve collider that has more of a bounce.

Hi Daniel,
Thank you very much!

A question regarding your C# addition: is there a way to make the curve boundaries stronger? The last sphere still runs off… (sorry, i don’t yet know C#, but if you add a note in the code i’ll read to learn from it). Here’s my last version. Thanks again! G. (29.6 KB)