3

Does anyone want to help me with some GeoNodes? I have a keyframed ball moving on a plane. I figure it should be possible to make the ball roll correctly using simulation nodes. In my mind, you compare the ball's location with its location on the previous frame, and that's the motion vector. The roll axis would be perpendicular to that and the floor.

I've figured out the first step. I compare the location with the previous one and get a motion vector. How do I translate that into the correct rotation?

Thanks in advance!

Robin
  • 45
  • 3
  • 1
    need more details. As the problem is stated currently, you don't even need a simulation zone: Rotate Spheres according to height of other points – Markus von Broady Oct 04 '23 at 17:39
  • @Markus Hard disagree. The linked answer relies on clear constraints where each position is unique. However, with a ball rolling around on the floor (think of a soccer ball passed around by three players), the position does not dictate the orientation. The question appears to be clearly structured, doesn't contain superflous details, details the authors attempt and narrows down the problem. Only the title could be a bit more specific. – Leander Oct 04 '23 at 19:14
  • @Leander my point is: we don't know if the keyframed animation isn't a movement in a straight line. I didn't mark as a duplicate but it could as well be depending on OP's requirements. – Markus von Broady Oct 04 '23 at 19:23
  • @MarkusvonBroady You're right; I could have been more clear. But Leander's (and your) interpretation are right on the money. Thanks for helping! – Robin Oct 05 '23 at 09:24

1 Answers1

6

Here's a simple setup for a 100% friction ball:

Markus von Broady
  • 36,563
  • 3
  • 30
  • 99