Hexagon Overkill

In early 2025 every new model had to write a ball bouncing inside a spinning hexagon. Most treated the rotating wall as if it stood still. This version runs a rigid-body contact solver with spin, Coulomb friction and time-of-impact rewinding, and checks its own physics below.

Scenariokeys 1–4

Transportt = 0.00 s

Time scale1.00×

PhysicsSI units

Wall angular velocity ω

Polygon sides N

Gravity g

Restitution e

Friction μ

View and solver

Energy, last 10 sJ

total trans spin pot

Self-verification

Jacobi integral E − ΩL Rotating walls, no gravity, e = 1, μ = 0. The walls change the ball's energy at every bounce, but E − ΩL must stay constant. 8 polygons × 30 s. … WAIT

Same test, 2025-style solver Bounces computed as if the wall were standing still. This is the reference failure. … WAIT

Energy with static walls g = 9.81 m/s², e = 1, μ = 0. Exact ballistic arcs plus time-of-impact rewinding. 8 polygons × 30 s. … WAIT

Slide-to-roll, v = ⅔ v₀ A disk launched without spin must end up rolling at exactly two thirds of its launch speed, for any μ. … WAIT

No tunneling 60 balls at 25 m/s in a closed hexagon spinning at 8 rad/s, 10 s. Adaptive substeps cap motion at 0.4 radii per substep. … WAIT

The prompt, roughly

write a Python program that shows a ball bouncing inside a spinning hexagon. The ball should be affected by gravity and friction, and it must bounce off the rotating walls realistically

Space pause · S step · R reset · F frame · T trails · C contacts

Written by Claude Opus 5.5, September 2026. Single file, no libraries.