A spinning disk can wobble faster and faster just before it falls flat. That accelerating rhythm does not mean it is gaining energy. Two different kinds of rotational motion are easy to confuse.

Faster wobbling, less energy
In Keith Moffatt’s 2000 analysis, the point where the rim touches the table travels around a circle. Its rate of travel describes the disk’s precession: the changing direction of its tilt, not simply the disk turning about its own axis.
As the disk leans closer to horizontal, the idealized rolling model predicts faster precession while its mechanical energy decreases. The center of mass is lower, and the motion changes geometry. Moffatt proposed that viscous resistance in the air beneath the disk explains its abrupt settling. His mathematical speed diverges near the predicted stopping time, but the model breaks down before a real disk reaches infinity.
The tabletop changes the result
Hervé Caps and colleagues tested metal disks and a ring on different surfaces in 2004 experiments. They measured decreasing tilt and rotation about the disk’s own axis alongside increasing precession. Rougher contacts shortened the motion.
Their comparisons supported an important role for rolling friction. Small excursions in the disk’s path also suggested slipping, although the apparatus could not resolve a specific stick-slip sequence. The authors argued that contact losses dominated their experiments, challenging an explanation based mainly on air resistance.

The rising spinning egg is another tabletop motion in which the contact matters. An ideal frictionless calculation and a real object on a particular surface need not follow the same history.
Different losses can dominate different stages
A 2026 preprint by Benjamin Thorne and colleagues reports high-speed measurements across disk sizes, masses and surfaces. The authors argue for rolling friction earlier in the motion but air drag dominating near the end. Partial-vacuum tests reduced late-stage energy loss; a ring lacking the solid center did not show the same crossover.
This is an unreviewed report, not a settled verdict for every spinning coin. Its proposed air-boundary-layer mechanism also differs from Moffatt’s original thin-gap model. Even in these experiments, rolling resistance accounted for most energy lost over the whole run. The fastest wobbling and the largest overall energy drain need not belong to the same stage.

Leave a Reply