The Coin That Screams Toward Infinity: The Strange Physics of Euler’s Disk

Spin a coin on a kitchen table and you already know how the story ends. It wobbles, it leans, it starts to rattle, and the rattle climbs into a frantic metallic buzz that gets faster and faster until, with a final shiver, the coin slaps flat and goes silent.

Now do the same thing with a heavy, polished steel disk on a curved mirror, and the ending stretches out to a minute or more. The disk leans lower and lower. Its edge seems to blur into a hovering silver ring. The whirr rises in pitch like a tiny engine revving toward the red line. And then it stops, all at once, as if somebody hit a switch.

That toy is called Euler’s Disk. It sells as a desk curiosity, but it also started a real argument in one of the world’s top science journals, about a question that sounds almost too simple to fight over: what actually stops a spinning coin?

A lunch break, a box of hockey pucks

The toy began as a distraction. Joseph Bendik, a physics graduate of UC Berkeley, was working at Hughes Aircraft in Carlsbad, California, when, around 1987, a box of heavy polishing chucks arrived in his lab. They looked like steel hockey pucks.

Bendik had always liked spinning coins, so he gave one of the chucks a spin on his desk. He was struck by how long it kept going before it fell flat, and by the whine some of the chucks made, a sound that kept rising in pitch the longer they went. He had a word for the motion: the disk wasn’t just spinning or rolling, it was “spolling.”

Over the next few years he tinkered. He tried disks of different sizes and weights, different coatings, different surfaces to spin them on. A mirror-like glass surface gave the longest runs. Then he found that a holographic sticker on top produced a striking illusion as the disk turned, a shimmering shape that seemed to float above the metal.

The last piece was the name. He chose Euler’s Disk, after Leonhard Euler (pronounced “oiler”), the 18th-century Swiss mathematician whose work underpins the mechanics of spinning and rolling bodies.

From mall shelf to sitcom

The finished toy, a thick chrome-plated steel disk with a slightly rounded edge and a rigid, gently concave mirrored base, went on sale in 1996. It was first introduced at The Nature Stores in shopping malls and later moved online. For years it was a small but steady side business: Bendik told the San Diego Union-Tribune in 2017 that he had been selling around 1,000 disks a year since launch.

It also had a quiet movie career. According to the same interview, sound designers on the 2001 film Pearl Harbor used the disk’s whirr for the motors of Japanese torpedoes.

Then television gave it a shove. In a February 2017 episode of The Big Bang Theory, the astrophysicist character Raj tries to hold his breath for longer than an Euler’s Disk takes to stop spinning. The morning after it aired, Bendik opened his email to find hundreds of orders. “It was flooded,” he said. “I thought I’d opened up the spam folder by accident.” He didn’t have enough stock, and ended up refunding hundreds of them.

There is a nice footnote to the origin story. Bendik had started inventing at 13 with a school geometry classmate, Larry Shaw, who went on to study physics at Cornell and invent Astrojax, the three-ball string toy. Two kids from one geometry class, two physics toys on the market.

Why the whirr speeds up

To see what’s strange about the disk, it helps to separate three motions that look like one.

First, the disk is tilted, leaning on a single point of its rim. Second, that contact point travels around in a circle on the mirror, which is what makes the sound: every lap is a tick, and the ticks blur into a tone. Third, the disk itself turns slowly about its own axis, which you can see if you draw an arrow on its face.

For an idealized disk rolling without slipping, the physics links the speed of the contact point to the tilt. Written out, it says the square of that whirl rate, Ω, equals gravity divided by the disk’s radius times a shape factor times the sine of the tilt angle: Ω² = g / (a·k·sin α). The smaller the tilt, the bigger Ω.

So as the disk leans lower, the contact point has to race around faster to hold it up. That’s the rising whine. Meanwhile, the slow turn of the face actually drops toward zero, and so does the disk’s total energy. The disk is running out of energy and buzzing faster at the same time, which is exactly what makes the ending feel so dramatic.

If nothing drained energy, the tilt would never change and the disk would spoll forever. Real disks always stop. The real question is what’s doing the draining, and how it makes the stop so abrupt.

The Nature paper: a singularity on your desk

On April 20, 2000, the Cambridge fluid dynamicist Keith Moffatt published a short paper in Nature titled “Euler’s disk and its finite-time singularity.” It turned a toy into a research topic almost overnight.

Moffatt’s idea was elegant. As the disk leans lower, the gap between its underside and the table becomes a thin wedge of air. The disk’s motion keeps squeezing that air in and out. Air is slightly sticky, or viscous, and pumping it through such a narrow gap costs energy. The flatter the disk gets, the narrower the gap, and the more energy is lost.

When he ran the mathematics, the result was a finite-time singularity. The whirl rate didn’t just climb, it climbed toward infinity at a definite moment, following a power law: Ω grows like (t₀ − t) raised to the power −1/6, where t₀ is the moment the disk would settle.

Of course, a real disk can’t whirl infinitely fast. Moffatt showed that the theory breaks down a tiny fraction of a second before the end. For the commercial toy, that cutoff comes roughly a hundredth of a second before t₀, when the disk’s acceleration would demand more than gravity can supply. At that point, it loses contact with the table and drops flat. That is the “switch” you hear at the end.

Then the coin went into a vacuum

A theory that blames air makes a clean prediction: take the air away, and the disk should spin much longer. Experimenters soon tested exactly that.

In November 2000, Ger van den Engh, Peter Nelson and Jared Roach replied in Nature under the wonderful title “Numismatic gyrations.” They spun a Dutch coin, a rijksdaalder, whose magnetic properties let them launch it at a precisely controlled rate, and they compared runs in air and in a vacuum.

Removing the air barely changed anything. They argued that slippage between the coin and the surface could explain what they saw, and pointed out that Moffatt’s air-only picture would predict a very long spin in a vacuum, which simply didn’t happen.

Moffatt’s published reply, in the same issue, offered a more general theory that could let experiments sort out which mechanism dominates. He also held his ground on one point: in the very last instant, when the tilt angle is tiny, viscous air dissipation should still win.

Friction takes the lead

Over the next few years, more labs piled on, and the evidence kept pointing the same way.

At the University of Guelph, D. Petrie, J. L. Hunt and C. G. Gray ran the disk in a vacuum at a pressure of just 0.1 pascal, about a millionth of normal air pressure. In their 2002 paper in the American Journal of Physics, “Does the Euler Disk slip during its motion?”, they reported that the vacuum didn’t significantly change how fast the disk lost energy. Swapping the disk for a ring, which has a very different underside, also made little difference. And they found the disk wasn’t slipping at tilts above about 10 degrees.

That same year, K. Easwar, F. Rouyer and N. Menon published “Speeding to a stop: The finite-time singularity of a spinning disk” in Physical Review E. Their disks behaved differently depending on the friction of the surface underneath, which is what you’d expect if rolling friction, not air, does most of the work.

In 2004, a Belgian team led by H. Caps came to the same conclusion in “Rolling and slipping motion of Euler’s disk.” Air was a minor energy sink. The main loss came from the disk rolling and slipping on its support. They also measured the tilt angle, the precession rate and the angular velocity all following power laws on the way down, so the singularity was real even if the cause was different.

So who was right?

Mostly the friction camp, with a twist that keeps Moffatt in the story.

In 2009, Remco Leine, writing in Archive of Applied Mechanics, studied several kinds of friction for a rolling disk. His experiments fit a model with two kinds of “contour friction,” which resists the contact point as it travels around the rim. One kind dominates most of the run. A second, “viscous” kind takes over in the final second or two.

The consensus picture today looks like this. For most of the performance, the disk is bled dry by friction where steel meets mirror. Air drag barely matters. Right at the very end, with the disk nearly flat, a viscous-type loss can take over, much as Moffatt said. And either way, the motion heads toward a finite-time singularity before contact is lost and physics cuts it short.

Moffatt’s particular mechanism lost the main argument, but his paper did something more lasting. It got a crowd of physicists to take a coin trick seriously, and that crowd answered the question.

The floating-ring illusion

One more effect is pure perception. Near the end, a well-lit Euler’s Disk often looks like it is hovering above its mirror.

What’s happening is a flicker too fast to see. Each part of the rim alternates between being lifted slightly off the surface, catching the light, and being down in contact, in shadow. As the whirl speeds up, those flashes come too quickly for the eye to separate. Your brain merges the bright reflections into a steady ring and drops the shadows, so the disk seems to float. The mirror base helps, hiding the surface and showing a gap between the disk and its reflection.

Watch one go

Seeing beats reading here. This video shows a long run all the way to the final buzz:

▶

Try it with a quarter

You don’t need the toy to see the basics. Any coin spun on a flat surface does the same thing, just for a shorter time. The commercial disk lasts longer because of its proportions and its polished, slightly rounded edge.

For a better home version, spin a clean coin on a flat hand mirror and watch from the side, close to the glass. You’ll get a few seconds of the rising whirr and the flicker. Then try a rough tablecloth, and see how much sooner the coin dies. That’s friction, the winner of the argument, showing up on your kitchen table.

It’s a nice reminder of how science often works. A lunch-break toy from an aerospace lab led to a Nature paper, a run of rival experiments, and a better answer than anyone started with, all about a sound most of us first heard spinning pocket change.

Further reading