Why Does a Moving Bicycle Stay Upright When a Parked One Falls Over?

A Bicycle Stays Up by Steering Toward the Side It Is Falling To

The first thing to know is what “staying up” actually involves. A bicycle that starts to lean has its wheels in the wrong place, off to one side of the weight on top. A rider (or a rolling bike) fixes this by steering toward the lean, which brings the wheels back under the weight [1].

Riders normally do this with their hands on the handlebars. With hands off, bending your body to one side can also make the bike steer. The surprising part is that many moving bicycles with no rider at all can steer themselves this way, and the same goes for a stiff dummy rider with hands off the bars [1]. The 2011 paper behind most of this note points out that a 1949 film, Jour de Fête, even shows a riderless bicycle balancing for long distances [1].

So the real question is narrower than it first looks. What makes a bicycle turn the right way, by the right amount, when it starts to fall?

The Spinning Front Wheel Was the Usual Answer, and It Seemed to Make Sense

A fast-spinning wheel resists being tipped, and when you do push on it, it responds sideways rather than the way you pushed. That is the gyroscopic effect.

Here is how the idea was supposed to work on a bike. When the bicycle leans to the right, the spinning front wheel pushes the handlebars to turn right, toward the lean. The 2011 paper traces the most careful version of this argument to a four-volume work on gyroscopes by the German scientists Felix Klein and Arnold Sommerfeld. They concluded that without the gyroscopic effect, “the speed range of complete stability would vanish” [1].

Notice that the claim is not that the wheel holds the bike up by brute force. The front wheel is light compared with a bike and rider. The idea is that the spin works as a trigger: a lean makes the bars turn, and the turning does the real work of catching the fall [1].

In 1970 a Bicycle With Its Wheel Spin Cancelled Out Could Still Be Ridden

If the spinning wheel was essential, then cancelling its spin should make a bike impossible to ride. The British scientist David Jones set out to build exactly that, which he called an unridable bicycle. On one version he attached an extra wheel next to the front wheel, spinning the opposite way, so that the two spins cancelled out [2].

It did not work out the way he expected. He wrote in Physics Today in 1970 that the bike could be ridden easily [2]. The 2011 paper adds a detail: with no hands, Jones could still ride it, though only barely [1].

So the spinning wheel was not the only thing linking leaning to steering. Jones pointed to something else, which sits in the way the front wheel is mounted [1].

Trail Means the Front Tire Touches the Ground Behind the Point Where the Steering Line Lands

Look at a bicycle from the side. The fork turns around a slanted line, the steering axis. Extend that line down to the ground and it lands a little in front of the spot where the front tire touches the road. That gap is called trail. On modern bikes it is typically between 2 and 10 centimeters (about 1 to 4 inches) [1].

Jones compared this to the wheels of a shopping cart, which trail behind their pivots so that the cart ends up straight when you push it. Roll a bike forward while holding it upright, and the front wheel lines up the same way [1]. Jones also worked out that when a bike leans, the weight pressing on the front tire tends to turn the bars, and that for a typical bike this turn is toward the lean, but only when the trail is positive [1].

To check, Jones changed his bike so the tire touched the ground in front of the steering line, which is negative trail. With that change he could no longer ride with no hands [1]. So there were now two favorite explanations: the gyroscopic effect and trail, working alone or together. By then it was widely believed that at least one of them was needed [1].

In 2011 Researchers Removed Both the Spin and the Trail, and the Bike Still Caught Itself

A team from Delft University of Technology in the Netherlands, the University of Twente, the University of Wisconsin–Stout and Cornell University in the United States began from the math rather than a hunch. They used the equations that describe a simplified bicycle. With the gyroscopic terms set to zero, the example bike they started with did lose its self-stability, just as Klein and Sommerfeld had said. But the equations also showed other bike designs that stayed stable with no gyroscopic effect [1]. They also found sign errors in Klein and Sommerfeld’s key calculation, so that claim could not be applied to bicycles in general [1].

Then they built one. Their experimental bike used small wheels to keep the spin low, and extra disks spinning backward to cancel what was left. Its front tire touched the ground 4 millimeters (about 0.16 inch) ahead of the steering line, so the trail was negative. It looked like a folding scooter, but it still counted as a bicycle: two wheels, two frames and three hinges [1].

After a push, it stayed upright while it coasted. When the researchers hit it sideways at a speed above about 2.3 meters per second (a little over 5 miles per hour), it swung back to straight, upright rolling. As it slowed below about 2 meters per second (about 4.5 miles per hour), it began to fall. That fit the speed the calculations had predicted [1]. A Cornell student newspaper covered the result that April, quoting Cornell’s Andy Ruina: “A bicycle’s not all that simple” [3].

The Weight of the Front End Matters Too, and It Interacts With Everything Else

If neither spin nor trail was doing the job, what was? In the experimental bike, the front assembly (handlebars, fork and front wheel) has its weight lower than the rest and ahead of the steering axis. When the bike starts to fall, that low front mass tips over faster than the tall rear part, for the same reason a short pencil balanced on its end falls faster than a tall broomstick. Because the two parts are hinged together, the faster-falling front steers into the fall [1].

The authors do not claim to have found one neat rule. They write that they found no simple physical explanation for why a bike steers the right amount at the right moment, and no simple condition that every self-stable bike must meet. They do not say spin and trail are unimportant: both are often important contributors. But almost any self-stable bike, they found, can be made unstable by changing only one of three things: the trail, the front-wheel spin, or the position of the front assembly’s center of mass [1].

One thing does hold in their equations. If you let go of the handlebars while a self-stable bike is in a steady turn to the right, the first thing the bars do is turn further right [1]. This matches the older picture: a bike that is falling turns toward the fall.

The Short Answer Is That No Single Part Does It, and the Details Are Still Being Worked Out

Putting it together: a moving bike stays up because leaning and steering are linked, so that a lean makes the front steer toward the fall. The gyroscopic effect, trail and the weight of the front end can each help build that link, and none of them is required on its own. Whether a real bike is self-stable depends on how they are combined [1].

One limit is worth knowing. The experiments and equations are about a bike rolling by itself. A person on a bike is a much more active part of the system, steering and leaning all the time. The researchers say that rider-controlled balance and self-stability are related, which is why self-stability is worth studying, but they are not the same thing [1].

Try It Yourself: Watch Your Handlebars Steer Into a Wobble, in a Wide Open Space

For this you need a bicycle, a helmet, and a wide, flat, empty place with no cars, such as a quiet parking lot or a school playground on a weekend. Ask a grown-up to come along.

1. Put the helmet on first. Buckle it so that it sits level and does not slide back, every time you ride.
2. Ride slowly in a straight line and watch your handlebars. See whether you notice small turns toward whichever side you start to lean. That is the steering-into-the-fall that keeps you up.
3. Try a bit faster, then a bit slower. Notice how the wobbles change. Keep both hands on the bars the whole time.

If you would rather stay on your feet, walk the bike beside you while holding the seat and keeping it upright. Tilt it gently to one side as you walk, and see which way the front wheel turns. Do not try to ride without hands.

Sources

Update history