Shimmy, or speed wobble, is a term for a shaking of the bicycle which usually occurs at high speed. Shimmy can be very scary, and can lead to loss of control. The helmet-cam video below records a shimmy incident. I was riding a borrowed bicycle at 35 mph on a long, steep downgrade, Market Street in San Francisco, California, USA. The bicycle had a suspension front fork and a rather heavy front wheel. I stopped the shimmy by braking lightly to reduce my speed.

All bicycles are subject to shimmy under the right (wrong) circumstances, but it is more pronounced with some than with others.

The rider feels shimmy mainly through the handlebars, so it is often assumed that the cause relates to the front wheel or the headset. This is most often an illusion.

Jobst Brandt gives a scientific explanation, and I agree with him. His crucial words are "[t]his steering action twists the top tube and down tube, storing energy that both limits travel and causes a return swing. Trail (caster) of the fork acts on the wheel to limit these excursions and return them toward center."

Let's expand a bit in this description. There is a mass -- the front wheel, fork and handlebar assembly of the bicycle. Thanks to the rake of the front fork, the center of this mass is ahead of the steering axis. Moving the front end of the bicycle to the right tends to steer the front wheel to the left, and vice versa. You can see this if you hoist the bicycle over your shoulder and swing the front end from side to side: the wheel steers the opposite way. If you get the wheel spinning before you hoist the bicycle over your shoulder, you can see that the wheel's rotational momentum (gyroscopic action) strengthens the tendency of the front wheel to steer opposite the direction in which it is moved. Move it to the right, and it steers to the left; move it to the left, and it steers to the right.

A mass supported by a spring makes a resonant system: A simple example is a weight hanging from a spring: if the weight is pulled down and let go, it will oscillate up and down around its rest position. A car with worn-out shock absorbers offers another example: it will bounce up and down if you push down momentarily on the bumper. Yet another example is the balance wheel of a mechanical watch or clock, rotating back and forth with its spiral spring providing the restoring force.

The spring which supports the mass of the front end of the bicycle is (simplifying slightly) the front triangle of the frame. This can twist, tilting the front fork and front wheel to one side and the other.

A momentary push will set up an oscillation which dies out over time. This happens with the weight hanging from a spring, once you let it go, or with the car that has worn-out shock absorbers when you push the bumper down and then release it. With many bicycles, when riding at a moderate speed, you can feel an oscillation that dies out if you bump the top tube with your knee, or give the handlebars a quick shake.

For an oscillation to keep going, though, energy must be replenished, producing a forced oscillation. This is the case with the balance wheel of the watch, turning back and forth, back and forth, driven by the pawls of the escapement, and with a violin string: the bow pulls the string until friction can no longer hold it, then the string snaps back beyond its rest position and the bow grabs it again. This happens over and over, producing a musical note.

The bicycle's shimmy is also is a forced oscillation. The forward movement of the bicycle provides the energy to keep the oscillation going. The front wheel's turning away from straight ahead converts forward motion into sideways motion at the bottom of the wheel. As the front wheel passes through its center position (directly under the frame) from right to left, it is turned to the left and its steering angle is pulling it to the left, but its trail and momentum are steering it back to the right. As the twist of the frame shoves it back to the right, the trail and momentum turn it back to the left. The wheel takes time to steer opposite the direction of the frame's twist, and so the front tire's road contact patch weaves rapidly from side to side as it moves forward along the road.

Shimmy works very much like the way a fish's tail drives the fish through the water, only in reverse: the side-to-side motion of the tail at the rear of the fish drives the fish forward; the forward motion of the front of the bicycle drives the side-to-side motion of the front wheel.

Violent shimmy typically starts suddenly when the bicycle is above a certain threshold speed, usually above 25 mph (40 km/h). One way to avoid shimmy is to hold speed down.

Certain bicycles are notorious for shimmy: one is the legendary Peugeot PX-10 road-racing bicycle. It is OK when used for its intended purpose, but it can develop a violent shimmy when carrying baggage on a rear rack: the luggage "anchors" the rear end of the bicycle, so twisting of the front triangle can more easily provide the spring restoring force for the mass of the front fork and wheel. Cyclists who opted for the high-end PX-10 and used it for bicycle touring often met with this problem. Less expensive Peugeot frames were much more suitable for touring.

I have found that a bicycle with a frame that is stiff against twisting of the front triangle -- for example. one with large-diameter aluminum or carbon-fiber frame tubes -- is generally less prone to shimmy, because the resonant frequency is higher. A steel frame with plain-gauge frame tubes tends to be less prone to shimmy than one with butted tubing, for the same reason. A tall frame is more prone to shimmy than a shorter frame, because the tall frame is more flexible. A bicycle with a suspension front fork may be especially prone to shimmy, due to slop in the fork, and its large mass, which lowers the resonant frequency. A heavy front wheel also can increase the tendency toward shimmy.

So, all in all, the best bet to avoid shimmy is a stiff frame with a light front wheel and fork. If you need an especially tall frame, look for an unusually stiff one. These considerations are especially important if you are carrying luggage on the bicycle.

That shimmy occurs at high downhill speeds makes it especially dangerous. A violent shimmy starts suddenly, and so it can panic the cyclist. The shimmy reduces the traction of the front wheel, so leaning into a turn will very likely cause a fall. Steering to keep a shimmying bicycle balanced is generally still possible while traveling in nearly a straight line. On a winding downhill run, you may be faced with the choice to keep the bicycle balanced and go off the road, or lean it over and crash. This is one more reason to brake before entering a turn.

Light braking is possible while the bicycle is shimmying. Slowing will stop the shimmy. A light grip with relaxed arms is best, to keep your hands on the handlebars, and to operate the brake levers. Stiff arms will propagate the shake through your entire upper body and make it harder to keep control.

Jobst Brandt reports that unloading the saddle without standing up will stop shimmy. I haven't had the occasion to try that, but it makes sense because it reduces the "anchoring" of the back end of the frame. Placing a knee against the top tube also may help, by changing the resonant frequency, but I haven't tried that either.

Also see Jobst Brandt's explanation of shimmy.

For extra credit, name another forced vibration which occurs on a bicycle. If you don't have an answer, here's one. (And it isn't the only one, but we'll leave it at that for now.)