Technical FAQ: High speed shimmy

Dear Lennard,
My brother rides a 1996 Cannondale SR-800 road bike daily and he noticed a bad shimmy on a 25mph descent. It is worse at 30mph and he hit the front brakes gently to reduce the wobble. It was okay on the same descent before this at even 40mph. The headset is 10 years old. He inspected the bearings and the racesand re-lubed the headset. He has about 20k miles on the bike.
Adjusting the headset helped. Is the wobble due to the resonant frequency of the frame? The weight of the rider, the fork geometry and material? Where do you recommend he look and fix to eliminate the shimmy?
-Chet

Dear Lennard,
I recently did a fit upgrade on my Colnago Master Olympic which included putting a significant angle rise on my handlebars, extending the stem from 100 to 120, and replacing my narrow handlebars with a wider set (I am getting old, fat and stiff and my previous set up was making greater-than-50 mile rides a drag).

I notice on sharp downhills at speed of 35 mph or so that a pronounced shimmy is now on the front wheel. I have checked to make sure the wheel is true (it is) and the fork is set tight(it is) and the bearings are fine (they are). It is more pronounced as I brake so I try to only brake with the rear pads and the shimmy stays.

So I am perplexed at the shimmy. Could it be due to the increased stem length, the increased angle, or the wider handle bars? Did something else happen in the shop I am unaware of that could cause this? I bought the frame new and it has never crashed.
-John

Dear John (and Chet),
These things are often hard to track down, but it is strange that it did not happen before and now it does. While it was in the shop, it is possible something could have happened to the bike, but let’s at least assume that nothing happened there since (1) we have no idea and could not change it if it did, and (2) I really doubt anything did, if you trust the shop. Things that could change the ride characteristics require some force – a change in fork alignment (or cracking inside of underlying carbon layers in the fork), a change in front wheel spoke tension, rim alignment or tire straightness.

And first off, check your tire pressure– too soft is not good for shimmy-prone bikes. Shimmy is a resonance phenomenon and depends on the resonant frequencyof the entire bike/rider system. It is certainly dependent on the weight distribution of the rider over the wheels, so changing your position could have pushed a previously reliable bike over the edge into the shimmy range.

As an example of how rider weight distribution can affect shimmy, I’m including instructions from my friend Rolf Dietrich on how to ride out of a shimmy safely, once it occurs, by changing your weight distribution over the wheels:

“Through the years I’ve experimented with speed shimmy and Calfee’s suggestion that defective forks can cause the onset of shimmy sounds reasonable to me as a source among other variables to which you alluded. You made a suggestion as to what a rider encountering this condition should do and I’m convinced that your suggestion can help. The trouble is that it may not stop the shimmy. However, I believe there is one sure way that will always stop the shimmy regardless of its cause. My solution is premised on the fact that with a bike there are two contact points with the ground and these are physically displaced from each other along a line (at the bottom of each wheel, the Road Contact Point –RCP- of each wheel) and connected to each other through a laterally semi rigid structure, the bike frame.

Onset of shimmy requires a lateral force input that in its effect arrives at the two RCPs at different times and in so doing causes a flex in the connecting structure between the two RCPs. In certain circumstances a positive “feedback” condition between the two RCPs can cause the oscillation of the flexing frame to diverge, become larger with time, setting up the condition we call shimmy. The oscillations are similar to a vibrating string held between two rigid anchoring points. In the case of the oscillating bike, the two anchoring points are the respective RCPs.

Eliminate one of the anchoring points and the vibrations stop instantly because the feedback is instantly eliminated. Easy enough with a vibrating string. On a vibrating bike the same can be achieved by reducing and even eliminating the road contact friction between the front tire and the road. If the rider reduces the weight on the front wheel by dragging the butt off the saddle to the rear and thus shifts his/her center of gravity to the top of the rear wheel, the vibrations will stop as soon as RCP friction is reduced or eliminated at the front wheel.

This is not difficult to do because on most bikes the rear axle sits only about 7 or 8 inches behind the center of the saddle and if one slide sfar enough back to place the center of the chest on the saddle one can actually totally unload the front wheel. To counteract the frontwheel totally lifting, one simple feathers the rear brake and the associated forward weight shift will keep the front wheel planted until one comes to a safe stop. With no – or reduced – RCP friction at the front wheel, the bike can no longer vibrate or shimmy. I’ve safely recovered from high-speed shimmy with this technique and it works every time and it allows total control through the maneuvermto a safe stop.”
-Rolf