[CR]Elementary Control-Theory Mistake in Jobst's article on Shimmy


Example: Framebuilders:Alberto Masi
From: "Donald Gillies" <gillies@cs.ubc.ca>
Date: Thu, 15 Jan 2004 11:34:57 -0800 (PST)
To: classicrendezvous@bikelist.org
Subject: [CR]Elementary Control-Theory Mistake in Jobst's article on Shimmy

> It cannot be fixed by adjustments because it is inherent to the
> geometry and elasticity of the components.

Proper headset damping enables shimmy. We all know from Control Theory that a system can either be under-damped (it oscillations are larger and larger, going out of control), over-damped (oscillations are reduced until they disappear), or critically damped (oscillations perpetuate themselves but do not grow or shrink).

Headset tightness is the main variable affecting damping in a bicycle that shimmies.

Clearly, you can crank down a headset so tightly that the bike only rides in a straight line and shimmy is impossible. You will fall off pretty quickly, but this is clearly an example of an over-damped system.

It should always be possible to reduce shimmy by increasing headset tightness, no matter what speed the bicycle is going. You will take a bike from the under-damped state, to the critically-damped state, to the over-damped state, which will eradicate shimmy.

The damping from a properly adjusted headset, and from a headset that is "too loose", is nearly the same. In both cases the handlebars will spin freely. Therefore, Jobst's assertion that properly adjusted headsets shimmy just as much as loose ones is correct.

I have experienced shimmy in riding no-handed on only one of the bikes I have owned. On my oldest ten-speed, a raleigh grand prix, when a shimmy appeared (it was not there when the bike was new), i eradicated it by increasing the tightness of the headset. Importantly, this appeared only after the fork had been straightened and bent. Moreover, I think the shimmy disappeared again when a new fork was installed. This is the key to my next theory about shimmy.

I wonder if it is possible for a bike to shimmy if the head tube is not faced or aligned correctly so that the lowest energy state for the headset occurs when the handlebars are not pointing directly forward. Then it may be possible for the system to oscillate between the two low energy states (one where headset is at lowest energy state, one where the trail of the wheels causes a lowest energy state). I will have to think more about this possibility.

Imagine that a bike shop customer tried two bikes of the same model, and they were adjusted more or less the same, and one of them shimmied while the other one didn't. This would support my theory.

- Don Gillies
San Diego, CA