Monday, April 07, 2008

Why are large frames more prone to shimmy?

Over on the Serotta Forum the subject of shimmy was being discussed; this subject is probably discussed on bike forums more that any other. One member posted the following:

“Am I nuts, or do all shimmy prone bikes have one thing in common?
Large size frames.
I ride a 52cm - 53cm frame. I've never experienced shimmy in any of
the many bikes I've owned. It seems like every shimmy story has a tall person in the starring role.”


No, you are not nuts, large frames are more prone to shimmy. First of all, shimmy is a natural occurrence on two wheeled vehicles, bicycles and motorcycles have the same problem.

At a certain speed a spinning wheel will start to nutate, That is, not only is the wheel spinning about its axis, (The wheel spindle.) the axis itself is constantly moving. To demonstrate this to yourself, hold a wheel by the spindle in your outstretched hands, and move your hands in the motion of pedaling a miniature bicycle.

Even a child’s spinning top will hold steady with its axis vertical if it is spinning at speed. One would think it would spin slower and slower until it gently falls over. However, as it slows to a certain critical speed it will start to wobble then fall. In other words, it starts to nutate at a certain speed.

“Rotation” means an object spinning around a fixed axis; “nutation” means the axis is also moving as the object spins. The front wheel is not only fluttering back and forth in the vertical plane, but in the horizontal plane also. Therefore, the head tube of the bicycle or motorcycle is shaking violently from side to side.

The rear wheel does not shimmy because it is fixed withing the frame. Just as on three or four wheeled vehicles the wheels are not prone to shimmy because the axels are held in one plane. Although on older cars, for example, when wheel bearings and steering joints start to wear allowing the wheel’s axis to move, they too will shimmy or wobble at a certain speed.

On a bicycle, most of the rider’s weight is towards the rear. The rider’s weight on the saddle and the rear wheel in contact with the road provide two anchor points holding the rear of the bike steady, while the front end can start to shake violently.

Because the seat tube slopes rearwards, as the frame gets taller the rider’s weight is more directly over the rear wheel. On a large frame this makes for a near vertical pivot line between the rider’s weight on the saddle, and the rear wheel’s point of contact. (See picture below.)

It is about this pivot line that the bike will start to shake, and if the rider then grips the handlebars tighter, his body will also start to shake along with the bike and a crash will probably ensue.

When nutation and the resulting shaking starts, it will only get worse unless speed or weight distribution changes; especially if the rider becomes part of the shaking mass.

On a motorcycle shimmy, (Or tank slapper, as they are called.) they are often so violent the rider is thrown from the machine. In this motorcycle tank slapper video, you will notice the front wheel is not just fluttering side to side about the steering axis, but the wheel is also moving side to side about a horizontal axis, throwing the whole bike and rider side to side.

On a smaller bicycle frame, the rider’s weight is more forward and the pivot line is less than vertical. (See picture below.) This means that even if the rider is riding “no hands” there is still a certain amount of weight on the front wheel.

This is a clue to avoiding shimmy if you are tall and ride a large frame. When descending at speed, move your weight forward and keep your back low so that air pressure on your chest is not forcing more weight to the rear.

Transfer weight from the saddle to the pedals, thus breaking one of the solid anchor points. Often a knee pressed against the top tube will dampen a shimmy. A loose headset may not cause a shimmy, but tightening a headset very slightly may have a dampening effect. Don’t over tighten a headset, or this in itself will make the steering erratic.

This is the third time I have written on this subject, and I don’t want to keep repeating myself. However, it is a subject that will always be around and so will continue to be discussed, and continue to surprise those who experience it for the first time.

In my first article “High Speed Shimmy” I called it a design flaw. This may have been a little strong, but frames I built did not shimmy as a rule, even the larger sizes; so, design and construction do play a role. The only time I was told a Fuso shimmied was when rear pannier bags were fitted, and the frame was not designed for this purpose.

I built my frames with slightly more trail than most other bikes; maybe this factor was enough to prevent shimmy. It doesn’t take much to alter a bike’s handling characteristics. Sometimes different wheels or a slightly heavier tire is all that is needed to stop a bike from shimmying.

I went into the subject in depth in my second piece called “Shimmy Re-visited.” I am now of the opinion that this is not so much a flaw, but a natural phenomenon inherent in any two wheel vehicle. Simply because the front wheel’s axis is free to move in any direction.

All the years I built bicycles, I never gave this subject much thought; I didn’t have to because I never had this problem. I do not have an engineering or science degree, and those who do will no doubt correct me if I am wrong.

I have not written about harmonic vibrations and the reasons why wheels start to wobble, what is needed is not more theories as to why this happens, but ideas to minimize the problem.

Bicycle designers and manufacturers should be concerned, and be looking for a cure. In the mean time, all an individual can do is get to know the limitations of their bike, and ride within those limitations.


8 comments:

Anonymous said...

Years ago, I experienced shimmy on a few scary as hell occasions. As a tall young to late teen, I owned two basic 25” road bikes. It was “the cool thing” back then to ride a bike that was two or three inches too large, and I suppose there was some “he’ll grow into it” factored in to the size choice. In the early eighties, as a more knowledgeable 20 year old, I began riding a properly fitted 58cm bikes of higher quality, and I have not experienced shimmy since.

I did not ride those old 25” frames with an overly high saddle for my leg length, so the fore/aft weight distribution wasn’t changed relative to the rear wheel, but the bigger frame had a longer top tube, head tube, and seat tube. In fact, since the top tube/stem was too long, I’d shove the saddle forward to compensate, resulting in a more forward center of mass.

So, was it was the inherently less stable long frame tubes that contributed to my shimmy experiences? Combined with the lower quality bike? Seems there are several hard to nail down factors that contribute to the problem, but I agree large frames are a big one.

Ron George said...

Nice post.

There are several universities across the world who are investigating the non-linear stability mathematics of bicycles.

From a practical standpoint there's valuable advice out there on how to reduce shimmy's. I like to understand the phenomenon from a model aspect.

Mathematical models are invaluable to understanding the real behavior of systems. Only if folks get the numbers right down on paper do they send a space craft to the moon or start the LHC at Geneva, which is big news now.

I hope computers can come to our help in understanding the shimmy.

Anonymous said...

nice post. do you post on the serotta forum? if so, what is your posting 'name'?

Yokota Fritz said...

I've had one bike that shimmies starting at about 35 mph. A little scary but just touching my knees on the top tube dampened things enough to make it stop. I've taken other bikes to over 50 mph without any problems, though.

That motorcycle video was dramatic -- wow!

Anonymous said...

Dave,

As a long time motorcyclist as well as cyclist, this is one problem I try very hard to avoid. I have taken a few clues from the motored version and incorperated a few "insurance" steps. First, keep an eye on tire wear. If the tire begins to square off to a significant degree I replace it. Second I actually balance my bicycle wheels just as one would balance any other rotating assembly. I use solid core solder wrapped around the spokes up against the nipples. Start by weighting the light or top of the wheel where it come to rest. Once the wheel always stops in a different position it is balanced. With just these two simple steps and perhaps a bit of luck, shimmy has never been a problem. BTW, I ride 60-61cm frames.

Brent said...

Dave,

I just returned from a month of work in Europe and had a chance to ride a bit in Italy. One of the guys I met there stopped riding his new carbon frame, 55cm, because of high speed shimmy on the downhills. He's now riding a steel frame, same size, solid as a rock. How much does frame weight or overall bike weight contribute to the likelihood of shimmy?

Redtaildd said...

I had a neighbor at one time who was a motorcycle tester. He told me about how he would be called as an expert witness in cases where motorcycle cops were suing the manufacturer for their injuries. Usually two broken arms from trying to stop the shimmy. He would take the offending bike out on a runway, get it up to speed, take both hands off the bars, and hit one side with the heel of his hand to induce shimmy. No hands. He would then lean forward and the shimmy would stop... Still no hands. He told me that if you try to stop it you make it worse. As a side note, he was in his late 60s or early 70s, ex-CHP.

Bujiatang said...

at the store I where I work we have two high-speed buffers. And we just put a new wheel on one of them. And it shook the whole bench it was attached to. As it turns out (har har) the hole the spindle goes through was slightly off center. (I was the one who noticed the off center hole, we all were afraid the motor was dying).

Since some bikes do have shimmy and others don't, leads me to wonder: are those the ones where the wheel is not out of true and the hub is off center.