Scary moment on ride yesterday

[Jdm]

Originally Posted by unterhausen

Every bike has a speed it will shimmy at. The better ones only shimmy at speeds you will never reach. But I can see how deep rims will hurt a bike's shimmy characteristics.

Why does a bike shimmy and what determines the speed at which the shimmy happens? I've been noticing this more on my bikes at speed and assumed it was gusts of wind or something off balance in a tire.

[blakcloud]

I had a problem with shimmy at the beginning of the year. Like the OP I thought it was the deeper dish wheel getting crosswinds, but I have two other sets of wheels, less deep and the bike did the same thing. It started at 50 km/h.

I found a solution and it was my error. When reinstalling the ISO speed front coupler system I had the sleeve turned 90 degrees and this was enough to make the bike unstable. Fixed this and now it starts to shimmy at 80 km/h which is a speed I will never do again.

[Turkle]

Little changes can make a big difference. I had a bike that had zero shimmy with 650b wheels, but shimmied dangerously with 700c wheels.

On that bike (it was steel), resting my left knee against the top tube was generally enough to eliminate the shimmy entirely.

[Mark McM]

Originally Posted by Jdm

Why does a bike shimmy and what determines the speed at which the shimmy happens? I've been noticing this more on my bikes at speed and assumed it was gusts of wind or something off balance in a tire.

https://velo.outsideonline.com/road/...-to-stop-them/

Long story short, unlike popular belief, a shimmy is not caused by something being "broken" or out of alignment or out of balance - a shimmy is an inherent system property that occurs within a particular speed range. Which is not to say that a shimmy can't be "fixed" - but the fix is to change system parameters to move the shimmy speed into a speed range the bike will never travel.

[deluz]

I was reading about this and one thing suggested is to press your knees against the top tube and unweight the saddle. So if it ever happens I will know what to do.

[benb]

IIRC (I am no expert on this) and in my experience what I have seen is that if you have a well designed bike that is designed to not shimmy until a very high speed you can see wind + bumps initiate a shimmy but the bike should not start oscillating out of control and increase the severity of the shimmy.

I believe Dave Kirk posted some very interesting stuff about this in the past about studying this at Serotta.

Most of my bikes including my Serotta Concours and even my motorcycles have exhibited this kind of behavior where a bump and/or wind at the right time would initiate a shimmy but it would resolve on it's own. On motorcycles the big initiator has always been high acceleration and you're either getting the front wheel light or you've coming down from a wheelie and the front tire hits something. Race replica bikes will have a damper built into the steering to prevent it.

We have had a ton of really high winds the last month or two. My Domane does not really shimmy at any reasonable speed I've experienced (but I have never gone above 50mph on it) but it has Ksyriums on it and those really don't do that great in extreme cross winds... I have be very cognizant of it to not tense up my body and there have been a handful of rides I have stopped pedaling and slowed down it's been so windy.

I would have thought a Tarmac would be so well engineered at this point it's shimmy speed would be at some outrageous speed like 75-100mph. Pretty much every carbon frame I've had that had a huge BB and a tapered 1 1/8" to 1 1/4" fork has been almost completely immune to it.

My gut feeling is newer bikes have been getting more and more resistant to this as tires have gotten bigger and at least on carbon they're doing tricks with compliance that let them keep the frame/fork extremely stiff in all the right places.

[unterhausen]

Originally Posted by Jdm

Why does a bike shimmy and what determines the speed at which the shimmy happens? I've been noticing this more on my bikes at speed and assumed it was gusts of wind or something off balance in a tire.

I don't think the speed that any given bike will shimmy at can be determined in any other way than experiment.
It's actually the nature of bicycle steering geometry. The stability of the geometry going straight is not that robust. If you think of it that way, many small issues with the bike can reduce that stability. The steering does self-center, but in a situation like a wind gust, the forces that self-center can push the wheel too far past center. Then it re-centers and goes too far the other way. And repeat. That's shimmy.

[benb]

Originally Posted by unterhausen

I don't think the speed that any given bike will shimmy at can be determined in any other way than experiment.
It's actually the nature of bicycle steering geometry. The stability of the geometry going straight is not that robust. If you think of it that way, many small issues with the bike can reduce that stability. The steering does self-center, but in a situation like a wind gust, the forces that self-center can push the wheel too far past center. Then it re-centers and goes too far the other way. And repeat. That's shimmy.

I think there are people who have actually put the hard engineering time in to understand it but this stuff really really well but they're on the moto side and they're not necessarily sharing it, they're protecting it very carefully as a competitive advantage and racing advantage.

They are at the level on motorcycles where they also think about how much the fork or frame can oscillate side to side (not on the axis of the suspension) and tuning that for extreme lean angles.. and it apparently changes with tire grip levels as one of the variables. They can do extremely stiff on a motorcycle but it sounds like it's not optimal.

You're not going to do that without actually engineering it. Especially if you're Ducati/Honda/whatever and the bike is a million dollars+ for Moto GP and trial and error testing could be very dangerous.

It's pretty fascinating to look at the history of motorycle frames and how they developed out of bicycle frames. The early ones were basically bicycle frames, no different than eBikes are today, and frames on cruisers are often still centrally located just like a bike frame.

[Mark McM]

Originally Posted by benb

IIRC (I am no expert on this) and in my experience what I have seen is that if you have a well designed bike that is designed to not shimmy until a very high speed you can see wind + bumps initiate a shimmy but the bike should not start oscillating out of control and increase the severity of the shimmy.

I believe Dave Kirk posted some very interesting stuff about this in the past about studying this at Serotta.

Most of my bikes including my Serotta Concours and even my motorcycles have exhibited this kind of behavior where a bump and/or wind at the right time would initiate a shimmy but it would resolve on it's own. On motorcycles the big initiator has always been high acceleration and you're either getting the front wheel light or you've coming down from a wheelie and the front tire hits something. Race replica bikes will have a damper built into the steering to prevent it.

A speed shimmy is an oscillation, but not all oscillations are a speed shimmy. An oscillation that starts with a bump or wind gust but then dies away on its own is most like a system natural frequency oscillation. But a speed shimmy will continue as long as the bike is traveling within a certain speed range. These types of shimmies are usually caused by a Hopf Bifurcation.

Originally Posted by benb

I would have thought a Tarmac would be so well engineered at this point it's shimmy speed would be at some outrageous speed like 75-100mph. Pretty much every carbon frame I've had that had a huge BB and a tapered 1 1/8" to 1 1/4" fork has been almost completely immune to it.

My gut feeling is newer bikes have been getting more and more resistant to this as tires have gotten bigger and at least on carbon they're doing tricks with compliance that let them keep the frame/fork extremely stiff in all the right places.

Both system natural frequency oscillation and Hopf Bifurcation critical speed are dependant on the system parameters, and a change in system parameters can change the frequency and/or critical speed. The system parameters includes things like wheels, tires, stem length, etc. that will probably be changed for every rider, and also the big variable of the rider themselves (size, shape, mass) - a bike that shimmies with one rider may not shimmy with another rider. It is difficult for bike manufacturers to account for all possible system parameters. While they can design bikes that is unlikely to shimmy for most use cases, there will always be corner cases in which they can shimmy.

[unterhausen]

Originally Posted by benb

I think there are people who have actually put the hard engineering time in to understand it but this stuff really really well but they're on the moto side and they're not necessarily sharing it, they're protecting it very carefully as a competitive advantage and racing advantage.

Now you've made me wonder if there is a rider model out there I can get my hands on. I think it can be analyzed successfully. The wikipedia article about shimmy had a good paper about it. Interestingly it's devoid of mathematical woo.

[smontanaro]

Originally Posted by unterhausen

Now you've made me wonder if there is a rider model out there I can get my hands on. I think it can be analyzed successfully. The wikipedia article about shimmy had a good paper about it. Interestingly it's devoid of mathematical woo.

I've got a tab open in my browser with this treatise:

https://ciechanow.ski/bicycle/

I never made it all the way through, but keep meaning to get back to it.

[bshell]

My two cents for anyone where the 1 knee (or 2) against the top tube doesn't work.

In 35+ years of riding all disciplines I've had speed wobbles one time (@4 years ago). And that's using my bikes repeatedly between 0-53mph(never again).

I was in a fast descent on a steep road and went to the top tube move I'd heard over the years but it did NOTHING. Neither did shifting weight fore or aft. It was terrifying.

What felt like an eternity had to be under 10 seconds. There was a curve ahead that I knew I wouldn't make so I started gently applying my rear brake. If I were going to eat it over a guardrail and into a canyon I wanted to be going as slowly as possible. That worked, and I was able to make the turn.

There are a lot of variables in things like this but I'd def not rule out equipment (funky tire, funky wheel, the slightest loose headset etc.)

Strong gusts are a separate issue. Crosswinds gonna crosswind and it won't matter which company made your 35mm - 40mm - 60mm rim.

[Andy340]

Originally Posted by tellyho

Can't really see the value of 40mm wheels. I don't go fast enough for it to make a difference.

I was thinking of getting a set of 50mm wheels in BF sales (for the primary benefit to me that they look better) but reading this thread I will stick with 30mm deep Levantes as my rides go up and down quite a bit.

[fogrider]

The velo article seems to cover the bases. The point is that its a combination of frame, wheels, components, rider and speed that determines speed wobble. And stiffer bikes will require higher speeds to induce it. But since weight is a damper, the lighter a the combination the more prone it is. I have a set of wheels that is balanced, which one would think should help.

Originally Posted by benb

IIRC (I am no expert on this) and in my experience what I have seen is that if you have a well designed bike that is designed to not shimmy until a very high speed you can see wind + bumps initiate a shimmy but the bike should not start oscillating out of control and increase the severity of the shimmy.

I believe Dave Kirk posted some very interesting stuff about this in the past about studying this at Serotta.

Most of my bikes including my Serotta Concours and even my motorcycles have exhibited this kind of behavior where a bump and/or wind at the right time would initiate a shimmy but it would resolve on it's own. On motorcycles the big initiator has always been high acceleration and you're either getting the front wheel light or you've coming down from a wheelie and the front tire hits something. Race replica bikes will have a damper built into the steering to prevent it.

We have had a ton of really high winds the last month or two. My Domane does not really shimmy at any reasonable speed I've experienced (but I have never gone above 50mph on it) but it has Ksyriums on it and those really don't do that great in extreme cross winds... I have be very cognizant of it to not tense up my body and there have been a handful of rides I have stopped pedaling and slowed down it's been so windy.

I would have thought a Tarmac would be so well engineered at this point it's shimmy speed would be at some outrageous speed like 75-100mph. Pretty much every carbon frame I've had that had a huge BB and a tapered 1 1/8" to 1 1/4" fork has been almost completely immune to it.

My gut feeling is newer bikes have been getting more and more resistant to this as tires have gotten bigger and at least on carbon they're doing tricks with compliance that let them keep the frame/fork extremely stiff in all the right places.

[redir]

My Moots Compact is the only bike I have that doesn't high speed wobble. The first time it happened to me I was doing about 50mph down a mountain road and it was wild. I saw the face of one of the drives coming up the other way he was shocked at what he was seeing. I don't know how I recovered from it because it was such a violent wobble but eventually I slowed it down. I posted it on Bikeforume.net and the guys thought me about the knee to the top tube trick and I've never had the problem since. As soon as I feel and oscillation the knees clench the TT and it's over.