Herbert did a whole thread devoted to this.

---

2024: Bevoman, Galveston, Alcatraz, Marble Falls, Santa Cruz

It is a vibration, a resonance.
Happen sometimes.

I had that on the track at hight speed sometime (50km/h) when doing quick direction change (sprint prep.), or with my P3, at lower speed, when having an big bottle between aero bars, and removing my hands from handlebar.
Both bikes are rigid (Dolan DF4 and Cervelo P3), wheels also (carbon 5 spokes and 55mm SLR0).

It is a mix of aero forces and mechanical forces (tires, wheels, fork, frame, things attacked to the bars, your hands and arms) resonating with significant amplitude, generally sound and feel like around 10 hz.

Probably comes from aero wheels resonating with fork and frame. In the case of P3 with bottle between aerobars, clearly this bottle weight and aero drag (creating torque in the fork) was key.

Scary, yes.

Yikes, I've had this happen before too. What wheels were on your Cannondale when you got the wobble?

It’s a new System six so the Knot 64mm deep wheels.

Hmmm, I just filled both bottles at the store on top of the hill. I wonder if their weight contributed.

I expect one of the biggest factors is rider/bike geometry setting up a positive feedback loop rather than resonance as such.
I doubt the bike's stiffness is much of a factor at all.

Consider a frame movement starting such that the top of the steerer moves suddenly left relative to the rider. The cause could be many things, hitting a bump, wind gust on the front wheel, whatever...
As the steerer moves left, perhaps before the rider has even identified the unexpected motion, the right handhold moves further from their right shoulder and they most likely apply an involuntary, maybe very small, tug to the right. Depending on the frequency and amplitude of the original disturbance, it seems likely to me that this could cause a positive feedback system where as the bike moves the rider inadvertently overcompensates. Not necessarily through an effort to get the bike under control but simply because they are holding both end of the bars and have a large mass moving relative to the bike.
Bracing the top tube with a knee and resisting the urge to grip the bars too tight is likely to disrupt the problem and get things back under control. It re-couples the rider and bike somewhat from a mechanical perspective.

A similar sort of situation is known in aircraft flight. It's referred to as pilot induced oscillation and is caused when the plane pitches slightly up or down, lets say up for the sake of this description: As the plane pitches up a little the pilot is pushed forward slightly in the seat and can't help but apply a small amount of pressure to the stick. If the control and stability characteristics of the plane are such that this pressure is enough to pitch the nose back down sufficiently, a feed back loop is established and the plane oscillates nose up and then nose down repeatedly until something changes. If memory serves the usual frequency where this can become an issue is around 1Hz and it normally happened with planes that had low damping in pitch like a tailless aircraft ("flying wing"). This has the potential to be very dangerous and is best tackled by not holding the control input too firmly and then finding a way to break the cycle. With a plane I think that could be a matter of deploying flaps, airbrakes, landing gear or changing throttle.

It's not really a failure of the pilot/rider that initiates it but techniques are needed to stop it.
I've had moderately bad speed wobble on three occasions I can recall, all on different bikes. The first was a very long time ago on a mountain bike set up with rear panniers and loaded down with 20kg of gear (I was bike packing across NZ just out of college). I can't remember if i did something specific to stop it. I think I just braked slightly and got lucky. The other two happened on road bikes while descending at speed and in both cases the oscillations were getting severe by the time I got my knee against the top tube and then it faded quickly. On was on a Ridley Orion rental bike and the other on my Canyon Ultimate. I don't think stiffness of any of the bikes was a significant factor.

Had this happen to me last summer on some mountain in Austria. It was a steep decline, 14-13% and a straight road for at least 2 or 3 kilometers. At 92kph/ 57.5mph my bike started shaking uncontrollably. It was shaking so badly, and with the high speed, I couldn't use my breaks very efficiently. Lucky enough it was a straight road without any corners, or else... I did manage to just barely evade a herd of cows passing on the road. Eventually I slowed down enough so the shaking would stop. But it took more than 2km to finally come to a stop. I had to recover a couple of minutes from that. Weird thing was that at 20kph it was still shaking...

This was the most scary thing I ever experienced on my bike. I already made peace with the fact that I would hit the pavement. Looking back at Strava this was where I hit my HR peak. First time that happened on the decent.

Still not sure as why this happened.

The fact that it persisted as speed dropped, IMO supports the argument that it's rider induced oscillation rather than true resonance. It takes a sudden significant disturbance to set up the motion but after the amplitude grows it persists even as speed reduces because speed is now only a small factor. Speed dictates how much the bike moves laterally for a given inadvertent control input, but after the oscillation amplitude grows, a much lower speed is sufficient to maintain it. It's also possible that after it has persisted for a short time, the rider is anticipating the oscillation and it becomes self fulfilling. Just a hypothesis, but a reasonable one I think.

I’ve had that a few times (never at such a high speed!) and it was terrifying. Glad you stayed up.

Did you try any of the tricks like clamping the top tube with your knees? If so, did it help?

Cat 1 MTB is a better bike handler than I’ll ever be. Shows it can happen to anyone.

---

Aaron Bales
Lansing Triathlon Team

Yes. It didn’t stop it but kept it from getting worse. It was the longest 20 seconds of my life lol. And I’ve done some really stupid shit over the years

I used to get speed wobble on my Norco CRR but since I dropped my stem and got a longer stem I haven't had it happen. My theory is that with more weight over the front end the front end isn't being "lifted" at high speeds so it's more stable. Whatever it was, it worked.

The only time it had ever happened to me, it felt the same - must have only last 15s but I had to get off the bike and gather myself for a minute or so, my whole body was shaking. Don’t have anywhere near the bike
handling skills you do, have been careful since then using the logic that not crashing will always be faster than squeaking out an extra 5mph downhill for a cyclist of my standard...

Longer stem reduces steering sensitivity which would reduce likelihood of positive feedback control situations occurring. Both a longer stem and a lower stem will tend to lower your "centre of gravity" which I suspect will also help by reducing the riders inertia during oscillations and reducing geometrical issues associated with moving relative to the frame.

I've been through it to varying degrees many times when I lived in the mountains. I think its a combination of a positive feedback loop, and some resonance of the bike/rider system. But, given the complexity and variety of human reactions there are probably a myriad of manifestations.

My own experiences and observations during the events were that the initial oscillations probably come from rider input, and a positive feedback loop exactly as you describe. Basically the rider reacting out-of-phase, which due to the time delay magnifies the oscillation instead of damping it. However, I think the sustainment of the oscillation comes from a resonance setup between the front end, and the rider mass with the arms acting as the springs. Eventually, the rider stiffens up and just tries to hold the handlebars still rather than actively countering the steering input. So, I think the steady-state phase is more resonant and less feedback-driven.

I've tried every method I've ever read of countering the wobble. I agree that the knee to the top-tube method mostly seems to lessen the effect UNLESS it is enacted very early on, ie in the transitory phase which I think is the feedback-loop portion of the event. Once steady-state is reached, which I'm positing is resonant, its much less effective. Relaxing or even limited release of grip on the bars is similar. But, that seems nearly impossible psychologically in later stages.

Unweighting the seat has been 100% effective within a cycle or two---something like left, right, swoop, done. It doesn't take much, just enough to allow the seat to move under the thighs without coupling the motion to the torso. Just get the weight off the saddle and onto the bottom bracket, and allow the bike to rock underneath the rider.

While descending I got into the habit of leveling the pedals, and bracing my thighs against the top tube whenever a gust or other perturbing force was experienced...just as a precaution. Also, once terminal velocity was reached I would do the same. I level the pedals so I can quickly unweight if the thigh-trick doesn't work. It wouldn't surprise me if I instinctively transfer a bit of weight to the pedals when I level them...but, that wasn't ever intentional.

I would be dedicated to Zwift following this event!

That's attitude makes me sad. I mean that sincerely.
As much as I like Zwift as a training tool, if it becomes a crutch for those afraid to face their fears it's net effect will be a negative one IMO.

Ever tripped while walking down the street? Better stay home.
Ever been rejected by a man/woman? Better stay single.
Ever been somewhere you didn't like, or eaten something you didn't like? Better stick with what you know.
Ever seen a car? Better just Zwift
Ever had a speed wobble? Better just Zwift
Ever had a puncture? Better just Zwift

Sad in the extreme

Worried about speed wobbles? Learn how to deal with them or don't descend so fast in conditions where they're more likely. I have plenty riding buddies nervous of descents (not specifically because of speed wobbles) who stick to 50-60km/h while I'm doing 75. We regroup at the bottom and carry on. No big deal. has anyone ever had a speed wobble on the flat? I doubt it! There are much better solutions than running away.

It happened to me once and the bike used is tolds to have a very stiff steering. It was racing and i was not super confident, i was riding fast on a dangerous route and i was wondering if i had a flat or a lose skewer just before. I guess my mind was quite messed up by the effort, the racing stress, the potential failure, an early race problem and the bad road and i had that death wobble mid a fast corner. Pretty sure it's rider induced.
I think heavier rider on flexy frame would exacerbate the problem but for most of us get your s*** together and it'll be fine.

I think we agree in parts but not sure if we're on exactly the same page. I'm not saying the positive feedback from the rider is conscious, active or avoidable. I think my hypothesis would work with a mannequin if you could set it up right. Stiff arms, ass on the saddle and even minor flex at the wrists, shoulders and torso would suffice I think.

Imagine you have a bike with a fairly rigid mannequin attached. You're holding the bike and the mannequin is sitting on it. give the seatpost a wobble and the mannequin, most of whom's weight is resting on the saddle, moves with it. Because the mannequin is attached to the handlebars, the motion of the mannequin will likely turn the bars. I'm saying the resulting motion can be such that it provides positive feedback by causing an over-correction so that an oscillation is set up which grows in amplitude.

Brace the mannequin's knee against the top tube and the mechanics change and the motion of the bars in response to the wobble will change. For a moving bike, where forward speed was previously causing the steerer yaw to sustain the oscillation, this would likely eliminate the feedback or change it's frequency or phase. This may eliminate the oscillation or perhaps just remove the source so that additional damping is still needed to eliminate the oscillation in an acceptable period.

Now back to the mannequin. Consider if you were to lift the it off the saddle by extending it's rather stiff, but not entirely rigid legs a fraction, so it's supported on it's legs just above the saddle. The system changes completely. It's now virtually impossible that the oscillation will continue as it was before. Maybe a different oscillation is possible but it's very unlikely to perpetuate the existing one, and there's loads of damping available. So in a real situation, this it can be used to eliminate existing oscillation.

really more of a joke...I personally do not descend faster than 30-35 mph for the reasons of the wobble, etc..but nothing beats a fresh-air ride (I don't even use Zwift, lol)

I can tell you from an observers perspective a speed wobble is almost as terrifying because you know what the outcome can be. I was heading up a long straight hill once, a group was heading down, but a lady in the back of the group looked to suddenly lose control. Text book speed wobble, I could only watch as she ate it at high speed. I stopped to help her and get her bike out of the traffic lane. Aside from cosmetic scratches her bike seemed totally fine, she on the other hand was not, lots of road rash and shook up badly.

---

--------------------------
The secret of a long life is you try not to shorten it.
-Nobody