We dont know how a bicycle works Really?

[DivergentSpectrum]

http://www.newstatesman.com/science/2013/08/we-still-don’t-really-know-how-bicycles-work
This is really kinda embarassing. Forget quantum mechanics our top scientists can't even figure out how a bicycle works. Is this true?

I always figured it wasnt so much the gyroscopic effect of the wheels, but more like the inertia of the entire bike.

 

[Astronuc]

http://www.newstatesman.com/science/2013/08/we-still-don’t-really-know-how-bicycles-work
This is really kinda embarassing. Forget quantum mechanics our top scientists can't even figure out how a bicycle works. Is this true?

I always figured it wasnt so much the gyroscopic effect of the wheels, but more like the inertia of the entire bike.

I think that the gyroscopic effect is part of it, but so is the balancing of the cyclist, i.e., keeping the center of mass/gravity over the line of points of contact of the tires/wheels with the solid surface - kind of like a tightrope walker. Skilled cyclists can stay upright at a stop (standstill).

 

[Tribo]

Reminds me of a Philip K. Dick novel. The undercover-junkie protagonist fails to understand something about how a bicycle works and the government assumes he's losing his mind.

 

[Doug Huffman]

David Gordon Wilson's Bicycling Science (2004 MIT) explains it perfectly. The analogic demonstration is balancing a broom in the palm of your hand. The gyroscope is well discredited by a device with a counter rotating front wheel.

 

[leroyjenkens]

I think that the gyroscopic effect is part of it, but so is the balancing of the cyclist, i.e., keeping the center of mass/gravity over the line of points of contact of the tires/wheels with the solid surface - kind of like a tightrope walker. Skilled cyclists can stay upright at a stop (standstill).

If they're going fast enough, they can stay upright without a rider.


So if there's no rider, and all the bike has is Newton's First Law to keep it upright, then obviously that must be enough to keep it upright.

 

[lavinia]

I'd be interested in an explanation especially why the gyroscopic effect is irrelevant.

On the surface it seems that any spinning wheel must be a gyroscope and that the gyroscope stabilizes the bike - this because a gyroscope resists a change in its plane of rotation. Additionally, precession of the the front wheel allows the rider to turn the bike. One needs to lean to one side in order to turn the bike so that the front wheel will precess. In some sense a bike does not really turn say the way one turns a tricycle does but rather precesses from the torque of gravity on the front wheel.

Why is that wrong?

 

[DivergentSpectrum]

as i understand the gyroscopic effect is just a manifestation of "an object in motion will continue in motion" So therefore its harder to tilt a spinning wheel than a stationary one, because youre changing the direction of the rotation/motion.
One thing about gyroscopes, is the wheels are intentionally heavy. This makes it require more force to change the trajectory.
Most of the time bicycle wheels are intentionally lightweight- just a few spokes and an aluminum rim.
So yes, the gyroscopic effect of a bicycle is negligible.

 

[leroyjenkens]

as i understand the gyroscopic effect is just a manifestation of "an object in motion will continue in motion" So therefore its harder to tilt a spinning wheel than a stationary one, because youre changing the direction of the rotation/motion.
One thing about gyroscopes, is the wheels are intentionally heavy. This makes it require more force to change the trajectory.
Most of the time bicycle wheels are intentionally lightweight- just a few spokes and an aluminum rim.
So yes, the gyroscopic effect of a bicycle is negligible.

Negligible compared to what? There has to be a main force that's keeping bikes upright if there's other forces that can be said to be negligible in accomplishing that task. How much gyroscopic force would be required to keep a bicycle upright? More than the wheels provide? Is there evidence of this?

 

[DivergentSpectrum]

well, I am not sure how they did it, but in the article they mentioned a device that cancels out the gyroscopic effect, and the bike remained stable. So I am guessing there must be some kinda other force at work. In my experience riderless bikes don't go very far, but apparently there must be something to it.
Definitely worthy of an (ig)noble prize.

 

[jtbell]

I've moved this thread to the General Physics forum. Over the years, there have been a number of threads here and in the Classical Physics forum, about bicycle stability. Use the forum search (top of the page) to find them. Simply searching for "bicycle" in the desired forum should do it.

 

[CWatters]

It's certainly worth looking at studies and experiments.. It seems NEITHER the trail or the gyro effect are mandatory (but may help)..

http://io9.com/5792341/engineers-overturn-physics-but-keep-a-bicycle-upright

Traditionally, two physics phenomena were considered necessary for keeping bicycles upright. Turns out neither of them are. And humans aren't necessary either. If anyone had gone up to their physics professors a few days ago and asked what keeps bicycles upright, they would have gotten two answers; gyroscopic stability and the trail.

Scientists have built a bike that has neither property, but not only does it stay upright, it stays upright without a rider.

 

[Doug Huffman]

John Forester addresses 'Steering and Handling' (and stability) in Chapter 3 of Effective Cycling (MIT 1993)

 

[SteamKing]

You shouldn't rely on a publication like the New Statesman for the latest news in what science does or does not know.

 

[A.T.]

The gyroscope is well discredited by a device with a counter rotating front wheel.

By this logic, the rider's steering inputs are discredited by self stable bikes.

 

[jedishrfu]

There is some discussion from a 1970 article on bicycle physics:

http://socrates.berkeley.edu/~fajans/Teaching/MoreBikeFiles/JonesBikeBW.pdf

from this web page:

http://socrates.berkeley.edu/~fajans/Teaching/bicycles.html

 

[rcgldr]

To summarize the previous threads on this, self stability of a bicycle relies on a geometry that steers the front tire into the direction of a lean. The conventional method for doing this is trail, where the extended steering axis intercepts the pavement in front of the contact patch. The alternative used in some exeperiemental bikes locates a mass ahead of and above the front tire, with the front tire mounted so it's free to rotate about it's steerring axis, resulting in a yawing torque on the frame when leaned, which ends up steering the front into the direction of lean, without requiring any trail or caster setup.

Gyroscopic forces are reactions to a change in lean angle, not to the amount of lean, so they dampen lean rate, but do not correct an existing lean.

There is a small gyroscopic steering reation due to the roll torque related to the lean angle of a bike (gravity effectively pulls down at the center of mass, pavement pushes up at the contact patches), but generally it's insufficient to result in self stability, and the main effect is that gyroscopic forces dampen the lean rate.

There also is a tiny corrective roll torque in response to the rate of yaw while a bike is turning, but it's also insufficient to result in self stability.

 

[A.T.]

Gyroscopic forces are reactions to a change in lean angle, not to the amount of lean,

The front wheel reacts to roll torques, by steering into the lean that the roll torque tries to achieve.

so they dampen lean rate, but do not correct an existing lean.

Which is a crucial element of a good control mechanism:
http://en.wikipedia.org/wiki/PID_controller

There is a tiny corrective roll torque in response to the rate of yaw while a bike is turning, but it's insufficient to result in self stability.

The key is front wheel response the roll.

 

[rcgldr]

The front wheel reacts to roll torques, by steering into the lean that the roll torque tries to achieve ...

Take the case where the bike just happens to end up in a coordinated turn, the lean angle combined with the speed and steering angle of the front tire resulting in zero net torque about the roll axis, so no gyroscopic related tendency to return to a vertcial orientation. In the same circumstance, trail or other self correcting steering geometry would steer the front tire further inwards, resulting in a correction to a vertical orientation (but in a new direction).

Generally the gyroscopic reaction to roll torque is insufficient for self-stability. Even with sufficient trail geometry for self-correction to vertical orientation within a range of speed, if the speed of a bike exceeds what is called "capsize" speed, then the combined effect of gyroscopic reaction and trail results in the bike falling inwards at an extremely slow (virtually imperceptable in real world examples) rate due to the dampening of lean rate, and in the case of racing motorcycles, the sense that a rider gets from a racing motorcycle at high speeds is that the bike tends to hold it's current lean angle as opposed to tending either fall inwards or return to a vertical orientation. At these speeds, it takes the same amount of counter steering effort to return a bike to a vertical orientation as it does to lean the bike from a vertical orientation.

 

[sophiecentaur]

This is all pretty well understood (not by everyone, though - as happens with the Moon Landings Conspiracy and other bits of non-Science). There are several different factors at work that keep a bicycle from falling over. Proponents of each factor get very precious about it and claim it's the only relevant one, But you can arrange a counter rotating wheel, turn the front forks the other way round, put an incompetent rider on the bike etc. etc. It will fall over. You could improve just one of those parameters and the bike would stand a chance of not falling over. However, most bikes have all factors working in their favour and they usually don't fall over.
Angular Momentum = Magic for many people, which accounts for a lot of the misunderstandings.

 

[A.T.]

Generally the gyroscopic reaction to roll torque is insufficient for self-stability.

Yes, just like generally the derivative term alone, doesn't make a good controller.

 

[DaveC426913]

Rider balance is by far the major factor.

Consider:
Take a good rider, get him going up to some nice speed, enough to convince the gyroscopists that the wheels are keeping the bike upright.
Now drop the bike's wheels into a streetcar track.
Wheels are still spinning nice and fast, but the rider cannot balance.
How long will he remain upright? A dozen yards? then Wham!

 

[sophiecentaur]

Rider balance is by far the major factor.

Consider:
Take a good rider, get him going up to some nice speed, enough to convince the gyroscopists that the wheels are keeping the bike upright.
Now drop the bike's wheels into a streetcar track.
Wheels are still spinning nice and fast, but the rider cannot balance.
How long will he remain upright? A dozen yards? then Wham!

But the gyroscopic action in question is not keeping the bike up directly.It is just causing the front wheel to turn into the fall, which causes a 'righting' moment. The necessary torque for this is a lit less than the rote needed to pull the bike and rider upright.

 

[A.T.]

There are several different factors at work that keep a bicycle from falling over.

Exactly. Claiming that one factor "has been discredited", because it can also work without that single one, doesn't make sense. It would mean that all of them have been "discredited".

 

[A.T.]

Rider balance is by far the major factor.

enough to convince the gyroscopists that the wheels are keeping the bike upright.

As sophiecentaur noted, you completely misunderstood the issue.

 

[rcgldr]

Turn the front forks the other way round.

Since most forks reduce trail (they curve forward), turning them around increases trail, and such a bike will be self-stable even at very slow speeds.

Exactly. Claiming that one factor "has been discredited", because it can also work without that single one, doesn't make sense. It would mean that all of them have been "discredited".

Gyroscopic precession isn't required, since a bike could be make with two rounded skate blade (non-rotating), and with sufficient trail would be self stable on ice (within a range of speeds). What is common to all self stable geometries, is some method to turn the front tire into the direction of lean (even if the turn is coordiated, in which case, there's no roll axis torque to invoke a gryoscopic reaction). The main method for conventional bikes is trail: the contact patch is "behind" the point where the extended steering pivot axis intercepts the pavement. An alternate method is to locate a weight above and in front of the front tire without trail or castor but free to rotate about a steering axis, with the weight inducing a yaw torque in response to a lean which steers the front tire into the turn. I'm not aware of any other self stable methods.

 

[DaveC426913]

As sophiecentaur noted, you completely misunderstood the issue.

I did not misunderstand the issue.

I said rider balance is the major factor. There is some gyroscopic effect but it is greatly superceded by rider balance.

Yes, an empty bicycle can stay upright for a short time without a rider. So what? It is not a meaningful example of a bike "working".

Put a 150lb deadweight on the bike, three feet above the ground. How long do you think it'll stay upright then?

 

[A.T.]

I did not misunderstand the issue.

You obviously did misunderstand the gyroscopic steering mechanism, as your misguided proposal of putting the bike's wheels into a streetcar track shows. Or what do you think that would demonstrate?

Put a 150lb deadweight on the bike, three feet above the ground. How long do you think it'll stay upright then?

If you provide it with propulsion, so it says within a certain speed range, it can stay upright indefinitely.

 

[A.T.]

Gyroscopic precession isn't required

And neither is trail, and neither is rider steering input. That's the point. None of them is, by itself required.

 

[rcgldr]

Gyroscopic precession isn't required ...

And neither is trail, and neither is rider steering input. That's the point. None of them is, by itself required.

From my previous post, what is required is "some method to turn the front tire into the direction of lean", sufficient enough to return the bike back to a vertical orientation (within a range of speeds, I'm not aware of any passive geometry that works at all speeds). Adding a weight high enough and forward enough on a bike without trail or caster works, at 1:30 into this particular video, a random lean occurs soon after release, but the bike recovers.



The test no trail bike in this video appears to be more stable at 9:20 into the next video. I'm not sure if any changes were made to the bike from the previous video to result in what appears to be a more stable bike.



One advantage of the caster effect that's part of the trail geometry (versus the weight mounted above and in front of the bike) is that reaction to a side force applied relatively low on the bike (so a relatively smaller roll torque versus a side force applied high on the bike), such as a gusting side wind on a faired motorcycle, results in the tire steering downwind, leaning the bike into the wind, which then results in the tire steering into the lean, reducing the amount of downwind drift due to the wind.

I have yet to see any test bike made stable relying only on gyroscopic reactions to roll and/or yaw torque(s).

TU Delft also did a study on a conventional bicycle, and did some treadmill tests. For some unexplained reason, although the mathematical model for this bicycle shows that it should be in capsize mode (falling inwards at a very slow rate) at 8.00 meters / second or faster, the actual bike is shown to be "very stable" at 8.33 meters / second (30 kph) in the last video on this web page:

http://tudelft.nl/nl/actueel/laatste-nieuws/artikel/detail/treadmill-measurements

The graph of the model for this particular bicycle is show on page 4 of this pdf file. At the time the pdf document was created, it appears that they had only tested the model up to 6 meters / second, and that the treadmill test at 8.33 meters / second was done later.

http://www.tudelft.nl/fileadmin/UD/MenC/Support/Internet/TU_Website/TU_Delft_portal/Onderzoek/Wetenschapsprojecten/Bicycle_Research/Dynamics_and_Stability/doc/Koo06.pdf

 

[Wes Tausend]

...If you provide it with propulsion, so it says within a certain speed range, it can stay upright indefinitely.

That seems like a pretty good insight A.T.

If it is entirely true, we could theoretically experiment with a riderless electric bicycle to prove your statement holds. An option would be to RC (radio control) control the riderless experimental bike to guide it within a reasonable "outdoor laboratory" range. This scenario would not greatly differ from RC controlled toys such as this demo:
.

The actual observable operation begins around 2 minutes. These toys do use a gyro assist, perhaps as low speed necessity, but I do not believe it provides all the amazing inherent stability.

The "head angle" seems to me to make the most amount of difference in the bike handling, i.e. loss or gain of self-righting degree. One can exaggerate this "fork angle" in ones mind to imagine different effects. For example, reversing the head angle will cause the instability of a bike rolling backwards. Kicking the head angle out (raking the fork) seems to me to more-or-less vary the tendency of self-steer correction at differing speeds.

I have a couple of older off-road bikes that are contrasts in purpose. One is designed for slow nimble Trials riding with a quick-steering, steep fork angle. The other is designed for stable, fast high-speed desert riding, with generous head rake and trail. Manufacturers still tinker with the various settings of new models today. It must still be more of an art than science. The geometry understanding is mostly science now, but the remaining art skill must be estimating the average forces encountered on the various tracks to be used. Yet it seems tracks could be easily measured with the right dynamic bike sensors during practice, and become all science, as are stability systems on modern autos.

Wes
...

 

[A.T.]

If it is entirely true, we could theoretically experiment with a riderless electric bicycle to prove your statement holds. An option would be to RC (radio control) control the riderless experimental bike to guide it within a reasonable "outdoor laboratory" range.

All control you need is speed. The rest can be left to the bike. But I'm afraid no length of stable run will convince people, who believe that some rider input is required.

This video shows not only a long stable run, but also how the bike recovers from a significant perturbation:

 

[sophiecentaur]

I don't see why that comment is worth making. We all realize that it's a mechanism with many different factors so why should it be at all remarkable that there are limits to some of the parameters? A given design of aircraft only flies within a certain range of speeds but that is 'obvious' and the fact is not used to discredit the standard treatment of an aerofoil.
The whole business of what keeps a bike upright is very interesting but the topic always seems to trigger dissension rather than co operation between the proponents (experts?) of certain factors. More than one person can be right here without others being necessarily 'wrong'.

HAHA- my auto spell-correct just replaced "bike" with "bile". The iMac made my point.

 

[A.T.]

The whole business of what keeps a bike upright is very interesting but the topic always seems to trigger dissension rather than co operation between the proponents (experts?) of certain factors.

I'm certainly a proponent of a combination of many different factors.

More than one person can be right here without others being necessarily 'wrong'.

Claiming that a bike cannot be bike self stable and needs a balancing rider, as Dave seems to suggest, is almost certainly wrong, given all the existing theoretical and practical evidence to the contrary.

 

[rcgldr]

The whole business of what keeps a bike upright is very interesting but the topic always seems to trigger dissension rather than co operation between the proponents (experts?) of certain factors.

The only dissension here is whether gyroscopic reactions help or hurt self-stability (defined as a tendency to return to a vertical orientation). Take the case of a bike with zero trail: without going into the math details, note the rate of precession along the steering axis of the front wheel is proportional to the roll torque, but as the front wheel precesses inwards, the roll torque is reduced, and the roll torque becomes zero once the fornt tire steers inwards enough to produce a coordinated turn (assuming that this even occurs), leaving the bike leaned over (in a coordianted turn) as opposed to steering inwards enough to return to a vertical angle. Take the case of a bike with a conventional amount of trail. Once the front tire steers inwards enough to start correcting the lean, an "outwards" roll torque is produced, and gyroscopic reaction would tend to oppose the inwards steering needed to correct the lean angle.

It's my belief that the gryoscopic reaction is always opposing the corrective action of trail in most circumstances, since the trail geometry attempts to steer the front tire inwards at a greater rate than the precession rate related to roll torque during the lean in stage, and this is why bikes go into capsize mode (fall inwards at a very slow, almost imperceptible, rate) above some critical speed where the gyroscopic reactions dominate over the trail reactions.

 

[A.T.]

The only dissension here is whether gyroscopic reactions help or hurt self-stability (defined as a tendency to return to a vertical orientation).

A better definition of self-stability is "a tendency to return to a vertical orientation, an stay there, not oscillate around it."

It's my belief that the gryoscopic reaction is always opposing the corrective action of trail in most circumstances,

That is the point of "damping", or the derivative term in control theory: preventing over-correction and oscillation around the optimum.