A bike with a reverse-spinning wheel | Hacker News Reader

A bike with a reverse-spinning wheel

www2.eng.cam.ac.uk

Why would spinning the wheel backwards cancel out the gyroscopic effect? The gyroscopic effect makes it hard to tilt the wheel left or right (from the perspective of the rider). Spinning the wheel backwards makes it hard to turn the wheel right or left, which has the same stabilizing effect.

I get that there is some effect other than the gyroscopic action of the wheel (how else could you track-stand, after all), but I don't feel like this device proves or disproves anything.

Also, the "begs the question" and the footnote explaining why it's OK to misue that expression was annoying. (If you have to write a paragraph about why you want to use incorrect English, wouldn't it be easier to just edit the misused construction. "This leads us to ask, ..." Write that, and you save yourself a footnote, and people will think you're not bad at writing. </rant>.)

The torque is

    tau = I*domega/dt

When we have two wheels spinning in opposite directions the torque is

    tau = I*domega/dt + I*d(-omega)/dt = 0

The gyroscopic effect doesn't actually make it harder to turn a wheel. It's just that if you turn it in the xy-plane, it automatically turns in the direction perpendicular to the push (the yz-plane). When a human is physically turning a wheel he will try to stop that from happening, thus the feeling that it's hard to turn the wheel. Note that in particular the gyroscopic effect does not produce any force in the direction opposite to the pushing force.

You have a wheel spinning one way, and say that if you try to turn it clockwise in the xy-plane then it will turn clockwise in the yz-plane because of the gyroscopic effect. If you have a wheel spinning the opposite way, and if you turn it clockwise in the xy-plane then it will turn counterclockwise in the yz-plane because of the gyroscopic effect. Adding the two effects cancels them.

You are exactly right. This article suffers the common misconception that two wheels spinning in opposite directions and same speed cancel out the gyroscopic stability. And this is simply not true. In fact two wheels spinning in opposite directions and same speeds will actually double the gyroscopic stability (as opposed to one wheel of the same speed).

However, two wheels spinning in opposite directions may cancel out other spinning wheel effects.

It is a common belief that two identical wheels spinning in opposite directions and the same speed have a net zero gyroscopic stability. The basis for this belief is usually false, but the assertion is true.

I've done both the math and the experiment - it's in one of the lectures I give on "Spinning Things".

In short, I believe you to be wrong, and this is based on my personal experiments and calculations. I would be interested to know the basis of your claim.

When computed in detail, gyroscopic stability is really the effect that torque applied in one plane manifests itself as movement in another. The direction of this movement depends on the direction of spin, and they cancel each other out when you have two wheels spinning in opposite directions.

My independent experiments and calculations were done before I read anything formal about it, and it agrees with the work of people who design devices that run on satellites and the ISS. I'd need a lot of convincing that I'm wrong.

This is independent of the other effects that help to keep a bike upright, such as the steering geometry. I have experience of that too, having raced bikes with very small rake, and toured on bikes with a larger rake. The difference is unmistakeable, and unrelated to the gyroscopic issues.

Well here is an experiment that I have done. I used a toy double blade helicopter, like these ones:

http://www.rc-airplane-world.com/coaxial-rc-helicopters.html

And it sure seemed to have a lot of rotational inertia. If you were right, then I could touch one side of the helicopter slightly and since it has no rotational inertia it would tilt to the side and completely change direction and fly off somewhere into the wall or the floor. This did not happen however, even when I tried to tilt it the helicopter remained horizontal. And please do not tell me that a $50 toy has an active stability system with internal gyros and all the required electronics and super fast servos.

And now that I think of it, I remember that the russians have been making counter rotational helicopters for ages. See for example this one:

http://en.wikipedia.org/wiki/Kamov_Ka-25

How does this helicopter stay horizontally stable in the air and does not fly off on a tangent when tilted by the slightest cross wind? Note that this one was designed in the late 50's before any electronic active stability systems could have been invented.

I have limited personal knowledge of RC helicopters, but a great deal of experience with rotational systems. Colleagues tell me that the counter-rotational systemson helicoptors are stable because the CoG is hanging from a point quite a long way above, and cross winds are applied more-or-less at the CoG because of the cross-sectional area distribution.

I don't doubt that your experience of a toy double blade helicopter is that it seems to have a lot of rotational inertia. I am certain that it's not due to the gyroscopic stability of exactly matched counter-rotating blades.

If you think about it, this explanation cannot be true. The main force the helicopter exerts on its surroundings is along the axis of rotation. If you tilt the axis of rotation significantly in any direction, the direction of the force will change significantly and the motion of the helicopter will also change significantly. Again, this does not happen.

It's unclear what you mean by the "this" in the "this does not happen"

However ...

A helicopter changes direction by using the cyclic setting on its rotors. Then you get more lift from one part of the rotor cycle and less from another. That causes the plane of the rotors to tilt, and then the helicopter goes off in the appropriate direction.

For example, having more lift in the rear portion of the cycle and less in the front means the helicopter gets tipped forward, effectively like lifting one part of a plate. The down-draft now has a significant rearwards component, so the rotors get pushed forward, taking the helicopter with it.

If you then change the cyclic to "flat" so there is identical lift everywhere, the weight of the helicopter dangling from the rotors causes it come come back upright (after swinging a bit)

It's a weird experience sitting in a helicopter when it's doing this, especially if you're mostly accustomed to ordinary aeroplanes.

I don't see any contradiction in any of this, and it really doesn't seem to have any bearing on the fact that contra-rotating disks have no net gyroscopic effect.

This is accomplished by way of a swashplate, as can be seen here: http://en.wikipedia.org/wiki/Swashplate_%28helicopter%29

When you say "this" you mean the translation of the pilot's controls into the angles of the blades at the various stages of their journey, mixing collective (giving overall lift) and cyclic (giving pitch and roll)

Yes.

Which other spinning wheel effects?

Another effect is the tendency to tilt to the side if you accelerate or decelerate the rate of spin. This effect will be canceled by counter spinning wheels.

Can you elaborate on this? I have never seen a wheel tilt when you accelerate or decelerate it.

Precession

Precession is caused by the gyroscopic effect.

Wait, so if you have two identical wheels on one axis, spinning in opposite directions, and you rotate the axis, are you saying you'll experience precession or you won't? My intuition (and physics understanding) says you won't. But the top-level post is saying you will still experience "gyroscopic stability". Is that true?

You don't. That's because the gyroscopic effect cancels out in that case. But if you use a single spinning wheel you will. I'm just saying that precession isn't a different effect than the gyroscopic effect; it's a specific case of it. The same forces that cause a wheel to feel like it's hard to turn (i.e. the "gyroscopic effect") cause precession.

I asked:

> Which other spinning wheel effects? [other than the gyroscopic effect]

And he replied:

> Precession

To which I replied that precession is not a different effect. I did not mean to imply that precession happens when you have two wheels spinning in opposite directions, in fact I believe it does not because the forces that cause a single spinning wheel to precess cancel out if you have two wheels spinning in opposite directions.

You're right, your understand is right, you won't experience "gyroscopic stability." I've done it. You don't.

I believe that people who believe otherwise are working on misunderstandings of perceptions. It's nototiously easy to confuse several effects in this. Careful experiments are required with tight controls on error bars, etc.

It's easy to be fooled. Eric Laithwaite is a classic and high profile example:

http://en.wikipedia.org/wiki/Eric_Laithwaite

The gyroscopic effect does not _directly_ make it harder to turn, it causes a reaction force (torque, really) at right angles to the applied force. With two counter-rotating wheels, these reaction forces work against each other.

No they don't. Any spinning wheel will create a force that tries to preserve its axis of rotation. If you have two counter-spinning wheels that share the same axis and try to tilt them, they will both try to preserve the same axis of rotation and thus exert a force in the same direction.

Have you done the experiment? "Syd12107" has claimed to in this thread: http://www.physicsforums.com/archive/index.php/t-173215.html

The result was that counter rotating gyroscopes had no net effect.

The only thing that matters is total angular momentum.

The resistance to being tilted is a _secondary_ effect. The push gets translated into motion at right angles to the push, and this motion gets translated into push at one more right-angle, exactly countering the original push. For two counter-rotating gyroscopes, this first-stage motion of one is exactly countered by the other, leading to no net effect. Really, truly.

That exactly matches my experience and calculations.

He's Australian, they misuse English wilfully all the time. ;)

a simple experiment can demonstrate what really keeps a bike up. Hold your favorite bike by the seat and give it a hard push forward (even better run a few steps with it to get it going and then let go). Then repeat, pushing it backwards.

A bike moving forward will self correct to stay up until it loses speed. A bike moving backward will quickly go into a hard turn right or left and fall over. Gyroscopic moments for wheels in both scenarios are the same. The difference is that as a forward-moving bike falls to one side, the front wheel straightens out underneath it. As a backward moving bike (or any bike with a turning rear wheel), starts to fall, the front (now rear) wheel turns further to the side as it falls rather than straightening out, and the bike goes into a very sharp turn and falls.

As an addendum, that's why recumbent bikes stick with a rear drive and front steer, which requires very long chains, rather than driving the front and steering with the rear (which at first blush seems like a much simpler configuration - the pedals are up near the front wheel, and the hands are near the rear).

The one-directional stability in steerage of wheeled vehicles (bikes, shopping carts, cars and non wheeled vehicles like boat rudders) is affected by more or less asymmetry between the center of pivot and the center of force. See:

http://en.wikipedia.org/wiki/Caster_angle

Actually it shouldn't matter to stability whether a recument (or other) bike is front or rear drive. In fact it appears some recuments are front drive. I would think that rear drive is prevalent because that's the wheel carrying most of the weight.

Also I think the larger effect on your experiemt is the "trail" thing the author mentioned. If you spin the front wheel 180 degrees (possible for some bikes) and try again I think you'll mostly see the reverse result: pushing the bike backwards will let it roll a while, but pushing it forwards will result in immediate crash. But the leaning effect you mention could be important too, I don't know.

there are two effects, the lean and the trail. A front-steering bike with trail is self-correcting through both effects. A rear-steering bike with trail will correct itself to a straight line as long is it is perfectly vertical, but will be unstable as soon as it leans. When it leans, it will go into a hard turn and collapse. There are front-drive recumbents, but there are no rear-steering two-wheeled recumbents (it is possible to address the lack of lean stability on a tricycle configuration).

this leads me to wonder - has there been a bike design where the rear wheel is allowed to freely "steer" as well as the front wheel? sorta like a crabwalking bike able to side pedal when needed.

Modern automobiles have incorporated rear wheels that turn together with the front wheel (I believe in opposite directions during a turn) in order to decrease the turning circle and improve maneuverability - I'd kill for a bike that has an electro-servo controlled rear wheel that turns to help the biker corner, etc.

The reason some vehicles have steerable rear wheels is because there are two rotational effects required when turning. One is to make the vehicle rotate, the other is to make the vehicle move in a circle.

If you only steer the front wheels then when the centripetal force is correct for turning the corner, the rotational force is much larger than necessary. This leads to excessive scrubbing of the tires. Making the rear wheels steer slightly in the direction of turn the rotational force is reduced, similarly reducing the scrubbing.

As a side benefit the wheels can be turned in the opposite direction at low speed, increasing the manoeuvrability.

See "swingbike" -- they're fun: http://blog.makezine.com/archive/2008/01/sweet_swing_bike.ht...

In this next picture, note the seat post (not the sissy bar) is actually inserted into something like a typical headset (where a fork and stem would go on the front of a bicycle) that pivots when the pin on the underside of the frame is removed. http://www.swingbikerider.com/4-24-05%20003.jpg

That looks pretty cool! and insane too!

Yup -- they're pretty fun... The bicycle has had nearly 200 years of evolution, though, and the reason most bicycles aren't built like this is it's simply impractical. If the fun/coolness is worth it to you though, it's worthwhile :)

I had a bicycle like this[1] as a kid--with the rear wheel and pedals on a hinge. It wasn't significantly better at cornering--a good understanding of countersteering[2] should explain why.

[1] (sorry, can't find a better reference/illustration with a cursory googleing) http://www.faqs.org/patents/app/20090014981

[2] http://en.wikipedia.org/wiki/Countersteering (this article may not be sufficient for a thorough understanding of countersteering, but it's a good start)

The article seems to be something of a joke. Yes, caster angle is important for stability, but the Gyrobike training wheel seems to illustrate how gyroscopic forces can also be helpful.

See: http://www.youtube.com/watch?v=Cbfe2_2DDc0

Additionally, stationary trainers like rollers that don't allow for the steerage correction described by the author illustrate how at speed something other than castor angle is assisting in stabilizing the bike.

See: http://www.youtube.com/watch?v=xaYDZ-RSOTA

I've been meaning to set up some gyroscopically neutral rollers and see if it's possible to ride them...but I've been lazy (and my neighbors hate my rollers).

A related note is the reverse steering bike which is impossible to stay up on unless you 'relearn' how to ride a bike.

http://www.youtube.com/watch?v=9gCnBBeV2ws&feature=relat...

I think that proves things a bit better - we stay up on a bike because we learn to notice any slight chance we're falling, and counteract that with steering to bring us back to upright.

I've spent quite some time racing Downhill Mountainbikes - and believe me - the gyroscopic effect IS there - and it is very noticeable (DH bikes tend to have MUCH heavier wheels).

Riding lighter wheels has a lot of effect on acceleration and braking. Handling also improves ALOT with lighter wheels.

It also contributes to stability. But you do need to ride faster than this "wuss" (pun intended) moving at snail speed.

Couldn't that effect be simply because of the inertia of the wheels that they also have when they're not spinning?

That might be the case... I mounted a heavy U-lock on the front forks of a mountain bike, made steering very heavy. Felt pretty similar to a heavy wheel to me.

Of course this is all subjective (my observation, and the parent's downhilling observation).

Definitely easier to steer lighter wheels regardless of speed.

Although, when the total weight of the bike is high, it is better to have weight up front too (front+rear panniers). Otherwise, steering feels flimsy.

If this kind of stuff interests you there is an excellent book on the physics of bicycles http://www.amazon.com/Bicycling-Science-David-Gordon-Wilson/...

If you want to ride an "unrideable" bicycle, try a flevo bike (like http://home.ultramagic.net/bicycles/knakker/). It has a giant joint in the middle, instead of steering bars.

The castor effect is what makes a bicycle self-steer. Bicycles with small fork rake are very twitchy-bikes with big fork rake are unresponsive.

Yep. Sports motorcycles have small rake (should be responsive on the race track) while choppers have big rake (long, straight journeys on Route 66). I prefer a KTM 990 Adventure, just in the middle, though :)

Reply on news.ycombinator.com