Hacker News Reader

Comment by matsemann | Hacker News Reader

I wrote my master thesis on optimizing bicycle wheels / spokes. I actually see I'm cited in the phd he cites, quite a fun surprise!

This is a great article. It showcases lots of the "simple, but surprisingly advanced" things surrounding bicycles. Which was what got me hooked in the first place. The visualization of how you have to turn right to go left is excellent. I've mentioned that fact multiple times here on HN, it's not commonly known, you just "do it" when you bike! And it explains why you sometimes can feel the curb "sucking" you towards it when you try to avoid it: you unconsciously avoid turning the wheel towards it, but that actually makes it so that you're unable to actually steer away from it!

View on HN

Veritasium did a great video on the opposite direction turning. They even show a bike that prevents you from turning the wheel one way to show that its essential.

https://www.youtube.com/watch?v=9cNmUNHSBac

When I took control theory at university, in the last lecture, the lecturer took out a bike with rear wheel steering and challenged us to ride it in the hallway outside the lecture hall in the break. No one could get it to roll more than a couple of meters. The second half of the lecture was spent proving that a rear wheel steered bike is in fact (almost) impossible to control.

And in opposite world: Rockets with steering thrusters near their top are highly unstable.

At some point (some) people though rockets with the thrusters near the top would be more stable. See e.g. this rocket by Robert H. Goddard: https://en.wikipedia.org/wiki/Robert_H._Goddard#/media/File:...

But it turns out that it doesn't matter whether you place the thrusters high or low, stability-wise. What matters is that the center of mass is in front of the center of the center of pressure.

Aren't they just as stable as regular rockets, but it's very inconvenient to put the thrusters on top? It's called pendulum rocket fallacy and from what I can tell that's the case.

Huh, wouldn't this be similar to riding a regular bike backwards, which a lot of people (although not myself) are able to do?

Yes, it was pretty much a reverse bike with the saddle on the frame and a handle bar just behind where the seat post should be, connected to the steering axle. The details is a little bit fuzzy since it was a couple of years ago but iirc it is actually possible bike backwards at slow speeds but require a lot more active balancing from the rider, and at some point it becomes impossible as the speed increases. Also it concerned the case where the bike is riding straight ahead, it is possible that it is easier to control if riding in a curve.

In the extreme, it's also equivalent to a unicycle - just lean back to put all your weight on the steering back tire. Unicycles can be controlled.

Both are undeniably less easy than a normal bicycle though.

> opposite direction turning

The term used in motorcycling is countersteering. A lot of people think they are using their body to change direction but that would not be sufficient. Also it helps to be deliberate about the handlebar pushing motion for safety and performance.

I also recommend the SmarterEveryDay video[0] where they rig a bike to turn the opposite way than it is steered.

[0] https://www.youtube.com/watch?v=MFzDaBzBlL0

A fair bit of the stability of a bicycle (especially at low speeds) is due to the fact that you have a human holding onto a bar that behaves exactly like a thing that a human would grab onto to steady themselves. If you grab this bar and rotate it clockwise, then you yourself will rotate anticlockwise - and that is true whether the bar is the handlebars on a bike or a random bar fixed on a wall. Holding ourselves steady by grabbing onto something is something that humans have a remarkably effective and quick feedback loop for, which is why bicycle riding comes naturally once you get over the fear of falling off and just do what feels right.

Rigging the handlebars to turn the wheel the other way cancels out that automatic feedback loop.

> ..you sometimes can feel the curb "sucking" you towards it when you try to avoid it

I was shopping for a kids carrying bikes, called bakfiets[1] (I think) in the Netherlands. The salesperson offered me a trial ride, his advice is something I still remember. Don't bother about the front wheels (they are far out in front), just look where you want to go and your hands will take care of steering and balancing.

[1] https://www.bakfiets.com

Edit: fixed thanks to the correction by “isoprophlex”!

Drop the space between "bak" and "fiets". Just "bakfiets", otherwise you don't mean the noun for cargo bike, but the imperative "bake a bike" ;)

Fixed, thank you :-)

Target fixation - you'll go where you look, for better or worse. Inexperienced riders will often stare right at the obstacle they're trying to avoid and ride right into it.

You went bakfiets shopping as a novice rider?

Sounds brave. Did it work out?

I'm sure the poster meant it was their first time riding a bakfiets, not that they never rode a bike before. The length of the bakfiets makes it a bit of a challenge.

The counterintuitive turn is more pronounced in motorcycles. You are literally pushing against the turn as you lean into the turn, and to lean in you have to turn out. Even more pronounced in sport motorcycles. Maintaining the centre of gravity during a lean at those speeds is a lot of fun to master. It’s one of those heart racing moments, literally no room for error.

Indeed, if you want to make fast turns at speed on a motorbike the trick is to first push/pull the bars in the oppositive direction - that gets the bike leaning over and then the bars come back and take the natural position for the turn. It's amazing how quickly even a big, heavy bike (e.g. BMW tourer) willl lean over with the right technique. At very highspeeds it takes quite a bit of muscle too.

I took the Motorcycle Safety Foundation basic course a very long time ago. One of the instructors brought his Honda Goldwing to the course just to help us get over the fear of falling over by showing that even a massive heavy motorcycle could go through all the tight turns without any problem.

I have to say, that course was some of the most fun I've ever had!

One of the best decisions in my life was to take this course. I started riding a Buell after this and boy what a thrill it was to lean this bike in the Pacific Northwest hilly roads.

Yes. On a motorcycle at high speed you have to actively maintain force on the wheel towards the outside of the turn to keep in the turn. If you stop the bike will straighten up.

Unicycles have a similar counter-intuitive control method where you need to accelerate the wheel to slow down and vice versa. Turning doesn't work quite the same way though as you can do sharp turns by twisting your hips and thus changing the direction that the wheel is pointing.

Never thought about, but yeah I guess it makes sense. If you were to suddenly stop you would be jolted forward, so need to get a bit behind the wheel first.

(too late to edit my original comment, but here is a link to my thesis discussed at the time: https://news.ycombinator.com/item?id=10410813 )

Many times I've observed myself as closely as I can taking a turn. I can't observe any steering in the opposite direction. I start by leaning, then turn into it (if I turn the wheel at all, which I don't unless the turn is very sharp). Maybe I really do steer the other way some imperceptible yet vital amount? This summer I plan to weld a bicycle unsteerable and see what happens when I try to steer solely by leaning as it feels like I generally do.

The welding the bars experiment you are talking about is done in a motorcycle school I once took with Kieth Code (California Superbike School), in order to prove to riders that countersteering is real.

I held an American Pro Superbike racing license in my 20s and countersteering is the only way I ever steered. Here is an experiment - go into a flat and wide open space like a parking lot, ride straight, and then turn to the right hard while keeping straight up and down. Hold the pressure to the right and I promise you will turn left or will initiate a turn to the left. The harder you turn the bars right the quicker the bike will fall left and begin arcing left.

In chicanes on a circuit track you can flip the bike over from one side to the other extremely quickly doing this. I've done this on everything from mountain bike to superbike, the latter being a more pronounced effect.

The only time I don't countersteer is when doing a 180-ish degree turn on a dirtbike, like what is common on an SX course. Also, when doing an extremely slow turn on a sport motorcycle, like what's common in USA motorcycle safety courses. But when moving at speed I always countersteer.

Knowing this phenomenon and using it may save your life.

Countersteering helps initiate that lean more quickly than just organic falling over. It's possible you're exceptionally patient but it's also possible you're just countersteering a little bit to initiative the lean, without noticing it.

Interestingly, I read about this a long time ago and intentionally turn the handlebars the wrong way to initiate the lean. (Not because it's necessary, but because it's weird. Riding your bike can be boring at times.) Very smooth.

It's totally necessary to push the handlebars in the opposite direction to initiate the turn because bikes are not minimal phase systems (at least that's what i remember from control systems).

It is unconscious and you have no idea you're doing it.

Watch the Veritasium video linked earlier. It's good stuff.

I've watched it multiple times and I'm not entirely convinced I don't tend to turn in some different way from the average person. One person managed the turn correctly by accident in the video because they'd happened to lean the right way beforehand.

It's easy to see countersteering in action if you have a quiet damp road. Try riding in a straight line and then turning to the right. Then dismount, walk back to where you started your turn, and have a look at your tyre tracks.

How thick are your tires? Initiating the turn requires just the tiniest bit of countersteer on very thin tires, in my unstudied experience. I did the same experiment as you and found it was so subtle as to be basically unnoticeable.

Yeah on a motorcycle leaning doesn’t do anything to engage a turn. “Push the direction you want to go” is what’s drilled into you and also just feels natural.

There's a YouTube video where they do exactly this. It doesn't end well for the rider.

I read once that when draisiennes first appeared in 1817, it was a surprise to the inventor that you could in fact lift your legs up and balance.

I forget completely where I might have read or heard this, but I love it so much I'm going to keep telling it. Can anyone confirm?

I can at least confirm that my daughter was very much surprised when her balance bike [0], which is essentially a small draisine, stayed upright when she lifted her legs.

[0] https://cdn.thewirecutter.com/wp-content/uploads/2017/12/bal...

That’s gotta be a cool feeling to see your work cited in other research. Kudos!

Similar concept that gets a lot of people in trouble on the OneWheel, "lean forward to slow down". When you're going to fast, the you intuitively lean back, which actually accelerates. There's even been a handful of lawsuits around this physics issue.

But why can't the inability to make a sharp turn be explained by the need to arrest forward momentum? The momentum has to change direction. Turning 90 degrees causes sudden loss in velocity. Turning against the arc of the turn i.e. along the orbit of the turn transfers the forward momentum into angular momentum. Thus, a counterturn is just an efficient way for the bike to follow the arc.

Higher the velocity, the larger the arc required.

Gyroscopic forces are fun. That phenomenon kicks in around 20-25 mph but that’s when I was on a crotch rocket. I’ve experienced it at like 15 mph on a bicycle.

This is an important concept when riding motorcycles at higher speeds. This was taught in the class I took to get my motorcycle license.

> I wrote my master thesis on optimizing bicycle wheels / spokes.

Any practical outcomes for hobby wheelbuilders with conventional parts (like some sort of novel lacing pattern or something like that)?

It was mainly a CS thesis on multi-objective optimization algorithms, and wheels were my chosen application. So me not being a mechanical engineer I didn't exactly push that side of the science any forward.

But my algorithm did end up "inventing" the 3x pattern perfectly, which I think was cool. Both as a confirmation that it's really a good versatile pattern being pareto optimal in multiple objectives, and the algorithm finding it also verified that my approach did have some merit.

Of the "unconventional" ones it found, my favorite I ended up using for the cover: https://i.imgur.com/b1ImCo8.jpg

Beautiful! Thank you for sharing :-)

(Just saw that you quoted the n+1 rule in the preface, hehe.)

That is a beautifully bound thesis.

Please send me hard copy of thesis.

$ I have coin if you have warez.

HI!

Thank Odin I found you!

I have a conceptual wheel design idea that I feel that only you can accomplish a successful design...

Before I email you some rambling wall of text, would you be open to hearing about some crazy concepts?

TL;DR:

The idea is to use Toroidal Propellers as 'spokes' in various light-weight, 3d-printable 'turbines', along with wind-shrouds to force vector air current to dynamos... specifically in various scaled applications for objects which already have a rotational input (shaft, wheel, spinny thing, etc)

Open to hearing from the loony bin?

Where read thesis?

Also, I am Norwegian! (but from Ballard, Seattle)

---

People tend to think of 'scaled' as in "LARGER" -- but it can also be used to refer to smaller...

Think of toroidal pumps in tiny bio tubes (veins, maybe, distributed heart pumping/nutrients pumping to isolated bio-assets (simulate pumping of heart of external bio-fluids to individually separated muscles connected to a biovascular pump system that can mimic the actual heart pattern of a donor to keep tissue happy) perhaps?)

Anyway -- its the evolution of Davinci's first documenting the importance of eddies, which we later discovered is how pumps work... (We knew pumps, but we didnt understand how they worked (documentedly) in Archemedies time (we also 'know' he did "discover" this, he documented it...) Anyway... (Sorry for the rant)

I want to develop a way to capture the eddies around certain objects.... If we examine Whales (the animal) they have a symbiotic relationship with barnacles... the barnacles attach to the leading edge of their fins. Whales eat off of plankton, small critters...

The barnacles create eddies along the wing surface..

BLAH BLAH BLAH

And the eddies feed both.... And I would like to talk to you about how this impacts flight! (passive extendable props that are fed off eddy wash) and pumps, and fluidic dynamics... and a bunch of cool boring shit.

Get a copy of the Bicycle Wheel by Jobst Brandt.

(and then try to map that experience onto current spokes/rims, which are somewhat different)

With motorcycles, you gotta learn “push right to go left” or you’re gonna end up in a bad sitch! But the physics of it are pretty mind boggling.

Similar to the 'bees can't fly' thing. I never, ever crashed my bike until i learned how not to. And where not to look.

Never heard of it. Show us this... the wheel

The “kerb sucks you in” phenomenon is target fixation, you’re getting closer to the kerb because you’re staring at it. It’s a common cause of motorbike accidents, among other things.

https://en.m.wikipedia.org/wiki/Target_fixation

You often see these two concepts (target fixation and counter-steer) discussed in threads on steering.

Not denying that target fixation is real and operative in some circumstances ... but counter-steer is even more "basic" in that it is solely a function of two-wheel dynamics, not rider psychology. You really do need to steer slightly right in order to initiate a turn to go left.

See the video linked up-thread.

Here's a page from a motorcycle POV which has a "truck pulls alongside" anecdote similar to your "riding parallel to curb" example: https://www.insurance.harley-davidson.com/the-open-road/tips...

No, it's because there's not enough room to counter-turn to shift your balance.

Well, it's probably also because of target fixation.

Wow, they really don't like being told "you go where you look".

Reply on news.ycombinator.com