Hacker News Reader

Comment by local_issues | Hacker News Reader

Ding ding ding! We still don't know how bikes work, we have a pretty good but incomplete model for lift on airplanes, and no one has the faintest why Tylenol works.

We should figure that out! But we can definitely keep using all the applications until then. (Except for Tylenol, we keep learning how bad that stuff is)

EDIT: It won't let me post more, so here's the answers to responses.

For sure!

https://www.cbc.ca/news/science/science-of-cycling-still-mys...

http://www3.eng.cam.ac.uk/~hemh1/gyrobike.htm

https://www.newscientist.com/article/mg22730370-400-how-does...

It's a fun little fact. We know a ton about BUILDING bikes, which is a more useful tool anyway

View on HN

Nonsense. There's nothing incomplete about our model for lift on airplanes.

Oh really, so why do the Navier-Stokes equations work?

Conservation of momentum?

There's no a priori reason conservation of momentum would apply; for that matter in 0-angle-of-attack aviation it doesn't.

Maybe to put it differently, the "Newtonian" model of flight comes out of assuming CoM applies, which we know isn't universally true.

EDITING: of course it always "applies" in the sense that there is a definable system in which total momentum will be unchanged, the point being that there is no guarantee that's a system in which the plane gets lift greater than its weight

A symmetrical airfoil at zero angle of attack doesn't generate lift because it doesn't turn the fluid. A non symmetrical airfoil will generate lift even at zero AOA because air has viscosity and the lack of symmetry causes it to turn around the shape. When you bend fluid flow, COM comes into play.

If you're asking how to apply the Navier-Stokes equations: apply boundary conditions

If you're asking how to derive the Navier-Stokes equations, use continuum / conservation principles.

Do you have a link handy for "we don't know how bikes work" ?

I'm assuming they mean the balance mechanism, and specifically what allows us to balance. How much is it the rider shifting their weight, how much is micro steering adjusts as we move forward, how much is the gyroscopic forces of the wheels, how much of it has to do with the angle of the handle bars to the wheel verse the center of weight.

That said, I'm guessing this one is well understood by experts, but more complex than someone would assume at first glance, and many who have some understanding likely have an incorrect or at least incomplete understanding of how balancing works.

I wonder what learning the specific system by which bikes are balanced would teach us about the world or human beings?

It's such a banal thing to be so fascinating.

We don't know until we know.

The smart phone is the culmination of understanding a million facts about materials sciences (applied and theoretical), some of which were obvious, some of which were non-obvious. Starting from a transistor you could see from across the room down to ones you can't even see with a magnifying glass.

It's the reason I got a degree in physics, if you have good professors - discussions like this cause you to break down the problem quite quickly in your head in a working model. Think force diagrams, but with a ton more math backing it up.

I actually find the ice skate a better example than a bike. We have all the physics solved for bikes, it's a complicated system but so is everything in motion. Hence we assume a spherical cow for the sake of the problem.

But ice skates... Now that's a funky one. Why do ice skates works? Ice skates aren't sharp bladed, they actually have flats. Ice is not slippery, it's when something is on ice in between our shoes and the ice that cause it to be slippery. Some people think it's the localized pressure of the blade that causes ice to locally melt. Hard to really wrap your head around. But it works :)

Ice skates aren't sharp bladed

Not sure how you meant that, but ice skates are sharp. Each edge of the blade is sharpened by grinding a hollow out of the center.

https://weekendwarriorshockey.com/how-sharp-should-my-skates...

I thought it was not just about ice skating but the question of why is ice very slippery is the hard part to explain.

>We have all the physics solved for bikes, it's a complicated system but so is everything in motion. Hence we assume a spherical cow for the sake of the problem.

This depends upon the question one is trying to answer. If one is trying to create a bicycle that can self balance, that involves considering different factors compared to trying to determine why certain injuries result in a person losing the ability to balance on a bicycle while others do not. Is the focus the bicycle or the human?

We have self-balancing bicycles. We clearly know it well enough to replace the human with a computerized machine.

Maybe we don’t know how humans use the bicycle but we know how bicycles balance, we can write programs to balance them physically.

0: https://youtu.be/Ya7iacmVjUM

1: https://youtu.be/2Z67NkvXIF4

You are correct, a lot of forces are canceling out on the long term(instead of instantaneously) it can easily be manipulated into increasing periods of unbalance in one direction until a point is reached then a separate mechanism is used to force it to a balanced state. Conservation of energy is always in effect. Gyroscope effects that bike wheels can be added with other separate gyroscopes. Thus, a self righting bike. The effect called precession is understood well enough.

How a human is able to manipulate it is simply by using the force of gravity from shifting their weight(moving the center of gravity). However, the movement of the center of gravity has to be perpendicular to the wheels axle. The steeper the angle of attack the wheel has to the ground, there will need to be an exponential increase in distance to move the center of gravity. Once the wheel is parallel to the ground, there will be an undefined distance needed to move the center of gravity.

e.g. What keeps bicycles balanced with or without a rider is still an active area of research, and even the seemingly basic idea that, for a bicycle to be self-stable, it needs to turn the handlebars into the fall, has not yet been proven.

[*]https://ciechanow.ski/bicycle/

"Active area of research" is quite different to "we don't know how they work".

We know how they work. We might not have fully characterised the stability conditions, but that's not the same thing.

Cmon, man

You can write out the equations of motion for a bicycle that will very accurately predict the dynamics. You can put these equations into a numerical simulation and predict motion very accurately. You can change the parameters of the model and do simulations with high confidence. Just because there isn't some neat little equation that says exactly what each parameter change is going to do (without doing the simulation) doesn't mean that we don't understand bicycle physics. It's a silly line of reasoning. Those articles are hyperbolic.

You are contradicting yourself:

> You can write out the equations of motion for a bicycle

> there isn't some neat little equation that says exactly what each parameter change is going to do

I mean, if you have the equations you can see what each parameter is doing !

Algebraic versus differential if that wasn't clear

Something smaller than may help you grasp this idea: https://www.youtube.com/watch?v=5nW3nJhBHL0

There are things we can characterize to any desired degree of accuracy but that e don't get cute little equations out of... and some of those things are so simple they've been staring you in the face since middle school and you just didn't ever notice their absence from your formula sheets.

This is not the exact same situation being described but it's a similar thing. Being able to put a complex system into a computer and arbitrarily manipulate it still doesn't mean we can extract some simple explanation.

On the other end of the scale, see all the AIs coming out. They're 100% computer artifacts with theoretically no mystery in them whatsoever... but they're just tables of billions of opaque numbers and doing anything with the numbers beyond just running them is amazingly difficult.

I think their point still stands, even if "we don't know how bikes work" is a flowery exaggeration.

But then we don't know how anything works...

Yet we still have Insane Clown Posse refrigerator magnets.

https://www.insaneclownpossemerch.com/collections/insane-clo...

Scientists Try To Teach ICP Fans How Fucking Magnets Work

https://metalinjection.net/av/scientists-teach-icp-fans-fuck...

But their point is we don't need a 100% complete scientific understanding of something in order to engineer it into something cool or useful.

That is true, but it's still incorrect to say we don't understand how bicycles work.

As a bit of an aside, most people don't know how bicycle wheels work. There's a whole section in https://en.wikipedia.org/wiki/The_Bicycle_Wheel that talks about how they actually work. It's not tension at the top, it's compression on the bottom.

what do you mean it's not tension at the top? did you misspeak? bicycle spokes are solely under tensile forces. they can't support compressive forces at all. i'm a hobbyist wheel builder and a once upon a time professional bicycle mechanic during hs & college.

if you want to test this take nearly all the pretension out of your spokes and sit on your bike. feel which ones are taught and which ones are completely loose. or just go to walmart. those bikes hardly have any pretension in their wheels.

They even make spokes that are rope

https://berdspokes.com/pages/technology

I wonder if those have less or more breakin issues compared to spokes. With a spoke you have to reset the angle of the bend to the shortest distance between the hub and the rim. But with cables they have to settle in along their entire length.

Doesn’t change the answer. They’re still compressing. They’re just pretensioned.

you sound highly confused, what is compressing? the rim? the spoke?

You need to reread Brandt.

There are two situations when you can push on a string. One is when it’s frozen, and the other when it’s tensioned. How do you unload a bow string? You push it off the notches.

bro. you're wrong. just build a bicycle wheel some time. it becomes very obvious how they work as you take it through the various stages of tensioning.

Why do you think someone would talk about Brandt’s work if they hadn’t used it?

I’ve built more than half a dozen bicycle wheels. My set, my spare that my brother road (into the ground - my first set and practically the only problematic ones, but he road over bumps without getting out of the saddle), a set my dad commissioned from me, and a pair that he had me build for a friend. All by age 17.

I then worked as a mechanic for two summers of college. I was never the fastest, but if we had a customer we could not afford to disappoint, I or the senior mechanic got the job because my repairs did not come back.

I saved three or four wheels that would have been scrap by unwinding the spokes halfway and building it back up again like a wheel build. Only added an extra ten or fifteen minutes but it works a charm. When a good customer comes in on Wednesday before an out of town bike ride you can’t afford to fuck it up. I think I only built a couple professionally, and usually singles. That’s a lot of labor and few will pay.

There's a lot more creep early on, which is compensated for by a staged tensioning over the course of a few days. They may require the spoke holes in the hub to be radiused, which can be a warranty issue between you and the hub manufacturer. Windup is controlled by a flat on the small bit of spoke used for threads.

They're ok (i.e., made it through Tour Divide with no issues), done well they're certainly better than badly done steel spokes, but it's not clear if the best builds are better than the best steel builds.

That book I linked has another name, “the wheel building bible”. Jobst Brandt earned an obituary in Bicycling magazine including quotes from his friend Tom Ritchey (one of the original mountain bike makers). Jobst was a bike fanatic and a mechanical engineer.

Bike spokes are not loose, they’re under substantial tension. Bolts, I just learned a couple weeks ago, work in the opposite way. A tightened bolt compresses the two pieces of metal together, and when you tug on them, the bolt doesn’t stretch more. The tension instead first cancels out some of the compressive force on the two pieces of metal, before the bolt ever feels more load.

Conversely, all the spokes on the wheel are under tension. When you put the wheel on a surface and push down, the compression cancels out some of the tension on the bottom of the wheel. Cancel out all of the tension, and the wheel turns into a potato chip if you don’t reload it exactly, perfectly on axis. IIRC, none of the prior models or theories for how a spoked wheel works could adequately explain how potato chipping happens. His does.

I used his book to build half a dozen wheels or so and the information it contained to fix many more.

i own the book.. and also The Art of Wheelbuilding - Gerd Schraner

of course in a properly built wheel usually all the spoke are under tension...

i was just demonstrating the fact that the spokes on the upper half of the wheel are supporting the hub and are under greater tension than the bottom ones, the spokes on the bottom half of the wheel should remain in tension, but only through the fact that they are already under tension applied during the building of the wheel.

the fact that the wheel works by tension of the spokes becomes obviously apparent when you start to remove the pretension and then the spokes will feel loose on the bottom half. of course you'd never want to ride a wheel like that because it will quickly become out of true.. just like a walmart wheel.

I think one of us needs to reread that book, because he emphatically denies that tension at the top of the wheel increases. It’s tension at the bottom that decreases.

> of course in a properly built wheel usually all the spoke are under tension...

No, a properly built wheel all of the spokes are always under enough tension you can bounce a penny off them. Always.

no. you're assuming the rim has no deflection which is untrue. if you build a rim out of schedule 80 steel pipe then yea. but 300-400-500g rims on high performance bikes do not act like that. the spokes are constantly loading and unloading tension as they bash through rocks and over jumps. the point is that the pretension on the wheel needs to be high enough the spokes do not loosen too much under these forces. if they do in fact loosen too much the nipples will begin to loosen and unwind and the wheel will become out of balance.

I’m not assuming the rim has no deflection. What makes you think that.

Anyway, this sums it up pretty well. Someone has a longer memory than I:

https://news.ycombinator.com/item?id=36891231

If you’re talking about twisted spokes unwinding, you don’t have to reach zero load for that to happen. You just need to reduce the load enough so the rotational force overcomes friction. Tension will also try to unwind a screw as well. But the thread pitch on spokes is very fine, which lessens that force. If you build spokes like wood screws we would have problems and that has nothing to do with reaching 0 newtons.

You can release a lot of those tensions by squeezing the spokes mid build. Just don’t wait until they’re too tight to do it. I had a pulse in my rear wheel that probably came from doing that wrong the first time. Unless it was a factory defect, I must have overtensioned and warped a brand new Mavic aero rim ever so slightly. Expensive lesson, but it could have been worse.

This dialogue is reminiscent of rec.bycycles.tech arguments with Jobst, ca 1993.

A bike wheel is a linear elastic system, that can be thought of as a superposition of a uniformly set of tensioned spokes as one state, and a set of spokes in compression in the loaded zone (bottom of the wheel) as the other state. So long as the superposition of the two states obeys the limiting conditions (i.e. spokes in tension) they can be analysed separately.

The size of the loaded zone is related to the relative stiffness of the spokes (axial) and the rim (bending), and can be calculated using beam on elastic foundation methods. For typical rim/spoke combinations, this is approximately 4 spokes.

Outside of the loaded zone, spoke tensions essentially don’t change.

And thread. Thereby proving my point: Almost nobody knows how bicycle wheels actually work.

Well. I wish we could sit in front of a bicycle wheel and discuss it. Because I have a feeling we are shooting arrows at different targets. As a mountain biker I'm more interested in what happens when an extreme amount of load is applied to the wheel, not the model with assumptions applied. Definitely a difficult concept to discuss with only text. Anyways.. glad to have a good discussion with you about bicycle wheels. Don't find many people like you. :)

I think I would have been a mechanical engineer if I were born 20 years earlier or 20 years later. Software was just new and shiny enough and it let me build things with my mind, at a time when I believed I was clumsy (I actually have always had excellent fine motor skills, it's macro motor control I lagged behind in). One of my better friends in college was an ME. Learned all sorts of things about metal fatigue and oddly enough picosecond lasers from him.

I don't know if I found Lego or Lego found me, but I definitely think in terms of shapes. I was past my midlife crisis before I realized that I don't have a large working memory (smaller than average in fact) it's just that I've been doing mind palaces without pictures since I was very small. When I'm thinking of large computer systems I'm essentially thinking of them as physics problems.

I really should figure out space to have a bike again. I never rode when I lived in Seattle (Seattle drivers are nuts) but I don't live there anymore and I need to catch up on 20 years of tech.

I think you're both in violent agreement using different terms.

You're looking at the macro "It's all in tension" (superposition of two states) and hinkley is looking at the "bottom is a compressive change" (dynamic portion of the load).

What I'm not clear of is if you think that the upper spokes change tension between the unloaded case and the plain gravity load case (force on hub down, ground on rim up at the bottom), or if you expect the top half spokes to increase and the bottom half to decrease in tension. I think this is what hinkley thinks you think.

That was indeed my interpretation of that half of the conversation. That they were claiming that the axle is suspended (tension increases with downforce) by the spokes above the midline of the circle, which is what Brandt vehemently contended was false.

Nah they both understand it. One of them is just arguing to show off.

The real winner will be who comes up with application for LK-99 to improve bicycles

Electric bicycles baby.

Hey do you happen to recall, did Brandt prove this theory with a tensiometer or did I just hallucinate that?

Pretty sure he did, but I don't have r.b.t. archives. He was definitely a proponent of them, preferred a specific brand/style, and would have easily been able to do the experiment.

Not to appeal to authority, but I wrote my thesis on bicycle wheels. You got it the wrong way around.

And have you read Brandt’s work?

https://news.ycombinator.com/item?id=36891231

The world is full of papers that are wrong. Including maybe the one this whole thread is about. It’s okay, it happens. Science doesn’t find right or wrong, though a lot of people think so. It finds more wrong and less wrong.

Yeah, that's complete crap. Bike dynamics is well researched and the instability and stability mechanisms are numerous but well understood.

We don't know how photosynthesis works but we know a ton about GROWING things. And life as we know it depends on photosynthesis.

I thought we had a pretty good understanding of this up until the molecular and electron level.

Am I missing something?

My daughter did her research in this field while an undergrad, and explained to me that photosynthesis is not yet well understood. Some of the theories of the mechanism are discussed here: https://physicsworld.com/a/is-photosynthesis-quantum-ish/

How photosynthesis works at a cell machinery level is high school level biology.

That's not an understanding of how it works

I disagree. It very much is an understanding of how photosynthesis works, it isn't a complete understanding sure, but an understanding nonetheless.

Unless you're going to argue that we don't know how an MRI works because we don't fully understand the physics behind superconductivity.

Reply on news.ycombinator.com