Very few things surprised me as much as rail design when I was studying structural engineering. Surface tension and fluid dynamics were both trippy too, but I expected those things to be complicated and while surface tension blew my mind due to the relative simplicity of the proof, and fluid dynamics blew my mind because it was somehow 100x more complex than I estimated.
Walking into rail design was hilarious. I worked on motorcycles and did some car stuff. I figured it was obvious, and sorta dismissed this assignment as a joke. Nope. My dismissive intuitions were just flat out wrong. It kinda leaves an impression on you to sorta avoid saying you know for sure before putting in some amount of work.
I worked on motorcycles for years using DIY guides and YouTube tutorials. Opening up real engineering books was an eye-opening and humbling experience that made me a better mechanic, driver/rider, homeowner and software engineer.
(For the curious motorcyclist, I recommend "Honda Common Service Manual" as a starting point.)
I know right? I also loved the times where I got to the math and all that internal intuition lined up completely. Torsional deformation for example. I almost blamed myself for not inventing the math myself it so obviously matched my intuitions. It's kinda fun no matter which way it goes.
Black body radiation never quite sat right with me. One of the few subject areas where I just resigned myself to memorizing the formulas and moving on with life. Same with non-integer dimensional spaces for the most advanced partial differential course I took. I can visualize 2 million dimensional spaces just fine professor. But one and a half? What does this even mean?
It just means that the degrees of freedom aren't used fully. This is usually seen in fractals, where there is a level of redundancy that is respective of the fractional dimension missing.
Think about it like this. If I have a 2d field (x,y) and I enforce every point's x value to be 0, I pretty much just made the x degree of freedom redundant, and can now call the field a 1d field. If instead, I enforce every 3rd point's x value to be 0, I've now got a 1+2/3 dimensional space. Because there is some redundancy, I no longer get the full entropy that 2 dimensions provide.
Never encountered this idea before, does it have a formal name? Most fractional-dimension spaces encountered refer to either the Minkowski or Hausdorff Dimension.
Here's a paper which discusses fractional degrees of freedom:
Effective degrees of freedom of a random walk on a fractal
by AS Balankin · 2015 · Cited by 42 — This allows us to define the fractional dimensional space allied ... number of effective dynamical degrees of freedom on the fractal
To hammer the relationship home, consider the holographic principle, which is based on observation of black holes, and states that our reality only needs 2 spatial dimensions instead of 3. Both Hawking and Susskind have eluded to this being the case solely because of the symmetries in the laws of physics. The symmetries cause a massive redundancy/pattern in the field values over the 3d space, such that we should theoretically be able to predict the state of the entire field given only the values at two-thirds of the volume.
Therefore, you can imagine a 2d surface which contains the state of our universe, and some kind of computational (possibly geometric/algebraic) projector, which understands the redundancies, reads the 2d surface, and renders a sparse 3d volume. In the case of our universe, the projection operation might be extraordinary complex, requiring a deep understanding of the laws of physics and the redundancies they induce into the underlying state that they operate on.
Math is more interesting, purely as math, as I get older. I found the same is true for history as well. The trick is to find good history authors for the particular time you are interested in. Same with math.
I graded very well in math through college, but later in life, I went back and explored more how all the concepts relate. In college, you are sort of fed calculus through a fire hose, and you just have to 'accept it' and move on. And you are left wondering, how were these ideas, these conclusions, reached? Until you go back through and see the long history of infinite series and see the various attempts to codify solutions. The problem is, as a student, you cannot possibly spend that much time deriving the whole solution from scratch and still hope to finish a degree in four years. As Carl Sagan said “If you wish to make an apple pie from scratch, you must first invent the universe.”
Which is why you can never stop learning. I am in my mid 60s, and I still learn something new regularly. Something big at least once a year, something smaller at least once a month. Never stop learning.
I meant that mathematics has been used for nuclear bombs, breaking cryptography, and other nefarious purposes. Physics is a sort of applied mathematics, where reality is the testing ground for models created on paper.
It was originally thought that Einstein's theory of special relativity was just a cool mathematical idea. But it was much much more. Mathematics has consequences that are so deep that even now it is rearing its head in AI..differentiable loss functions.
Isn't black body radiation a problem that drove scientists crazy for many years before Einstein found an explanation that got him a Nobel prize. That explanation is what started the huge mess that is quantum physics.
So I don't think anyone can be blamed for not getting it intuitively.
I'm not 100% sure whether or not you're attributing the right physicist here ;).
Max Planck wrote about quantized energy emissions from black bodies first in 1900. With this assumption, the spectrum of black body radiation could be derived successfully. That won him the Nobel Price in 1919. Albert Einstein postulated that light itself was quantized in one of his famous series of papers in 1905. This paper won him the Nobel Price in 1922.
[Side note: confusingly, Max Planck was awarded the 1918 Nobel Price and Albert Einstein was awarded the 1921 Nobel Price. This happened because the committee decided in 1918 and again in 1921 that none of the candidates met their standards and withheld the price for later.]
I have seen a couple ways to assign meaning to the words "one and a half dimensional" and neither of them are actual spaces (i.e. have a basis, are full of vectors).
I don't understand the first one, the second one isn't about continuous spaces (cantor dust has gaps), but the last one seems pretty interesting and I'm working through it. Thanks for the links!
Edit: The last paper only describes fractional dimensions in a limited sense, because the way the author constructs the coordinate system still involves a positive integer number of coordinates.
I worked professionally as a motorcycle mechanic from mid 1970s through the early 1980s. Honda's service manuals were very good, but my favorites were the early manuals from the late 1960s to mid 1970s, mostly for their Japanese to English translations.
One of my favorite lines from an early CB750 manual was "Pleased to be applying the 26mm spanner to the castellated nut. Thank you."
How can you not follow instructions when they are worded like that?
I learned about train wheels decades ago, when I briefly ran through train history as an interest. What I found interesting at the time was how thoroughly this was understood right at the advent of the age of steam. Rail engineers understood this from the beginning.
Another thing is that since the wheels ride an a narrow portion, going straight most of the time, wheel wear distorts the conical section, which leads to the wheels rubbing (instead of rolling) in corners, which accelerates the wear. So rail wheel wear is non-linear. Once the wear reaches a point wear it causes slipping/rubbing, it leads to positive feedback on the rate of wear. A failure mode that isn't obvious until you thing about it.
I wonder if a similar phenomenon is the reason my final drive chain and sprocket seems to wear a little after install, be stable for 10,000+ miles and then suddenly wear a lot.
Turns out designing a new system to fit requirements is orders of magnitudes harder than fixing a system somebody else designed.
You see this in software all the time. Anyone can follow a tutorial. But can you start from scratch and build something novel? Can you build it such that others can maintain long after you’re gone? That’s hard.
Same with cooking. Anyone can follow a recipe. But can you design a recipe?
This brings to mind my amusement when friends would bash Ikea cabinets as too flimsy. I was struck that they were just strong enough to do what they claimed. No more.
Everyone loves to shit on value engineering but the reality is that it's a lot harder and takes a lot more craftiness than designing things under less price pressure. When you start talking about revisions to complex things where you have to both work fast and avoid revising too much of the thing (to keep costs down) it gets really crazy.
You do know that the architects hand off their work to structural engineers on anything more complex than a single pavement building before it gets built, do you?
It has a resonance mode at 2Hz, now successfully damped. Bridge designers know to look out for resonances at those frequencies, because pedestrians tend to identify and accentuate them by walking, especially in large numbers immediately after opening. The resonance mode wasn't identified during design because unusually it is in the horizontal plane - most bridge designs have very slow resonances in that plane (sway) but the millennium bridge is a side mounted suspension bridge, with the pylons alongside the roadway.
Road curves can be quite surprising too. There are always exceptions due to topography and obstacles but optimal curves are not always obvious.
A simple corner on a motorway/autobahn/highway/etc has at least three curves in plan - in, out and shake it all about. OK, in and out and the main bend itself. The idea is that you need to transition the various forces into and out of the turn as safely as possible. A right angle turn is not a simple: straight -> quarter circle -> straight. As well as that, we have to consider line of sight and overtaking and water runoff and the effect of wind on high sided vehicles and ... and ... . Oh and of course these turns happen in 3D. I'm quite a fan of Holden Hill in Devon on the A38 - get your speed right in either direction and it feel effortless but you need enough power too. Purists in the UK will probably point at Snake Pass and most of Welsh roads and the like but I know Holden Hill.
This is sorta related to the small principle in visual design, that you shouldn't actually make rounded corners by joining a straight line (of infinite radius) to an arc of a fixed radius—because things don't work that way in nature. Instead, the radius should change gradually in and out.
If only Arrow Dynamics understood that when they designed their roller coasters. Maybe they'd still be in business instead of losing to B&M and Intamin.
AD seem to have been able to make the pitch flow quite alright, but dropped the ball on the roll movement. Shame—the ride per se looks exciting just from the vid.
The other problem Arrow had was the lack of heartlining. That is, the axis of rotation when banking.
For optimum comfort of the rider, when the track banks, the center of rotation should be at the rider's chest. Arrow fails to do this, and places the center of rotation between the rails. This causes any banking to create a significant lateral force on the rider, which whips the rider's head into the over-the-shoulder retraint. In a proper heartlined roll, the track appears to "slip" from under the train, like this: [0]
In the video I linked, the turnaround after the two vertical loops was particularly brutal and painful.
What makes it even worse is a poor design of the wheel assemblies. The wheels on the inside of the rail that keep the car centered are designed to allow a small gap between them and the rail, and they're not spring-loaded or dampened in any way, which causes extreme levels of hunting oscillations [0].
Their roller coaster Drachen Fire [1][2] was so rough and painful that they stopping running the ride out after only 6 years.
That's interesting you mention Haldon Hill. They may have got the A38 bit right, but if you go up the other road, the A380, then there is a quite sharp blind right bend at the top that clearly hasn't been done right.
That's due to topography I think and probably history too. Those roads have been around for quite a long time. I don't go down the A380 very often these days but I know the the A38 between Plymouth and Exeter very well.
Holden and Telegraph hills and that area in general have roads that are millenia old or their course was generally decided quite a while back and repurposed every now and then. Starting off life as livestock drovers trails and the like. The old Britons may have cut back the trees and the like a bit. The romans did proper engineering. After them things went a bit vague for a while. etc.
Have a look at this: https://www.romanobritain.org/7-maps/map_counties_roads_town... - the Fosse Way runs "under" the A38 for quite a way and may be the A380 too, south of Isca Dumnoniorum (Exeter.) I'm now in Yeovil and the Fosse is the A37 here. The stretch from here to Ilchester (Lendiniae) is called Roman Road and is straight and flat and prone to flooding!
I think the Holden Hill roads have been done quite well. The hill is a nightmare shape! Each square metre will belong to someone, who would have to be bought out or worked around. There's a race course at the top and a fuel station, houses, and who knows what else on it. The traffic volume can be huge, especially in summer.
My mum was from Devon and her description of driving from Newton Abbot (Ipplepen) to London in the sixties is pretty ... different to now. It often took two days by the time you decided to stop after sitting in traffic for hours (now: about three hours.) She also drove to Edinburgh and beyond quite often and that really took some time. I used to commute from Plymouth to Chertsey (near enough London) and that took around four hours in the mid nineties.
Next time you drive out that way or anywhere for that matter, cast an eye on the place. See the boundaries, think of the history. In Devon, look at the depth of the hedges - its called "Devonshiring": the dense hedging "walls" with lanes running through them. Many of those hedges are bronze age or older. Modern Devon and Cornwall are roughly what was Dumnovaria (according to the Romans) which is suspiciously similar to modern Brittany and rather closer to Wales (Brythonic) than what becomes modern England eventually. The Corns wave a flag and a language that sadly was only properly native up until about 50 years ago when the last native speakers died. It is being revived and is a Brythonic language, like Welsh (Cymraeg), Scottish (Gaelic), Irish (Gaeilge) and the rest. There is also Cumbric (Cumbria) and others. Devon was largely subsumed by Wessex (West Saxons) earlier than Cornwall and hence is a bit more English (whatever that means.)
The reason that I'm wittering on about history is those roads are seriously old and have a context. Imagine who else has trod those roads back in the day. For example the patron saint of Germany is St Boniface. He's from Crediton.
Almost anything when done at the fine edge of engineering for optimum price point without sacrificing reliability is going to be an amazing thing to do a deep dive in. Windmills are another, they seem so simple and obvious until you dig in.
Agreed. I still don't really understand it, I imagined there would be lots of people working on off the grid houses who would be more than happy to pick up a free windmill but apparently that wasn't the case. So much work down the drain. Anyway, I shouldn't dwell on it, I've been lugging it with me from one residence to the next since 2008 so it wasn't for lack of trying.
Motorcycles turn at speed by a similar principle, though the cones arrangement is a bit flipped around.
I usually demonstrate it with two solo cups put mouth-to-mouth, to make a pair of facing cones that represents the motorcycle tire. The starting condition is that you're above parking lot speeds, and the bike is stable and is dynamically inclined to stay perfectly upright. To go left, you turn the bars right to upset the stable bike onto the left cone, and it goes left. To go right you turn the bars left and it upsets the bike onto the right cone, and goes right.
This can also be easily experienced with a bicycle. Just push the handlebar forward on one side and watch/feel it tip over to that side instead of the other side where the wheel is pointing to.
In fact even for a bicycle at the lowest speeds counter-steer is required. Now that people have read this there will be a whole new group of people who on their next bike ride will think "According to theory i must be subconsciously turning the handlebars left in order to perform a right hand turn?"
And then will you notice yourself doing it. It's quite remarkable. All those years you thought you turned the handlebars into the turn. You've actually been turning them the other way subconsciously in order to lean into the turn.
Last time I saw this discussion I checked what I was doing and found that I don't usually countersteer when turning so it does not actually seem to be required. It was a while ago so I don't remember the details but I do remember making a very low speed turn with my hands positioned to not be able to countersteer and it worked fine. I'm guessing the alternative is to lean before turning.
At very low speeds (tight turns, i.e. high curvature of the bike's path), the highly-curved path requires steering into the turn. However for most turns (higher-speed) the curvature is less significant so the handlebars don't really turn at all - you do however need to apply a force to the handlebars to establish and maintain a lean into the curve - this force being in the opposite direction to the turn (i.e. when turning right, you push forward on the right handlebar).
I'm still not convinced countersteering is ever necessary since it is obviously possible to lean a bicycle without turning the handle bars at all. Why wouldn't lean first then turn the way you are turning work? Or even lean while turning the direction you want to go. I am slow in general and for me most turns are sharp and low speed.
Any kind of serious turning involves lateral acceleration. The bike must lean (well, the centre of gravity must be offset laterally from the wheels contact patches) to compensate. It's an inverted pendulum, with a feedback mechanism (caster angle of the forks) and the combined effects of gravity and centrifugal force (from the rider's perspective). There are also minor secondary things like gyroscopic effect of the spinning wheels.
The equilibrium position (lean angle) of this feedback system can be shifted by the rider either by applying force to the handlebars, or shifting the centre of mass away from the middle (leaning your body to the side).
It's possible to steer by only shifting weight, but a bike with locked handlebars would not stay upright (no feedback mechanism). Applying force to the handlebars is the way to get maximum control authority over the system.
Edit: more detail... the tilt angle is defined by the lateral offset between the centre of mass and the wheel contact patch. The rider shifting his weight achieves this by moving the centre of mass, while using the handlebars moves the wheels to the side while leaving the centre of mass (mostly) unmoved. Of course, the wheels' contact patch can be moved further and faster laterally (by steering) than the rider's weight can be shifted (which is limited by the rider's flexibility).
Sounds like you are agreeing with me then? I'm not saying no handlebars at all, just that if you want to turn right it doesn't seem to actually be necessary to slightly move the handlebars left first, which is the claim that is often made when this discussion happens.
I agree it's possible in principle to ride a bike without the handlebars (have done this myself on many occasions, including motorbikes), but more dynamic maneuvers do require use of the bars (e.g. swerving to avoid an obstacle).
Sudden maneuvers are an essential survival skill on two wheels, and it's important to understand how to steer rapidly. A rider who just steers intuitively is vulnerable to a phenomenon known as "target fixation": becoming mentally focused on an obstacle and steering into it. Having a rational understanding of how to steer allows the rider to exert conscious control over the bike's direction in these situations.
>However for most turns (higher-speed) the curvature is less significant so the handlebars don't really turn at all
Is it possible to turn a bicycle at higher speeds without leaning? Ie turn like a car by just turning the front wheel. I've tried it a few times and it's been very difficult.
Horizontally, a bicycle is an inverted pendulum: to accelerate it in one direction it must lean in that direction. It's still possible to keep the frame of the bike upright in a turn but as the rider you would have to shift your weight into the turn to compensate.
If you consider the whole system (bike and rider) as a point mass on a stick, it's still leaned over (centre of mass of not above the tyres) in that case.
That is actually not what I experience at all. After I first read about the counter-steering thing on bicycles, I tried to observe exactly how I corner. I never was able to observe counter-steering, and forcefully counter-steering felt very wrong. I just somehow succeed in reaching the correct leaning angle for whatever turn I want to make.
Whatever happens must be too subtle to notice, and/or completely subconscious.
One of the ways I keep myself occupied on long empty highway rides is to turn my cruise control on and keep myself in the lane by "punching" my handlebar. Punch the left side to turn left and the right side to turn right.
Yes, I'm talking about motorcycles. Most touring bikes have cruise control these days. Ducati's newest Multistrada even has radar cruise control and auto-braking.
The risk is definitely quite great and requires personal responsibility. Avoiding drug and alcohol use and using modern safety gear (not DOT junk) greatly reduces the probability of serious injury, but cannot eliminate it. It is a fantastic pastime, though.
There's always dirt riding, track riding, and mountain biking to scratch the itch without traffic!
I loved motorcycling, but gave it up because I could not accept the risk (about 7 to 10 times more likely to die per mile driven vs a car). The most common (about 45%) cause of a motorcycle fatality is a car making a left turn into a motorcycle. I took a lot of training classes, had good gear and always practiced my situational awareness, but after my advanced riding instructor with over 35 years of experience and the best gear money can buy was killed instantly by a driver in a minivan texting while making a left turn, I gave it up.
The feeling of riding a motorcycle is truly amazing and I absolutely miss it, but the risk factor is simply too high for me to be worth it.
DOT is the 'minimum' requirement. There are certainly better helmets. That said, the DOT requirements came into being in response to the number of junk helmets in the 1960s/1970s. It provided a minimum bar that a rider could rely on. After some states actually mandated DOT helmets, there was a huge market for fake DOT stickers from riders that were still wearing crap helmets for 'style' points. Mostly cheap half-buckets.
I started riding in the dirt at age 15, racing 3/8 mile dirt track at 16, and motocross at 17. Got my first street bike at 19. Rode about 100K miles until I moved from a rural part of the country to a heavily urban area, and gave it up because it was no fun in heavy traffic.
Avoiding alcohol and drugs, definitely. Wearing good safety gear, also definitely. I trashed one helmet in my life, walked away with zero injury. Equipment includes proper shoes, pants, jacket. I had one high speed slide, head never touched the ground, but I lost a couple patches of skin, and that was with leather. Modern materials are better. Denim is worthless. Also, carefully choose who you ride with. I had a number of peers that I would not ride with, they were accidents waiting to happen, and I didn't want to get caught up in their dumb ass mistakes.
This can be demonstrated easily to someone who doesn't believe they are doing it. Just ask them to ride their motorcycle with only the right hand on the throttle and their left hand off the bar.
What happens when turning a bike while riding with no hands? Does the lean induce the handlebars to counter-steer briefly, or is the counter-steer not necessarily because the lean is already happening?
I think you are right (but we could both be wrong).
The way I experience countersteer is that it pulls the bottom of the bike towards the outside of the curve, in order to lean or maintain a lean. If you lean with balance it’s not needed.
With no hands, the equilibrium lean angle is controlled by the rider shifting weight (centre of gravity) from side to side. The caster angle on the front wheel is the negative feedback mechanism that keeps the bike balanced.
Not sure the counter-steer makes you actually turn fully the opposite direction, it might be a force you apply in opposite but overall the handlebars still turn into the turn?
When you turn the handlebars to the right, it caused the bike to lean to the left. When you stop applying force to the handlebars, the front wheel “falls into line” with the left hand lean and they turn to the left.
However, it’s so subtle that in practice you don’t feel them turn back to the direction of the turn and it fully feels like you’re steering right to turn left.
Just to add this explicitly, to expand on some specifics, motorcycles have the added complexity of angular momentum and its effects. As you mention, turning the wheel effects your lean angle through angular momentum conservation. The cone then is the dominant factor causing the bike to turn.
The speed matters because as the speed is faster, a smaller angle change in the handlebars corresponds to a bigger sideways tilt motion of the bike.
The irony is that the majority of people will think you're making it up when describing counter steering, yet they are already doing it without even thinking about it all the time.
The first time I read about counter steering, I thought "man, this'll take forever to practice" until I realized it's really the only way one ever does it.
I blame that more on the people who "explain" it. Bad explanations of relativity or quantum mechanics have nothing on countersteering - everyone who talks about it does it as "woo, this super mysterious thing".
In particular people say you can't turn left without first turning right and vice versa, which is obviously nonsense - if that were true as stated then it would be impossible to ever turn at all.
The main advantage to knowing is that you can stop wasting energy with footpeg weighting or unnecessary leaning and use bar pressure. (weighting and body position have their uses but for most riders in most situations countersteering is the most efficient technique.)
That's when they call it push steering. But it's really the same thing. When you push, you are causing your bike to counter steer. You are literally pushing the bar forward, even if you think you are pushing downward, which is by definition a counter steer.
I think a lot of people assume those are different things. But they aren't. You simply can't turn at speed without counter steering, regardless of how you visualize the mechanics.
You are spot on for people who think pushing with their foot on the inner foot peg has an effect. I'm sure it does, but it's a lot of wasted effort given that the end result you are looking for is the bar turning the opposite direction, and you'll be doing that whether you are conscious of it or not. It takes far less effort to simply not think about it, and do what comes natural since your arms and hands will inevitably do the right thing without any "different" theories interfering.
It's also why 3 wheel bikes are notorious for throwing the rider high-side if they corner too quickly. It stops you from leaning, so there is no ability to counter steer. You end up behaving more like a London double-decker bus in the turn.
The only time I ever notice it is if I'm riding up really close to a curb or something and it feels like it's somehow pulling me in, like there's no way to steer away from it.
Intuitively it feels weird, but logically it makes sense - if I'm too close to something, I can't turn briefly towards it in order to get away from it.
Same thing in mountain biking and getting up on the edge of the trail. It can look like you just rode straight up it for no good reason, but in reality you intuitively felt that you had no room to counter steer without running the front wheel off the track.
You realise that motorcycles existed for literally decades before it had a name, how where they turning corners at speed, hell how was I doing it in the 90s on scramblers when I was a kid, no one taught me, it’s intuitive.
Totally correct. At speed, you simply cannot turn without counter steering. If you tried to turn by turning your bars in the direction you wanted to go, you simply won't go there.
At some point, someone noticed racing bikes counter steer really severely so you can easily see the wheel is actually pointing the opposite way. But the reality is that even at "won't fall over" speed on a bicycle, you're already doing the exact same thing. When you lean, you counter steer. Otherwise you'll high-side like a missile.
One of the Wright brothers has a quote about this (before they were airplane makers they were bicycle makers, the bicycle being the latest mechanical marvel of their time). The summary of the quote is that the action is intuitive but nobody realizes (or even admits) that they're doing it:
> I have asked dozens of bicycle riders how they turn to the left. I have never found a single person who stated all the facts correctly when first asked. They almost invariably said that to turn to the left, they turned the handlebar to the left and as a result made a turn to the left. [...] I have never found a non-scientific rider who had particularly noticed it and spoke of it from his own conscious observation and initiative.
The existence of counter-steering is still controversial to some riders, to the point where machines like the "No B.S. Bike" were created to demonstrate it as a necessary effect: https://soundrider.com/archive/safety-skills/nobsbike.aspx
Another way to demonstrate it on a bicycle, you can do something pretty much every child has already done numerous times: ride without your hands on the handle bars (hands free).
To turn, you lean in the direction you want to go. But what way does the handle bar turn when you do that? It counter steers! You will fall if it doesn't (which is essentially what the No B.S. Bike demonstrates).
It is a bit of a brutal watch. FWIW it says the rider survived, and it does clearly demonstrate how countersteering can be non obvious in some situations. (In the video the rider is already in a corner and needs to tighten his line (sharpen the corner). Instead of counter-steering, he tries to simply turn the bars into the corner more, which has the opposite effect, and causes him to hit the truck.)
This is a good, relevant video and should not be so heavily downvoted. Yes, it is NSFW, but we're all adults here and this is an informative video aimed at adults, to teach the subject and demonstrate where it _applies_.
I realise this isn't a very helpful comment in itself, but I hope Hacker News comments don't go the way Reddit comments have, where they routinely post and upvote content that's already in the linked article (usually because few people read beyond the title).
In this case - just reposting the Feynman video - it's fine, but in other cases it leads to a lot of uninformed or unnecessary discussion, sometimes speculating on some hypothetical that the article already answers.
Particularly good parts are the explanation of fire and trees ("trees come out of the air"): https://youtu.be/nYg6jzotiAc?t=440 and the explanation of the mirror problem, i.e. how does a mirror know to reverse left and right but not up and down: https://youtu.be/nYg6jzotiAc?t=1976
Elsewhere in this thread, jcrawfordor explains that hunting oscillation from conical wheels can cause “chatter” at higher speeds which led to BART’s experimentation with cylindrical wheels [1][2]. Others also note that the wheels of high-speed trains are closer to cylinders than cones. [3][4]
BART in general was designed by aerospace engineers, because people in mid-century America thought that rail was an old outdated thing to replaced with monorails, personal rapid transit and maglev. The people who designed BART from the ground up on first principals (and ignoring everything that rail had done previously.)
The problem with this is that newer technologies were not competing with rail purely on the technological merits, but with rail's massive amount of competition in suppliers and economies of scale. It turns out that with not a lot of effort, you can apply much the same savings and technological improvements to traditional rail, except without the huge added expense of converting to a standard you are currently incompatible with. Newer technologies have the reverse problem, in that they have very few, sometimes even a single supplier, and are basically custom projects with all the expense that entails. So the monorails, PRT, maglev, and other weird system are mostly unique specimens, or very few in number.
This is the explanation I've heard over and over, but it seems a bit simplistic, stemming from an unknown article decades ago, becoming nearly myth or legend at this point.
I've yet to see a photo of a BART wheel up close, nor an explanation as to why they supposedly chose a cylindrical wheel. I've looked. I can't imagine the design decision happened without some knowledge of why conical wheels were used in the past. Call me skeptical.
I think this may be the primary reason why narrow-gauge railways are better at tighter curves: the shorter axle means the same wheel radius difference (caused by lateral displacement) causes a smaller turn radius versus a standard-gauge axle.
Are the cones mainly for keeping the train centered, or for allowing the outer wheel on a turn to be effectively larger? Not sure if the differential effect is an added benefit that isn't really necessary, or if it is the main goal of the design.
Both are factors in good centering, but mostly the change in diameter. In turns, there is a natural tendency for the train to shift towards the outside of the curve due to inertia. The wheel diameters become asymmetric which helps to re-center the train. It's usually not sufficient on its own, which is why superelevation is used as well - the outside rail is somewhat higher than the inside rail which shifts relative gravity to pull the train back towards the inside as well. The relationship between these two effects is a bit complex (depends on weights and speeds of trains) so it's usually all a bit approximate.
The conical section of the wheels is mostly intended to prevent hunting on straight track, and the shape can't be made too aggressive without increasing the wear on wheels on rails. So on curves the superelevation is added to provide the extra force required.
Because conical wheels do increase wear and can contribute to oscillation in their own way, there have been experiments with cylindrical wheels especially on higher-speed trains---BART is a well known example. It ultimately didn't work very well and so they have been re-trueing the wheels to a non-cylindrical profile, although still not quite a traditional conical one. Basically in higher-speed operation the re-centering effect is too significant and causes one wheel to "chatter," which over time creates a significant vibration in the rail. Trouble is cylindrical wheels tend to cause the same thing to happen on the other side. It was a very hard problem before computer modeling became available.
> It's usually not sufficient on its own, which is why superelevation is used as well - the outside rail is somewhat higher than the inside rail which shifts relative gravity to pull the train back towards the inside as well. [...] So on curves the superelevation is added to provide the extra force required.
I've never heard about that theory as for why superelevation/cant is supposedly being used until now.
Given that most of the time you'll end up with a remaining net force to the outside of the curve even after application of cant, it doesn't seem to make that much sense, either.
That's the conventional explanation of superelevation, although I worded it in sort of an odd way. But I'm describing the same thing that e.g. Wikipedia does. Superelevation directs the force of the car more "straight down" in relation to the rails which improves centering and balance of the load by the same token. The thing I said about "shifting gravity" is unnecessarily confusing because it depends on reference frame.
I think for low-speed freight the balance needs to be pretty close on to ideal to meet regulations, e.g. FRA regulations give calculations for acceptable ranges. But since it's dependent on running speed it's hard to get correct for freight and passenger mixed operation which is the subject of this FRA report that has a lot of detail on the calculations: https://railroads.dot.gov/sites/fra.dot.gov/files/fra_net/19...
Hmm, well the conventional explanation that I know of is that it's simply to reduce the lateral forces acting on your train and more importantly on the payload you're carrying - especially with passenger trains it's passenger comfort that's the limiting factor by far, not safety against derailment or overturning (which is how tilting trains can work, since tilting the train body only reduces the forces felt inside the passenger compartment, but not the forces acting at the wheel-rail level).
I see what you mean with regards to how it's also described on Wikipedia – only I've got some currentish (European) literature in front of me which claims that cant and the resulting cant deficiency/excess are only of secondary importance with regards to wheel and rail wear (the main factors are simply the curve radius itself and the construction of the running gear of the trains operating over the curve), and as such the main importance of cant is simply ride comfort. Likewise it also claims that according to some practical experiments done by some infrastructure operators, no link could be found between occurrences of cant excess for slower moving heavy freight trains and increased maintenance requirements (Which interestingly somewhat contradicts the corresponding supposition given in your FRA document...).
This also matches the evolution of the design rules on the German national railways – in the 80s there still used to be a relatively elaborate system of determining the allowable cant excess for slower moving trains depending on the annual tonnage of that kinds of trains, but since then at some point that system got dropped and has been radically simplified:
The regular cant is simply 55 % of the equilibrium cant and it's up to the design engineer to deviate from that value if necessary (when the speed distribution varies from that of a normal mixed-traffic route).
Interestingly all of that somewhat contradicts the statements given in your linked FRA document. To some extent this can probably be explained by European freight trains being shorter, somewhat lighter (lower axle loads) and also nowadays slightly faster than their American counterparts, and also due to traditionally using somewhat higher allowable cant deficiency values, especially with regards to passenger rolling stock.
It likely doesn't explain everything, though, but I don't know enough, either, to reconcile those two differing points of view.
They keep the train centered by the differential effect. The angle of the cone is very slight, nowhere near enough for gravity to overcome friction to cause the train to slip laterally into the center. And then keeping the train centered as the track turns results in the train turning with the track.
Sure, but I guess I'm wondering if that is just a convenient effect of the centering, or if it is actually necessary to prevent the wheels from skipping.
To demonstrate this more fully, consider the case of having the wheel flanges on the outside, with the conicity of the wheels pointing the other way. Gravity would still tend to centre this arrangement, but I'm told that if you build such a system in practice, then it won't run nearly as smoothly.
(PhD was 'Residual stress in rails', for what that's worth. Judging from the profiles of the rails I saw, direct contact with the wheel flange plays a substantial role in keeping the train in place on curved track. But on roughly straight track, I'm satisfied that the argument about conicity applies).
> direct contact with the wheel flange plays a substantial role in keeping the train in place on curved track.
The London Underground has some lines that are horrifically loud. The squealing must surely be at dangerous sound levels. I’d always assumed it was the flange against the rail, and you appear to be confirming that?
That's called "flange squeal". Yes, it can be ear-shatteringly loud.
But it mostly (totally?) happens on very tight curves. It shouldn't happen much or at all on gentler curves.
(Of course, this is circular, because I'm kind of defining "gentler" and "tight" based on whether they cause flange squeal. Still, there's a point - there is something like a threshold of curve tightness where flange squeal becomes much more probable.)
It can happen on straightrail on an incline, too. It's hard to assess precisely what's happening in the locomotive, but under traction I believe the running gear will toe out, and align based on the path of least resistance. I believe this leads to the flange pressing, with immense force, against the rail. You also get a lot of wheel slip in this condition.
I surmised this running 2 motors up a 3%(?) grade with 20k ton gross at 10mph. It's about the only explanation I could come up with is that the running gear was twisting under the gravity and the energy being put down to work against it. It might also just be a stringline sort of effect dragging the motors to one side of the track and pressing the flange. Maybe one of the rail engineers will come holler at me for my poor trainhandling skills.
The goal is for the train to NOT be centered when cornering, the cones allow for this. If the train was centered during corning, or simply with non-coned wheels, the outer wheel would necessarily slip.
Yes, the cars' weight rests on the end of each axle via a "bogie" that holds the suspension and brakes and such, and then the multi-axle bogie itself rotates on a center pin:
They are, and for non-driven wheels, the bogey and train cars basically sit on top. There's only a few minor things like brake hardware that need to be removed to remove a wheel. When there is a derailment, many of the wheel sets fall off.
The axles are tough. Each axle weighs about 1 ton if I remember correctly. Each wheel can be reworked on a lathe several times (either with the wheel set removed or in situ on a drive-through floor-mounted lathe). After a few years, the diameter of the wheel is out of spec, and new ones are pressed on the axle. Axles can last about 75 years.
Because the differential effect requires the wheels to be rotating in sync, so the conical sections can do their thing. If the wheels could independently rotate at their own speed, then there would be nothing stoping the train from derailing itself (or bouncing off the flanges that you’d have to incorporate)
It exists. They are called fast angle wheels in the modeling world.
http://cs.trains.com/ctt/f/95/t/79912.aspx
To quote:
Fast angle wheels first came out when MPC took over Lionel. The wheels are not squared off where they ride on the rail. They are angled to the flange. "Fast angle" is a toolmaker's term for adding an angle to a surface so the part can be quickly removed from the tool without marring the surface during manufacture. Hence the term "fast angle wheel" was coined by Lionel employees.
The fast angle did more than benefit manufacture. Because the wheels are fixed to the axel, it benefits them on curved track. The wheelsets can drift to a point where one wheel diameter point touching the rail is slightly larger than the opposite wheel diameter point touching the rail. This reduces friction because the outside rail is longer in circumference than the inside rail. Especially sharp 031 or 027 curves. If you look closely, you can see the cars lean into the curves as the outside wheels drift to a larger diameter.
a similar thing becomes true for motorcycle and bicycle tires in a curve, without the differential effect of two wheels on one axle: when leaned over, the contact patch of the tire deforms conically and the effect is like rolling a solo cup on the ground: it “wants” to keep turning.
Of course, pneumatic tires have cones that adjust their shape on the fly...
But yes, historically the awful screeching around corners was because BART used cylindrical wheels. It's also, apparently, why they can't run all night -- the tracks need nightly maintenance due to the grinding.
I wonder what the reason for not making them conical in the first place was, given that this knowledge has been around for over a century now. Maybe they were worried about https://en.wikipedia.org/wiki/Hunting_oscillation ?
If you're curious about the constraints of BART and the history involved in its development that led to why it is the way it is, I can certainly recommend Michael Healy's book BART: The Dramatic History of the Bay Area Rapid Transit System.
Heh... I don't think I have the patience to read a book about BART but I'd read the cliff notes on the cylindrical wheels decision. There are so many mentions of the cylindrical wheels on the internet but none of them explain why they were chosen when conical wheels were already well-understood at the time... I'm sure the engineers weren't just ignorant.
Cylindrical wheels were expected to reduce hunting oscillation and rail wear, which were particularly significant problems for BART because of the high speeds it operated at. The basic problem is that computer modeling was not yet available, and so the new design was validated experimentally using a set of instrumented test carriages on a short rail section built for the purpose. This found positive results on improved ride, but failed to detect the long-term problematic track wear. BART wheels have mostly been re-trued to a new profile which is not cylindrical, but also not quite a traditional conical section, and was designed with extensive use of computer modeling.
The cylindrical wheel decision is closely related to the decision to use Indian/broad gauge, which was expected to provide a smoother ride as well as allowing more support equipment to be mounted under the car where it would produce less vibration.
Both are decisions that have not stood the test of time, although the choice of Indian gauge cannot practically be reversed. But I think the discussion around this often pays the original designers far too little credit: BART was intentionally a highly innovative design with numerous aspects that were somewhat experimental. BART's automated control system, for example, was such a debacle that BART initially operated with signal towers and the control system required nearly complete replacement. But it was a completely trailblazing design, and the same missteps would have to be made somewhere. BART was used once again as a test platform for an innovative radio control scheme in the 2000s, evidence of which can still be seen mounted trackside on the SFO wye.
Many lessons learned from BART's performance have contributed to later designs around the world, including notably the DC Metro which was built just shortly after by some of the same contractors.
There was absolutely nothing "trailblazaing" about the BART engineering, rather they were just completely ignorant of the world around them. The St Petersburg metro had ATC and platform screen doors way back in 1961. The spread-spectrum radio research project was also a gigantic waste of taxpayer dollars (radio-based signaling was a common off-the-shelf technology by then).
> although the choice of Indian gauge cannot practically be reversed.
It would probably be very expensive, and might not have ROI, but couldn't you lay narrower gauge rail in between the current rails, then modify or replace rolling stock to use the smaller gauge... and once done, remove the old broad gauge rail?
Looking at some random BART rail images, laying the new rail would be difficult; some places have equipment between the rails, other places have concrete between the rails. It would probably need to be a very long project; early stages could just be verifying which sections would be feasible to add narrower gauge to and making sure new construction would allow for it and when rework is already happening, consider working room for narrower gauge into the maintenance. You could really only progress sensibly once at least one line was nearly ready.
I guess the question would have to be, would all the expense and time it would take to switch to a narrower gauge, be justified the improvement in user comfort (I assume) and benefits of using more standard equipment.
I can't seem to find this online, but I would assume that BART cars are wider in general than most standard subway cars, not just in track gauge. If this is true, you would also need to close platform gaps to continue to meet ADA requirements (a large enough gap is considered inaccessible).
It would be a lot of expense, for not a whole lot of benefit in the short term, and BART probably doesn't have enough money with reduced COVID commuting trends to even think about paying for something like this when they could instead build more rail and get more butts in more seats.
The presentation linked in https://news.ycombinator.com/item?id=28355136 says the BART car width is 3.2m, which seems to be the typical North American passenger car width [1]. The BART cars certainly look wide, but that's because they're not very tall.
This is wider than most metro trains (especially 19th century systems), but BART isn't a metro system anyway.
> The cylindrical wheel decision is closely related to the decision to use Indian/broad gauge
My recollection is that this was also related to running (relatively) high-speed trains on elevated tracks in windy areas, and so wanting additional lateral stability.
The Paris metro's noise does not compare with the incredible screeching of the BART.
Also, I think the Paris metro has special requirements, including quite sharp turns which I'm not sure trains can handle. Some of the lines in Paris actually use tires with side rails to channel the train to handle this.
They are also pressure fitted onto the axles. They are not bolted or otherwise fixed. The only thing which keeps them on the train is friction and the rail itself.
I was on a commuter train which derailed because it LOST a wheel. I don't understand why they aren't at least locked in mechanically. That wheel went rolling at 80MPH and blasted straight through trees near the tracks.
The wheels also have to have the right size to not get resonance. This have been a problem in Norway when the train reach 200km/h, because they forgot (?) to factor this in.
This feels like a really bumpy road at high speed, and stop if the train driver reduce the speed just a little bit.
High speed rail wheels are not cones, or much less slanted cones, but they are turning in a much larger radius, and sometimes having independent wheels.
I can’t find anything from either - though did read about the institutional racism dated by Feynman. Imagine being the person who questions his suitability for a Phd.
Yeah I was referring to Feynman. At the end of the video embedded in the article, he ends a story by saying something like "these were the sorts of things you had to know in the fraternity."
Meta observation: The top two comments are indicative of quality drift in HN. The first one, from 3pt14159 is inquisitive, interested, and humble. The second one, from the aptly named garbagetime, is dismissive and rude. Let's all please try to be more like the former, and less like the latter.
meta meta observation: You commented rather early in the posts history. Currently garbagetime's posts is near the bottom, while 3pt14159's is at the top.
How many posts were there at the time of writing? Did your comment influence the subsequent voting? Would garbagetime have been naturally downvoted if given enough time? Is hacker news actually declining in quality, or is it just tendency to favor good things when remembering the past?
Walking into rail design was hilarious. I worked on motorcycles and did some car stuff. I figured it was obvious, and sorta dismissed this assignment as a joke. Nope. My dismissive intuitions were just flat out wrong. It kinda leaves an impression on you to sorta avoid saying you know for sure before putting in some amount of work.