9 ms·
All of the Hyperloop development efforts are ignoring the elephant in the room: thermal expansion. If you do the math, the ends of a viable hyperloop track wil
by lisper 9y ago
All of the Hyperloop development efforts are ignoring the elephant in the room: thermal expansion. If you do the math, the ends of a viable hyperloop track will have to move hundreds of meters [1]. No one has yet advanced even a viable idea for how to deal with that, let alone an actual design. Until that happens, the Hyperloop is vaporware.
[1] It's a trivial calculation. The thermal expansion coefficient of steel is about 10^-5. A typical run of, say, SF->LA is 600 km. Temperatures in the central valley range over about 100 degrees. Multiply everything together and the result is 600 meters.
- iamgopal 9y agoExpansion joints are available since decades.
- lisper 9y agoNot ones that will hold a vacuum.
- madamelic 9y agoI am not an engineer. Is it impossible to create one? If a tube is a vacuum, then the tube moves, I don't think that negates the vacuum or does expansion and contraction make a vacuum impossible? I wouldn't believe so.
- dx034 9y agoImpossible not. But expensive. And hyperloop is supposed to be cheaper than HSR.
- LeoPanthera 9y agoNot with that attitude.
- dclowd9901 9y agoYour comment may seem flippant, but honestly it seems like people are more eager to be right than aspirational these days.
- TeMPOraL 9y agoIt's not about being right, it's about pure attitude - "why this can't work" vs. "let's talk how can we make this work".
- dclowd9901 9y agoI don't think people are actually that defeatist. If it came to a difficult situation in their own lives, I bet they'd rise to the occasion, but they'd rather see others fail than succeed greater than they did.
- Robotbeat 9y agoDid you try googling? Here's a whole bunch: http://www.metraflex.com/metal-expansion-joints/ http://www.metraflex.com/metal-expansion-joints/
- lisper 9y agoLet me be more precise: not ones that 1) will hold a vacuum and 2) are available in the 3+ meter diameter that the hyperloop requires and 3) provides the smooth inside surface that the hyperloop requires and 4) are economically viable.
- nkoren 9y agoThis kind of Hyperloop criticism is commonplace, but strikes me as intellectually lazy. Did you read Robotbeat's link? Particularly with respect to to the MNLC Bellows Expansion Joint. > 1) will hold a vacuum Plain wrong. > 2) are available in the 3+ meter diameter that the hyperloop requires This is a simple manufacturing problem, although calling a "problem" seems like an overstatement. Obviously not everything to build a Hyperloop is wholly off the shelf. If there is some reason why this can't be built in larger diameters, then that's certainly not obvious. > 3) provides the smooth inside surface that the hyperloop requires The Hyperloop runs on an air cushion. It's a hovercraft. Millimeter-scale bellows ought not to be a problem. > 4) are economically viable And here, you could be entirely correct, and is why I'm not yet a Hyperloop true believer. I'm not wholly convinced that this kind of machinery can be economical over its whole life-cycle in its intended service environment. But proving that this is the case requires more than glib, hand-wavy assumptions. You need to run the numbers. I have yet to see any Hyperloop critic do that in a remotely convincing fashion, whereas there are many Hyperloop engineers who most assuredly are running numbers. Maybe those calculations are wrong, but the only way to refute them is with better calculations. Anything else is just dogma.
- mrguyorama 9y agoHyperLoop does NOT run on an air cushion (remember, the entire tube is supposed to be nearly evacuated!), but instead floats on a maglev track. He's also missing another requirement: 5) The entire track, including any possible expansion components, valves, inspection hatches, etc etc etc have to be at least mildly tolerant to intentional or accidental damage. Running into a sudden wall of air at 700mph is a good way to destroy any passenger craft in the system.
- dmoy 9y agoI am curious, does an expansion joint not help here? Is it impossible to make an expansion joint that deals with the kinds of pressure involved?
- olegkikin 9y agoExpansion joint is likely the way to do it, but I'm not aware of vacuum expansion joints of that size. Even if it's possible, it sounds very expensive to make and to maintain. Just like everything in this project. That's the thing about Hyperloop. It's not technologically impossible. It's just not economically viable.
- TeMPOraL 9y ago> That's the thing about Hyperloop. It's not technologically impossible. It's just not economically viable. That's probably why this is an idea from Elon, not from the industry. The guy has a habit of forcing the market itself to accept beneficial ideas, and in the process they become viable.
- ErikVandeWater 9y agoJust to clarify - is the thermal expansion issue both in terms of the length of the hyperloop and the inside diameter of the tube, or just the length? Relevant discussion of thermal expansion with high speed tracks: http://boards.straightdope.com/sdmb/showthread.php?t=471152 http://boards.straightdope.com/sdmb/showthread.php?t=471152
- lisper 9y agoIt's a combination of things: the length, the fact that it has to hold a vacuum, and the fact that it can't make tight curves. If any one of these things weren't true the problem would be easy. But they are all true, and that makes it really, really hard. (Oh, and it has to have a smooth inside surface, and it has to be manufacturable at a reasonable cost.)
- pasta 9y agoYeah, Elon Musk never thought about that because he was busy landing rockets. Come on.. why do you think he did not think about that? By the way, curves in the track can eliminate this when they can slide a little. And I can imagine there are more and better solutions.
- oconnor663 9y agoRelevant: http://www.commitstrip.com/en/2016/06/02/thank-god-for-commenters/ http://www.commitstrip.com/en/2016/06/02/thank-god-for-comme...
- kpxxx3 9y agoJust like how all those experts at Tesla had no trouble scaling up mass manufacturing.
- lisper 9y ago> why do you think he did not think about that? Because this issue has been brought up many, many times since the Hyperloop paper was first published and no one has yet proposed a viable solution. > curves in the track can eliminate this when they can slide a little But you can't curve the track. The track has to be very nearly straight or the G forces become intolerable. This is the thing: lots of people glibly propose what they think is a solution, but none of them actually work.
- cma 9y agoElon can get things wrong. He already said after he had announced and hyped hyperloop he had to delay the white paper because it turned out his initial idea just didn't work and he had to come up with a new one. In 2016 he said we would have full level 5 autonomy in self driving cars within two years. That means in all driving conditions a human could handle. That seems exceedingly unlikely. https://electrek.co/2017/04/29/elon-musk-tesla-plan-level-5-full-autonomous-driving/ https://electrek.co/2017/04/29/elon-musk-tesla-plan-level-5-...
- drzaiusapelord 9y agoMusk has addressed this: Similarly, the proposal briefly discusses thermal expansion: as the steel of the tubes heats in the hot California sun, the metal expands. That expansion needs somewhere to go. In high-speed railways, rails are allowed to overlap at the ends, but that’s not possible in the Hyperloop, and so Musk has a different solution: “Specially designed slip joints at stations will be able to take any tube length variance due to thermal expansion,” he explained. “This is an ideal location for the thermal expansion joints as the speed is much lower nearby the stations. It thus allows the tube to be smooth and welded along the high speed gliding middle section.” https://www.theguardian.com/technology/2016/may/12/hyperloop-or-over-hyped-latest-demo-does-little-to-ease-doubts https://www.theguardian.com/technology/2016/may/12/hyperloop... I think the hand-wringing over expansion is a bit over the top right now. Seems like its a relatively solved problem and one the Hyperloop team is taking seriously. I imagine the cost of engineering and building giant slip joints is just part of the overall cost package and probably a non-trivial part. I find it hard to believe someone as relatively trustworthy and technical as Musk is selling this concept knowing full well its impossible. Most likely, this is a solvable problem the same way many difficult problems were solved for cars, planes, and rails during their inception. I read an analysis somewhere that the Wright brothers solved 4 or 5 'hard' problems with their first plane. Their competitors at the time weren't able to solve even one. I'm not saying Hyperloop is guaranteed to work, but declaring it 100% impossible seems overly pessimistic. My worry is a bit more prosaic, if we gain progressive leadership in congress, we may be looking at Euro-style high speed rail in many US regions, which would invalidate the hyperloop concept. This seems less likely, imo, but by far the saner move.
- mrguyorama 9y agoDoes he not realize that you can't just take up all the slack at the ends? That would mean the pylons nearest the station would be pushed over around 300 feet. How would the hyperloop stay elevated without pylons?
- daurnimator 9y agoWhy would the pylons need to move? You can essentially have tube sitting on top of the pylons (which I expect could have ~50 degrees of contact area). You could even have it sitting on rollers between the pylons and the tube.
- cjsuk 9y agoIs there any material available with a negative expansion coefficient that can be used to offset this? I ask because that’s how you tend to work around thermal effects in electronics.
- lisper 9y agoAn electronic device contains many orders of magnitude less material than a hyperloop track. There are all kinds of materials that are economically viable when measured in micrograms. Hyperloop track material has to be bought by the kiloton.
- clavalle 9y agoCompressable joints between each section?
- indescions_2017 9y agoWe could replace steel with a low coefficient metal matrix composite without sacrificing stiffness or long range strength. It would blow out the cost structure however ;) Have also heard anecdotally that condensation in the tube could be a real problem. Even at near vacuum, the rapid pressure differentials can cause buildups in front of and behind the speeding vessel.
- lisper 9y ago> It would blow out the cost structure however Sure, all of these problems are solvable if you are willing to expend arbitrary resources. But the whole point of this exercise is to provide an economically viable mode of transportation. (Isn't it?)
- DiThi 9y agoThis point is clearly addressed in the paper: The ends of the tube can slide 300 meters next to the station, and the tube is not firmly attached to the pylons but can move to accommodate the expansion.
- lisper 9y agoNo. The paper gets this badly wrong. It looks only at the incremental thermal expansion between pylons, and neglects the fact that in a sealed tube the thermal expansion will be cumulative. "These would absorb the small length changes between pylons due to thermal changes, as well as long form subtle height changes." [emphasis added] The cumulative expansion is dealt with in a single sentence: "A telescoping tube, similar to the boxy ones used to access airplanes at airports would be needed at the end stations to address the cumulative length change of the tube." But that's not enough. The entire track near the ends is moving by this amount. That means that the ends of the track are advancing and retreating over multiple pylons (unless you can figure out a way to cantilever the track over 300 meters). This is a completely unsolved problem.
- inetknght 9y agosealed telescoping tube between pylons then?
- losteric 9y agoI'm not this kind of engineer, but couldn't the pylon problem be addressed by a sliding rail mechanism between the pylon and the sealed tube? Or replicating the station mechanism at each pylon? Maybe I just don't see the problem you're pointing out. edit: in retrospect, I can clearly see the problem with expanding shaped tracks. What happens to a long S-shaped track? Think about expanding each segment of the top curve... the turn radius increases and as does the length of the turn. How do you support a tube that can get longer, move outwards along the curve, and change angles at various points? S is just an example, it seems to be a problem with any number of turns (including 1)... I suppose one option is to simply put it underground to stabilize ambient heat input, and use heat extractors to manage heat generated by the train. If the system shuts down, there might be some thermal contraction but that's easier to manage (it's OK for track segments to separate when trains aren't running - just warm them up before operation)
- Gustomaximus 9y agoI dont know if this an elephant in the room. I'm guessing engineers have thought about this and just not discussed it publicly. They have some pretty smart guys. Also to nitpick, from "the ends of a viable hyperloop track will have to move hundreds of meters" - this doesn't have to be true. There can be absorption points along the track. Overlaps. Or maybe they tunnel the entire way where its cooler. Or wrap the tunnel in something cooling and reflective. We dont have train tracks moving hundreds of meters do we? I dont profess to know the solution, but do know there are solutions somewhere.
- lisper 9y ago> There can be absorption points along the track Sure, but these have to hold a vacuum. No one has figured out a way to do that.
- nkoren 9y agoI am assured by those in a position to know that thermal expansion joints which hold a near-vacuum are an entirely solved and essentially trivial problem. I do have some concerns about whether they're a trivial problem within this context. An engineering solution which works in small numbers for PhDs in a lab isn't necessarily scalable to mass low-cost manufacturing, deployed in the field and irregularly serviced by workers of uncertain provenance. But evidently there is a way to do this, at least.
- dx034 9y agoAnd, more importantly, can be certified for real world use. Considering the speeds used, scrutiny will be extremely high. Even if you use it for freight at first, the track will be close to settlements and any accident could end up in a disaster. The behaviour of trains on tracks is very well understood (learning from accidents), the behaviour of a vacuum train tube is not.
- nkoren 9y agoThis is a very fair point. There might be components that are routinely used in non-life-safety-critical applications, but this is no guarantee that they will pass certification scrutiny when they are in the critical path for thousands of lives every day. I worked on an autonomous vehicle project here in the UK about 10 years ago (the Heathrow Pod). The certification process was absolutely intense. And this was for a system that was capped at 25mph; had we gone any faster than the regulatory burden would have become very significantly more onerous (below 25mph it wasn't necessary to test vehicles and infrastructure to destruction; above 25mph, it is). In this domain, there's a huge difference between a cool engineering testbed, a private pilot project, and running actual service for the public. If Hyperloop is going to be doomed by anything, failure to appreciate this fact is probably at the top of the list.
- frahs 9y agoThis is only if the tube sections all move together. This could be handled at the joint between each tube section, and then the joints would only have to move a fraction of that distance.
- goodcanadian 9y agoMany replies mention expansion joints. There are also other techniques that can be used: for example pre-stressing as is done with continuous weld rails: https://en.wikipedia.org/wiki/Rail_stressing https://en.wikipedia.org/wiki/Rail_stressing Overall, it is not a particularly difficult engineering problem.