That’s mostly right. A certain amount of expansion around a curve could be handled by the pillar supports–the expansion would increase the radius of the curve, and push it outward. A certain amount of this is acceptable, but there’s just too much in the design to handle this way. So it also calls for telescoping tubes at the end points–but you need something like a thousand feet! If the middle of the path warms up, all that extra tube needs to be pushed out to one of the endpoints, possibly hundreds of miles away.
If the path were perfectly straight, and you had truly frictionless bearings, that just might work. We don’t have either of those two things, though. At this scale, it’s like trying to push a cooked spaghetti noodle through a series of Cheerios.
So local expansion joints are the answer. I think they’re doable, but they aren’t trivial, and will definitely increase the cost. It might require some small redesign of the cars, too.
Maybe the expansion joints could be at the vacuum pumping stations… Incidentally the PDF lists $10 million as the total cost of the vacuum pumps for the entire system both tubes.
I don’t know, honestly. In principle, a $100 pump would be sufficient if the whole thing were perfectly sealed… but it would take a while to pump down (I’m ignoring outgassing, but I don’t think that’s relevant at these pressures).
If I had a bit of time I could write a simulator that would take the total CFM capacity of the pumps and compute the total opening area they could handle. You could then use that to estimate how good your welds and such need to be (say, <1 mm^2 of leak area per segment). That calculation is a bit of work, though, and I have no real starting intuition for it.
What difference will grade and elevation differences make to these tubes? Can the propulsion system handle, say, either a 5% rise or descent over many miles? What will that do to the required air pressures within each tube?
For the sake of argument, say that the California high speed rail will get built before a Hyperloop system can be demonstrated with a viable prototype. That is, take the LA to SF corridor out of the current discussion. Is it likely that an LA to Las Vegas system would be feasible? That’s a route often discussed for fancy transportation systems, but nothing ever seems to come of it. The LA to LV system will face more topography challenges than the LA to SF system, and at 265 miles is shorter than the optimum route length proposed by Musk, but it might provide a decent sample system that would flush out a lot of yet-to-be-found operating problems.
If the Hyperloop system were really to be built next to an interstate or state highway there’d need to be at least some state government cooperation required - another great can of worms.
Even if there’s some master permit granted to use state right-of-way for the support pylons and system tubes themselves, there’s going to be land needed for whatever stations, maintenance yards and access roads are needed. That right-of-way acquisition is going to take time and money. Further, if you think you can build a station in one city without the adjacent cities having something to say about it, think again. So - how long will it take and how much will it cost to overcome the legal and political challenges that are sure to come?
Last question: where can I get a job helping to build this thing? It sounds COOL.
It would basically have to be a bellows-type joint as mentioned upthread. The tubes themselves would probably need an interlocking pattern (tangential and radial) but allow axial play. I’m thinking like a square-wave sort of pattern. If the machining of the tube ends can be mass-produced on a custom milling machine and the bellows joint mass produced, then these joints can be glued straight to the tube on-site. This whole operation would have to be done by a giant machine that is at least two column-spans long and rolls from column to column like a tank, grabbing the pipes from below, raising them, placing them on the bearings atop the columns, slipping the bellows on, gluing it to one tube, raising and placing the next tube, gluing the second part of the bellows, then moving on the the next section along the track and repeating this whole process.
The install could be done in a matter of months assuming that steel mills churn out enough plate to be welded into tubes.
The proposal is to have boosting stations spread out throughout the tube.
I think that the elevation change would cause a negligible change in pressure since it is already at 0.001 atm (the equivalent of being at 48km altitude).
As far as state government involvement, I think it’s pretty safe to bet that, since they are the ones who will be financing the construction of this, they would be involved. It would be this way no matter which state or country you build this. And I believe that the right-of-way issues are already covered since it would be built alongside I-5. Sure there will be need to use eminent domain in certain sections, but that is to be expected, yet at a much lesser extent than a traditional high-speed train.
One thing is for sure: Andersen’s Pea Soup will be deserted in the future.
The examples you made are only “feasible” if infinite money were available, and material resources mean nothing.
Moving highways. Meaning every section of pavement is mounted on-top of some kind of tractor so you can rearrange the sections? or it’s all a series of conveyer belts? Either way, that would increase the cost by a factor of about 100.
Rockets full of mail? That would cost $10,000 a kg with current technology.
Space elevator has a critical weakness in that a failure at any point along the cable, and you lose the entire investment. It’s a bad technological idea, even if we some day developed strong enough materials. There are other ways to get to orbit that would not really be any more expensive per trip.
The point is, the 3 examples you gave have problems that are extreme with them, to the point that they are de facto impossible with current technology.
There are hypothetical technologies that would make all 3 practical, but we aren’t there yet.
I think a better example that would make your point would be to say that there is no technical obstacle that would prevent us from powering everything with nuclear reactors. The cost barriers are mostly caused by excessive government regulation and lack of economy of scale. Safety is an issue, but we could in principle install all the reactors in places where if they melt down and leak, few are affected.
Care to sketch up this concept in MS paint? The mental image I have is that in the middle of each concrete pillar there is a junction ring between 2 steel pipes that has accordian folds and is only a couple inches long. This means the skis won’t have to jump a gap of more than an inch or 2. The ring is made of spring steel, seam welded to the 2 pipes like everything else, and is there to keep the air out - the structural load is born by the concrete pillar underneath.
Frankly, this sounds like a relative inexpensive solution to the problem, one that won’t affect the cost estimates much.
What WILL affect estimates are big things like need for more access roads, fenced facilities, etc among some of the most expensive real estate on the planet.
Or, again, I didn’t see an entry in the .pdf for R&D costs to develop the tube cars or build the factory to make them. They might only cost a million each in materials and labor to build, but what about fixed costs? Those things are comparable to a jet aircraft.
Here is what I had in mind. Not to scale. Ideally, the accordion section would be pre-stressed to mostly counteract the air pressure, but that’s not strictly necessary.
The pipe would be supported on both sides of the accordion by the pylon. The center portion would be somewhat floating.
This hyperloop seem like a nice idea. But I think the cost is severely underestimated. 15 tonne capsule is estimated at $1.35M. Looking at pressurization and the big compressor, it is essentially an aerospace grade tech. $13.5M seems more likely.
The cost of the line also seems very low (about $7.7M/km). The savings are supposed to come from using pylons. Yet if this was feasible, couldn’t you put a regular rail line on pylons too? Historically, $100M/km was a good estimate for rail projects, over and above the cost of land.
Please note that my reasoning for saying it won’t happen don’t lie in the technology. We can build it if we want to. We just don’t want to.
I have a trip scheduled from my city to San Jose next month. I’ll get up at 5 am for the flight. Stop in Denver to change planes, then on to San Jose. I’ll land at 11am local time and spend an hour getting out of the airport and driving to my destination (noon local time; 2pm home time). Total flight time is about 4.5 hours. But my total trip time is 9 hours. I’m paying $450 for this trip.
If I took the tube, it would cut my flying time by what? 1/2 maybe. So now my ‘flight time’ is 2 hours, but my entire trip time is still 7 hours. How much am I willing to pay extra to save two hours of my trip?
If planes cost $450 for the trip…how much are YOU willing to pay to reduce your trip from 9 hours to 7 hours? Seriously…think about it. How much will YOU PERSONALLY be willing to pay to make a trip like that where you save 2 hours out of 9. I think I’d pay, maybe and extra $50. But I’m sure as heck not going to pay $1000 when I can take a plane for less than half that.
THat is the main problem. Its not the technology, its not the materials. Its your personal willingness to use the tube. Its economics.
THis is where it will fail. Even the biggest supporters on this thread won’t even use it.
I agree that there are huge amounts of technical issues to work out; far more than the hand-waving design seems to think. And even then, I don’t see the economics working out (7.3 million passengers/year assumes first of all zero downtime, and second full cars every moment of every day and night – 2:24 AM, 2:26 AM, 2:28 AM, etc. all exactly 28 people ready to go each time. )
But I’m not sure that thermal expansion is the biggest technical challenge. Engineers seem to have already worked it out successfully for long-distance gas and oil pipelines. And I think oil pipelines may be a more difficult case, as unlike the hyperloop, they face pressures greater than one atmosphere (going outwards rather than inwards, but I don’t think that’s a huge difference).
The only reason I’m optimistic about the space elevator is that the carbon nanotubes needed would have a host of other applications, and thus are likely to be developed with or without the elevator. The first things built out of it will likely be sporting equipment and military hardware, then it’ll gradually start showing up in more accessible consumer projects, and in civil engineering like bridges. Eventually, once it’s a mainstream material, then we can start talking about the many additional challenges in using it to build a space elevator.
Actually, the difference may be quite small. The pylons are supposed to be 30 meters apart. So each pylon needs to carry one capsule (15t) and 30 meters of the tube (38 t) for a total of 53 tonnes. With a high speed train, each pylon would carry the rails (3 t) and 30 meters long section of the train (about 68 t) for a total of 71 tonnes.
I also checked that only $100k is budgeted per pylon. This is a severe underestimate. Just the concrete required will exceed this.
