Hyperloop Alpha [pdf] | Hacker News Reader

Hyperloop Alpha [pdf]

spacex.com

513 pointssoupboy193 comments

Location is a primary reason why Musk can claim Hyperloop will be cheaper than high speed rail. As the paper correctly points out, right-of-way acquisition is very expensive. Bypassing major population centers is an easy way out of using eminent domain or expensive tunneling to build the system. The HSR project is required by Proposition 1A to pass through several population centers in the state and have a couple dozen stops. Notably, Hyperloop proposes to bypass all of the SF peninsula, San Jose, and Central Valley cities like Fresno and Bakersfield. HSR was politically viable because these regions were included in the project. Furthermore, the route and station map appears to hint that Hyperloop's stations are positioned away from city centers. For example, the "San Francisco" station appears to be located in the East Bay and the "Los Angeles" station is closer to Burbank. This is quite significant from an urban planning perspective as it means that short haul transit or car trips have to be accommodated for a majority of those people who want to use the system. It also adds to the total travel time needed to get from door to door for a greater number of people than does HSR.

I'm not saying that Hyperloop is an inherently flawed project or that HSR is the perfect solution to the state's long-haul transportation needs. I just think that Musk is oversimplifying the situation.

>>I just think that Musk is oversimplifying the situation.

Yes. After all, the project whitepaper has the word "alpha" in the title, and the authors specifically say they are open to constructive feedback. Have you thought about emailing them about your concerns? I think they are valid.

I'm fairly certain they've realized this. My guess is that "Proposed SF-LA tube costs billions less than HSR" makes for a better headline.

For example, the "San Francisco" station appears to be located in the East Bay and the "Los Angeles" station is closer to Burbank. This is quite significant from an urban planning perspective as it means that short haul transit or car trips have to be accommodated for a majority of those people who want to use the system. It also adds to the total travel time needed to get from door to door for a greater number of people than does HSR.

I'm not an urban planner but it seems to me that accommodating one station in each city's boondocks might actually be easier than accommodating transit to several stations in the most central areas. San Leandro has a lot more room to build than SF and a lot of people will be coming to the SF HSR station if it's built there (even with a San Jose station).

Burbank isn't even the boondocks. It's just part of the San Fernando Valley sprawl. You literally couldn't tell where it ends and Los Angeles begins without the signs.

Well, ok, and the nicer road surfaces.

Having said that, the map actually looks like it terminates at the northern tip of the valley, where there are current no mass transit rail lines, so you'd be looking at buses or cars from there.

Understandable as land acquisition costs skyrocket once it hits the valley. Still, it's a little misleading cost-wise for that reason.

That's correct. California waited almost 50 years to build a high-speed rail, and now the "that stuff only works in Europe" mentality is going to be painful to swallow. When I was a kid in the 1970's, I'd see shows with Californians bragging about how they loved their cars, and mass transit wasn't needed. It's really kind of embarrassing how Americans couldn't plan for the obvious.

Unfortunately, It's not going to get any cheaper, and only becomes more difficult as the state becomes more crowded. If "Governor Moon Beam" had build his trains in the 1970's like he tried, we'd now be planning for a 300mph maglev by 2030, just like Japan.

> For example, the "San Francisco" station appears to be located in the East Bay and the "Los Angeles" station is closer to Burbank.

They did this a lot in Japan as well for the Shinkansen. There are a lot of Shin<City-Name> stations built expressly for the Shinkansen because the city center was too crowded. Once the train line is in, a new city blossoms up in the activity around the station.

It's interesting that the biggest line item in the costing is not the tube, nor the capsules, nor the vacuum pumps, nor the solar panels, nor the land, but the concrete pylons that support the tube - about two-fifths of the total. The pylons are planned to be 30m apart, making for about 25000 of them.

But that's the beauty of it! You can make those pylons anywhere, truck them in, and install them in whatever order you want. It's construction....that scales up almost effortlessly.

I suppose that's true, though I don't know if that's a real difference compared to building other very long things that are the same all the way along, like roads and railways and the Great Wall of China.

It's just fascinating to me that the most expensive thing (in aggregate) is also the "dumbest" part.

The most expensive thing in aggregate is the part which surely received the most attention when cost cutting. I think it makes perfect sense it's the dumbest part.

Also much easier to maintain and replace

perhaps you can pour sections and truck those in, but there's no way in hell you're moving a precast that size on anything that doesn't float.

For scale, you know those 4' by 4' by 8' blocks of cement you see all over the place, used for erosion control and the like? They're called ecology blocks, they're about $20 each FOB the cement plant. You want them shipped? A lowboy can carry two or four at a time without exceeding load limits. When I looked into buying a couple hundred, having them shipped about sixteen miles, it was going to cost hundreds per block to truck it out plus renting a crane to unload/place them.

Those are tiny blocks with almost no rebar, no engineering, no preload, nothing.

Ecology blocks: http://www.tworiverssandandgravel.com/wp-content/uploads/201...

They are also colloquially known as 'mafia blocks' for some reason...

While I certainly can't say anything for the economy of long distance shipping of giant concrete pylons, I can confirm they do get trucked around a lot in the Dallas area. The overpass ribs are basically hooked to a small truck attached to the cab with a small truck on the other far end.

Again, dunno if it's feasible over long distance, but it's certainly done. (Though Texas has a certain affinity to overpasses that borders on the pathological, fun as they are to bike over...)

Pretty sure you can truck pylons if you can truck in precast bridge decks:

https://www.google.com/search?q=precast+bridge+deck

Concrete is about two tons per yard. The legal limit on most trailers is about 20 tons (40,000 pounds, the trailer is about 40,000 for the legal limit of 80,000 pounds). So you can get about 10 yards of concrete in one go; you can use specialty trailers to get higher numbers, but even if you double it to 40 tons, that's still only 20 yards.

a 4' x 4' x 8' ecology block is about five yards. So let's picture the pylon having a 4' x 4' base that sits on granite, requiring no foundation. The pylon is going to max out at 16 or 32 feet tall if we're required to stay within the load limits.

furthermore, when you see those ribs moving around, they're moving around on interstates or state roads of solid construction. If there was an interstate running directly from San Francisco to Los Angeles along this route, we wouldn't be having the discussion. Building such a road to deliver precast sections is quite the endeavor.

I stand by my original claim that these would need to be cast-in-place.

> a 4' x 4' x 8' ecology block is about five yards

That's just brilliant. (for the rest of the world 1,201,202.4m is 3.5m^3 and density of concrete is around 2400 kg/m^3)

If there was an interstate running directly from San Francisco to Los Angeles along this route...

Umm, that's kind of the point. It's called I-5.

I guess we can agree that we don't need to have this discussion, as I-5 runs the entire route where the pylons would need to be installed.

You'd still have to pour the foundation in place though, right? Or can you make foundations off-site and then drive them into the ground?

They don't seem to break down the labour costs of building the tube/pylons very well. In my experience watching big government projects, that's where the massive overages start coming in.

It's easy to say that something will cost (eg.) $1bn if it's built in 2 years, but when most of that is labour and bureaucracy starts getting involved, the labour gets put on hold for a year or two whilst still getting paid. Suddenly the overall cost of the project has tripled without anything actually changing, except some bureaucrat sticking their oar in.

Maybe I'm just cynical, but the labour costs seem to be one of the least well defined aspects of this proposal. It's easy enough to make it look cost effective compared to existing rail projects when you don't factor that in.

Afterall do you really think that the HSV with two steel rails is costing $80bn+ because of the price of the components? Although I don't really know the specifics, I find it hard to believe that the rights of way cost that much either. It's usually down to contracting an entire company worth of engineers for 10-20 years that ramps up the costs.

The focus from people on bathrooms is puzzling to me. I've been on buses, without facilities, that go for hours without stopping. This is a 35 minute ride with restricted movement, if you greatly fear being able to hold it for that long, there are other, less efficient methods of transport to accommodate you.

On a bus you have the chance to get off every 5-10 minutes, that's a big difference. There will be thousands of passengers on that day. Just imagine if you have to do this trip 10 times a day every day for 100 days, will you still take risk?

Update (since I've been downvoted):

All these buses analogy isn't exactly the same case because they still have the option to stop somewhere, which give the passengers the assurance that if they really can't hold it's not the end of the world. Taking that into consideration, I think it could solve most people's bathroom concern by adding bathroom stops to the loop, in the case of semi-emergency, ideally a pod should be able to make a stop upon request within 5-10 minutes without disrupting other pods. Some financial incentive might be needed for encouraging people to hold longer but I do see legal risks in that approach.

That depends where in the world you are. I've been on many long distance buses in Asia that might not stop for 4 or 5 hours.

I have taken a 40min commuter bus everyday for nearly 2 years and I never had to get off early to use the bathroom.

You're thinking city buses. I'm thinking long-haul busses (like greyhound in the US). Greyhound buses likely have facilities, what I'm actually talking about are the buses used a core parts of the transportation infrastructure all over South America. Many of these also have facilities but I've been on many where the bathroom isn't functioning or just plain isn't there. You can easily travel for 30 minutes to an hour without stopping.

We're largely (70%) talking about commuters here, they're taking this twice a day, most days out of the year. Again, I'm not sure why for 90%+ of people that this would be an issue.

Headline: Musk reveals Hyperloop. Americans too fat and incontinent to use.

All these buses analogy isn't exactly the same because they still have the option to stop somewhere, which give the passengers the assurance that if they can't hold it's not the end of the world. Taking that into consideration, I think it could solve most people's bathroom concern by adding bathroom stops to the loop, in the case of semi-emergency, ideally a pod should be able to make a stop upon request within 5-10 minutes without disrupting other pods. Some financial incentive might be needed for encouraging people to hold longer but I do see legal risks in that approach.

Financial incentive? You're traveling at 700mph through a partially evacuated, elevated tube in order to get to work. That isn't incentive enough?

Not really. Take the Staten Island express bus for example. Between SI and Manhattan, there is about 35 min on the Verrazano Bridge and Brooklyn-Queens Expressway, without any opportunity to get off (BQE does not even have shoulders for the most part). That's on a good day with decent traffic. In bad traffic and bad weather, that could be an hour or even more without any opportunity to get off.

EDIT: a clarification for non-New Yorkers -- people take these daily to get to work, if they work in Manhattan. The alternative is SI railroad-ferry-subway, which allows more bathroom safety but takes quite a bit longer.

Interestingly, my DC Metrorail trip to home or work takes between 45-60 minutes one-way... and there are no facilities on the train itself, and very few stations have any public-accessible facilities. "Holding it" has honestly never been an issue, you either go before you leave, or sack up and wait until you get home. Worst case is probably that you get sick en route and well... that alone wouldn't wreck Hyperloop.

Exactly, my experience with regional rail commuting in Philly was the same. Regional rails frequently have the sort of trip lengths that we are talking about here, and I have never seen a regional train that has facilities.

One of the most unpleasant (although memorable) experiences of my life was having to hold it for over an hour while sitting over the wheel well on a bumpy bus ride in the middle of Belize.

I'm not saying that it doesn't have the possibility of being unpleasant at times, I'm saying that it doesn't strike me as a serious objection to a brand new transit method that is superior to existing options in so many ways.

Yeah. Adult diapers! Problem solved.

"methods of transportation"

If not using the bathroom for an hour is an issue you should see a doctor. Not only are there good treatments to alleviate this, but it may be an indication of a serious medical problem. Get it fixed then enjoy life including perhaps the Hyperloop.

I agree with the sentiment, but its part of the disadvantages of this transport. You cant travel with babies or little kids in them, and you certainly cant do it sick.

However its not such a big flaw, specially to warrant 63billion to alleviate.

Nappies.

I would gleefully slap another $50-100 onto every plane ticket I ever bought to ensure that there was nobody anywhere on the plane under the age of 16. I can't be the only one...

Only 50-100? I would go as far as 200! :)

Well, I probably would too, but depending on the airline, that is nearing first class levels.

Of course on many/most planes "first class" does not get you isolation from children... (certainly not for that low a price) but at least you get unlimited liquor and legroom. I'd be willing to pay more for a flight without children but I would want some unlimited liquor once I go over 200 or so.

On this note, what would be the booze situation on hyperloops? There isn't an aisle on those things as far as I can tell, which means no stewardesses... Maybe per-seat minibars could be installed on them? Strictly speaking it wouldn't be necessary for 30 minute rides, but I see no reason this idea shouldn't be scaled up past 30 minute trips (Musk points out that supersonic airliners are better at longer distances, but since we don't have those either I'd settle for a 30+ minute hyperloop. Both a hyperloop and a return of Concord/Tu-144 would be great of course.)

Don't worry -- once the hyperloop is built and running and we still don't have a practical supersonic airliner, Musk will sigh and design one of those for us too.

IIRC, he's actually talked about electric airliners a few times, though I'm not sure if those ideas were for supersonic electric airliners. I suspect not, but at the least it seems to be an industry he has thought about.

I'm sure that kid will feel the same way about paying for your Social Security someday.

I certainly don't count on that scenario ever playing out... By all means though, those children should be able to pay $50-100 for the privilege to never pay into social security. That is a great idea.

What is your thesis here, I shouldn't be annoyed by children because one day I will be old? Not sure I understand that logic. "Being annoyed" isn't even a proper choice; if I could turn it off, I certainly would. In fact, I would say that the ability to turn if off for a few hours at a time is worth about $50-100 to me......

Yeah; actually you are the only one in control of your feelings. Turning it off is a matter of maturity; one day you may (god willing) smile when you hear children.

That kid will be paying for their OWN Social Security. Any given retiree has already paid for theirs. That's what the lifelong taxes are for. Remember?

There aren't bathrooms on public transportation either and yet many people still take subway and bus rides that are 35 minutes or longer.

There are conditions with no cure that make it difficult to go an hour without using the bathroom. I think the people with those conditions are well aware that it's a "serious medical problem" without your advice.

Your post was about as useful as telling someone in a wheelchair to see a doctor to get their legs fixed (hey, it may be an indication of a serious medical problem!)

It's a half hour. Not an hour. Furthermore, from a given location on an average freeway, what do you think the real-life time is, in heavy traffic, to get to a bathroom? Could be over a half hour, easily.

There's no new problem here.

No, he said an hour. See:

> "If not using the bathroom for an hour is an issue "

"An hour". It's right there in his post. I didn't say there was any new problem. I didn't say I objected in any way to the Hyperloop. I specifically addressed his post which said "an hour". OK?

I would point people towards the Stadium Pal, described by David Sedaris. Essentially, it is a urine bag strapped to your leg attached to an external catheter. I am not sure if it works for women, but I am sure Musk could figure that out.

Similar, though less invasive, solutions exist for 'emergencies' on private planes and such. They come with support for both male and female genitalia.

Well it's not just a matter of bathrooms.

Even on a commuter bus or plane, if someone has a medical emergency and needs to get off right now, the bus will pull over or the plane will land. In this design you'd have to just sit there, helpless, until the vehicle reaches its destination. Heck, even if you're a doctor in the next seat you wouldn't be able to get up and help your fellow passenger.

They discuss this in the paper. In the event of a medical emergency the pod will proceed to the destination as normal, and paramedics can be waiting at the end. In the absolute worst case you are looking at ~30 minutes before you can get to medical attention. That's not obviously worse than the time it would take a plane to descend from cruising altitude to the nearest airport, land, and taxi to a gate.

Except you can get first aid assistance in a plane (they ask if a doctor is here, and they have a number of drugs on board). Keeping the person alive until paramedics arrive is key to handling emergency situations, and I don't see how you do it with the Hyperloop.

Landing a plane will guaranteed take more than 35 minutes. With the hyperloop, it's simpler and more efficient to just have EMTs called and on station when the car arrives.

Actually, an airliner can usually descend and land within 10-15 minutes in an emergency, assuming a suitable runway is close by. There are more smaller airports with 5000 ft runways that don't typically handle 737s or larger.

The problem with emergencies at major airports is that between EMTs and fire services inspecting the jet, that runway is out of service for at least 30 minutes.

On a normal flight a 737 flies an approach at under 180kts but the approach limit is a lot faster. In a cardiac arrest type situation, the pilot may decide to fly a faster approach into an airport with medical services available. They would still touch down at the normal speed.

Yeah, I believe this is the best point in the argument against pod transport. Even taking doctors out of the equation, if someone were to go into cardiac arrest on a plane, pretty much anyone on the plane would be qualified to operate an automatic external defibrillator and all flight attendants are required to be trained in CPR. The opportunity for first aid in this situation is the difference between life and death and it is unavailable in this pod design. 3 minutes without oxygen to the brain and serious brain damage is likely; 15 minutes without oxygen and you're pretty much guaranteed never to recover if you aren't already dead. If you have a cardiac episode on the hyperloop and need to be resuscitated, there's a good chance that the only good that those EMTs will do when the train arrives at its destination is to remove your dead body from the train.

It's a very small space, with no apparent facility to handle not just #1 and #2, but severe gastric distress. The other passengers then get to look forward to up to 25 minutes in that closed space. There's no stewardess, no way to stop the vehicle. An "emergency evacuation module" shouldn't take much room.

The design incorporates very powerful air compressors and water pumps, and generates, transfers, and isolates a great deal of heat. The cabin will be very well-ventilated, because if it weren't then passengers would die of something other than methane poisoning.

If the U.S. doesn't take this on, I hope another country with a big appetite for enormous public transit projects takes this on. Japan, Korea, and China, I'm looking at you...

China seems set on a different path, but I guess you never know. Korea seems too small to benefit. And Japan perhaps too mountainous?

Looking at California: http://bit.ly/1eFevxA And Japan: http://bit.ly/19ZXwbJ I don't see any roads in Japan that are nearly as straight as I-5.

The places where this would make sense have destinations fairly far apart, but not too far apart, with somewhat amenable landscape between them. The Midwest would work, Australia, Russia. Probably Brazil, Argentina? Probably a bunch of connections in Africa or maybe the Mideast... but that wouldn't work for other reasons. East coast seems a bit cramped.

In Australia this could potentially be viable between Melbourne and Sydney, but probably only if it doubles as a cargo train. I doubt there is enough daily people traffic to warrant it otherwise?

It is around 900km though, almost a straight line along the Hume highway and the busiest city-pair in Australia by a long way.

>I doubt there is enough daily people traffic to warrant it otherwise?

Sydney<->Melbourne is the third busiest air route in the world with 8 million passengers in 2012[1]. I'm sure there's more than enough "potential" passengers.

[1] http://en.wikipedia.org/wiki/World's_busiest_passenger_air_r...

The public in East Asian countries see public transit in a different light than most North Americans, partly due to a lack of cars and the population density. They are probably the some of the few governments that can afford to foot the bill for such an unproven project with potential for dramatic benefits (and cost savings over building HSR), but also with the risk that the 10 billion or so could go down the toilet. Europe would be another ideal target, but I doubt any governments there have the appetite for big capex infrastructure projects right now.

> Russia

Yes, 630+ km Hyperloop between Moscow and Saint-Petersburg would be a great project. Problems:

1. Winter. Would solar panels give enough energy during winter?

2. Highly-inflated cost due to corruption. I would speculate corruption is a big problem for California HSR too, given its cost.

3. "Russian Railways" company monopoly for all intercity railways in the country.

The Republic of Texas would be a good fit. But Southwest Airlines is located here so it would never get through the legislature.

IMHO Musk (or anyone who follows up) should drop the "passenger" version and go with the "passenger plus vehicle" version only.

Because right now what people and the press are going to focus on is how small the cabins are, why there are no toilets, etc. On the other hand if you can imagine putting a car in there the claustrophobia argument disappears and you can think straight again...

It would seem to make the stations a lot more complicated – they'd need to handle the car loading, the accompanying traffic, and the security would be really challenging.

Can we not be thinking about adding the inefficiency of the TSA to such a nice system? The security theatre needs to end.

I disagree. The "passenger plus vehicle" option is definitely more complicated, and in the first version of any project you need to keep things as simple as possible. I mean, the system in question is as complex as it is. I think the "passenger plus vehicle" version is included only to make the case for the additional possible uses of the system.

Disagree completely. I'd ditch the vehicle option. If the earliest forecast is for this to be built in seven years, how much will the driving landscape change in that time? Robo-cars even? This is primarily for commuters who can walk/taxi/rent at the destination. Even travellers will do the same.

If you want to take your car, drive it IMO. When you fly, there is no expectation of taking your car along with you and this should be similar - focus on what we will need rather than on what a minority could fancy.

I agree to your point about taking your actual car. Loading cars inside would indeed complicate the system a lot.

What I tried to say was more about the psychological impact of imagining a car fit in the capsule, even if the same space is just taken by larger seats.

Interesting quote:

    The intent of this document has been to create a new open source form of 
    transportation that could revolutionize travel. The authors welcome feedback 
    and will incorporate it into future revisions of the Hyperloop project, 
    following other open source models such as Linux.

I would really like to see people starting to build small scale models from this design. It could become the new version of building a rocket in your backyard.

Strange PDF title -- "Falcon 9v1.1 - Reentry Nozzle Study" -- is there some design reuse or just document reuse?

I bet they titled it like that to hide it from prying eyes. That way even if you found the account where it was being stored, you would have no idea it related to hyperloop; and there were probably people looking for this sort of document.

A funny mistake most likely, probably document reuse for basic layout. Download it again. It was just updated from "Nicholas Liggett" and "Falcon 9v1.1 - Reentry Nozzle Study" to "Elon Musk" and "Hyperloop Alpha"

I worry that the necessary straightness of the tube will be difficult to achieve. The pads float 0.5 to 1.3 mm from the inside of the tube, so deviations from straightness on that scale will cause the pads to crash. Any contact at 760 mph would probably damage the pad, plugging the air holes.

If someone shot the tube with a medium sized gun, would the impact cause an internal defect on the order of 0.5mm?

The tubes for the passenger-only version would be 20 to 23 mm in thickness. I'm not sure what damage a medium-sized gun would do to steel that thick.

According to some quick youtube searching, armor-piercing rounds versus regular steel would create problems: http://youtu.be/q_zZd0NsYG4 .

Sure, so would high explosives directly on the tube. I am considering a situation that could conceivably be accidental because it is incredibly difficult to engineer something that can stand up to malice.

The bar to malice is important to consider. If the bar is "high explosives or shaped charges", then you are probably okay. If the bar is "a redneck who is bored with shooting stop-signs", then that might be a problem.

Bored people don't shoot guns at trains today. Why would they start?

Because there's no possibility that shooting at a train would result in a fireball a thousand yards high, whereas...

How thick are propane tanker cars?

Back of the bar napkin calculation with Barlow's formula, assuming 200psi (about what bbq tanks are at iirc?), somewhere around half an inch.

Because the tube isn't a train?

Are you sure? I imagine most trains these days are pretty tolerant of that, it could easily go unnoticed.

Bored people shoot track, put pennies on tracks, drop water balloons from freeway overpasses, put potatoes in car exhaust pipes, and on and on. Its important to consider the most random variable in the safety equation - bored people.

I'd say all these concerns warrant putting the damn thing underground.

Isn't there some kind of inverse square law affecting the force between the walls and capsule? The closer you get to contact, the harder it pushes back?

It should be largely self-correcting.

He doesn't properly address thermal expansion of the entire system. A long straight steel tube from SF to LA will expand a total of 300 meters over a 40C temperature change. Where does it go?

If I read it right, the movement is absorbed by shock absorbers that need to be installed for earthquake protection anyways. You just need to make sure that the segment between the lateral movement/shock absorbers are sized such that the differential thermal expansion doesn't violate the deflection requirements.

The paper describes having a telescoping section on each end to accommodate thermal expansion, but the previous commenter is stating that it is likely this telescoping method will not be sufficient, as the variation in length of the structure will cause significant bending of the pylons themselves. Between the ends, there is no mechanism to allow for thermal expansion, which will either cause massive axial stress on the steel tube, bending of the pylons, or both. The shock absobers will not have sufficient travel to address this, (100s of meters); axial bearings on the tube may be required.

Each 30M long tube expands 15mm over 40C of temperature change. You could imagine a 15mm expansion joint at each connection, which is roughly what oil/gas pipelines do, but the paper explicitly says tubes will be welded together for straightness.

I don't know how to do the math, but couldn't it be absorbed by the horizontal play at each plyon? Imagine the tube "zig-zagging" a tiny bit at each column.

To absorb the 15mm expansion over each 30M long tube, you'd need to zig-zag nearly a meter. At full speed you're passing through 10 tubes per second. It would literally shake your head off.

"By building a system on pylons, _where_the_tube_is_not_rigidly_fixed_at_any_point" [...] "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"

So, the tube is lying loose on the pylons, and he assumes it will keep its shape while expanding.

I think that is highly optimistic for a tube that is a straight line, even more so for one that is curved.

If he manages to accomplish that, the end stations will move over 300 meters, in total. Hopefully, when the tube contracts again, the motion will be reversed perfectly. Otherwise, that end station in LA could slowly move north or south over the years.

>If he manages to accomplish that, the end stations will move over 300 meters, in total.

Where are you getting this idea from? That seems like an insane implementation compared to using accordian-like flexible section[s].

From what I cited: "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"

I guess that's because accordion-like flexible sections and carriages moving at 1000 km/h on a tiny air cushion don't go well together, or because such sections, being a lot leakier then welds, would let too much air in.

Right, so the tube changes length. The entire station doesn't move. Also, the trains aren't going 1000 km/h at the station.

How does rail address this?

The gaps between each section of rail can be quite large.

Prior to the use of continuous welded rail there were gaps between each 39 foot section of rail.

Nowadays with welded rail, careful compliance with construction standards is required. Slow orders are issued during heat waves and frequent track inspections are conducted to look for potential or actual sun kinks.

A good summary of the issue can be found in the Heat Orders FAQ from Virginia Railway Express. [1]

[1] http://www.vre.org/feedback/frequently_asked_questions/faq_h...

Does it have to be steel? Any other materials we can be considering? Some interesting new forms of carbon...?

If you want it building for a reasonable cost then it's steel or nothing.

Why is everyone so negative?

Musk just posted a innovative, open-source, thorough, alpha design study which addresses a real problem.

Personally I think it's great, and I hope it gets taken seriously by the right people.

It's interesting that the Shinkansen has raised sections built on pylons, especially the new bits they built in the last few years in the built-up areas in the north of Japan.

As does the Chinese system and for the same reasons Musk addresses.

When I was a kid I had a similar idea, a low pressure elevated concrete tube transport system. In my idea cars would get rails underneath them and freely enter/exit the tube at some key points. The cars would get a compressor to help keep the not-so-large pressure differential at bay; all inspired by evacuated tube documentaries.

In other words, this has been proposed several times before. However, there are some key innovations here - namely the compressor/nozzle pair and the air suspension system - which may indeed change the game and make this viable.

This could be considered a "next generation train". Elon Musk as we know likes to consider things from "first principles". It looks like here, he has taken the first principles of high-speed "rail" travel on land, and come up with this.

I think our children might enjoy using it ;-)

Looks like a good idea, but the major glaring technical point that seems to be missing is the issue of horizontal stability within the tube.

When you get very low friction, high speed and no tracks, you're going to start having attitude problems without the right control surfaces available. Don't want to be spinning passengers around the roof of that tunnel!

The train could be just very bottom-heavy.

Perhaps on a straight, but in the bends or slopes you'd still need some sort of control. I'm visualising the kind of oscillations that luge or bobsled riders get in the winter olympics when they come out of the corners.

High speed and low friction could mean a lot of potential lateral movement that needs to be controlled somehow. Maybe releasing some of the stored compressed gas in controlled bursts like used in various space endeavours. That would have the added benefit of off-loading some of the potentially dangerous steam.

I've never operated a bobsled, but if they're like other machines I've operated with physical controls that have mass or that move in response to things that have mass, then the oscillations are caused by the driver. Which makes sense because the better bobsled drivers are observed not to suffer from these oscillations as often as their competitors. (In fact we might say they are "dampers" rather than "drivers".)

In the absence of any controls, we would expect the train to climb the side of the tube on a corner, without ever reaching a "horizontal" orientation. When exiting a corner, perhaps there would be a pendulum situation, but I'm thinking that a train that wasn't round in cross-section would damp that pretty quickly. If not, the system generates compressed air. That can be released through nozzles that are oriented and controlled to dampen the rolling.

But you're right, of course. A bottom-heavy train is just a more effective pendulum.

Nice font. But looks good so far.

EDIT: There's a bit of minimally technical details, but this is neither highly polished (public-perception-focused) nor highly technical (engineer-focused). Is it a good balance or would one of the extremes have been better?

It says right at the beginning that the first section is an overview while the second section is more technical...

So? It needed a section with more polish than the first section for not HN, and a more detailed technical section.

Both would have been ideal - Ideally the webpage would have been a public-perception-focused with simple explanations (syringes with compressors!) and the concept sketches front-and-center, with a link for a more technically oriented whitepaper. That would have taken a lot more time though - the linked pdf is still impressive.

I think the occasional mention of safety on high-speed rail is a bit unfair, considering that good systems like the Japanese Shinkansen have never had a fatal accident:

"Risk of derailment is also not to be taken lightly, as demonstrated by several recent fatal train accidents."

There were two fatal high speed accidents just this summer in Europe-- the Spanish one killed 80 people.

Cool concept, but a 500 mile bridge? Elevated transportation routes are expensive [1,2,3]. The utopian dreams tend to give way to an ugly reality [4,5].

The $1 billion budget for land and rights of way also seems wildly optimistic for razing a path through multiple heavily populated metropolitan areas with astronomical real estate prices. And let's not underestimate the tenacity and ingenuity of a determined opposition [6,7].

[1] The 3.2 mile, mostly elevated [1a] BART/OAK connector with dinky little cable cars will cost $361mm for the capital construction alone[1b].

[1a] http://bart.gov/docs/oac/OAC%20Project%20Alignment%20Map%20S...

[1b] (slide 36) http://bart.gov/docs/oac/Microsoft%20PowerPoint%20-%20OAC%20...

[2] http://www.nationalaffairs.com/doclib/20080528_197905505cost...

[3] http://www-pam.usc.edu/volume2/v2i1a3s2.html

[4] http://www.youtube.com/watch?feature=player_detailpage&v=DkZ...

[5] http://en.wikipedia.org/wiki/Embarcadero_Freeway

[6] http://en.wikipedia.org/wiki/Highway_revolts#California

[7] http://dc.streetsblog.org/2012/06/25/indianas-big-dig-raises... (this one is hilarious)

In Singapore all short haul trains (called MRT) are either underground or on Pylons. Seems to work out well enough there.

Yes (I lived in Singapore for 2.5 years) but they go quite slow. Top speed of the most recent, unmanned MRT (yellow line) is about 65 Km/h.

Yes. I was just talking about pylons being viable.

http://en.wikipedia.org/wiki/Overseas_Railroad

In 1905, they built 128 miles of railroad connecting the Florida keys. Since then, engineering has advanced significantly. Overland is still easier than overseas.

It's too bad the tubes and trains can't be made transparent. Imagine the view!

Use the technology from Google Streetview to composite the views from cheap cameras mounted at 1/24 second intervals at the top to the tube. You use a cluster of these for a 360 view.

What will probably happen is that you get a transparent car and advertisements on the inside of the tube.

That fly past a metre from your face at 1,000 km/h!

http://www.youtube.com/watch?v=CUia8HnNZrM&t=10s

Fair solution! Might be cheaper to just cover the interior with advertising though?

"Have you seen the 200 foot long billboards in the country beyond town? Did you know that once billboards were only twenty feet long? But cars started rushing by so quickly they had to stretch the advertising out to make it last." -- Ray Bradbury, Fahrenheit 451

On the Beijing subway there are adverts on LED displays on the tunnel walls that move along with the train, keeping pace outside the windows.

Cows on I-5. Greatest benefit is it's an enclosed tube, odors not allowed.

They should fly a helicopter down the California coast with a 4k camera pointed out towards the ocean and then mount 30" LCD screens next to each seat and play back the video at high speed, with image stabilization, correlated with the position of the train.

Each passenger gets an Oculus Rift headset and can choose their virtual route.

This is supposed to be for commuters? I worry not many would commute daily from LA to SF.

Also, travel time to the station has to be taken into account, even more so if we compare it to driving.

Yes they don't because its not feasible _yet_. If your residence/office is within 30 mins from the stations at points of origin/exit, then with this, it'll be about 1.5 hours end-to-end (I'm sure auxiliary services would spring up to support this like shuttle buses etc). I know people who change buses, trains and do some walking, chalking a total of that much right now everyday while working and staying in the same city.

I wonder what they have in mind for how stations will offload/onload capsules onto the loop. That will be heck of an engineering problem to solve - ultra high safety, pressurized environment, fast turnaround times...

My initial thought was they would have switches on the track like a normal train and capsules would veer off. But I bet they would come to a stop on there own and then get lifted vertically off the train and into a decompression area where they'd cue up.

This is addressed in the document. Essentially, each station is a terminus, so airlocks are transited at low speed. A turntable reorients trains for their next trip. The only switching required takes place entirely within the evacuated zone. (This is for when spurs are built to serve e.g. San Diego or Las Vegas.)

Claustrophobes will not be eager to travel in this fashion.

Travelling in any kind of tunnel is no more claustrophobic than this is it?

"The maximum width is 4.43 ft (1.35 m) and maximum height is 6.11 ft (1.10 m)."

Yeah, that's a little bit more claustrophobic than your average underground train.

But less claustrophobic than sitting in a car.

Cars have windows and stuff.

The passenger compartment mockups indicate there won't be enough room to sit upright. I'm sure the seats will make this comfortable for people of average stature who aren't afflicted with claustrophobia, but there will be much less room than tunnels that accommodate e.g. upright humans or subway trains.

Strap on an Oculus, and simulate a train car. Or, project windows onto the interiors. (Since we're talking future technology)

Yes I expect every visible interior surface would have display capability. A popular mode would be the view outside the tube, which could be constructed from fixed external cameras, and wouldn't have to be precisely "live" anyway. Some masochists might prefer the view inside the tube, which could be prerecorded or even animated.

Why worry initially about a disorder that only affects 5–7% of people. Fix the transportation problem for a large majority of people first, anyone else still has existing transportation as an option.

Comfort is proportional to room, there are plenty more than 7% of the population that would refuse to ride in a windowless box they can't even sit upright in.

Remember there are two passenger compartment mockups, this is the smaller one we are referring to: http://i.space.com/images/i/000/031/716/i02/hyperloop-passen...

Unscientifically I'd bet less than 25% of people would be willing to ride in that for more than 10 minutes, much less 30-40.

Exactly. I think this fails because of comfort requirements. If you stacked people inside airplanes like this, I'm sure they would be 10x more efficient and competitive with this proposal.

I think this would be great for journeys of _under_ 30 minutes. And it wouldn't be for everyone, for sure -- so starting off as a niche transportation mode is not a great start.

Elon mentions in the paper that even on planes today landing procedures take over 30 mins, and during that time people can't get out of their seats, and everybody is ok with it. For longer routes you would definitely need more space, but then he also describes the model that would allow for carrying cars, which is a lot more space.

Then use http://www.bbc.com/future/story/20130812-planes-that-can-pic.... Compared to that, this Hyperloop idea is best described as "No pilot. Less space than a 747", by teenage boys.

On the other, more important hand, at firt view, this looks feasible to me (taking into account who published it)

I do wonder about the "little noise" claim, though. That fan a few meters in front of you that compresses air so that you can float on it will not be noise-free. I also think those tubes could start resonating at times.

Finally: only about a thousand passengers per hour or, alternatively, going at close to the speed of sound, on guaranteed collision course, with 30s intervals?

Oh the irony if you try to access the bbc link http://www.bbc.com/future/story/20130812-planes-that-can-pic... from the UK :

«BBC Future (international version) We're sorry but this site is not accessible from the UK as it is part of our international service and is not funded by the licence fee. It is run commercially by BBC Worldwide, a wholly-owned subsidiary of the BBC, the profits made from it go back to BBC programme-makers to help fund great new BBC programmes. »

For those in the same boat: the BBC article talks of a design where maglev trains, starting in city centers, accelerate to a few hundreds km/h and then get picked up by a huge flying wing that swoops on it like a bird of prey. Later, much farther away, that flying wing will put the trains back on a maglev track, without ever landing.

That truly is a toy for boys (as in "looks cool and dangerous") unlike the Hyperloop ("it may or may not be dangerous, but you can't see that, so who cares?")

The technical challenges posed by that design seem quite significant compared to those of TFA.

Cool stuff here, but I'd rather see them focus their efforts on moving cargo rather than people.

Find a way to remove the cargo (trucks) from the commuter transportation system (roads & highways) and you'll improve the lives of many more people.

Moving cargo is a lot easier than moving people because it doesn't complain or need to go to the bathroom. It also won't be missed by most people if it's destroyed.

Could someone buy land in the potential path of this hyperloop then block access or sell the rights or land to make a profit?

No, because of the principle of eminent domain, meaning the government of California can forcefully purchase the land for public use: http://en.wikipedia.org/wiki/Eminent_domain

Legally, whoever tried this would be forced to give over the land for "fair market value", and if they were difficult about it they'd probably receive much less.

I was trying to figure out where I'd seen this before.

http://www.thecomicking.net/2012/06/25/the-core/

The (terrible) movie "The Core" has a vehicle -- the Virgil.

Anyone has the resources from this pdf and willing to commit to github? Lets start forking!

I sent a request to their feedback email seeing if they would be open to something of that sort. Copy pasting formatting to Markdown/ReST wouldn't be too bad, but it seems to me polite to ask permission.

If this system will evolve, how switches will work between different routes? Will it be only from station to station?

Eeeh, that looks awfully cramped. Especially for anyone a little overweight.

No more so than flying coach. It actually looks more like flying business class on BA.

So what happened to section 5? Is that redacted by the NSA?

How would you rescue when pod breaks down mid tube?

The pods have something similar to landing gear for use at speeds less than 100 mph. Still, what happens to pods that are sandwiched between two malfunctioning/mangled pods? I think they should have contingencies where sections of tubes may be excised safely.

Isn't the pod going to be extremely loud?

Honestly, this paper should be made as a mandatory reading in schools. Honoring modern science and economics.

What's the ideal format for an open source whitepaper? Certainly not PDF. Straight up HTML on Github?

latex.

What disadvantages do you perceive for PDF?

The same ones that we perceive for "open source development" on compiled binaries.

With that said, it's not hard in this case to C&P to something more suitable for GitHub and really get things kicked off, but LaTeX would be more suitable.

Although it was effectively the case before, since 2008 PDF is an open format. It's quite misleading to compare it to a binary for which source is unavailable. Since the format specifies compression, not every part of a PDF file may be easily observed in a text editor, but there's a bunch of open software that can read (and edit, and produce) PDF.

Producing and reading PDF isn't the issue.

Again, we have open source tools that will generate, mangle, introspect, disassemble, etc. to both ELF and PE binaries.

What PDF is not is the "preferred format for development", no matter how good the open-source tooling is.

Sometimes practical considerations preclude using the most "preferred format". Will editing this white paper really be a significant obstacle to deploying a system like the one it describes? Conversely, does anyone ever release a white paper in LaTeX only?

What are you on about?

No one here is saying that an open-source project to assist with Hyperloop could not be launched. Nor that the ability to edit the whitepaper that was released has anything to do with this proposed open-source project. Nor even that people typically release whitepapers only in LaTeX format.

However if an open source project will be launched with Hyperloop then its success will be at least somewhat dependent on the ability of participants to meaningfully contribute.

If "contribute" just means discussing key points then a simple Wiki might suffice. However if "contribute" means technical designs, then the actual drawings and data will be what's needed, not glorified screenshots of AutoCAD and MATLAB.

mpyne you're hilarious.

I converted the PDF to a rough markdown: https://github.com/pooyak/hyperloop/blob/master/document/hyp... Needs more clean up, but still easier to work on.

How does one maintain oxygen ratios?

So where is the 'loop'. Seems like this should be called 'hypertube'.

the "loop" is likely at the end of each line so the pods can turn around.

did you read the paper?

Not every word, I have a day job. If you have something specific to contribute please share.

The proposal is to use turntables. Since emphasis is made in the paper on maximizing turning radii, it could be that the train is not intended to be able to turn sharply enough to negotiate a feasible "loop".

loop /lo͞op/ Noun A shape produced by a curve that bends around and crosses itself.

Reply on news.ycombinator.com