Hacker News new | past | comments | ask | show | jobs | submit login

This is very serious first proposal.

As a professional mechanical engineer I have worked on high quality steel tubes (nuclear submarines) before. The immediate thing that sticks out to me in this proposal are the tight mechanical tolerances that have to be maintained. Talking about tens of thousandths of an inch tolerances on a 10' diameter tube is not to be dismissed lightly. That is going to be tough to maintain - especially with welding heat distortion. I would image the tubes will be joined with automated friction stir welding or something similiar, but that will still require a fair amount of post weld machining which has its own pitfalls. Not to mention simple thermal expansion and contraction as the temperature changes could change the circularity and inner diameter.

I would be more interested to see a tolerance stack up of those considerations than an FEA model of the concrete pylons. I can gaurentee that we can build concrete pylons capable of holding up a steel tube, that is done all over the country dozens of different uses cases. But can we build a multi-hundred mile long steel tube to the required tolerances?

I would be inclined to trade off efficiency for manufacturability. I.e. maybe a higher internal pressure or larger diameter to make it less sensitive. There should be plenty of power from the solar panels so it doesn't have to be perfectly efficient.

I'm also surprised that the I-5 plan is cheaper than buying private land. I may be naive here, but the pylons really do take away most of the objections from farmers and installing tubes over farmland has to be a lot cheaper than doing construction above a highway. I just look at boondoggle that was the SkyTrain in NYC (tram running over a highway out to JFK airport) and wonder if that is a great option.




> I would image the tubes will be joined with automated friction stir welding or something similiar, but that will still require a fair amount of post weld machining which has its own pitfalls

The document mentions standard orbital seam wlelding, plus specialized machining equipment that travels along the tube to smooth out the gliding surface.

The capsules are only 60% the diameter of the tube, or 36% it's area (68/47% for the vehicle-carrying version), it doesn't seem to require tight tolerances for operation. I got the idea that tube distortion and movement is taken into account into the system.


> it doesn't seem to require tight tolerances for operation

Well, it's true that the top and sides of the tube don't get too close to the capsule, so those parts seem relatively low-tolerance, as you say.

But the load-bearing skis ride on an air bearing of 0.5mm to 1.3mm (see page 20), moving at over 1000km/h. As rossjudson notes, the skis are on mechanical suspension, to smooth out shocks to the riders, but it's not clear (to ignorant me) how much of a bump those skis can glide over. 0.1mm, no problem. What about 1.0mm?

On a related note, Musk seems sanguine about the sag you'd get in any structure supported by pylons (see e.g. page 27). Even with inch-thick steel walls, with pylons an average of 30m apart (100'), you'll see some sag, right? Any engineers want to comment on the deviation in 30m of inch-diameter-wall steel tubing? Let's see, 1200kph, 30m, so you pass a pylon 11 times per second. So in 0.09 seconds, you have to go from the top of one pylon, to the valley between two pylons, and back to the top of the next pylon. I guess that's all absorbed by the mechanical suspension?


You could just build the tubes bending upwards slightly, so that the sag will pull them straight.


Bridges and overpasses are already build that way.


No you could not. The steel tube will expand and contract as temperatures change. Along most of the route, that longitudinal movement of the tube will be in the 10s and 100s of feet, which means that the "valley" could be anywhere on the tube according to temperature.


Also, tubes will resonate and vibrate as vehicles go by. This could be actively damped at the pylons though, like they already considered doing for ground subsidence.


Offtopic, but please consider writing some blog posts or articles about your experiences working on nuclear submarines. That just sounds so cool, and while the work is probably dull, I'd nonetheless be fascinated to read some detailed stories.


Unfortunately I walk a very fine line about what is sensitive and what is not. I really couldn't go any deeper than what I posted here.


I will say as a former nuclear submariner myself, how appreciative we all were of the work the designers put into the boats. I was on an Ohio-class myself and got to appreciate the design first-hand. I've also heard simply marvelous tales about the Seawolf-class (pity she was so expensive).

So thanks for whatever you did in keeping the boats safe (both in design and construction), it was certainly much-appreciated in the Fleet.


Well, is it cool?


It's definitely cool. I left about a year ago to join a startup, but that had more to do with wanting to live in NYC again than the submarine company. Two interesting books: http://www.amazon.com/Dark-Waters-Insiders-Account-Undercove... and http://www.amazon.com/Blind-Mans-Bluff-Submarine-Espionage/d...


As a fellow mech eng, what startup environment are you involved in and is it utilising your mech skillset or something else, eg: programming abilities?


My background is BS Mechanical Engineering, minor CS, primarily focused on robotics. Grew up programming and building robots. Worked at Bug Labs which bridged the gap pretty well, then went to General Dynamics for 100% mechanical engineering, now at getlua.com for 100% software. There are some mechanical startups around NYC (MakerBot is the obvious one) but I really liked the team and the product at Lua and I'm quite happy here.


OK, then more like this please! ;)


If I read the proposal correctly, the system is less sensitive than the tolerances you describe. There is a tight tolerance to maintain between the "air skid" and the wall of the tube, but the air skid is mounted on a suspension system that is probably intended to deal with small variations away from the ideal tube shape. Those variations will manifest to passengers as "bumps" and will also result in transient additional drag, I suspect.

It would be cool if you could take another look and see if the tolerances are truly what you describe, or if there's actually more room for variance.


The tube is considerably larger than the vehicle, so really it's only the smoothness of the bottom part of the tube that is important, not the precise dimensions. And the surface the vehicle flies over doesn't need to be steel - you could line the bottom half of the tube with a wax, or something similar, and that would be much easier to smooth to the tolerances needed, and re-smooth if it suffers any damage.

In fact you might be able to handle some thermal expansion (at least seasonal, not daily) that way - release some clamps on a telescopic section, adjust, reclamp, melt and re-smooth the wax.


Back of the envelope, SI units.

13 × 10^-6 thermal expansion coefficient of steel

45 - -40 max temp difference

570k length of tube in meters

line1 × line2 × line3 = 629m (~1890') of thermal expansion over a 570 km tube.

Did I read it right that they will take out all that expansion at the ends? To get that down to say 30m between pylons you have to have about 30 expansion joints. Can they go full speed through an expansion joint?


They are taking it out at the ends. Assume a fixed point in the middle, and you have some 1000' of movement at either end. That should be feasible, but the tube needs to be designed for very significant longitudinal movement through the pylons, and the design needs to take into account the possible lateral loads on pylons from having to bend the tube away from the geometry it has at the baseline temperature.


Thanks for writing! I can also imagine that when there's one very long tube there's going to be a lot of the length change with the temperature, ideally one wouldn't want to have any hard joints at all to allow the individual pieces to extend and shrink. One hundreds of miles long welded tube does sound potentially problematic.


The biggest problem seems to be that Musk's costs don't include most of the costs actually included in High Speed Rail proposal that allow you to actually travel between SF and LA and not SF minus bay (they omitted that cost) and Sylmar (where you've still got over an hour to go on the Antelope Valley Metrolink Line to reach the LA city):

http://stopandmove.blogspot.fr/2013/08/hyperloop-proposal-ba...

It seems that the proposal is actully much less serious than on the first glance:

Amusingly enough, the California HSR budget for the Central Valley is under $10 billion. Ie, in the same ball-park as this proposal. The reason the HSR project is going to cost $60 billion is because it has to face an uncomfortable truth; actually getting to LA and SF is expensive. Very expensive.

Moreover, would you like to be sitting in the tube on the chair from which you can't even stand up for a half an hour without the chance to do anything but remain sitting, not getting any help the next half hour on the occasion when you get sick? Would you take your kid there? Your parent?


You can't stand up in a car either. In any case the ride is going to get you to qualified medical personnel within half an hour :)


You're in control of your car though so if you need to you can stop and get out to deal with any bodily functions or sickness. You can't do that in this train. There's a lot to be said for the effect of just having the option even if you don't exercise it. Control is very comforting to people.


Not only comforting, the seemingly "rare" cases actually happen all the time and the advantages of being able to stop, go out, help the person etc. are effectively used all the time.

Designing mass transportation while ignoring such aspects is like programming without handling the limit cases "it works for N between 2 and 100 but not for 0, 1 and MAX_INT, even if these are allowed inputs." It's bad, very bad for the user, only an "astronaut developer" can love such a solution.

The first time a kid dies because it got sick the first minute of a 30-minute ride, the project is dead for good. Imagine the press, imagine the public response.

I can imagine entering the capsule to reach the international space station, preparing a whole year before. I can't imagine doing the same preparations for the ride between SF and LA. Give me the real train, thank you.


The Channel Tunnel between the UK and France works fine, and if someone got sick at the start of the tunnel section then there isn't too much that could be done quickly.

The public seems to have no issue with that concept, it is a simple risk that goes with getting on the train.

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


Yes, that's why we built that airport and hospital on the North Pole. Now intercontinental flights can make emergency landings in under 30 minutes at all times.


With multiple stations between SF and LA, this could be mitigated with an emergency system that pulls the train off at the next stop. Similar to how some buses or trains have emergency stop systems, except instead of stopping, the capsule would divert to the next station.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: