7 ms·

There's one thing that keeps bothering me about all the Hyperloop competitions and startups: everyone seems to be using maglev. This does not match up with the

by NickM 10y ago

There's one thing that keeps bothering me about all the Hyperloop competitions and startups: everyone seems to be using maglev. This does not match up with the original Hyperloop proposal, which used an air compressor to suck up oncoming air and blow it out the bottom, creating a fluid bearing. Maglev trains have been proposed before; there's nothing particular new or novel about the idea. One of the bigger problems with maglev is that the tracks are very expensive. One of the exciting things about the Hyperloop concept was that the tracks should be much cheaper, since all you need is a steel tube. Is it really right to call it a Hyperloop if it eschews the original concept in favor of building a maglev train in an evacuated tube? Feels like they're taking a totally different idea and just using the Hyperloop name to generate hype.

jandrese 10y ago
I'm not sure how a gigantic many mile long pressure vessel is supposed to be cheaper than maglev rail... In my opinion the Hyperloop is an insane idea and these companies are suspect. It seems more likely that they're the 2010s version of the Moller Aircar, something to keep "almost there" for decades while you milk investor money.
- continuations 10y ago
  The ultimate investor milking machine would be "fusion-powered flying car moving inside a hyperloop." It's almost there and it'll be truly revolutionary!
  - xkcd-sucks 10y ago
    "fusion-powered self-driving flying car moving inside a hyperloop."
    - Udik 10y ago
      This sounds very xkcd.
nickff 10y ago
One of Maglev's biggest problems is that it requires a great deal of power to keep at high speed because of atmospheric drag; Hyperloop's evacuated tube mitigates this issue. Hyperloop is also faster than most Maglev concepts, and having the tube helps to guide the vehicle (without requiring complicated curve construction).
- SteveNuts 10y ago
  How much energy does it cost to evacuate many miles of tubing?
  - snuxoll 10y ago
    Depends on how much atmosphere is leaked back into the tunnel and how much vacuum you need. The former is a much bigger issue with the concept, if it takes a lot of energy to evacuate the tube once it's not too big of an issue, but ongoing maintenance of the vacuum would be problematic with leaky or damaged seals over the lifetime of the tunnel. Of course, most of the vacuum tube proposals require a vacuum of something like 0.001 atmospheres, which is incredibly problematic as well....
  - nickff 10y ago
    Quite a bit, but the power does not have to be stored aboard the vehicle, or transmitted to it while the vehicle is travelling at hundreds of kilometers per hour.
- throw12032121 10y ago
  Evacuated underground maglev has been proposed at least for 30 years now. I've first read about it in a popular magazine. Speed figures up to 10000 km/h were thrown around (assuming near perfect vacuum).
pstuart 10y ago
Stupid question: why evacuate the tube? The ends of the cars could have props that provide thrust and "lift". I know it's not as efficient as gliding through a vacuum but it would be a much simpler solution.
- zymhan 10y ago
  Because the drag from normal air pressure would waste a significant amount of energy.
  - pstuart 10y ago
    I recognize there's a penalty there, but it also translates into a zero cost for creating/maintaining a vacuum, and the thrust could be used for levitation as well. It just seems like if it was done this way it would be so much simpler and therefore cheaper to build.
    - Someone 10y ago
      If you do that, you can also leave out the tube to make it even simpler and cheaper. My guess would be that it would be more energy efficient, too, as it gives the air more room to 'escape' from the front of the train. Only problem is that it would turn the idea into 'just another maglev train'.
schiffern 10y ago
Good question! Many teams are using induction maglev. Regular maglev technology required either permanent magnetic tracks or expensive electronically controlled tracks with embedded electromagnets, but induction maglev requires only a conductive plate on the floor (the hyperloop test track uses aluminum)[1] and a "hover module" which induces eddy currents in the floor and repels off them.[2] Based on their spec sheet the current generation of technology consumes 66 watts per kg of lifted mass, and since it's a constant power draw the faster you go the less energy it uses per mile. [1] https://www.badgerloop.com/documents/TubeSpecs.pdf https://www.badgerloop.com/documents/TubeSpecs.pdf [2] http://arxpax.com/ http://arxpax.com/
mulcahey 10y ago
I'm on one of the hyperloop teams (UCSB) that's using maglev. We aren't using the type of maglev that's used in maglev trains where the track is magnetized and super expensive. We're using ArxPax hover engines[1]. They just spin a Halbach array[2] super fast and generate a hovering force off of an aluminum track. The winning team of Design Weekend last January, MIT, is also using magnetic levitation but they are doing a passive version. The track still isn't magnetized, but their magnets generate their eddy currents just by the movement of the pod. It sort of "takes off." The first reason we -- and others in the competition, though not the majority I believe -- went with maglev is that the we (well, the Mechanical Engineering students, I'm CS) couldn't find air-skis that worked at high speeds. Elon mentioned at Code Conference that they'd been tested up to Mach 1.1 but when I told the ME's of that they said they hadn't found anything they could use nor any research showing that the air being pushed out of the ski "keeps up" with the pod at high speeds. In addition to the speed problem with air skis, they also hover much lower off the ground than the maglev and require much higher tolerances on the track (more expensive). Part of the reason to have the competition is to try a bunch of different ideas at once and see what works best. I do believe there are plenty of teams that will be testing air-skis on the track at competition weekend. We'll see how it goes. [1] http://arxpax.com/product/hover-engine/ http://arxpax.com/product/hover-engine/ [2] https://en.wikipedia.org/wiki/Halbach_array https://en.wikipedia.org/wiki/Halbach_array
- zwieback 10y ago
  In your opinion, how complex are the civil engineering challenges, e.g. keeping the tube straight?
  - jcoffland 10y ago
    42
  - mulcahey 10y ago
    I'm honestly not properly educated to answer that. My responsibly on the team has been controls & networking. I don't think the tube necessarily has to be perfectly straight as the pod could bank on the turns like a racecar or roller coaster. I think the tube itself could be made reasonably cheaply compared to the CA high speed rail system. We transfer oil in huge pipelines with high pressure differences so I think it should be pretty feasible to keep it properly sealed. I'm more worried about the cost of the safety mechanisms and them being able to secure the land rights. Elon mentioned at design weekend that initial versions could even use wheels if they aren't going over ~250MPH.
Animats 10y ago
The original "flying" hyperloop plan has a very low flying height - 0.5 to 1.3 mm, wrote Musk in his original paper. This requires a very smooth tube. Welded joints would need inside polishing. Expansion joints would be complicated. Tube sections can't have any sag, which probably means a stiffening beam will be required. All this pushes up the cost of the tube/track. The whole point of the "hyperloop" was that the tube/track would be cheap, on a par with pipelines. There are also safety concerns with such a tight clearance. The Transrapid maglev, the system used in Shanghai, has about 150mm of ground clearance, and about 10mm rail clearance.[1] The Hyperloop competition (not Hyperloop One) had an aluminum rail down the track center, and the winning team (MIT) levitated magnetically against that at 15mm height. Maglev against an aluminum rail is energy-intensive, but they were only making a trip of a thousand feet or so, and could do it on batteries without using superconducting magnets. Maglev in vacuum is an old idea. That's called "Vactrain". There have been many proposals, none built. [1] http://www.transrapid.de/pdf/tri_engl.pdf http://www.transrapid.de/pdf/tri_engl.pdf
semi-extrinsic 10y ago
> the Hyperloop concept was that the tracks should be much cheaper, since all you need is a steel tube [Technical nitpick first: tube is a structural element, while pipe is something that can hold pressure.] The price of pressure-rated steel pipe is roughly proportional to the diameter squared. At 132 inches in diameter, the hyperloop pipe itself would cost at least $600/ft. It would also weigh > 1000 lbs/ft, so the support system would be absolutely massive and probably double that price. That's just the cost of raw materials. For comparison, Elon Musk's total cost estimate is $1400 per foot per pipe [$6 billion total cost, 380 miles, two pipes], so I'm inclined to say that his estimate might be a bit optimistic. From what I can tell, total cost per foot for maglev is estimated around $6000/ft. [US highway construction cost per lane per foot is also around that number.] But I understand those numbers as the total cost including also engineering work, shipping/logistics, actual building work, buying the land etc. While my $1200/ft estimate is just raw materials. So if we as a very coarse estimate say raw materials is 1/4 of the price, hyperloop is not significantly cheaper than maglev.
- schiffern 10y ago
  Small correction: you used the diameter of the large Hyperloop design, but the cost estimate of the small one. It should be $7.5b or $1870/tube foot. Fyi there is a cost breakdown on pp55 of the Hyperloop PDF, which lists the tube construction costs as half what you specify. What kind of "pressure tubes" are you referring to for these cost estimates? http://www.spacex.com/sites/spacex/files/hyperloop_alpha-20130812.pdf http://www.spacex.com/sites/spacex/files/hyperloop_alpha-201...
  - semi-extrinsic 10y ago
    I was estimating based on interpolating between carbon steel and 316L stainless, ASME Schedule 40 pipe. You can have that up to 144 inches.

by NickM 10y ago

jandrese 10y ago

I'm not sure how a gigantic many mile long pressure vessel is supposed to be cheaper than maglev rail... In my opinion the Hyperloop is an insane idea and these companies are suspect. It seems more likely that they're the 2010s version of the Moller Aircar, something to keep "almost there" for decades while you milk investor money.

continuations 10y ago
The ultimate investor milking machine would be "fusion-powered flying car moving inside a hyperloop." It's almost there and it'll be truly revolutionary!
- xkcd-sucks 10y ago
  "fusion-powered self-driving flying car moving inside a hyperloop."
  - Udik 10y ago
    This sounds very xkcd.

nickff 10y ago

One of Maglev's biggest problems is that it requires a great deal of power to keep at high speed because of atmospheric drag; Hyperloop's evacuated tube mitigates this issue. Hyperloop is also faster than most Maglev concepts, and having the tube helps to guide the vehicle (without requiring complicated curve construction).

SteveNuts 10y ago
How much energy does it cost to evacuate many miles of tubing?
- snuxoll 10y ago
  Depends on how much atmosphere is leaked back into the tunnel and how much vacuum you need. The former is a much bigger issue with the concept, if it takes a lot of energy to evacuate the tube once it's not too big of an issue, but ongoing maintenance of the vacuum would be problematic with leaky or damaged seals over the lifetime of the tunnel. Of course, most of the vacuum tube proposals require a vacuum of something like 0.001 atmospheres, which is incredibly problematic as well....
- nickff 10y ago
  Quite a bit, but the power does not have to be stored aboard the vehicle, or transmitted to it while the vehicle is travelling at hundreds of kilometers per hour.
throw12032121 10y ago
Evacuated underground maglev has been proposed at least for 30 years now. I've first read about it in a popular magazine. Speed figures up to 10000 km/h were thrown around (assuming near perfect vacuum).

pstuart 10y ago

Stupid question: why evacuate the tube? The ends of the cars could have props that provide thrust and "lift". I know it's not as efficient as gliding through a vacuum but it would be a much simpler solution.

zymhan 10y ago
Because the drag from normal air pressure would waste a significant amount of energy.
- pstuart 10y ago
  I recognize there's a penalty there, but it also translates into a zero cost for creating/maintaining a vacuum, and the thrust could be used for levitation as well. It just seems like if it was done this way it would be so much simpler and therefore cheaper to build.
  - Someone 10y ago
    If you do that, you can also leave out the tube to make it even simpler and cheaper. My guess would be that it would be more energy efficient, too, as it gives the air more room to 'escape' from the front of the train. Only problem is that it would turn the idea into 'just another maglev train'.

schiffern 10y ago

Good question! Many teams are using induction maglev. Regular maglev technology required either permanent magnetic tracks or expensive electronically controlled tracks with embedded electromagnets, but induction maglev requires only a conductive plate on the floor (the hyperloop test track uses aluminum)[1] and a "hover module" which induces eddy currents in the floor and repels off them.[2] Based on their spec sheet the current generation of technology consumes 66 watts per kg of lifted mass, and since it's a constant power draw the faster you go the less energy it uses per mile. [1] https://www.badgerloop.com/documents/TubeSpecs.pdf https://www.badgerloop.com/documents/TubeSpecs.pdf [2] http://arxpax.com/ http://arxpax.com/

mulcahey 10y ago

I'm on one of the hyperloop teams (UCSB) that's using maglev. We aren't using the type of maglev that's used in maglev trains where the track is magnetized and super expensive. We're using ArxPax hover engines[1]. They just spin a Halbach array[2] super fast and generate a hovering force off of an aluminum track. The winning team of Design Weekend last January, MIT, is also using magnetic levitation but they are doing a passive version. The track still isn't magnetized, but their magnets generate their eddy currents just by the movement of the pod. It sort of "takes off." The first reason we -- and others in the competition, though not the majority I believe -- went with maglev is that the we (well, the Mechanical Engineering students, I'm CS) couldn't find air-skis that worked at high speeds. Elon mentioned at Code Conference that they'd been tested up to Mach 1.1 but when I told the ME's of that they said they hadn't found anything they could use nor any research showing that the air being pushed out of the ski "keeps up" with the pod at high speeds. In addition to the speed problem with air skis, they also hover much lower off the ground than the maglev and require much higher tolerances on the track (more expensive). Part of the reason to have the competition is to try a bunch of different ideas at once and see what works best. I do believe there are plenty of teams that will be testing air-skis on the track at competition weekend. We'll see how it goes. [1] http://arxpax.com/product/hover-engine/ http://arxpax.com/product/hover-engine/ [2] https://en.wikipedia.org/wiki/Halbach_array https://en.wikipedia.org/wiki/Halbach_array

zwieback 10y ago
In your opinion, how complex are the civil engineering challenges, e.g. keeping the tube straight?
- jcoffland 10y ago
  42
- mulcahey 10y ago
  I'm honestly not properly educated to answer that. My responsibly on the team has been controls & networking. I don't think the tube necessarily has to be perfectly straight as the pod could bank on the turns like a racecar or roller coaster. I think the tube itself could be made reasonably cheaply compared to the CA high speed rail system. We transfer oil in huge pipelines with high pressure differences so I think it should be pretty feasible to keep it properly sealed. I'm more worried about the cost of the safety mechanisms and them being able to secure the land rights. Elon mentioned at design weekend that initial versions could even use wheels if they aren't going over ~250MPH.

Animats 10y ago

The original "flying" hyperloop plan has a very low flying height - 0.5 to 1.3 mm, wrote Musk in his original paper. This requires a very smooth tube. Welded joints would need inside polishing. Expansion joints would be complicated. Tube sections can't have any sag, which probably means a stiffening beam will be required. All this pushes up the cost of the tube/track. The whole point of the "hyperloop" was that the tube/track would be cheap, on a par with pipelines. There are also safety concerns with such a tight clearance. The Transrapid maglev, the system used in Shanghai, has about 150mm of ground clearance, and about 10mm rail clearance.[1] The Hyperloop competition (not Hyperloop One) had an aluminum rail down the track center, and the winning team (MIT) levitated magnetically against that at 15mm height. Maglev against an aluminum rail is energy-intensive, but they were only making a trip of a thousand feet or so, and could do it on batteries without using superconducting magnets. Maglev in vacuum is an old idea. That's called "Vactrain". There have been many proposals, none built. [1] http://www.transrapid.de/pdf/tri_engl.pdf http://www.transrapid.de/pdf/tri_engl.pdf

semi-extrinsic 10y ago

> the Hyperloop concept was that the tracks should be much cheaper, since all you need is a steel tube [Technical nitpick first: tube is a structural element, while pipe is something that can hold pressure.] The price of pressure-rated steel pipe is roughly proportional to the diameter squared. At 132 inches in diameter, the hyperloop pipe itself would cost at least $600/ft. It would also weigh > 1000 lbs/ft, so the support system would be absolutely massive and probably double that price. That's just the cost of raw materials. For comparison, Elon Musk's total cost estimate is $1400 per foot per pipe [$6 billion total cost, 380 miles, two pipes], so I'm inclined to say that his estimate might be a bit optimistic. From what I can tell, total cost per foot for maglev is estimated around $6000/ft. [US highway construction cost per lane per foot is also around that number.] But I understand those numbers as the total cost including also engineering work, shipping/logistics, actual building work, buying the land etc. While my $1200/ft estimate is just raw materials. So if we as a very coarse estimate say raw materials is 1/4 of the price, hyperloop is not significantly cheaper than maglev.

schiffern 10y ago
Small correction: you used the diameter of the large Hyperloop design, but the cost estimate of the small one. It should be $7.5b or $1870/tube foot. Fyi there is a cost breakdown on pp55 of the Hyperloop PDF, which lists the tube construction costs as half what you specify. What kind of "pressure tubes" are you referring to for these cost estimates? http://www.spacex.com/sites/spacex/files/hyperloop_alpha-20130812.pdf http://www.spacex.com/sites/spacex/files/hyperloop_alpha-201...
- semi-extrinsic 10y ago
  I was estimating based on interpolating between carbon steel and 316L stainless, ASME Schedule 40 pipe. You can have that up to 144 inches.