German students win hyperloop competition, reaching 200 mph | Hacker News Reader

German students win hyperloop competition, reaching 200 mph

theverge.com

148 pointsrmason97 comments

Important to note, the WARR team doesn't receive any funding or direct support from the university - not even professors or Ph.D's.

This is a 100% student created and run initiative

Makes winning so much sweeter: no need to concede bullshit "acknowledgements" to absent profs and grant agencies and supervisors who sit the entire project out, only to inject themselves into the final credits and collect the post-victory interviews. Standard academic practice.

Great job, students!

Why wouldn't a grant agency get acknowledgements if they provided funding?

If you look at the winning pod, it is covered with logos of sponsors as acknowledgement.

It's one thing to have a line in a paper saying 'this research was funded in part by the NSF' and another thing entirely to have people talking to the media about said work.

It's not all that different from accelerators partially funding startups and associating themselves with the successes as long,as they want. I think it's arguable earned.

Are you sure? TUM is listed next to Airbus under "Main Sponsors".

I'm sure. They are officially incorporated under the umbrella WARR organization which also has the rocketry and satellite student projects of the TUM, but doesn't receive any funding either. They basically have financial compliance oversight and that's it.

Is WARR the cart in the image? Why does it have all those logos?

If that is the WARR vehicle, the GP doesn't mention not getting sponsors, just that the University didn't directly support it.

Though, I wonder where they built this and worked on it. My guess would be that they had lab space provided by the school.

There is lab space provided by university (we find it very weird that you call it school).

The main point is that the WARR teams (Hyperloop, space elevator, deep space concepts, rockets and satellites [I'm in the sat group]) are student teams. Some Ph.D students are involved, too, but the point is that we decide what we do. The teamleader is a student. The ideas are from the students.

The University, in the form of the chairs (Astronautics and Propulsion) are partners and friends, but we are in charge of what we do.

WARR is actually older than the Universities Astronautics chair.

(btw, high to all WARRians on hackernews)

In America, "school" is a generic term for any kind of educational institution

Well, but we Americans also say "where did you go to school?" meaning, in most adult contexts, what other English speakers mean when they say "where did you attend university?"

I agree it's a weird synecdoche, although, English is full of them if you look. https://en.wikipedia.org/wiki/Synecdoche#Examples

To be fair, America seems to change up a lot of conventional English, haha :P

Greetings from your friends at TUfast Driverless :)

Lab space is provided by TUM makerspace, which is a separate commercial venture, affiliated with but organizationally separate from the university.

I posted to distinguish the TUM team from for example Swissloop, who receive direct funding and R&D support from ETH.

Those are their sponsors. So I don't know if "no funding" is right. Also they directly list TUM (TU Munich) as sponsor.

http://hyperloop.warr.de/#sec_sponsor

I don't think that it's supposed to be read "no funding in general and no direct support from the university", but instead "the university offered them no funding or direct support"

The students went out and got their own sponsorship from external bodies. The university provides no money.

The logos are those of sponsors, not of the university ;)

Based on the number of logos on the vehicle. I'm sure they have sponsors.

The WARR team went out and got sponsorship on their own, not through the university.

It should be noted the California bullet train is planned to run at 220mph for its fastest leg, so this is within spitting distance already.

If you pump the (vast majority of the) air out, no reason you couldn't get supersonic. That's much faster than any non-rocket-propelled train has ever gone.

The fastest train in service is around 270mph, more experimental maglevs have reached 375mph. Supersonic is 767mph, so over twice the fastest maglev.

> If you pump the (vast majority of the) air out, no reason you couldn't get supersonic.

Doesn't this sound like a maintenance nightmare? I'm not clear on a couple of things. What happens when the vacuum fails? What happens to passengers if the seal fails on the vehicle, and then all the air is sucked into the vacuum? Maintaining long tunnels is by itself hard, but it seems like maintaining long tunnels that also need to maintain a vacuum would be that much more complex.

I don't see much discussion on this aspect of the hyperloop.

> What happens to passengers if the seal fails on the vehicle, and then all the air is sucked into the vacuum?

I fly a lot for work. Every week, someone gives me a demonstration of what to do when this happens.

Unfortunately, that won't work. Oxygen masks only work up to a certain altitude, which is roughly the altitude where airliners cruise. At pressures lower than that, even pure oxygen isn't enough to sustain you.

You don't die instantaneously when exposed to vacuum, although it is pretty quick. I wonder if it would be possible to repressurize the tube (by blowing a big hole in it?) within 15 seconds or so.

The trouble with blowing a hole in the tube somewhere to re-pressurise is that whatever is inside the hole is hit by a wall of air travelling somewhere around the speed of sound. There are a LOT of problems with hyperloop not limited to someone walking up to the tube and shooting it with a .50 Desert Eagle, earthquakes, and the expansion and contraction of the tube as the sun rises and sets.

It's definitely the most unlikely of Elon Musk's plans.

The hyperlink won't be a vacuum. It's just a low air pressure chamber. The plan has never been to maintain a vacuum since that's too difficult at such large scales.

Hyperloop is planned to run at roughly 0.1% of sea level air pressure. For the purposes of a discussion about human survival in the event of pressurization failure, that's effectively a vacuum.

We already run vehicles in 0.14 atm environments. The issue of running these vehicles in an environment hostile to human life has been mitigated pretty well.

In regards to there being more than expected air in the tunnel - the vehicle would be slower, but not much else right?

Less air than expected inside the vehicle could be life-threatening for breathers. For humans we could solve this with an emergency oxygen hose at every seat like in planes. Pets and other oxygen dependant lifeforms wouldn't be protected by this system though. An alternative could be to have an electronic hatch at regular distances in the tunnel. So when a vehicle detects a massive oxygen leak inside the cabin, it could signal the nearby tunnel hatch to open. The whole tunnel, and thus also the vehicle [1] would fill with air pretty fast.

[1] The vehicle would have a large hole in order for oxygen levels to suddenly go from life-sustaining to not. Thus that same hole would serve to refill the vehicle with oxygen coming from the tunnel hatch. However there could be an additional hatch for the vehicle to guarantee airflow. This could maybe even mount directly to a tunnel hatch, so that outside air would come to the vehicle and skip the tunnel.

Have you considered, that, in light of the fact that Elon wants to go to Mars, that research in having vehicles perform in a near-vacuum is a possible side-benefit for that goal?

My understanding is that labeling it a vacuum tube is a bit of a misnomer. It should be called a low-pressure environment or near-vacuum. Achieving that much reduction in air pressure will greatly reduce drag and allow for increased speeds. It doesn't need to be in a vacuum.

The maintainance of a vacuum tube this long and the consequences of a failure are a main theme in most hyperloop "debunking" videos on YouTube.

I'm guessing that any leak is likely to be small, and thus slow. And depending on how many vehicles are 'in flight' at once, and how easy it is to detect, they could be slowed down before hitting atmospheric pressure to avoid shocks or vehicle damage.

> The fastest train in service is around 270mph [...]

French TGV has broken the 350mph limit in non-commercial run.

Could "normal" wheeled train could go this fast reliably in a vacuum chamber? I suspect heat in the wheels and bearings would cause problems, especially without air cooling, but it would take much less energy to drive the train.

China seems like the place to build the first hyperloop. It would cost a lot less and they’ve already embraced HSR.

They’re increasing their trains to 248 mph.

https://www.nbcnews.com/mach/science/china-reclaim-title-wor...

Long routes like Beijing/Shanghai would greatly benefit from a hyperloop. They transport almost 1.5 billion people yearly on 12,000 miles of HSR, and they’re building another 10,000 miles.

That’s exactly why China is doing that, and why China in the past years bought from SIEMENS the entire IP for TransRapid (the only high speed maglev system that actually shipped in production, and is used daily by many travellers)

I agree -- in fact I'd wager that California is likely to be a laggard in this race. Highway expansion is very slow, causing clogged freeways, and local commutes in populated areas like LA and San Fran are among the worst in the nation. BART is embarrassingly old and dilapidated. The new Bay Bridge was way over schedule and budget (and with serious quality issues)...Given the track record with "traditional" infrastructure, I just don't see California being a leader in this area. Maybe some of the tech will be conceived here, but implementation would be different.

Except this is the whole reason Musk developed the Hyperloop concept. He was annoyed at sitting through California traffic.

Ironically, it is California's crappy traditional infrastructure that birthed Hyperloop.

The idea is to build the first one sooner rather than later. Then everyone who wants it can build it.

Sort of like Bill Gates working with China to get the next generation of nuclear energy developed in China.

https://qz.com/627113/bill-gates-says-china-is-the-best-plac...

Sadly it seems that (broadly speaking) bribery is the only thing holding human achievement back, across our planet.

If you want to go very fast you can basically only go in a straight line and you need a long time for accelerating and braking. G-Forces on passengers are fairly limited, typically to around 1m/s^2.

Much more than that (~4m/s^2) is experienced nominally by jet liner passengers.

Roller coasters do 50-60m/s^2.

I don't think this is a major problem. Just accelerate slower.

Source? From what I know 6g is the realm of anti-g suit, it seems quite strange to me that it is deemed normal in an attraction used also by children.

Google's search result blurb says:

> Tower of Terror holds the record for most G-force on any roller coaster in the world, an intense 6.3G. This is nearly twice the amount that the average astronaut experiences when making their way through the atmosphere.

So world record for a roller coaster was ~60 m/s2, and that was presumably for a very short duration.

On a transit system you don't want roller coaster style restraint systems so g-forces have to be much much lower.

Some people might be into "ride a roller coaster from NYC to Boston" but that sounds like a niche market to me.

To clarify the referenced ride is in Gold Reef City in South Africa and is not the same as the Twilight Zone Tower of Terror at Disneyland

> Twilight Zone Tower of Terror at Disneyland

Which doesn't exist anymore. It's now a Guardians of the Galaxy ride (much better IMO).

In California. It still exists in Florida.

Just to add, a competition Jetsprint Boat can pull 7G's in the corners... And that's pretty intense.

https://www.knfilters.com/news/news.aspx?id=1444

It highly depends on a direction, see https://en.wikipedia.org/wiki/G-force#/media/File:Human_line...

That's sustained 6g. Momentary 6g is easy to deal with. But it's not like most people will be experiencing such accelerations in a transportation setting normally.

6 _sustained_ gs would require a g-suit, but for very brief durations it's not bad. It would be very uncomfortable for passengers though.

Humans are pretty tough, it turns out. And people often are overly conservative in these kinds of discussions.

http://rollercoaster.wikia.com/wiki/Highest_G-Force_on_a_Rol...

Maybe you don't want to have a transportation system that is as uncomfortable to ride as a rollercoaster. You might want to be able to transport frail elderly people and small children as well.

This is a test track and a competition. Nobody is claiming to build a passenger hyperloop with these accelerations except for fun.

270mph is 120 m/s. Two minutes of acceleration.

Passenger jets take off at more like 4m/s^2.

"supersonic" relative to the sound speed at normal density I assume. With no air, there would be no sound!

Edit: corrected pressure to density

https://mobile.twitter.com/elonmusk/status/90197511779785113... @elonmusk: "No such thing as true vacuum"

BTW, speed of sound doesn't change (to first order) with changing pressure (edit: nor does it change (to first order) with density, assuming temperature is the same). Mostly effected by temperature.

Hyperloop uses a partial vacuum, of course. No true vacuum exists.

"No true vacuum exist."

How high is density of atoms in space? Could there be a 1mm x 1mm space without atoms?

According to Wikipedia:

"Plasma with a number density of less than one hydrogen atom per cubic metre and a temperature of millions of kelvins in the space between galaxies accounts for most of the baryonic (ordinary) matter in outer space; local concentrations have condensed into stars and galaxies." - https://en.wikipedia.org/wiki/Outer_space

I knew that intergalactic space could see densities of one atom per some human-relevant volume of space, but I had no idea that actually accounted for most of what there is!

My understanding is that there would still be quantum fluctuatons causing particles to pop in and out of existence.

Would this be protons? Can vacuum contain photons and still be a vacuum? Not sure :-)

https://en.wikipedia.org/wiki/Quantum_fluctuation#cite_note-...

The fastest train in service by the way: https://en.wikipedia.org/wiki/Shanghai_Maglev_Train

(Which is a version of the SIEMENS Transrapid Maglev system)

The WARR sled runs on, and is propelled by, wheels.[1] It's not flying on an air cushion, which was the original Hyperloop concept. Nor is it a maglev, like Hyperloop One.[2] (That's a vactrain.)

[1] http://hyperloop.warr.de/pod-ii/ [2] https://hyperloop-one.com/blog/how-and-why-were-levitating

So originally they wanted to have an air cushion, inside a vacuum enclosure?

Yes. The Hyperloop concept Musk published in 2013 [1] was based on a partial vacuum with air cushions like a plane flying at 40,000 feet rather than a satellite zooming through space.

[1] http://www.spacex.com/sites/spacex/files/hyperloop_alpha-201...

Read the original Hyperloop paper.[1]

[1] http://www.spacex.com/sites/spacex/files/hyperloop_alpha-201...

By the way, the only reason commercial air travel can go at transonic speeds (670mph) efficiently is because it's at near vacuum, i.e. at high altitude, just like Hyperloop.

That's also why conventional or maglev non-vacuum trains won't ever travel as fast as passenger air travel cruise speed. It takes a stupendous amount of energy to travel that fast near sea level. Even a lot of supersonic fighter jets can only reach supersonic speeds at high altitude.

The whole advantage of hyperloop (and vacuum-trains) is you're bringing that high altitude air down to near sea level. The other things are implementation details.

This is why hyperloops will fairly quickly eclipse other train speed records, currently 375mph for maglev.

Just a FYI, the Hyperloop One pod recently reached 192mph, so this student pod bested them (although the student pod is much smaller and the student competition hyperloop track is much longer, so the Hyperloop One pod must've been higher performance).

Air pressure at 10 km altitude is about 30% or so of sea level, it is much less but definitely not near vacuum.

If it were a near vacuum any airplane would stall...

747 max altitude is 45,100ft, or 16% sea level.

At Concorde's cruise altitude of 18000m, it's just 8% of sea level. Of course, Concorde had to fly very fast to keep from stalling. U2 spy plane has the same problem at 21,000ft, 5% sea level pressure. It's nearly stalled while cruising. I wouldn't call that "definitely not near vacuum." Still higher planes, one at 100,000 ft, have flown (NASA's Helios), 1% atm. Near Mars pressures, definitely near vacuum. (as much as this term is subjective)

Hyperloop One plans an altitude equivalent of about 200,000 ft, or about 0.16% sea level pressure.

Seen any Concordes lately? The last time I saw one it was sitting on a static display looking very good but going nowhere.

Concorde is not a representative sample of passenger air travel, neither is a U2 spy plane nor is Helios.

What Hyperloop One plans is not relevant, what they will manage to achieve is and there is no way that they will manage to run that network at 0.16% of sea level pressure.

The whole hyperloop discussion would benefit from an injection of some solid engineering principles, the linked article is a nice example of what can be done and even if they won a competition has nothing to do with the hyperloop concept as presented originally, the vast majority of the original engineering challenges still stands.

"Seen any Concordes lately?"

Nevertheless, that was the genesis of the Hyperloop idea. Musk mentioned Concordes as an example.

And you can't (realistically) have supersonic passenger travel without traveling at similar altitudes (or altitude equivalent pressure) because it uses too much energy. I think you're missing this key point: transonic (670mph) or supersonic speed travel requires low or near-vacuum pressures to work efficiently. Other aspects of hyperloop (or related ideas) are implementation details. Power losses scale with drag which scales with air density. And to the degree that Concorde failed because it was too inefficient, having an even lower operating pressure should address that as well.

"Concorde is not a representative sample of passenger air travel" It is, along with the Soviet jet, the /only/ sample of supersonic passenger air travel.

I don't see any solid engineering analysis in your comment, by the way, just assertions that this or that cannot be done.

The general idea here is not that something can not be done in principle or not but that it will be hard to do it economically. Little details like radius-of-turn, g-forces imparted on passengers, cost of maintaining infrastructure and so on are important.

Elon Musk so far has done spectacularly well at re-doing stuff that was already done in principle - other than the re-use of rockets, which is very impressive and a serious source of headache for SpaceX competitors. To create a whole new class of transportation from scratch is a different kind of challenge. I wish him well, it is just that these mere matters of engineering are hard and long distance travel is mercilessly dictated by economics. Hence my reference to Concorde, which worked well but was not competitive in the longer haul. So if you are looking for engineering reasons you will not get those from me.

For now Hyperloop does not exist, we will see if it gets done in the next decade. Or two. And once it has been done whether or not it can be operated profitably.

By this standard only way to satisfy your question of whether it can be done economically is if it already is being done, and being done en masse. There's no point in discussing any future technology if this is the standard, as there's no way this can be satisfied except in hindsight.

This sort of absurd standard of discussion is super common on the Internet and highly boring as it leaves zero room for actual insight or analysis, as everything that isn't already far in the rearview mirror is dismissed as "unconvincing."

I'm not sure your Concorde analogy is instructive.

You're saying because commercial supersonic flight (which caused sonic booms and was insanely expensive fuel and maintenance-wise) required a high altitude, that Hyperloop as a concept is a failure?

Holy moly that's fast. https://twitter.com/elonmusk/status/902039243601485824

Can you imagine traveling that way?

Sorry for the fluff comment, but wow!

> Can you imagine traveling that way?

Plenty of people can imagine it because they travel that way every day. Trains all over Asia and Europe reach those speeds and faster.

Maybe I have been taking the wrong trains (in Germany), but 200 mph (321 km/h) seems plenty fast to me.

That's higher than the maximum speed of most of Deutsche Bahn's high-speed trains: https://en.wikipedia.org/wiki/Intercity-Express

Even those that can go faster than 300 km/h tend to do so only on very short legs, because they need long straight stretches to stay on track and have to accelerate/decelerate at the ends as well.

Real high speed in Germany does not make sense, too many stops due to political reasons ("Wolfsburg" comes to mind and "Kassel"). That's the reason new ICE trains are slower than their predecessors. Other countries have far fewer stops on high speed routes and can therefor have faster trains (e.g. Japan and France).

Distances in China are enormous compared to those within Germany. Japan is in the convenient situation of being rather long and slim so that its railway network is almost 1D rather than 2D. So their settings are quite different.

Note how the ICE 3 operates at 320km/h and the ICE 4 only goes up to 230km/h. I would assume they’re optimising for efficiency now.

Hyperloops would also have those constraints on track layout.

I think this is another reason why musk is getting into tunneling. Reducing the cost of tunnels lets you place the hyperloop track in a more optimal location.

would you still use a steel containment pipe? poured cement seems like it would be cheaper, but every example I've seen develops cracks and starts letting in water.

https://en.wikipedia.org/wiki/Siemens_Velaro

Not in a 10ft wide, hundreds of miles long tunnel; this is making a big difference.

There are ~5ft diameter natural gas pipelines operating at high pressure (>1000psi) that are thousands of miles long: https://en.wikipedia.org/wiki/Yamal-Europe_pipeline

You could fit a hyperloop pod in one of those, just seat single-file. Larger diameter is optimal since you can get by with a lower pressure, but we're in the ballpark here already.

In a gas pipeline, the high pressures help to stiffen the structure. In a hyperloop, you have very low pressure inside the pipe which changes the pipe design completely.

https://youtu.be/T9bpUfWy8Wg?t=161

Certainly and good point, but vacuum is just -15psi while these gas pipelines are 1200psi, almost a factor of 100 greater. Simply the required thickness of metal for 1200psi may provide sufficient stiffness to withstand vacuum.

In fact, let's do the math. For a typical carbon steel tensile strength of 370MPa, internal pressure of 1200psi with a factor of safety of 2, and a radius of 2.5ft, you'd need 3.4cm thick walls:

2.5ft/(370MPa/(2 * 1200psi)) in cm

https://www.google.com/search?q=2.5ft/(370MPa/(2*1200psi))

According to this design guide for carbon steels, and using 30psi pressure differential for vacuum (i.e. a factor of safety of 2), we get: http://lss.fnal.gov/archive/test-tm/1000/fermilab-tm-1378.pd...

1.2cm thick walls, which is much thinner than the gas pipeline.

5ft * ((1-.28^2) * 30psi/(2 * 30 * 10^6psi))^(1/3) in cm

https://www.google.com/search?q=5ft*((1-.28^2)*30psi/(2*30*1...

So those gas pipelines should already be plenty thick enough for vacuum service. My point stands.

Bear in mind the camera is close to the ground (not to mention in a very narrow tunnel) which exaggerates the feeling of speed.

Details on the winning pod: http://hyperloop.warr.de/pod-ii/

IIUC it accelerates using a wheel pressed towards the track, powered by an electric engine... no fancy magnets or fans and air cushions.

It's cool, but it doesn't look a lot like the original Hyperloop white paper.

e: from the competition rules at http://www.spacex.com/sites/spacex/files/2016_0831_hyperloop..., it looks like they were at least using an evacuated tube.

>It's cool, but it doesn't look a lot like the original Hyperloop white paper.

That's the first thing I noticed. This seems like a competition for fast electric train.

From my understanding, it seems the main concept behind Hyperloop is a "fast electric train" in an evacuated tube.

Other design details from the original spec. don't seem as central to the definition.

And? The Wright Flyer had a terrible mechanism for flight controls, it's good that was moved away from. Innovation is about iteratively pursuing progressively better designs, even if they differ from your original concept.

Musk made comments about traveling at up to 800mph in this 0.8 mile tube.

At constant acceleration, you'd have to have about 10 gees (each way) to reach this speed in that short tunnel:

sqrt(2 * acceleration * distance) = speed

sqrt(2 * 100m/s^2 * 0.4mi)=358m/s = 800mph

https://www.google.com/search?q=sqrt(2*100m/s^2*0.4mi)

To accelerate at 100m/s^2 at the end when you're at 358m/s requires about 35.8kW/kg, which is /insane/. You'd need some sort of active cooling to get that to work. (Water flashed to vapor? Liquid nitrogen?)

So you'd probably have a really high acceleration at first (as fast as the sled could handle, maybe 100g?) and really hard deceleration at the end in order to reduce required electric motor specific power. Might be tough to do all that with a single motor, so you probably need either gearing or a two-stage design, with a larger low-speed, extreme acceleration stage pushing a lighter high speed moderate acceleration stage. I don't know if two stage designs are allowed, but it'd certainly help in this competition.

Reply on news.ycombinator.com