Engineering, Passenger Rail, Technology and IT

How we can make super-fast hyperloop travel a reality

Hyperloop. Artist's impression: SpaceX

With 35 new inter-city routes shortlisted for testing around the globe, it’s time to start taking hyperloop seriously, Hugh Hunt writes.

Across Europe and parts of Asia, travellers can enjoy some of the fastest rail services in the world. From Málaga to Madrid, Tokyo to Osaka, high-speed electric trains condense the travel times between major hubs by racing along at some 300kph. The fastest commercial service in the world is the Shanghai maglev – short for magnetic levitation, the method of propulsion it uses to glide along its tracks as rapidly as 430kph.

Of course, air travel is still much faster: an Airbus A380 aircraft has a cruising speed of over 1,000kph. But at a time when reducing emissions is a top priority across the globe, there’s an urgent demand for cleaner, more energy-efficient alternatives – especially in the US, which is by far the world’s biggest user of air travel, with almost 800m passengers each year. Enter, the Hyperloop – a train-like technology which has the potential to match air travel for speed.

Hyperloop is the brainchild of US business magnate Elon Musk. First proposed in 2013, the Hyperloop system consists of “pods”, which are suspended inside a tube by magnetic levitation and propelled using a linear electric motor. The environment inside the tube is almost a complete vacuum, allowing the pods to travel at great speeds without being slowed by air resistance. The tubes themselves can be placed underground, or run above ground, elevated by columns.

The race begins

Musk originally intended the Hyperloop to cover the 600km route from Los Angeles to San Francisco at an average speed of about 960kph, reducing what’s currently a 12-hour train journey to just 35 minutes. Although funding has since been channelled into a bullet train service for this route, the idea of the hyperloop has attracted interest elsewhere.

The wealthy city-state of Dubai has agreed to conduct a feasibility study for a 150km link with Abu Dhabi. There’s also a proposal to connect Vienna with Budapest and Bratislava. And US start-up Hyperloop One recently announced a shortlist of 35 potential hyperloop test projects, which included proposals for routes linking Sydney with Melbourne, London with Edinburgh and Mumbai with Delhi.

While these developments have sparked much excitement, some remain sceptical about whether they can work in the real world.

Too fast to function?

Hyperloop pods are designed to reach their top speed of 1,220kph (slightly less than the speed of sound) in about 70 seconds, when accelerating at 0.5G (the “G” refers to “G-force”, which is how we measure acceleration).

To put this in context, at 1G we are pushed into the back of our seat with a force equal to our body weight – it would be uncomfortable. But the acceleration of an aircraft during takeoff is typically around 0.4G, and most people are happy with that.

We also experience G-forces when we go around a curve. This “centrifugal force” is what flings you from side to side on fairground rides. Again, about 0.5G is the limit for comfort. Travelling at speeds of 1,220kph sets the minimum curve radius to about 23km, which means that the track has to be pretty straight. It must be very level, too, because vertical hills and bumps also give rise to G-forces.

With the right site, these constraints could be manageable. The real challenge for hyperloop will be dealing with earth movements. In all large-scale engineering, allowances are made for thermal expansion, ground water and seismic activity – things that make the ground shift around. Normally, these aren’t too much of a problem. There are expansion joints in bridges and pavements, and even when subsidence causes cracks to appear in a wall, we shrug our shoulders and say “so what?”.

But movement in the hyperloop track could cause real problems, when the pods are travelling at such high speeds. That’s why Musk favours a track on columns, so that it can be adjusted and realigned in the event of ground movement. Indeed, we already do this kind of realignment with conventional railway tracks: the rails on sleepers are loosely supported on ballast and regular “tamping” ensures that the track is kept straight.

With such demanding specifications, actually constructing a hyperloop will not be cheap. But the days of aircraft and ships are numbered, unless we can find a way to power them with electricity or hydrogen fuel. Perhaps we could even learn to live with nuclear-powered ships. Hyperloop offers a novel vision of the future of long-distance travel – one that might just catch on.

The Conversation

Hugh Hunt is Reader in Engineering Dynamics and Vibration, University of CambridgeThis article was originally published on The Conversation. Read the original article.


  1. Oops … what is a kph? k is short for kilo, a prefix meaning 1000. p is short for pico, a prefix meaning 0.000 000 000 1. And h is an hour. Put them together and one kph means one nano hour, a measure of time, not of speed. If instead you refer to speed, the correct units are km/h.

  2. its interesting how all the engineering issues surrounding hyperloop serves as a distraction from the real problem – the lack of a business case. In simple terms, hyperloop has a poor ratio of capacity (people transported per hour) to construction cost. Whereas a conventional HSR line could transport in the order of 20,000 people per hour per direction (using 16 car trains), hyperloop would struggle to transport a tenth of this number of people. The basic problem is that hyperloop is structured around rigid “pods” and there is a limit to frequency set by both airlock cycle time and more importantly emergency braking.

    Acceleration from 0 to 1200Km/hr at 0.5g takes 68 seconds. Emergency braking would have to be at over 1g. Not impossible, but clearly there isn’t much room for improvement as far as capacity is concerned. Even if you accept 0.5g acceleration into your seat (and I’m one of those who would enjoy this – but then again I like roller coasters), the minimum curve radii are far in excess of those required for a conventional HSR line operating at 400Km/hr – somewhere between 4Km and 8Km depending on design. In other words, you just don’t get as much opportunity to avoid terrain with hyperloop even if you accept higher g forces. Maybe this is why Elon recently (jokingly?) started tweeting about going into the tunneling business.

    Lets consider that we want to build a high speed transport system between Sydney and Melbourne. Conventional HSR (400Km/hr class) will get you there in about 2 hours. This would easily capture most of the present air travel market. Hyperloop would reduce this to 1 hour. Now we can capture nearly all of the air travel market. Problem is the added value in hyperloop over conventional HSR is only that incremental shift between “most” and “nearly all”. Worse still is this isn’t much use because hyperloop simply doesn’t have the capacity regardless.

    Stepping back a bit, part of the problem with this conversation is the general (mis)understanding that HSR is about air travel replacement. Actually, its more about connecting regional cities with capitals and replacing car travel over medium to longer distances (50 to 500Km or so). Journeys like Newcastle to Sydney for instance. The problem is that hyperloop’s lack of capacity really starts to hurt when you want to move tens of thousands of commuters.

    The other thing that bugs me is Elon’s assertion that hyperloop will be cheap. The reality is that there isn’t a lot of difference between a structure capable of supporting a modern high speed train (using aerospace materials) and a structure capable of supporting an inch thick steel tube that has a mass of roughly 3 tonnes per linear metre. Any advance you can think of to cost reduce a viaduct for hyperloop is also an advance in the construction of a conventional HSR line. This is partly why I don’t mind hyperloop – it will bring forth some innovations. It will also drive the development of magnetic drive systems which can only be a good thing.

    Bottom line here is that hyperloop is perfectly feasible, but as a mass transit system it just doesn’t have the capacity. The only place I can imagine it having a use is over very long distances (transcontinental). Sydney to Perth comes to mind. Problem is that then regular aviation easily competes on economics. One other comment. Regardless of the source of energy, we are still going to have aviation fuel in the future – whether that’s bio-engineered or hydrogen – the hydrogen fuelled SABRE hypersonic engine is quite possibly where we are headed.

  3. Yeah, not a good start to a proposal when the author can’t even use the correct units of measurement is it!
    You wonder about standards at Cambridge!

  4. Language is a moving standard.
    Dr Google shows that kph is in common usage as kilometres per hour. (kilometers per hour)