> - 2-3x speed of bullet trains. Can change the way we perceive geographic distance.
> - Cost isn't as much of a concern because a national hyperloop network would create a lot more value than a bullet train equivalent.
High speed brings with it some pretty severe problems. The most well-known is curve geometry: curve radii scale with the square of velocity. Transition track (how long the "S" of a S bend needs to be) scales with the cube. But it's not the only issue. Also scaling quadratically with speed is energy consumption and stopping distance. Speed only gives you a linear improvement in time, so things like capacity decreases [1] and stop penalties are only linear.
But the difficulty you face is switches. Here, you have curve radii constraints, which means your problems are in the quadratic issue. I've never seen any depiction of a hyperloop switch [2], but I imagine it's going to be more in the vein of move the entire tube than just swap a point. In any case, a high-speed switch capable of being traversed at 200 km/h is... 164m long. To go 1000 km/h would require it to be 4.1km. That means it will take a vehicle at least 14.76s to traverse it. So you can't schedule vehicles any tighter than that, but for safety reasons, you want the following vehicle to be able to come to a complete stop in case the switch breaks down before it can complete moving--which is going to slow you down to minutes between vehicles [3].
With these constraints, you might think people would be trying to design high-capacity vehicles, competing with the ~1000 you can stick on a normal trainset. But the design is more akin to a small airplane with the wings and tail cut off, fitting maybe a few dozen people per pod. Actually, the more you look at the designs, the more you see airport-style operational motifs sinking in. It starts to look a lot more like a long chain of point-to-point links, with the capacity issue of airplanes, but the infrastructural cost of railroads. Even the modelling seems to basically assume that people will treat hyperloops as airplanes.
[1] Yes, making a train network higher speed decreases its total throughput.
[2] A telling sign in itself.
[3] The highest capacity metro systems today can manage to schedule 45 trains per hour, and most systems are typically about 25 trains per hour. I believe the highest capacity HSR line is 6 trains per hour.
Smartrail and PRT are designed as a point to point separated-grade network carrying on av. 1 person or a pallette of goods. With a hanging rail you don't need heavy batteries, parking for the vehicles, you can run a pod straight into the factory to pickup goods, and large/rich places can pay to have track straight to their door.
You can prefab the rail and as land usage is just poles in the ground it can be rolled out over fields etc quickly. Track is one-way to eliminate junctions. Pods are on-demand ie. No waiting.
Accessibility improves, you can use the top of the rail to generate (solar) power, run highspeed internet cables in the rail to improve comms across a country, and save on distribution center logistics as you're going point to point.
The last mile may possibly be an issue, but forklift drones and bicycles can take most of the load I feel.
Drivers for this are that it would go fast (200mph+ as light pods so little wear), can go overnight (sleeper pods), you could buy track to your door, personal transport (like a cinema room if you want). The main real issue with cars is that there's a large lobby behind what is a legacy transport solution...
By way of example, UltraPRT has been running flawlessly at Heathrow airport for 10 years, was built on time, on budget, and performs exactly as predicted/modelled.
Couldn't the capacity issue be somewhat mitigated by stacking tubes vertically and/or increasing the tube diameter?
As for the curvature issue, there seem to be a couple ways around that. One is to straighten the route to the maximum extent possible (not trivial but not impossible for some routes). The other is to build massive curves and decelerate into them. Again, there are tradeoffs (cost,speed) but it's not impossible. There are just so many variables. A hyperloop could end up being great for some areas and unworkable in others.
Lastly, right of way/land clearance should never be a key constraint. That won't stop us if we really want to build a hyperloop.
> Couldn't the capacity issue be somewhat mitigated by stacking tubes vertically
Significantly increased costs
> and/or increasing the tube diameter?
Immeasurably increased costs. Because hyperloop already operates on the idea of operating in vacuum. So let's say you have a 10-kilometer tube that you need to keep in vacuum and you want to increase its diameter...
> - Cost isn't as much of a concern because a national hyperloop network would create a lot more value than a bullet train equivalent.
High speed brings with it some pretty severe problems. The most well-known is curve geometry: curve radii scale with the square of velocity. Transition track (how long the "S" of a S bend needs to be) scales with the cube. But it's not the only issue. Also scaling quadratically with speed is energy consumption and stopping distance. Speed only gives you a linear improvement in time, so things like capacity decreases [1] and stop penalties are only linear.
But the difficulty you face is switches. Here, you have curve radii constraints, which means your problems are in the quadratic issue. I've never seen any depiction of a hyperloop switch [2], but I imagine it's going to be more in the vein of move the entire tube than just swap a point. In any case, a high-speed switch capable of being traversed at 200 km/h is... 164m long. To go 1000 km/h would require it to be 4.1km. That means it will take a vehicle at least 14.76s to traverse it. So you can't schedule vehicles any tighter than that, but for safety reasons, you want the following vehicle to be able to come to a complete stop in case the switch breaks down before it can complete moving--which is going to slow you down to minutes between vehicles [3].
With these constraints, you might think people would be trying to design high-capacity vehicles, competing with the ~1000 you can stick on a normal trainset. But the design is more akin to a small airplane with the wings and tail cut off, fitting maybe a few dozen people per pod. Actually, the more you look at the designs, the more you see airport-style operational motifs sinking in. It starts to look a lot more like a long chain of point-to-point links, with the capacity issue of airplanes, but the infrastructural cost of railroads. Even the modelling seems to basically assume that people will treat hyperloops as airplanes.
[1] Yes, making a train network higher speed decreases its total throughput.
[2] A telling sign in itself.
[3] The highest capacity metro systems today can manage to schedule 45 trains per hour, and most systems are typically about 25 trains per hour. I believe the highest capacity HSR line is 6 trains per hour.