Hacker News Reader

Comment by hugh3 | Hacker News Reader

How's this for a simpler implementation without many of the problems which have been mentioned:

Approaching the station, there's an announcement on the train "Everybody getting off at the next station please move to the last carriage". Then the last carriage simply detaches from the train and coasts into the station to unload at the platform.

Meanwhile, another carriage has been loaded with the passengers who were due to get on at the station. After the train passes the carriage is shunted by a small locomotive to catch up with the train and it attaches to the end.

No need for crazy long parallel tracks and dangerous transfers between high-speed trains. Difficulty: most high-speed trains have locomotives at both ends afaik, but that's a minor issue.

View on HN

High speed trains don't have any locomotives (only German ICE1 uses this approach), powered axles (and other electric equipment) are distributed across the whole unit. (So you can't split it, only couple/uncouple two units together.)

But, if you think that those transfers are dangerous, I don't know what to say about coupling in high speeds. Insane, maybe :-) Coupling even in station when only the locomotive is moving is favorite source of accidents. Basically you'd have to reach almost same speed as the train you are coupling with in the exactly same moment when you are going to touch it, with only very small difference that buffers can absorb. Hard enough if that speed is zero.

Surely the trains would be specially constructed to perform this procedure reliably, and it would be entirely automated. This doesn't seem like it should be a hard problem if the system is engineered from the beginning to handle it.

Coupling two things together in 200 or 300 kph is not a hard problem? That's like people without any programming experience saying, hey, that traveling salesman can be easily solved for 10 nodes, so just add more power and solve it for 100 or 1000, that cannot be that hard!

It's quite hard to control exact speed of a train (at such high speed), even without any atmospheric effects and with ideal adhesion conditions that never change. I can imagine it in like 50 kph, when the worst thing to expect is a derailment, some deformation zones blasted, and few non-fatal wounds. But every problem rises at least exponentially with the speed.

Fair enough, I hadn't realised how difficult it was to control the exact speed of a train at high speed.

But if that's the case then it still seems like the original problem of running two trains right alongside each other, on different tracks, for several minutes while people cross from one to another, and with the consequence of any error being passengers getting sliced in half, would be even harder.

Ok, it's a hard problem, but there's gotta be a way.

Maybe the coupling mechanism could be something long that extends from the back of the train. The cars can connect can over a wide range of distance, kind of like a runway. The connection is made with a friction clutch that can be applied gradually and eased off if something goes wrong. The coupling also has a drive to pull the cars together after they are securely linked.

They could always slow down to make the link too.

Deleted comment

And still. Does anyone except military use in-flight refueling? Is every attempt of in-flight refueling successful? If you suddenly lose some speed or something (unfriendly wind or whatever), you have plenty of space to maneuver, and even in the worst case, you should have rather good chance to survive hit by that pipe.

ICE train in Germany got derailed by few sheep two years ago…

If this is a high speed train, and the front car of the train is always being rotated backward, how do you deal with making sure that this first car is aerodynamic, and remains aerodynamic once it is no longer the first car?

That wasn't what I was suggesting; I'm suggesting changing the back car. The first (N-1) cars remain the same throughout the journey and the rearmost one gets swapped at every station.

Reply on news.ycombinator.com