Hacker News Reader

Comment by throwaheyy | Hacker News Reader

The most limiting factor is the pantograph/catenery contact; as speeds go up the wire has to be kept at a higher and higher tension to prevent the train catching up to the standing wave it creates in the wire. If the tension is not high enough to prevent this, the train will lose electrical contact or even rip down the wire. This is also why high speed trains can only have one or two pantographs to collect power.

View on HN

In a more practical sense, the limiting factor is curve radii. The higher the speed, the larger the curve radii, and the more property parcels your curve will intersect with, blowing up the cost of construction. At top speeds of 350-380km/h Chinese HSR has a minimum curve radius of 7km. You can't really make curve radii smaller while keeping speeds high, or drinks will start sliding off tables and people may start vomiting.

If you know the approx speed the train will be going, you can curve harder and camber the track at the right angle to compensate. People will feel slightly heavier, but force will still be toward the floor.

Planes do this to the extreme, it's always weird noticing the plane roll 45 degrees and still feeling bum firmly in seat.

> ...it's always weird noticing the plane roll 45 degrees and still feeling bum firmly in seat.

Yes, that is a coordinated turn. It's one of the very first things you learn when you take flying lessons.

Airplanes aren't very good at flying sideways. You can make them do it to some degree, and it is an important technique for landing in a crosswind. But they are better at just climbing or descending - from the plane's point of view.

So the way you turn is to roll the plane in the direction of the turn and simultaneously start "climbing". Climbing straight "up" from the point of view of the airplane, not up according to the ground. The airplane doesn't know about the ground, it only knows about the air.

A coordinated turn is a combination of banking and climbing, such that you always feel like you're being pressed straight down into your seat. And the tighter the turn, the more you are pressed into the seat.

When you make a turn in a car on a flat road, you feel a normal 1G straight down and a force pushing you sideways. But if the road were perfectly banked for the speed you're going, you would feel like you were being pressed straight down in your seat. Instead of feeling a sideways force, that would be added to the downward force you feel inside the car - just like an airplane.

I suspect that making railways perfectly flat for high speed use is hard enough. making them banked at 20+ degrees for several kilometres sounds like a serious engineering challenge

The other caveat is that banking only works if you're traveling at the correct speed. Too fast and you still feel the turn. Too slow and you feel the bank. For airplanes this isn't a problem because there's never a speed restriction on air. For things that travel along the ground there is no end of reasons why they might have to travel slower than normal.

Banked turns spanning several kilometers are not unusual, even in standard rail systems.

This is true, but the camber you can use is limited by the fact that you may have to stop the train on the curve without it falling over.

There is / was a train design where the train itself could lean into corners on a flat (legacy) train track. Pretty sure that's not considered anymore because the tilting system adds a lot of weight and complexity.

Mag-lev trains seem to reach 'around' the tracks, so that could work. Of course if the levitation stops in a corner (or anywhere, really) on that one it'll cause major damage.

>There is / was a train design ...

Pendolino:

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

Up to 8-10° , i.e. 1.35 m/s^2 compensation https://it.wikipedia.org/wiki/Pendolino#Caratteristiche_gene...

Tilting trains are actually in wide use, see: https://en.m.wikipedia.org/wiki/Tilting_train

Even the lates shinkansen transets do it to reach higher speed, especially on older parts of the track with sharper turns: https://en.m.wikipedia.org/wiki/N700_Series_Shinkansen

You could implement a strategy of allowing evacuations only at scheduled stops.

But if the track is distrupted (a tree falls on it, or somehow it's just damaged), the train would have to stop right there and then.

Maybe they can make something so the train can grip the ground/tracks so it just doesn't fall off when the forces are no longer acting on it.

The tracks are not fixed to the ground. If a train grips the tracks in order not to fall over, it will probably still fall over and also rip the tracks out of the railbed as it does so.

This is not always true, most of Shinkansen track is concrete slab track, which is effectively fixed to the ground. And they actually do have special anti derailment guides build into the track to prevent a train from derailing during an earthquake: https://www.rtri.or.jp/eng/rd/seika/2006/01/safety_E05.html

True, as we've learned in this article. I guess it will have to grip to something else.

But in the theoretical banking the original comment was talking about (let's say 30 degrees), the track would probably have to be secured to the surface...

Well, how heavy is a bullet train, where is its center of gravity, and how much banking does it need if it wants to travel, say, 300 km/h in a (how sharp?) a curve?

Keep in mind that banking is only comfortable if you are traveling at the correct speed.

There are a whole host of reasons why a train may not travel at maximum speed for a given segment, and the rail line still needs to be reasonably comfortable at all of those.

Doesn't work if the train caught fire or something.

Planes also require you to strap in and put your tray tables up whenever the seat belt sign is on. There's generally not a meal service happening when the plane is tilting at 45 degrees.

One of the competitive advantages that train travel has over plane travel is that it is significantly more comfortable, and you can get up and walk around or have a coffee on your table whenever you'd like. Train travel would be less able to compete with plane travel if the experience was more akin to sitting in a roller coaster.

Even turns in the vertical direction are a problem, for uncomfortably pushing you into your seat, or lifting the train off the rails. Which results in more bridges & tunnels than would otherwise be needed. Unlike slow trains, there is plenty of power for going up hills, but you can't switch to going down the other side.

While ordinary horizontal curves are cambered, IIRC the resulting train-vertical acceleration is still one of the limiting factors.

Yeah, but in case of emergency stop the train will roll over...

Not unless the bank angle is so great as to put the center of gravity outside the wheels. Considering how passenger rail cars are build (a deck/chassis to take the push/pull forces with a tin can on top for the cargo) it shouldn't be an issue.

HSR trains usually do it as well. In a TGV, you normally don't feel curves at all.

The curve radius of new LGV is 7km and cant is 160-180mm. The curve is still absolutely massive and the cant still small.

The solution for this was tilting trucks. As the train rounds a curve, the cars themselves bank to counteract the outward forces, just like an airplane. The track doesn't have to bank, the vehicle does.

This works up to a point, but early tilting trains (e.g. British APT) were scrapped because the tilt was too high and it made drinks spill and people too nauseous.

One quote that stuck with me is "On reaching the 500 km/h mark, the camera controlling the pantograph condition already shows a continuous electric arc" - you can imagine that under these conditions the wire will wear down much faster than with "only" 350 km/h. So, before reaching a physical limitation, you already reach an economic limitation...

It is also possible to travel faster than some waves emitted in the ground. Particularly when going from a rigid to a soft foundation (like coming off a bridge). The air compression going into tunnels can cause drama as well. Everything gets more difficult the faster you go.

I guess it'd be a big safety headache but perhaps the train could be powered through the rails? To avoid accidents the rails could be split into sections that only get powered when a train is directly above them.

The rails are already used for ground/return current. The single overhead wire is at 25kV (or higher for this speed test).

It is only where the vehicle has rubber tyres (i.e. trolley buses) where there are two overhead cables.

Changing the current system would be prohibitively expensive. I also imagine that you have pretty bad losses at 25kV AC if your conductors are touching the ground and are potentially wet or covered by snow. It would also be a fire hazard I guess.

Plenty of metro rail systems use a third rail today, even outside. Maybe their losses are more manageable because their systems are relatively short compared to long distance rail, but it doesn't seem like an insurmountable obstacle.

Metro rail runs at a much lower voltage afaik.

They do this in South East UK using a dedicated third rail for power. As I understand it, the limit is 100mph for this technology, although I don't know the physics, so it's not even fast enough for classic High Speed

Part of the problem is the low voltage conventional third rail systems use. In the UK, third rail voltage is 750V, compared to 25kV for overhead lines, so third rail currents get pretty high. You already get get pretty good arcs when the contact shoe moves on and off the third rail going through junctions. And you can't really increase the voltage very much before the rail starts to arc to ground, especially when it's raining and the insulators get wet.

Reply on news.ycombinator.com