8 ms·

Tilting Trains and Technological Dead-Ends

71 points by luu 6y ago

Someone 6y ago
I don’t understand the first picture. I expect the sum of the green arrows to be zero or point inwards to turn the train with the curve. If the forces were as indicated, wouldn’t the train derail?
- dredmorbius 6y ago
  The diagram shows force vectors, not centre of gravity. Railcare rolling stock typically has low CG. Bogies, switches, and related equipment are located low in the car and have significant mass. The carriage shell itself is typically far lighter construction (less so in the US, due to higher crush-strength requirements based on the assumption that passenger and freight rail share trackbeds, significant enough that it's limited the maximum speed of US-based rail initiatives even presuming sections of dedicated track so long as any shared rights of way exist). For a CG diagram, I'd expect the CG to remain well within the wheel gague width, resulting in a stable vehicle.
- roelschroeven 6y ago
  There would be a force towards the outside of the curve, yes. But that doesn't mean the train would derail: in the extreme case the flanges on the wheels would prevent it, but what actually happens in normal circumstances is the same thing that keeps trains centered on straights. In a nutshell: since the wheels have a larger diameter on the inside than on the outside, and since the left and right wheels are on the same solid axle and thus rotate at the same speed, any deviation caused negative feedback. The required superelevation for a given curve to eliminate lateral forces depends on the speed of the train; since not all trains travel at the same speed, and you have to account for slow or even stopped trains, the superelevation is always a compromise.
- aidenn0 6y ago
  The vectors are the perceived force (which must be centrifugal) rather than the net force (which must be centripetal as you suggest).
londons_explore 6y ago
So this tilting is entirely for passenger comfort? When presented with "you can get there in 1 hour, but you can't put coffee on the table" or "it'll take 2 hours", do passengers really choose the latter? A coffee cup is at risk of sliding off a table around 4.5 m s^-2, so much higher than discussed in this article. And coffee would definitely slide off a flat surface in most journeys by car - that's why cars have cup holders.
- flohofwoe 6y ago
  I have the opposite experience (very low "passenger comfort") in tilting trains, and I always thought that this is what actually killed them: I'm slightly suspectible to motion sickness (e.g. "normal" trains are fine, but sitting in the back seat of a car or in a bus can be a bit of a problem, especially when trying to read something). When I was on a tilting train once (maybe about 10 years ago? AFAIK they are no longer in use) the motion sickness kicked in pretty hard. PS: Interestingly, a tilted track (which seems to be quite usual now on newly built high speed tracks) doesn't affect me at all. I guess it was the general "wobbliness" of the tilt train.
  - londons_explore 6y ago
    There is no difference between tilted track and tilted train from what you can perceive from the passenger compartment, provided the tilt is done correctly. I suspect tilt angle was perhaps rounded/approximated/delayed, which would feel the same as if someone had laid the track rather haphazardly on a bog...
    - pessimizer 6y ago
      No difference in respect to acceleration, but I've heard that motion sickness is largely visual.
      - labawi 6y ago
        AFAIU, the passenger compartment will indeed have identical angle and acceleration in both cases, as there is no other way to cancel lateral force. What is likely different is the stability of tilted tracks (reasonably easy to make nice continuous tilts and turns) and tilted train's uncannily small lateral force, while still having time and angle discrepancies and inferior smoothness causing very noticeable over and undershoots.
cnorthwood 6y ago
Pendolino's are still used on the West Coast Mainline connecting London, Birmingham, Manchester, Liverpool and Glasgow and I don't really have a problem with them. They're not going away any time soon, until we have HS2 (brand new route and alignment without the curves that need tilting for 125mph running). Interestingly, they are increasing the speed on some of the northern bits of WCML through changing the bank of the track more. My understanding is that if the bank was too steep, slower running freight trains had issues with it even if faster running passenger trains were fine. With a general increase in the speed of freight trains, the speed differential is less so a higher bank can be tolerated, and I think there are sections of WCML where non-tilting trains can hit close to the speeds of the Pendolinos.
- isthisnametaken 6y ago
  As an aside, after one was destroyed when it derailed in Cumbria, its replacement - a standard, non-tilting train - was nicknamed the "Pretendolino"
- chrisseaton 6y ago
  Do you not find the WCML a bit of a rollercoaster ride in places? There’s definitely a few points where you’re being buffeted and tilted all over the place unpleasantly. I can’t believe they’re going to make parts faster? Tea will be flying everywhere.
  - cnorthwood 6y ago
    few bits in the Thames Valley if I'm walking back from the buffet car or whatever, but the bits they're making faster are the bits through the Lakes and Borders and whenever I've been on the Transpennine 110mph trains I can't say I've found it particularly bad compared to the 100mph ones it replaced. (Not been to Scotland on one of their new 125mph trains yet and I don't know if they'll be authorised to run at full speed)
- Gravityloss 6y ago
  I'm not sure if I'd call Pendolino a dead end? From Wikipedia: "Pendolino is an Italian family of tilting trains used in Italy, Spain, Poland, Portugal, Slovenia, Finland, Russia, the Czech Republic, Slovakia, the UK, Switzerland and China." In Finland, they seemed to have had a lot of problems with the tilting mechanism in the winter, so new digital hydraulics system was developed for that. Digital hydraulics is a fun concept by the way. Sources: https://community.boschrexroth.com/t5/Rexroth-Blog/New-hydraulic-design-faster-through-the-curve/ba-p/237 https://community.boschrexroth.com/t5/Rexroth-Blog/New-hydra... https://en.wikipedia.org/wiki/VR_Class_Sm3 https://en.wikipedia.org/wiki/VR_Class_Sm3
rjmunro 6y ago
Maintenance costs are not a law of physics. Surely a better tilting mechanism could be developed that didn't require as much extra maintenance.
- bluGill 6y ago
  How much are you willing to pay? I'm sure it can be dine,but I'm not sure of the costs
  - jjk166 6y ago
    Well how are you making them, how many are you making and how much do alternatives cost? Economics vary wildly over time pending circumstances, but are irrelevant to the question of whether something is a technological dead end. To be a technological dead end there must be an alternative technology that accomplishes the same feat in a fundamentally superior way, such that regardless of the economic situation or future development it always makes sense to go with that alternative outside of niche applications.
    - bluGill 6y ago
      No, the problem here is there are only a few trains in the world that need this technology. Thus there are very few to amortize the costs of more R&D over. Tilting as a whole is a dead end: passengers get sick where it is used. The replacement is don't tilt your trains on corners at all - thought there are other technologies that are already more reliable as well if you must. It can still be done, but the ROI is very low.
      - jjk166 6y ago
        Path you don't care enough to go down != dead end.
jeffbee 6y ago
Sticking with Alon Levy's usual act of just dunking on American rail designers, I always laugh whenever I pass over this stretch of the SFMTA Muni light rail line between Townsend and 2nd, where the track is banked as if the streetcar was going to go around it at 80 MPH, even though the track is posted as 5 MPH max for LRVs and Muni's average speed is, last I read, 1.7 MPH, or half a human walking pace. Riding the streetcar over this track, or coming to a stop there which is very common, you see all the passengers leaning to the outside to counteract the ridiculous bank angle of the tracks. https://www.google.com/maps/@37.7811699,-122.3885358,3a,75y,179.8h,91.76t/data=!3m7!1e1!3m5!1sX7G7sQEj4Liddopi17VNBA!2e0!5s20110401T000000!7i13312!8i6656 https://www.google.com/maps/@37.7811699,-122.3885358,3a,75y,...
nickt 6y ago
You can’t talk about tilting trains with mentioning the ill-fated British Rail Advanced Passenger Train that started in the 1960’s. The article gives it a short mention. “Iwnicki thinks the biggest problem with the APT, devised in an atmosphere of "blue-skies thinking", was that it included too many innovations at one time. The casing, bogies, tilting and other features were all new - setting itself an impossibly high benchmark for success. But these have since become commonplace around the world, he adds.” https://www.bbc.com/news/magazine-35061511 https://www.bbc.com/news/magazine-35061511 https://www.crewehc.co.uk/apt https://www.crewehc.co.uk/apt
- rozab 6y ago
  Very good Mustard video: https://www.youtube.com/watch?v=nDN7PPW4AE8 https://www.youtube.com/watch?v=nDN7PPW4AE8
  - dharmab 6y ago
    6:17 is amazing. "Smooth, quiet and delightful" the spokesman says as the entire train shakes around him like an earthquake.
- bobthepanda 6y ago
  The tech got sold off to make the Pendolinos, so they do talk about its descendants.

71 points by luu 6y ago

I don’t understand the first picture. I expect the sum of the green arrows to be zero or point inwards to turn the train with the curve. If the forces were as indicated, wouldn’t the train derail?

dredmorbius 6y ago
The diagram shows force vectors, not centre of gravity. Railcare rolling stock typically has low CG. Bogies, switches, and related equipment are located low in the car and have significant mass. The carriage shell itself is typically far lighter construction (less so in the US, due to higher crush-strength requirements based on the assumption that passenger and freight rail share trackbeds, significant enough that it's limited the maximum speed of US-based rail initiatives even presuming sections of dedicated track so long as any shared rights of way exist). For a CG diagram, I'd expect the CG to remain well within the wheel gague width, resulting in a stable vehicle.
roelschroeven 6y ago
There would be a force towards the outside of the curve, yes. But that doesn't mean the train would derail: in the extreme case the flanges on the wheels would prevent it, but what actually happens in normal circumstances is the same thing that keeps trains centered on straights. In a nutshell: since the wheels have a larger diameter on the inside than on the outside, and since the left and right wheels are on the same solid axle and thus rotate at the same speed, any deviation caused negative feedback. The required superelevation for a given curve to eliminate lateral forces depends on the speed of the train; since not all trains travel at the same speed, and you have to account for slow or even stopped trains, the superelevation is always a compromise.
aidenn0 6y ago
The vectors are the perceived force (which must be centrifugal) rather than the net force (which must be centripetal as you suggest).

londons_explore 6y ago

So this tilting is entirely for passenger comfort? When presented with "you can get there in 1 hour, but you can't put coffee on the table" or "it'll take 2 hours", do passengers really choose the latter? A coffee cup is at risk of sliding off a table around 4.5 m s^-2, so much higher than discussed in this article. And coffee would definitely slide off a flat surface in most journeys by car - that's why cars have cup holders.

flohofwoe 6y ago
I have the opposite experience (very low "passenger comfort") in tilting trains, and I always thought that this is what actually killed them: I'm slightly suspectible to motion sickness (e.g. "normal" trains are fine, but sitting in the back seat of a car or in a bus can be a bit of a problem, especially when trying to read something). When I was on a tilting train once (maybe about 10 years ago? AFAIK they are no longer in use) the motion sickness kicked in pretty hard. PS: Interestingly, a tilted track (which seems to be quite usual now on newly built high speed tracks) doesn't affect me at all. I guess it was the general "wobbliness" of the tilt train.
- londons_explore 6y ago
  There is no difference between tilted track and tilted train from what you can perceive from the passenger compartment, provided the tilt is done correctly. I suspect tilt angle was perhaps rounded/approximated/delayed, which would feel the same as if someone had laid the track rather haphazardly on a bog...
  - pessimizer 6y ago
    No difference in respect to acceleration, but I've heard that motion sickness is largely visual.
    - labawi 6y ago
      AFAIU, the passenger compartment will indeed have identical angle and acceleration in both cases, as there is no other way to cancel lateral force. What is likely different is the stability of tilted tracks (reasonably easy to make nice continuous tilts and turns) and tilted train's uncannily small lateral force, while still having time and angle discrepancies and inferior smoothness causing very noticeable over and undershoots.

cnorthwood 6y ago

Pendolino's are still used on the West Coast Mainline connecting London, Birmingham, Manchester, Liverpool and Glasgow and I don't really have a problem with them. They're not going away any time soon, until we have HS2 (brand new route and alignment without the curves that need tilting for 125mph running). Interestingly, they are increasing the speed on some of the northern bits of WCML through changing the bank of the track more. My understanding is that if the bank was too steep, slower running freight trains had issues with it even if faster running passenger trains were fine. With a general increase in the speed of freight trains, the speed differential is less so a higher bank can be tolerated, and I think there are sections of WCML where non-tilting trains can hit close to the speeds of the Pendolinos.

isthisnametaken 6y ago
As an aside, after one was destroyed when it derailed in Cumbria, its replacement - a standard, non-tilting train - was nicknamed the "Pretendolino"
chrisseaton 6y ago
Do you not find the WCML a bit of a rollercoaster ride in places? There’s definitely a few points where you’re being buffeted and tilted all over the place unpleasantly. I can’t believe they’re going to make parts faster? Tea will be flying everywhere.
- cnorthwood 6y ago
  few bits in the Thames Valley if I'm walking back from the buffet car or whatever, but the bits they're making faster are the bits through the Lakes and Borders and whenever I've been on the Transpennine 110mph trains I can't say I've found it particularly bad compared to the 100mph ones it replaced. (Not been to Scotland on one of their new 125mph trains yet and I don't know if they'll be authorised to run at full speed)
Gravityloss 6y ago
I'm not sure if I'd call Pendolino a dead end? From Wikipedia: "Pendolino is an Italian family of tilting trains used in Italy, Spain, Poland, Portugal, Slovenia, Finland, Russia, the Czech Republic, Slovakia, the UK, Switzerland and China." In Finland, they seemed to have had a lot of problems with the tilting mechanism in the winter, so new digital hydraulics system was developed for that. Digital hydraulics is a fun concept by the way. Sources: https://community.boschrexroth.com/t5/Rexroth-Blog/New-hydraulic-design-faster-through-the-curve/ba-p/237 https://community.boschrexroth.com/t5/Rexroth-Blog/New-hydra... https://en.wikipedia.org/wiki/VR_Class_Sm3 https://en.wikipedia.org/wiki/VR_Class_Sm3

rjmunro 6y ago

Maintenance costs are not a law of physics. Surely a better tilting mechanism could be developed that didn't require as much extra maintenance.

bluGill 6y ago
How much are you willing to pay? I'm sure it can be dine,but I'm not sure of the costs
- jjk166 6y ago
  Well how are you making them, how many are you making and how much do alternatives cost? Economics vary wildly over time pending circumstances, but are irrelevant to the question of whether something is a technological dead end. To be a technological dead end there must be an alternative technology that accomplishes the same feat in a fundamentally superior way, such that regardless of the economic situation or future development it always makes sense to go with that alternative outside of niche applications.
  - bluGill 6y ago
    No, the problem here is there are only a few trains in the world that need this technology. Thus there are very few to amortize the costs of more R&D over. Tilting as a whole is a dead end: passengers get sick where it is used. The replacement is don't tilt your trains on corners at all - thought there are other technologies that are already more reliable as well if you must. It can still be done, but the ROI is very low.
    - jjk166 6y ago
      Path you don't care enough to go down != dead end.

jeffbee 6y ago

Sticking with Alon Levy's usual act of just dunking on American rail designers, I always laugh whenever I pass over this stretch of the SFMTA Muni light rail line between Townsend and 2nd, where the track is banked as if the streetcar was going to go around it at 80 MPH, even though the track is posted as 5 MPH max for LRVs and Muni's average speed is, last I read, 1.7 MPH, or half a human walking pace. Riding the streetcar over this track, or coming to a stop there which is very common, you see all the passengers leaning to the outside to counteract the ridiculous bank angle of the tracks. https://www.google.com/maps/@37.7811699,-122.3885358,3a,75y,179.8h,91.76t/data=!3m7!1e1!3m5!1sX7G7sQEj4Liddopi17VNBA!2e0!5s20110401T000000!7i13312!8i6656 https://www.google.com/maps/@37.7811699,-122.3885358,3a,75y,...

nickt 6y ago

You can’t talk about tilting trains with mentioning the ill-fated British Rail Advanced Passenger Train that started in the 1960’s. The article gives it a short mention. “Iwnicki thinks the biggest problem with the APT, devised in an atmosphere of "blue-skies thinking", was that it included too many innovations at one time. The casing, bogies, tilting and other features were all new - setting itself an impossibly high benchmark for success. But these have since become commonplace around the world, he adds.” https://www.bbc.com/news/magazine-35061511 https://www.bbc.com/news/magazine-35061511 https://www.crewehc.co.uk/apt https://www.crewehc.co.uk/apt

rozab 6y ago
Very good Mustard video: https://www.youtube.com/watch?v=nDN7PPW4AE8 https://www.youtube.com/watch?v=nDN7PPW4AE8
- dharmab 6y ago
  6:17 is amazing. "Smooth, quiet and delightful" the spokesman says as the entire train shakes around him like an earthquake.
bobthepanda 6y ago
The tech got sold off to make the Pendolinos, so they do talk about its descendants.