Recent

Author Topic: Railway signalling  (Read 1882 times)

MarkMLl

  • Hero Member
  • *****
  • Posts: 6646
Railway signalling
« on: December 05, 2022, 09:14:49 pm »
Does anybody have an interest in traditional semaphore signalling?

My understanding is that there's at least one simulator written in Delphi https://www.simsig.co.uk/ but that it doesn't interface with authentic peripherals...

MarkMLl

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

https://www.ekeving.se/tlv/Saxby/union.html
MT+86 & Turbo Pascal v1 on CCP/M-86, multitasking with LAN & graphics in 128Kb.
Pet hate: people who boast about the size and sophistication of their computer.
GitHub repositories: https://github.com/MarkMLl?tab=repositories

winni

  • Hero Member
  • *****
  • Posts: 3197
Re: Railway signalling
« Reply #1 on: December 05, 2022, 10:03:23 pm »
Hi!

Only a little bit near topic:

The worlds greatest model railroad layout is the Miniatur Wunderland in city of Hamburg

https://www.miniatur-wunderland.de/


Since the first day they use Delphi for the whole system: Railroad, cars, ships, airplanes,  and the light in the 15 minute night and day cycle. [ Exception: the software inside the cars is assembler]

Some years ago there was a report about this in german c't magazine:
 
https://www.heise.de/hintergrund/Delphi-im-Einsatz-bei-der-weltweit-groessten-Modellbahnanlage-1921899.html


A bunch of crazy girls and boys creating great technical solutions.

and one half of the twin-bosses is a great Delphi freak.

A lot of stuff can be found on youtube

Winni

MarkMLl

  • Hero Member
  • *****
  • Posts: 6646
Re: Railway signalling
« Reply #2 on: December 05, 2022, 10:07:52 pm »
Thanks, noted :-)

Although I'm talking 12-inches-to-the foot here :-)

MarkMLl
MT+86 & Turbo Pascal v1 on CCP/M-86, multitasking with LAN & graphics in 128Kb.
Pet hate: people who boast about the size and sophistication of their computer.
GitHub repositories: https://github.com/MarkMLl?tab=repositories

PascalDragon

  • Hero Member
  • *****
  • Posts: 5444
  • Compiler Developer
Re: Railway signalling
« Reply #3 on: December 05, 2022, 11:09:02 pm »
Does anybody have an interest in traditional semaphore signalling?

Considering that there's quite a different bunch of signalling in and around Munich (a few real semaphores on a cargo station in the east, an ever reducing number of light H/V signals, a growing number of Ks signals and PZB and LZB (on both S-Bahn and U-Bahn) as cab signalling systems as well as a mixture of different kinds of signalling controls from mechanical to digital (ESTW)), yes, I do have a passing interest in that. ;D

My understanding is that there's at least one simulator written in Delphi https://www.simsig.co.uk/ but that it doesn't interface with authentic peripherals...

Why the interfacing with authentic peripherals? And what kind of peripherals are we talking about?

Zath

  • Sr. Member
  • ****
  • Posts: 391
Re: Railway signalling
« Reply #4 on: December 05, 2022, 11:44:57 pm »
I live about 20 miles from Clayton Tunnel.
We always used to make a point of looking at it as kids when we were going to Brighton.

I have to say I'm more a motive power and rolling stock person but anything railway orientated is good for me.

You might like to read this.
https://www.bluebell-railway.com/brps/signalling-tokens/

Whether it would have stopped the Clayton tunnel crash is another matter. Semaphore or electric lights, a train can still proceed if the driver wishes.

If you spend time looking around older lines, you can usually find remnants of old signalling and the links between signals and changing of points, ground repeater signals etc.
« Last Edit: December 05, 2022, 11:48:36 pm by Zath »

Bogen85

  • Hero Member
  • *****
  • Posts: 595
Re: Railway signalling
« Reply #5 on: December 06, 2022, 01:16:50 am »
It’s completely unacceptable for trains on rails to have collisions with other trains. In general the best solution would be to have each train in each direction on a separate track.  I know this might not be possible for areas near stations.

And would be financially impractical in many areas to require 2 tracks where there currently only is one.
And would never happen in many of those areas due to the financially impracticability.

There are many things in mechanical and electrical systems that require semaphores, or something similar to enforce exclusive access.

Lock out / tag out... (to prevent electricians from getting injured or killed...) https://safetyculture.com/topics/lockout-tagout/
Power transfer switches (to prevent power line workers from getting injured or killed...)  https://www.cpwrconstructionsolutions.org/electrical/solution/988/transfer-switch-for-generators.html
Draw bridge semaphores: https://www.towerbridge.org.uk/discover/get-creative/learn-semaphore



Bogen85

  • Hero Member
  • *****
  • Posts: 595
Re: Railway signalling
« Reply #6 on: December 06, 2022, 02:56:26 am »
It’s completely unacceptable for trains on rails to have collisions with other trains. In general the best solution would be to have each train in each direction on a separate track.  I know this might not be possible for areas near stations.

And would be financially impractical in many areas to require 2 tracks where there currently only is one.
And would never happen in many of those areas due to the financially impracticability.

Especially if the train track is in a tunnel through a mountain (or under a channel of water, where a bridge is not really an option).
Requiring 2 train tracks in those cases would make many such projects cost prohibitive.

Proper semaphores would be far more cost effective in those situations.


dseligo

  • Hero Member
  • *****
  • Posts: 1177
Re: Railway signalling
« Reply #7 on: December 06, 2022, 07:41:50 am »
It’s completely unacceptable for trains on rails to have collisions with other trains. In general the best solution would be to have each train in each direction on a separate track.

Many train collisions were not head on, but second train hit first one from behind (including Clayton Tunnel crash). In those cases separate track in other direction won't help.

MarkMLl

  • Hero Member
  • *****
  • Posts: 6646
Re: Railway signalling
« Reply #8 on: December 06, 2022, 08:24:59 am »
Why the interfacing with authentic peripherals? And what kind of peripherals are we talking about?

This sort of thing https://signalbox.org/wp-content/uploads/2021/04/tunwellswestbNC15048.jpg

I find myself playing with an authentic Saxby and Farmer frame from which the original interlock bed has been removed and replaced with a much simpler later mechanism. Since it's not in a location where it's ever likely to control traffic again I'm wondering whether it would be feasible to mock up something similar to the original mechanism and then set it up with ghost signallers in (what would have been) boxes a mile or two away for demonstration purposes.

MarkMLl
MT+86 & Turbo Pascal v1 on CCP/M-86, multitasking with LAN & graphics in 128Kb.
Pet hate: people who boast about the size and sophistication of their computer.
GitHub repositories: https://github.com/MarkMLl?tab=repositories

MarkMLl

  • Hero Member
  • *****
  • Posts: 6646
Re: Railway signalling
« Reply #9 on: December 06, 2022, 08:27:01 am »
I live about 20 miles from Clayton Tunnel.
We always used to make a point of looking at it as kids when we were going to Brighton.

I used to be much closer to it than that, looking out at a line of craters that the Germans had made trying to block the Southern cutting in about '42.

MarkMLl
MT+86 & Turbo Pascal v1 on CCP/M-86, multitasking with LAN & graphics in 128Kb.
Pet hate: people who boast about the size and sophistication of their computer.
GitHub repositories: https://github.com/MarkMLl?tab=repositories

MarkMLl

  • Hero Member
  • *****
  • Posts: 6646
Re: Railway signalling
« Reply #10 on: December 06, 2022, 08:35:56 am »
In general it seems that for profit rail companies will get away with whatever they can including blaming employees.

...shut off power to electric trains until it’s safe to proceed.

I did not start this thread for it to turn into an anti-capitalist rant, and if you're sufficiently ignorant of the field that you think that electric power had any relevance to the railways in 1861 I suggest that you STFU.

MarkMLl
MT+86 & Turbo Pascal v1 on CCP/M-86, multitasking with LAN & graphics in 128Kb.
Pet hate: people who boast about the size and sophistication of their computer.
GitHub repositories: https://github.com/MarkMLl?tab=repositories

MarkMLl

  • Hero Member
  • *****
  • Posts: 6646
Re: Railway signalling
« Reply #11 on: December 06, 2022, 08:46:25 am »
Especially if the train track is in a tunnel through a mountain (or under a channel of water, where a bridge is not really an option).
Requiring 2 train tracks in those cases would make many such projects cost prohibitive.

Proper semaphores would be far more cost effective in those situations.

It was already a two-track line, but not four-track (which was extremely rare in the mid-C19th) and in any event multiple tracks- at least in the UK- are normally used for "slow" and "fast" traffic: any attempt to run similar-speed trains in parallel would have been ghastly in view of the primitive signalling of the day and would at best have led to trains queuing to get into stations.

Anyway, my interest here is in the fact that the Clayton Tunnel crash led to Saxby designing one of the earliest robust signalling systems with interlocked semaphores and points (switches, turnouts... language varies depending on what country you're in). The one I'm looking at is in a Grade-II listed building, adjacent to level crossing gates (not barrier arms) which are also Grade-II listed.

By now I've got a modest understanding of the older interlock mechanism which was ripped out, and hope to spend some time over Xmas with CAD and 3D-printer sorting out some of the dimensions for a model: some of the fiddlier buts won't show and can be replaced by stock parts from China.

But what would be /really/ interesting would be to set up a simulation equivalent to https://blockpostsoftware.co.uk/simulations.php , but using the real peripherals :-)

MarkMLl


MT+86 & Turbo Pascal v1 on CCP/M-86, multitasking with LAN & graphics in 128Kb.
Pet hate: people who boast about the size and sophistication of their computer.
GitHub repositories: https://github.com/MarkMLl?tab=repositories

PascalDragon

  • Hero Member
  • *****
  • Posts: 5444
  • Compiler Developer
Re: Railway signalling
« Reply #12 on: December 08, 2022, 10:49:46 pm »
It’s completely unacceptable for trains on rails to have collisions with other trains. In general the best solution would be to have each train in each direction on a separate track.  I know this might not be possible for areas near stations.

While a track for each direction is the optimum more often than not you have to work with what you have and especially in rural areas there might only be a single track available and adding more - especially if the landscape is more complex - it might be rather costly to add a second track or the people living there might protest against it.

Traffic signals for trains are not a guarantee of safety. Probably something mechanical that makes it impossible for trains to collide or that could shut off power to electric trains until it’s safe to proceed.

At least on continental Europe the signalling of trains not only consists of “traffic signals”. There are axle counters that count whether all of what had entered a protected block also left it again (usually employed inside stations) and the signals are protected by something that will trigger a full stop of the train if run over while showing a “stop” signal.
For example in Germany that is the PZB aka “punktförmige Zugbeeinflussung” which can be translated as “intermittent automatic train running control”: next to each signal and pre-signal there is an electro magnet that is read by the train's control system. If the pre-signal (which is usually located a kilometer in front of the main signal) signals that a signal with a “stop” symbol is upcoming it will control whether the speed of the train is low enough that it can stop at that signal. If not it will trigger a slow down of the train. If the train should pass the main signal while it shows a “stop” symbol it will trigger a full emergency stop.
In addition to this the signal operator is not allowed to change switches so that a route of a train will lead into a route occupied by another train.
The main cause of accidents with these systems is if the operator or train driver override the safety mechanism due to some mistake of the system and don't follow the proper procedures to do this. Or its due to shuting which might often circumvent these mechanisms as well.

In addition to the PZB there is also the LZB in Germany which essentially allows to get rid of the signals completely. It's usually used on high speed tracks and for example at the core track of the suburban train in Munich (S-Bahn) as well as all of the Munich subway (U-Bahn). The LZB allows to reduce the required distances between trains (due to the length of the way a train needs to stop from a given speed), for example in both cases of the S- and U-Bahn this can be as small as 80m. In addition to that in the Munich U-Bahn it also allows for automatic operation: the train driver merely triggers the train to start off and then it will proceed to the next station by itself. In Nuremberg there even already are subway lines that operate purely automatic.

Also please note that there aren't only electric trains as there are also diesel or hydrogen trains (or steam ones) where you can't simply switch off the power.

Local and express trains running on same track which will require extra tracks at stations for the local train to wait on whilst the express train’s wiz by. This definitely requires a lot of coordination of switching trains to different tracks and scheduling track usage.

This “coordination” is the bread and butter of the schedule creators (for the fixed plan) as well as the signal control operators (for quick fixes during operations). Here you can see one of the instruments that is used to plan such schedules (it's German only, but you can use Google Translate or whatever to translate it).

Why the interfacing with authentic peripherals? And what kind of peripherals are we talking about?

This sort of thing https://signalbox.org/wp-content/uploads/2021/04/tunwellswestbNC15048.jpg

That link gives a 404 ;D

I find myself playing with an authentic Saxby and Farmer frame from which the original interlock bed has been removed and replaced with a much simpler later mechanism. Since it's not in a location where it's ever likely to control traffic again I'm wondering whether it would be feasible to mock up something similar to the original mechanism and then set it up with ghost signallers in (what would have been) boxes a mile or two away for demonstration purposes.

Something like that should definitely be possible. I know of a group of people that used an old signal control from DB and hooked it up to a model railway.

MarkMLl

  • Hero Member
  • *****
  • Posts: 6646
Re: Railway signalling
« Reply #13 on: December 09, 2022, 08:53:58 am »
I'm grateful to people who made useful comments, but am locking this topic due to persistent trolling.

MarkMLl
MT+86 & Turbo Pascal v1 on CCP/M-86, multitasking with LAN & graphics in 128Kb.
Pet hate: people who boast about the size and sophistication of their computer.
GitHub repositories: https://github.com/MarkMLl?tab=repositories

MarkMLl

  • Hero Member
  • *****
  • Posts: 6646
Re: Railway signalling
« Reply #14 on: December 09, 2022, 10:10:05 am »
MT+86 & Turbo Pascal v1 on CCP/M-86, multitasking with LAN & graphics in 128Kb.
Pet hate: people who boast about the size and sophistication of their computer.
GitHub repositories: https://github.com/MarkMLl?tab=repositories

 

TinyPortal © 2005-2018