Route signalling vs. speed signalling - an oversimplification?

[Grovenor]

5 hours ago, jim.snowdon said:

As I understand it, all that a distant signal tells the driver is that, if On, one or more of the stop signals on the line of route  within that signal box’s are of control has to be presumed to be showing a stop aspect. It doesn’t tell the driver which one, so it must always be presumed that it is the next stop signal. If it isn’t that one, it must be the next one, and so on until the driver reaches the section signal.

 

Jim,

The SM just put it correctly, it is always the first stop signal that must be expected at stop, and this is enforced by Rule 39a and its successors.

There can be a different situation if leaving manual block and entering track circuit block where a distant, or its equivalent, may show a double yellow.

[micknich2003]

Following on from "Grovener;s" post. attached a couple of sketches illustrating a "Hybred" signal used at the transition from semaphore to MA Signalling. The model is one I made some years ago.

[The Stationmaster]

16 hours ago, Grovenor said:

Jim,

The SM just put it correctly, it is always the first stop signal that must be expected at stop, and this is enforced by Rule 39a and its successors.

There can be a different situation if leaving manual block and entering track circuit block where a distant, or its equivalent, may show a double yellow.

And = Rule 39a still applies as I have just checked it.

 

So the situation mentioned  by JIm was actually an Irregularity and should be reported.  It might be that for whatever reason the Section signal had been returned to danger before his train reached (or had not been pulled off for some reason) but it was still an irregularity. 

 

The only exception is if one of the stop signals is a colour light capable pf showinga yellow aspect to indcate that the next stop signal in advance is at danger.  

[bécasse]

If a distant signal is fixed, the following home signal could be on or it could be off, and, indeed, all the related stop signals could be off. A driver obviously has to approach each of them at a speed where he (or she) could stop but since they can often be seen well in advance (and the existence of the fixed distant suggests that the line is effectively speed restricted anyway) that may not slow the train significantly. Obviously if a box is switched out its fixed distant will remain at caution even though all the stop signals are off.

 

I subsequently realised that in fact splitting distants do provide a modicum of speed signalling since, when off, they do tell the driver which route he will eventually be following and his route knowledge will tell him what the speed restrictions for those various routes would be.

[Michael Hodgson]

10 minutes ago, The Stationmaster said:

And = Rule 39a still applies as I have just checked it.

 

So the situation mentioned  by JIm was actually an Irregularity and should be reported. 

Perhaps worth adding that 39a is really a procedural instruction to signalmen.  Mechanical interlocking cannot enforce it, since you have to be allowed to clear all of the stop signals before you can clear the distant.  However Compulsory Rule 39a can be achieved electrically: with power-operated semaphores or colour lights, levers can be pulled but the corresponding stop signals stay red until an appropriate track circuit has been occupied long enough for the driver to have been able to see it at Danger and reduce his speed appropriately.  As we tend to use electrical methods to power model semaphores, this is a technique that we can automate in modelling. 

 

The value of rule 39a is clear from numerous historic accident reports in places where trains normally had a clear run which say something along the lines of "Driver Bloggs passed the Distant at Caution and three stop signals at Danger ..." whereas the driver's evidence was that he missed the distant at caution because of some sort of visibility problems. 

 

If he's doing line speed and the first he knows of a train ahead of him is a semaphore stop signal at Danger it's time for clean underpants.  Drivers realising they have missed the distant would be alert to the state of the first stop signal and hope it was off.  If that other train happens to be waiting at the Advanced Starter (as was often the case), and he sees the Outer Home on, although he will probably SPAD it, he's still got a much better chance of it being a "near miss" and stopping in time or at least of significantly reducing the speed of impact.

[Michael Hodgson]

3 minutes ago, bécasse said:

Obviously if a box is switched out its fixed distant will remain at caution even though all the stop signals are off.

 

I subsequently realised that in fact splitting distants do provide a modicum of speed signalling since, when off, they do tell the driver which route he will eventually be following and his route knowledge will tell him what the speed restrictions for those various routes would be.

If the distant is fixed, it will also be at caution even though the stop signals are all off even when the box is open, so the lack of a signalman on duty makes no difference to the driver who has a clear road.

Except on a single line where token exchange might also be required, and t=if the box is switched out, he would have a long section staff.

Edited May 30, 2023 by Michael Hodgson

[Suzie]

On 29/05/2023 at 12:53, The Stationmaster said:

The angle of the Junction Indicators ('feathers') is a geographical indication of the relative positions of the routes they read to - not an indication of variation, if any, on the turnout speed to those routes.

 

As a simple example there have been, and probably still are, signals with Position 1 (top left) and Position 4 (top right) JIs  with completely different (and on some cases widely different) speeds through the relevant diverging pointwork.

I don't think that you will find many junctions where a position 1 feather (one divergence) will indicate a route faster than no feather (no divergence), and similarly a position 2 feather (two divergences) will be unlikely to be for a route faster than a position 1 feather.

 

I am merely noting this for modelling purposes, not to tell a driver how to do his job!

[big jim]

On 29/05/2023 at 12:53, The Stationmaster said:

The angle of the Junction Indicators ('feathers') is a geographical indication of the relative positions of the routes they read to - not an indication of variation, if any, on the turnout speed to those routes.

 

As a simple example there have been, and probably still are, signals with Position 1 (top left) and Position 4 (top right) JIs  with completely different (and on some cases widely different) speeds through the relevant diverging pointwork.

 

 

54 minutes ago, Suzie said:

I don't think that you will find many junctions where a position 1 feather (one divergence) will indicate a route faster than no feather (no divergence), and similarly a position 2 feather (two divergences) will be unlikely to be for a route faster than a position 1 feather.

 

I am merely noting this for modelling purposes, not to tell a driver how to do his job!

South Ruislip pre evergreen remodelling was something like no route indicator 50mph toward paddington and No1 indicator 75mph toward Marylebone 

 

bushbury jn has a position 1 toward oxley at 15mph, position 2 toward bescot at 20mph and a position 4 toward Wolverhampton at 60mph, the odd thing about that particular set of indications is the No1 indicator (facing left) will take you firstly over points that are 20mph  toward bescot but 15mph diverging toward oxley (those points are known as bushbury oxley jn) that line then curves round 180 degrees and passes below the line to Wolverhampton so ends up to the right despite getting a left hand indication! 
 

 

Edited May 31, 2023 by big jim

[5BarVT]

9 hours ago, Suzie said:

I don't think that you will find many junctions where a position 1 feather (one divergence) will indicate a route faster than no feather (no divergence), and similarly a position 2 feather (two divergences) will be unlikely to be for a route faster than a position 1 feather.

 

I am merely noting this for modelling purposes, not to tell a driver how to do his job!

Colwich Jn Dn Fast no RI 45 to Stoke, Pos 1 90 to Stafford.  Unusual, and difficult to signal but not unknown.

Higher speed for more divergent turnout with RI are much more common - just depends on the track layout: e.g Aynho Jn Up Pos 1 15 to Goods Loop, Pos 2 to Bicester 90; Stoke Gifford (Bristol Parkway) Dn Badminton Pos 1 25 to Avonmouth, Pos 2 60 to Filton.  Probably others that I can’t think of immediately too.

Paul.

[Michael Hodgson]

13 hours ago, Suzie said:

I don't think that you will find many junctions where a position 1 feather (one divergence) will indicate a route faster than no feather (no divergence), and similarly a position 2 feather (two divergences) will be unlikely to be for a route faster than a position 1 feather.

 

 

12 hours ago, big jim said:

South Ruislip pre evergreen remodelling was something like no route indicator 50mph toward paddington and No1 indicator 75mph toward Marylebone

 

Never say never in signalling - there's (nearly!) always an exception that proves the rule.

[bécasse]

7 hours ago, Michael Hodgson said:

 

 

Never say never in signalling - there's (nearly!) always an exception that proves the rule.

A few weeks ago I was examining some rather unusual signalling arrangements that had been put in place by the LSWR in 1914 and which remained in use until the line concerned closed to passengers in the mid-1960s. The most bizarre feature was an FPL, complete with bar, on a trap point at the exit from a goods line which bolted the point only when it was set for the trap! I realised eventually why it had been done, and the reasoning behind it was obviously quite clever, but it was certainly an arrangement I have never come across elsewhere.

[njee20]

23 hours ago, Suzie said:

I don't think that you will find many junctions where a position 1 feather (one divergence) will indicate a route faster than no feather (no divergence), and similarly a position 2 feather (two divergences) will be unlikely to be for a route faster than a position 1 feather.

 

I am merely noting this for modelling purposes, not to tell a driver how to do his job!

But that’s an inherently flawed assertion. It’s a bit like saying platforms serve as speed limiting devices, because lines with platform faces have lower line speeds. It may be technically accurate, but it’s not necessarily an inherent link.

[St Enodoc]

4 hours ago, bécasse said:

A few weeks ago I was examining some rather unusual signalling arrangements that had been put in place by the LSWR in 1914 and which remained in use until the line concerned closed to passengers in the mid-1960s. The most bizarre feature was an FPL, complete with bar, on a trap point at the exit from a goods line which bolted the point only when it was set for the trap! I realised eventually why it had been done, and the reasoning behind it was obviously quite clever, but it was certainly an arrangement I have never come across elsewhere.

Do tell!

[Michael Hodgson]

14 hours ago, bécasse said:

A few weeks ago I was examining some rather unusual signalling arrangements that had been put in place by the LSWR in 1914 and which remained in use until the line concerned closed to passengers in the mid-1960s. The most bizarre feature was an FPL, complete with bar, on a trap point at the exit from a goods line which bolted the point only when it was set for the trap! I realised eventually why it had been done, and the reasoning behind it was obviously quite clever, but it was certainly an arrangement I have never come across elsewhere.

Presumably somewhere you wanted to make really sure runaways went the right way, and towards which you might shunt?

My first thought was the lift shaft into the Waterloo & City, but that accident happened later!

A swing bridge somewhere perhaps?

[bécasse]

The attached downloadable pdf tells all, it doesn't make for a quick read!

boscarnejunction.pdf

[Grovenor]

Still a bit like overkill, it would work just fine without FPL 8, crossover 7 would still be held normal by 6. So still covering for a very unlikely circumstance. None of the other 5 traps being locked in this way.

[St Enodoc]

It's got me beat too but as I'm only a rolling stock engineer that's not surprising.

 

I almost follow Keith's argument, although I can't work out the "very unlikely circumstance".

 

The bit I really don't understand is "Plunger stands normally in & locks points normal only with lever reverse in frame" (Dunmere Junction GF 6/8 and Boscarne Junction GF 1/6), which sounds like a contradiction in terms.

[Jeremy Cumberland]

4 hours ago, St Enodoc said:

The bit I really don't understand is "Plunger stands normally in & locks points normal only with lever reverse in frame" (Dunmere Junction GF 6/8 and Boscarne Junction GF 1/6), which sounds like a contradiction in terms.

I took "Plunger stands normally in" to refer to the points being locked when the frame isn't opened with the tablet. "Locks points normal only" is clear enough, and "with lever reverse in frame" refers to the lever position. When not in use, Dunmere Junction Ground Frame levers 1, 6 and 8 are reverse.

 

I can't work out the reason for 8, though.

[St Enodoc]

6 minutes ago, Jeremy Cumberland said:

I took "Plunger stands normally in" to refer to the points being locked when the frame isn't opened with the tablet. "Locks points normal only" is clear enough, and "with lever reverse in frame" refers to the lever position. When not in use, Dunmere Junction Ground Frame levers 1, 6 and 8 are reverse.

 

I can't work out the reason for 8, though.

Interesting. I thought "Plunger stands normally in" meant that the point lock plunger/bolt is in the slot on the stretcher (i.e. the points are locked) when the lever is in the normal position. The idea of some ground frame levers being in reverse when the frame is not in use sounds unusual. With luck, @bécasse will put us out of our misery before long!

[bécasse]

Re Dunmere Junction.

 

I started off just as puzzled as everyone else not least because I didn't originally realise that the bolt was "in" when the road was set for the trap rather than when the road was set for access to the running line towards Boscarne Junction. It didn't help that, while I had dyeline prints of the linens for BJ box and both ground frames, the "linen" for Dunmere Junction GF was one produced by Reading S&T (and, judging by the drawing date, almost certainly produced specially in response to my request for a copy) which, unlike the other two which were updated originals from 1914, didn't contain a locking table. However, I realised eventually that the information on the two original linens for BJ box and GF contained between them sufficient intelligence to work out exactly how Dunmere Junction GF was locked - and that showed that the junction point for the Wenford Line was only bolted when set for Bodmin North (or Bodmin LSW as had been) which, of course, is perfectly normal practice for a route to a goods line that will never be used for passenger trains. That meant that bolting the trap point for the other direction would have been a nonsense but I had a photo showing quite clearly that there was an FPL complete with its locking bar so the diagram wasn't wrong. The answer proved to be that it was used to ensure that a train towards Wenford was standing well clear of the Boscarne Junction-Bodmin North running line, and the trap set, before all the levers in the Ground Frame could be restored to their habitual positions enabling the section Tablet to be released and returned (on foot) to Boscarne Junction box. The Wenford train, while all this was going on, was stood with its loco just short of the ungated level crossing at Dunmere Siding which, with a train the full permitted length, left just enough room for the guard to shut the gate behind it (the line was unfenced) when he returned from BJ box and boarded his van. The gate doesn't appear to have been lockable (although it was obviously bolted) although all the running line point levers along the branch were padlocked so the guard carried a "common" key.

 

One further oddity of this very early line was that the original Act of Parliament permitted passengers to be carried on goods trains, and right up to and including BR days one could just buy a ticket at Wadebridge booking office to travel in the brake van to Wenford and back, no permit being required.

[5BarVT]

7 hours ago, St Enodoc said:

The bit I really don't understand is "Plunger stands normally in & locks points normal only with lever reverse in frame" (Dunmere Junction GF 6/8 and Boscarne Junction GF 1/6), which sounds like a contradiction in terms.

In WR signal engineering terminology (and operations I think) it is a contradiction.  As shown in the locking table and in the latter half of the description, the FPL stands Normally Out.  But the ‘usual’ position i.e. with GF closed is FPL In.
The slots on the Boscarne Jn signals back into the two sidings are shown as Normally Off (so that the box can use them as single ended sidings when the GFs are closed) but the locking shows that to be with the lever reversed.  I wonder if that is a different use of the terminology between WR and SR.

On the SRS plan for Boscarne Jn, both GF are shown as covered GF so having levers reversed when the GF is closed would not have the same issues as with an open GF.  (There is an open GF behind Shrub Hill station which sits with both levers reverse.)

All very interesting!

30 minutes ago, bécasse said:

One further oddity of this very early line was that the original Act of Parliament permitted passengers to be carried on goods trains, and right up to and including BR days one could just buy a ticket at Wadebridge booking office to travel in the brake van to Wenford and back, no permit being required.

At first one might have thought that the potential for passengers was the reason for the bolt on the trap, but then it would have been with points reverse and 6 would have needed to both the crossover R too.

 

30 minutes ago, bécasse said:

The answer proved to be that it was used to ensure that a train towards Wenford was standing well clear of the Boscarne Junction-Bodmin North running line, and the trap set, before all the levers in the Ground Frame could be restored to their habitual positions enabling the section Tablet to be released and returned (on foot) to Boscarne Junction box.

That doesn’t quite explain the need for the FPL, the bar could have been driven by 6 to ensure clearance from the traps before closing the frame.

I wonder if it was due to gradients on the line: it appears to fall ~ 50m in c.6 miles (don’t you love mixed units!) so just steeper than 1:200.  Having an FPL adds extra confidence that the traps have made up properly and will be effective if a loaded clay train ran away down towards the junction.

 

All very interesting, thank you for posting.

Paul.

[The Stationmaster]

7 hours ago, St Enodoc said:

It's got me beat too but as I'm only a rolling stock engineer that's not surprising.

 

I almost follow Keith's argument, although I can't work out the "very unlikely circumstance".

 

The bit I really don't understand is "Plunger stands normally in & locks points normal only with lever reverse in frame" (Dunmere Junction GF 6/8 and Boscarne Junction GF 1/6), which sounds like a contradiction in terms.

To me that is a contradiction in terms.  The reference to the lever should not be needed - the FPL  bolt ('plunger') would either stand normally in,  or stand normally out.  If it stands normally in the lever which works it would be standing normal. with the FPL bolted.  If it stands normally out the lever which works would have to be pulled to its reverse position to bolt the FPL  (which was the usual arrangement on Western frames).  

 

On most small ground frames, and all that I have ever encountered, the common arrangement was for any FPL they worked to stand bolted with the lever at normal - it kept the locking nice and simple.  However I then look at George Pryer's drawing for Boscarne Jcn and Dunmere GF (which he shows as a covered GF)  and he notes that FPLs stand normally out - that obviously applies to Boscarne Jcn (and perhaps not to Dunmere GF but no reason why that should not happen in  a covered GF).  George consistently notes for all the LSWR 'boxes in that area that the FPLs stand normally out so that would appear to have been standard Stevens/LSWR practice in the early part of the 20th century and subsequently.

 

Coming to Ian's point about nomenclature alas my copy of SR Signalling Principles is filed away so not immediately to hand.  But George Pryer - who drew most of the SRS diagrams for the Southern and Western (plus no doubt others)  -was very much a Southern, South Western, man and used the terminology about the way FPLs stand in exactly the same way as we did on the Western. 

 

Similarly on the basis of the SR standard principle for the normal lie of points at a double line junction I would expect the FPL there to normally stand out because that would keep down the number of lever movements.  Whether the SR/LSWR followed that practice elsewhere I don't know but 'Railwest' of this parish might know.

 

Finally as to the use of a locking bar and FPL on the Wenford Bridge branch trap I suspect the idea was to help ensure that the point was and  remained fully closed in view of the very steep gradient  (1in50 falling towards Wadebridge at the connection).

[Grovenor]

The word normally is clearly used with two meanings in that one diagram note, hence Paul suggesting that one of them could be 'usually'. Usually is a bit vague and we don't really have another word to replace normally in this scenario. The ambiguily would have been avoided if the fpl levers had been arranged the opposite way so that they were left normal when the GF was locked up. I can't see any reason why it was not done that way. Only the slot lever needs to be left in Reverse to maintain the convention that a signal or slot lever is normal when the signal or slot is on.

My "very unlikely circumstance." is just referring to the probability of the drive for the Wenford end of crossover 7 being out of adjustment and the points thus standing open at the very time when the trap is needed for a runaway. It may well have been the gradient that drove the decision to have a bolt.

The drawings offered don't actually show whether the fpl also had a lock bar, since the points would never be locked when in use as facing points there would be no need for a bar  except for the theory advanced by Becasse. Bars are not usual for a GF as the operator is on the spot and expected to see that the train is clear. For the Main box here it appears that there is track circuit locking rather than bars.  Is there any evidence that a bar was provided for fpl 8?

 

[ruggedpeak]

I'd suggest anyone interested in trying speed signalling uses a train simulator package like Train Sim. Whilst I'm no expert on UK signalling I can drive UK trains on it ok, however I am still struggling to get my head around German signalling and systems like PZB. On one route driving an electric loco on a freight train at the same spot the train goes into full emergency stop as I think I am overspeed but I have no idea why. This video may help give a flavour....

 

 

[5BarVT]

6 minutes ago, ruggedpeak said:

speed signalling

We were a bit Off Topic in the immediately preceding posts, weren’t we!

Paul.