|
|
Post by commuter on Feb 2, 2024 7:16:18 GMT
Hoping someone can explain this for me —-
As an example, approaching Harrow-on-the-hill N.B fast you have signal JB2 / 13 / 14 which is the same physical signal on the ground but with multiple numbers depending on the route offered. There are multiple other signals numbered like this at Harrow.
Why are these signals numbered like this as opposed to the more common arrangement whereby you’d get one signal number which is would be described as, e.g. JB14 route 2?
Im guessing they’re distinct levers in the cabin/ I.M.R., as opposed to just a single points lever and a single signal lever, but why?
|
|
|
|
Post by Dstock7080 on Feb 2, 2024 7:38:31 GMT
Hanger Lane junction (District) signal is WM20/WM21 depending upon route set.
Before resignalling, Rayners Lane had route setting levers, the eastbound station starter was P32H for Piccadilly; P31H for MET; the H or L denoting whether the level should be pusHed or pulLed.
Harrow still retains route setting levers but in a different way, each lever is assigned just one route: JB2 pfm.1, JB13 pfm.3 and JB14 pfm.4, points would change automatically.
|
|
|
|
Post by PiccNT on Feb 2, 2024 9:21:42 GMT
And there are a couple at Arnos Grove EB PJ3/500 and PJ4/600 that depends on the level of protection required.
|
|
|
|
Post by tut on Feb 2, 2024 17:42:38 GMT
Why are these signals numbered like this as opposed to the more common arrangement whereby you’d get one signal number which is would be described as, e.g. JB14 route 2?
As others have said it's not necessarily more common per se, not where signals are identified by the number of the lever that works them. It's much more common across Britain's railways generally these days are it's tragically much rarer for interlockings to be based on signal levers, but where interlockings are based on signal levers, as they should be, this is not unusual at all.
So what's going on here then? Well let's start with the way things were done historically. In proper semaphore signalling you have a separate signal arm for each route. So for example here we have Shrub Hill's classic 5-armer:
On the left you have SH75 which reads to the Up & Down Branch which is to the left, in the middle you have SH26 which reads to the Light Maintenance Depot straight ahead, and higher up on the right you have SH78 which reads to the Down Main and on towards Droitwich. There's a right hand bend further on, but it's basically the main route and you can see it's higher up as well, but we don't need to worry about stepping. The point is you have three separate arms and so you need three separate levers. Now this is actually quite a good way of doing things as well, because if you want to send a train to Droitwich but you've left the points set for the Up & Down Branch then when you go to clear SH78 signal it won't come, because the route is set wrong, so it saves you from making a mistake. With ground discs they didn't always provide a disc for every route, they tended to be a bit stingy, but we don't need to go into that. But when you're setting the route from SH52 disc you have to be a bit careful because it reads over 4 possible routes and if you've left a set of points in the wrong place the train will go the wrong way and that can be very embarrassing so you don't want that.
But of course when it comes to identifying the signal it can be a little bit of a problem. Now historically, in the old days, signals would have been referred to by name, that signal at Shrub Hill is the Down Main Advanced Starting signal. That is its name and it is also its function and drivers would have been able to refer to signals by name. Plating up all semaphore signals on the national rail network is a comparatively recent development and plenty of ground discs etc. are not identified on the ground. Historically it would have been rarer, you'd have just said 'hello bobby, currently standing at your Down Main Home signal'.
Anyway, one thing that you can do, which you can see has been done at Shrub Hill, is designate a number for all three signals. So that signal at Shrub Hill is plated up as SH78 and that makes things easier for drivers. It does mean the signalman has to remember how the signal is plated up so as not to cause confusion. It's no good telling a driver to pass SH75 at danger, even if they are actually going round the branch. So the solution to that is to find some way to identify on the signal box diagram how the signal is plated up on the ground, for example by adding a blue sticker next to the appropriate number or making one of the numbers bigger or bolder.
You can see that that approach has a lot of advantages because the driver doesn't really care about which levers you're pulling. The driver doesn't want to have to refer to SH75/26/78 and nor does the signalman when he's doing a line blockage because it takes all day to get through the form. There's a bit of a downside, though, which is that signalmen and S&T and people need a way to identify specific signal arms. Also, if you're doing a line blockage and you want to block the Up & Down Branch but you want to be able to signal normally on the Down Main, then that can be done and the signal lever that you need to put a reminder appliance on is 75. 78 lever can be worked normally. The problem is the signal is plated up as SH78 and if you identify on the form that your protecting signal is SH78 that's a bit of a trap.So in some circumstances it's better to refer to the signal as SH75/26/78 and in others as SH78.
Another thing that you can do is plate the signal up and pedantically identify all the arms. You can see that this had been done at Wrawby Junction before they vandalised the magnificent signalling in the area, and it's also been done at Greenford.
I think that's a bit unnecessary, though, I don't think the driver really needs that information, but you can see the logic.
Now what's all this talk of National Rail semaphores got to do with the Underground? Well we must remember that early Underground practice certainly did make use of semaphores worked mechanically or electro-pneumatically (I think they were electro-pneumatic) and early colour light practice also sometimes continued the practice of providing separate signal heads for separate routes as seen at Arnos Grove on this page:
But what's quite interesting actually, as you'll see if you peruse Harsig's diagrams, is that often the electro-pneumatic semaphores had separate arms worked by the same lever, as we see here:
So while at Hanger Lane Junction you had WM.20 on the left and WM.21 on the right, at Ealing Common you had WM.5 A, B and C. But I should imagine this early form that signals took was where the practice of identifying a signal by all the levers that work it came from. And indeed it has been very common on the Underground. At Hammersmith (Met) you had OZ4A/3A/2A (sometimes (usually?) referred to as two three four alphas) as can be seen here, before they cut it down and replaced it with a sign
And you'll also notice on the ground the disc is separately identified as OZ29. This is completely standard practice on the Underground but a Network Rail colour light with a subsidiary signal would not separately identify the subsidiary. And indeed, going back to Shrub Hill, SH26 is a 3-foot semaphore reading to the Light Maintenance Depot and is actually a kind of shunt signal, but it is identified as SH78 on the ground along with the two main arms. Similarly SH83 Down Main Home signal has SH68 disc on the ground at the foot of the post, but the disc isn't plated up. But as I say, and as others have said, what you've identified at Harrow isn't uncommon on the Underground at all. And indeed, at King's Cross Met you even had the old OJ14E.19B. That was just one signal with just one route from it, but it was worked by two levers as that signal was either the 5th signal worked by 14 lever for trains on the inner rail or the second signal worked by 19 lever for trains crossing from outer rail to inner rail over 18 points. Then there was the old signal at King's Cross on the Piccadilly, definitely a favourite of mine:
RL.10A.10B.10C.22B/6B.6C.10D/6D |
|
|
|
Post by tut on Feb 2, 2024 19:27:26 GMT
So what is going on at Harrow-on-the-Hill in particular? Well that is a fascinating question. Let's start by looking at a proper mechanical frame and I'll show you what's going on. Here we have an excellent (albeit slightly out of date) picture of the signal box diagram in Droitiwch Spa Signal Box: photos.signalling.org/picture?/8104/category/279-droitwich_spaYou can see that one of the signals (the Down Main Inner Home, if you're interested) has two arms, 9 and 11. You can see which arm is worked by lever number 9 and which is worked by lever number 11. Two arms, two levers. If you want the train to go straight on along the Down Main to Hartlebury you will need to pull number 9 lever and if you want the train to go right along the Down Branch to Stoke Works you will have to pull number 11 lever (and obviously the points will have to be set correctly or the lever will be locked and you won't be able to clear the signal). Here you can see the actual signal itself with its two arms, DS9 higher up on the left and DS11 lower down on the right. But you'll notice that the plate on the signal post identifies it only as DS9, but it could say DS9/11, or there could be a plate under DS9 arm identifying it as DS9 and a separate one under DS11 arm identifying it as DS11. www.flickr.com/photos/jjm2009/52237703337/Now if you flick through this album photos.signalling.org/picture?/8056/category/279-droitwich_spa you can see the levers for 9 and 11 in the frame and it's quite obvious and intuitive how this all works and how it all comes about. One lever per signal. But now let's have a look at this disc here: photos.signalling.org/picture?/8011/category/279-droitwich_spaThis is not plated up but it is DS26 "disc at 32B" photos.signalling.org/picture?/8074/category/279-droitwich_spaNow there are actually two routes from this signal, you can either go left over 32 crossover reverse and up to DS9/11 or you can go straight on with 32 points normal up to DS53 disc. But there's only one disc so there's only one lever so you have to make sure you've got the route set properly beforehand or the train will go the wrong way because the driver has no indication of route. Now you can provide multiple discs, as was done quite extensively at Buxton for example: www.flickr.com/photos/jjm2009/44969950375/But that's quite an expensive luxury. Usually separate discs would be provided very much on an as required basis. But that's where our story begins and it's worth having a good picture of that practice in your mind. Now what happens when we want to make the move to power and especially to colour light signalling (assuming we want to do that and it's not clear why we would, but let's pretend for some reason we prefer fairy lights to proper signals). So what could we do? Well one thing that we could do, especially if we're still providing separate signal heads or separate signal arms (see WM.20 and WM.21 at Hanger Lane Junction also the colour lights at Arnos Grove) is we could maintain the practice of having separate levers for separate routes. This is a fabulous way of doing things for the signalman because what you do is you set the route you want and then you pull the lever for the route you want the train to take and if you haven't set the route correctly the lever will be locked. So we can see how that approach was applied at Edgware Road until tragedy struck www.flickr.com/photos/cookie68/53393676741/in/album-72177720313360954/Here you can see OP.2/3 signal and if you look at the frame www.flickr.com/photos/cookie68/53393988279/in/album-72177720313360954/You can see how number 2 lever requires no other levers reverse (particularly 13 must not be reverse). Number 3 lever, meanwhile, requires 13 (and 22) reverse. So if we want to send a train into platform 4 (Road No. 24) we pull 2. If we've accidentally left 13 points reverse it'll be locked and the train won't go the wrong way. Meanwhile, if we want to send a train into platform 3 (Road No. 23) we have to reverse 13 points (and 22 points) and then pull lever 3. And if we've accidentally left the route set for platform 4 then number 3 lever will be locked. But have a look slightly lower down at OP.8. Here one lever applies to multiple routes as you can see (shunt from outer rail to road No.23 or road No.22 or road No.21) and it's incumbent upon us to get the route set right and if it's wrong the signal will come off and we'll be done for. Now at least here the driver will get an indication of the route, but in some ways that's a bit worse for us because if the driver notices and we don't then we get all the blame, so that'll be unfortunate for us! So you can see why it might be a good idea to provide separate signal levers for separate routes in the traditional way. But it does add expense, obviously, and I couldn't tell you whether it makes the interlocking easier or harder to provide but I suspect it has its cons. So what we could do is provide just one lever per signal as was done at Edgware Road at OP.8 but as is also done with main aspect signals, e.g. at Leytonstone. From these two pictures: www.flickr.com/photos/cookie68/5976863503/in/album-72157627283838216/www.flickr.com/photos/cookie68/5976864175/in/album-72157627283838216/You can see that LF21 signal was worked by one lever: 21. If you wanted the train to go to Epping you needed 18 points normal and if you wanted it to go to Newbury Park you needed 18 points reverse. No.21 lever would come with the points in either position so you had to pay attention to what you were doing there. But you can see how both approaches make sense and in Underground practice the plate on the ground identifies any and all levers that work the signal. If two levers work it, the signal will have two numbers associated with it. If one lever works it, it will have one number. Let's look at some other situations where we can have two numbers. Let's look at Liverpool Street on the Central. You can see the box in action courtesy of a video of a forum member which is here Unfortunately video cameras in those days don't provide quite the level of detail that they do today but Harsig's diagram www.harsig.org/PDF/CentralLineEasternExtensions.pdfshows us that on the Westbound road there was an inner home signal LB19B.20B.22E. Why's that? Well if we look at the diagram you can see that to get a train out of No.1 Siding into the Westbound platform we will need to set the points (15 and 17 I would say) and reverse 19 lever to clear LB19A disc and LB19B(.20B.22E) colour light. To get a train out of No.2 Siding obviously we'll set the points and pull 20 lever and to route a train on the Westbound main it will be 22 lever which works LB22A, LB22B, LB22C, LB22D and LB22E. Why so many? Well it's a multi-home arrangement, you pull 22 lever and as a train departs the westbound platform LB22A will clear, then as the train leaving the platform gets a bit further on, LB22B will clear and so on so that a train arriving can follow a departing train nice and closely. But it's all on one lever as if you want to hold a train on the Westbound for one to come out of the siding it will have to sit at 22A. There's no way to drop a train from 22A down to 22B and keep it there. These are old interlocking requirements which are still in force. On the big railway where you have mechanical signals if a train is going over a junction you can't allow a train to approach the signal protecting the junction, you have to hold it at the one before in case of a SPAD. What you can do is bring a train up to the signal protecting the junction, hold it there, then run a train over the junction. But it's all trade-offs, I don't design interlockings, I don't know about all the ins and outs. I do know it's perfectly possible to quite simply have a situation like the one at Bethnal Green where you can see you have LC1A, LC1B and LC1C on the approach to LC6 crossover. And then at the crossover you have LC2 disc. And then after the crossover you simply have LC3, rather that LC1D.2B. But designing signalling is a dark art. Usually the witches speak to the designer and you just do what they say. Operators don't need to know why things are the way they are in that kind of detail. But there's another reason why we might want signals to be controlled by multiple levers. Let's look now at East Finchley. Now you can see here: www.flickr.com/photos/cookie68/52285422550/in/album-72177720301283770/that on the Northbound Road NP16.20 once proudly stood a sentinel of safety so much better than a silly sign with a red hexagon on it. Why are there two levers for one signal? Well 16 lever could not be reversed if 17 crossover was reverse. But if you pulled 20 lever that would come with 17 crossover reverse but the signal would only clear after a time out. So it was a way allowing a train to draw into platform 1 while a conflicting move out of platform 2 was being made, which can obviously only be done safely if the speed of the train is under control. If you clear 16 then that will lock 17 points normal and you won't be able to make the conflicting move but the train on the northbound will then not have its speed checked. So you the signalman can choose whether you want to check the speed of the train down or commit to giving a train on the northbound a clear run through. This is what's going on at Arnos Grove in a nutshell but obviously that's controlled by a computer, which is wicked. |
|
|
|
Post by tut on Feb 2, 2024 19:56:52 GMT
Now Harrow-on-the-Hill is something quite different. And it's very interesting. At Harrow-on-the-Hill you have push-pull route-setting levers. What does that mean? Well HOTH is a big, complicated location. Now traditionally to signal trains from a lever frame you would have moved all of your points levers and then you would have cleared the signal lever. Now in a big complicated location like HOTH that's quite a lot of work that you have to do so you would probably need two people to operate it. So one thing you can do is use push-pull route-setting levers. These are impressive in their way but it's not as fun as doing it yourself obviously, especially with nice mechanical points, but that's another matter. With route-setting levers the idea is that one lever movement will set up an entire route. So let's see how it works at HOTH. Let's look at signal JB61/62 on the Southbound Fast: www.flickr.com/photos/cookie68/53407530582/in/album-72177720313501913/www.flickr.com/photos/cookie68/53408786739/in/album-72177720313501913/The idea is that if you push the lever then 204 points lever moves automatically reverse and the signal clears for the Southbound Local and if you pull it then 204 points stay normal and the signal clears for the Southbound Fast. So push the lever and the train goes one way, pull the lever and the train goes the other way, no need to set the route up yourself. So how does that work? Well it's rather fun actually. At Harrow you have three IMRs. You have Harrow South www.flickr.com/photos/cookie68/albums/72177720313953834/Harrow North www.flickr.com/photos/cookie68/albums/72177720302419765/And Harrow Station. The IMRs are conventional lever frames with conventional locking. You have separate points levers and separate signal levers and the signal lever cannot move until the points levers are all in the right position. Now they can be worked manually, but normally they are worked by compressed air and what happens is the route-setting levers in the signal box start in the mid position. If a lever is pushed (or pulled) then that will send electrical signals to the slave frames in the IMRs that will move the points levers as required (using compressed air) and then move the signal levers and it's all controlled by electronics. Now electricity is the work of the Devil so the ins and outs of that I can't really explain, but that's what happens. When the train goes over the route the route setting lever has to be moved back to the mid position and when that happens everything is reset and the points and signal levers go back to normal. Then you can push or pull the route setting lever again and everything moves again. This is a bit inefficient that you can't just restroke the route, but I understand it's how it has to be done. So everything gets reset after the passage of every train. Now what's fun is the station slave frame is at the far end of the lever frame in the main signal box. So as you work the frame in the box the levers in the station slave frame moved backwards and forwards by compressed air, as you can see in this video Now the advantages of having slave frames like this is not only does it reduce the amount of lever pulling the signalman has to do, but it can also be a way of achieving remote control. For example, Ruislip Gardens was controlled from West Ruislip Cabin when the Central line was extended along there and it was worked by push-pull route levers. And the idea was that all of the safety interlocking was there at Ruislip Gardens and all of the cabling from the slave frame to the equipment on the ground was really expensive high quality safety critical communications cables. But all of the cabling form West Ruislip Cabin to the slave frame at Ruislip Gardens could be cheap telecomms cabling because the safety stuff was there on site. If you'd had to run high-quality cabling from West Ruislip Cabin to points and signals at Ruislip Gardens it would have cost a lot more. Interestingly at Grange Hill they also had a slave frame and it was usually remotely worked from Hainault cabin, but I believe route setting levers weren't provided and every lever at Grange Hill slave frame was duplicated in Hainault cabin. Anyway, to answer your question about numbering you can look at Harsig's diagram for HOTH and you can see how it all works www.harsig.org/PDF/HOH19621996.pdfNow at HOTH they used quite an interesting numbering scheme in that they numbered the pushes with odd numbers and the pulls with even numbers. So the actual physical lever No.1 is numbered 1 (push) and 2 (pull). Physical lever number 2 is numbered 3 (push) and 4 (pull) and so on. Signal levers in the slave frames are numbered to correspond with the push pull levers. So you can see levers 1 and 2 in the slave frame at Harrow South here: www.flickr.com/photos/cookie68/53453444246/in/album-72177720313953834/Points and slots seem to be numbered separately. This is actually quite an unusual way of doing things and it wasn't done historically. D Stock refers to Rayners Lane and you can see the old cabin here: www.flickr.com/photos/cookie68/53478344949/in/album-72177720301507189/www.flickr.com/photos/cookie68/53478172028/in/album-72177720301507189/In those days you had 31H (31 Push) and 31L (31 Pull) and then in those days 31 lever was 31 lever, it wasn't 61 push 62 pull. But times change. There's a great little collection of photos of High Street Kensington here: www.flickr.com/photos/cookie68/albums/72177720301142221/And a whole rabbit hole of diagrams on Harsig's website. There, aren't you glad you asked now? |
|
|
|
Post by zbang on Feb 3, 2024 6:05:59 GMT
I'm glad you answered! (This is the sort of thing I came here for.)
|
|
rincew1nd
Administrator
Junior Under-wizzard of quiz
Posts: 10,225 
|
Post by rincew1nd on Feb 3, 2024 19:47:03 GMT
There, aren't you glad you asked now? YES |
|
Tom
Administrator
Signalfel?
Posts: 4,100
|
Post by Tom on Feb 4, 2024 11:59:01 GMT
I mean I could have answered it along the lines of "it's an older practice where one lever was allocated per route, rather than one lever for all routes", but that would be too simple!
|
|
|
|
Post by tut on Feb 4, 2024 14:05:28 GMT
I mean I could have answered it along the lines of "it's an older practice where one lever was allocated per route, rather than one lever for all routes", but that would be too simple! It's easier to get carried away when you're on annual leave  But could you tell us why one way (one lever per route) might be preferred over another (one lever for all routes)? Obviously with mechanical signalling each signal arm is worked independently by its own lever. Also
So you can see why it might be a good idea to provide separate signal levers for separate routes in the traditional way. But it does add expense, obviously, and I couldn't tell you whether it makes the interlocking easier or harder to provide but I suspect it has its cons. So what we could do is provide just one lever per signal as was done at Edgware Road at OP.8 but as is also done with main aspect signals, e.g. at Leytonstone.
It occurred to me after I'd finished that of course with earlier forms of GWR locking conditional locking could not be applied and if you wanted two routes from a signal you had to provide two arms or two discs. In early GWR practice discs would and could only be lowered if the points they were at were reverse. Discs could be passed in the on position on the driver's own authority if a disc further up the road came off. So for example at Newland East:
15 disc could be lowered only if 14 points were reverse. If you wanted to signal a train from 15 disc to the Down Refuge Siding, you would pull 11 points reverse, lower 12 and the driver would pass 15 at danger on the authority of 12. Such discs were distinguished by being provided with white lights rather than red ones.
So with certain kinds of locking anyway, it is easier to provide locking if separate levers are used for separate routes. I don't know very much about LU mechanical locking though. To provide the conditional locking at Newland East (15 is released by 12 with 14 normal or by 14 reverse) required the use of a slider.
Personally I quite like having one lever per route, it's a nice little line of defence against a careless wrong route, but I would have imagined generally companies would have wanted to save the cost of additional levers in the frame.
Those who are especially interested may enjoy as much as I did this article by Piers Connor which the LURS has generously made available to download
It reminded me that the earlier push-pull route setting frames at Rayners Lane and High Street Kensington (and elsewhere) did not make use of mechanically interlocked slave frames like HOTH, but instead made extensive use of relay interlocking. The mechanical interlocking that was provided was seemingly rather secondary. I'm not altogether sure I approve of that, but LU certainly had an intriguing history with relay interlocking, being apparently very early adopters before seeming to move back to mechanical locking with their V frames persevering with good, reliable mechanical interlocking into the nineties if I'm not mistaken.
|
|
Tom
Administrator
Signalfel?
Posts: 4,100
|
Post by Tom on Feb 4, 2024 16:41:58 GMT
It basically goes back to semaphore days, and that principle continued until around the Second World War by which time there were no new semaphore installations. The scheme design for Harrow was more or less complete pre-war.
As I understood it, the push-pull frames were mechanically interlocked - Harrow certainly is! Mechanical locking was provided at the Deterflex slave frame sites for use in 'local control' mode.
Bit of an update now I've had some sleep - Using multiple levers probably reduced the amount of both ways locking at diverging junctions but then there would be a need to prevent the two (or more) signal levers being reversed at the same time - but that would probably come for free using signal to point locking.
Thinking of Hanger Lane Junction (pre 2015), it used to be that that 20 x 9 and 21 x 1.➈ but if we were to use one lever it would need to be something like 20 x (1w➈).9 ➈ which requires a degree of additional conditional locking. Not the most complicated but still more complex than the use of two levers.
Of course now we have powered points on the EB too the locking is now 20 x 9, 21 x ➈, and 9 x 10. The need to provide 21 x 1 is not necessary as 1 x ➉ therefore the locking 21 x 1 is provided indirectly via the point to point locking.
Had 20 provided both routes the locking would be 20 x 9 ➈ and 9 x 10.
|
|
|
|
Post by rdm on Feb 4, 2024 22:25:10 GMT
This thread is most interesting! Having read all of Piers' articles in that London Underground Signalling series I do wonder whether there is any possibility of these being put together in book form - in appreciating that this is quite a 'niche' subject I for one would be willing to pay for such an issue outside the LURS subscription if it was possible to produce the articles in this format (as has been done with the series on the trains themselves).
|
|
|
|
Post by tut on Feb 4, 2024 23:50:01 GMT
Bit of an update now I've had some sleep - Using multiple levers probably reduced the amount of both ways locking at diverging junctions but then there would be a need to prevent the two (or more) signal levers being reversed at the same time - but that would probably come for free using signal to point locking. Thinking of Hanger Lane Junction (pre 2015), it used to be that that 20 x 9 and 21 x 1.➈ but if we were to use one lever it would need to be something like 20 x (1w➈). 9 ➈ which requires a degree of additional conditional locking. Not the most complicated but still more complex than the use of two levers. Of course now we have powered points on the EB too the locking is now 20 x 9, 21 x ➈, and 9 x 10. The need to provide 21 x 1 is not necessary as 1 x ➉ therefore the locking 21 x 1 is provided indirectly via the point to point locking. Had 20 provided both routes the locking would be 20 x 9 ➈ and 9 x 10. That's exactly how I would expect it to be done on Network Rail with tappet locking. So going back to Droitwich I don't have the locking table in front of me, but I expect that there is no locking between 11 signal lever and 73 signal lever at all. Instead 43 will lock 73 normal and 73 will lock 43 normal because that's required anyway, you need 43 normal for the route to be correctly set beyond 73 anyway. Then 43 will release 42 so you won't be able to pull 42 points unless you've got 43 reverse, which will lock 73 (which it will have to do anyway because the route would be set wrong). With 73 reverse that locks 43 normal (which it has to do anyway) which locks 42 normal which locks 11. Typically on Network Rail as far as locking between signal levers goes, distant signals are released by the associated stop signals (and lock the associated stop signals reverse so in an emergency you have to remember to get the distant back first), opposing locking will typically be provided to prevent trains being signalled towards each other head on (e.g. between the discs on either side of a crossover, e.g. 26 and 33 at Droitwich), locking is provided between signals that read in opposite directions (e.g. 33 and 9/11 at Droitwich - if you're between those signals either you're going back over the crossover via 33 disc or you're going forwards towards Hartlebury/Stoke Works via 9/11, you can't do both, I know such locking is often provided on LU as well, although I don't find it that necessary and it's not often seen with modern colour light installations) and you've got sequential locking which is often electric. But conflicting routes over junctions and so on are not normally prevented by locking between signal levers. That all applies to standard traditional mechanical interlocking. Once electricity and colour lights and relays enter the picture things will be different I'm sure. Which brings us back to the early push-pull frames. At Harrow, as far as I understand it and I could be mistaken, the slave frames are classically mechanically interlocked, with standard mechanical locking between points and signal levers (although I note with interest that the points levers are not checked mid stroke until the points are detected, the points lever will reverse, the signal lever will reverse and it seems the locking is essentially done on the aspect, by which I mean if the points don't detect the aspect will be held at danger, even with the appropriate points and signal levers reverse, we do have that sort of thing on Network Rail and it was a factor in a derailment at Knaresborough which is a topic for another time). You then have a kind of compressed air signalman operating the slave frames in response to the movement of the push pull levers. I understand places like Rayners Lane were a bit different. I don't think Rayners had any points levers at all, I know Elephant & Castle didn't. The route levers called the points and then the signal was lowered subject to the points gaining detection, with all of that locking performed by the relays and no locking between points and signal levers because there weren't any points levers! At High Street Kensington you can see in the pictures That points levers were provided as a fallback, but you can see that they were normally in the mid position, so during remote or push-pull operation they don't seem to have had any role to play in the interlocking. And then further Piers Connor quotes Robert Dell in his article saying
So from that I understand there was mechanical interlocking between the push-pull levers but as soon as the train had taken the route you could put the route lever back and pull a conflicting one. Now it wouldn't be possible to pull that conflicting one until you'd put the original one back, and the original route would be held until the passage of the train over the track circuits released the route behind it, but you can see what Connor means when he writes
Even if I think the statement 'didn't perform the basic safety function as it had no backlocks' might be putting it a bit too strongly and I didn't mean to imply there was no mechanical interlocking, merely that well indeed the early frame at Elephant & Castle didn't even have points levers to interlock with signal levers, but then there was a bit of a move back from that with the standard frames on the Central Line extension and then LU's V frames, the one for Elephant I believe being commissioned in 1991? by which time geographical relay interlocking was being superseded on BR by SSI. Obviously we don't approve of that, this is not to disparage LU, I merely find it interesting.
This thread is most interesting! Having read all of Piers' articles in that London Underground Signalling series I do wonder whether there is any possibility of these being put together in book form - in appreciating that this is quite a 'niche' subject I for one would be willing to pay for such an issue outside the LURS subscription if it was possible to produce the articles in this format (as has been done with the series on the trains themselves). I wholeheartedly agree. Sadly I don't feel I would benefit very much from a subscription to LURS, but I would certainly be prepared market rates for such a book. His similar book on the trains themselves that you mention was an excellent read. |
|
