Jump to content
 

Mechanical Interlocking - How to learn the Lost Black Art?


Recommended Posts

Hi @Titanius Anglesmith - good point for the ground signal at 11, thank you - added here as 21.. As for the 8/13/x arrangement, I think I've got what you mean here:

image.png.6b7cf85a2f2cbee7d65dcb201c27b9bd.png

 

Ground signal 8 set to a yellow since at caution it will still permit access into the headshunt and goods shed at y. Presumably 1,2,3,4 and 18 would be on a signal gantry across the tracks, and the longer platforms (1 and 4) would have calling on arms, and the platform starters would have shunt signals (in the pre-group era) to permit movement out to 18.

 

That particular formation in 'real life' looks like this, with the standard 4 tank on P1 and the two sidings off to the bottom-right of the photo:

 

HZBqrX7.png

 

Originally I had depicted ?? as a double slip  - the headshunt providing a trap for both the loco and goods shed roads. In reality I found while signalling the plan that there is no need for a route from P1 into the headshunt over y, since the goods siding was only long enough for 2-3 wagons and the engine shed lead would be fine for that, so I changed it to a barry slip (where the two curved roads never join) - however if there's a compelling reason to choose one over the other I'm all ears.

 

 

Edited by Lacathedrale
Link to post
Share on other sites

Whether 13 is a Barry slip or a double slip you have a problem there, since Platform 1 isa passenger line it has to be trapped from the sidings and engine shed as any move over the slip will foul the passenger line. You can use the goods shed point as a trap hence it has to be worked from the box and a signal provided similar to your original design. The engine shed road needs a new worked trap and signal. the goods siding signal can be a yellow, the engine shed signal has to be a red. Personally I would reverse the normal lie of 10 but that can be done either way.

So far as the locking goes have another read of your green books. Always do the point locking first then add the signals so you are not duplicating lots of locking. It helps also, if you want it prototypical, to show facing point locks, you need a lot of them.

  • Agree 1
  • Informative/Useful 1
Link to post
Share on other sites

Thanks Keith, definitely more to learn - such simple rules feel so bloody complicated when it comes to a real implementation! Thanks for the tip on sorting the locking of points first, I'd been going by signals locking opposing signals first!

 

EDIT: thanks for the diagram, why did you change the lie of 10?

 

Any particular distinction between your recommendation of implementing 21 and 22 as box controlled and Mike's suggestion of hand signaled and hand-levers?

Edited by Lacathedrale
Link to post
Share on other sites

  • RMweb Gold

Right - needs a rather different treatment then -

1.  The engine shed sidings needs its own trap point plus exit ground signal.

2.  The headshunt outlet point 'Y' ill also need to act as a trap

3.  I'm pretty sure that the SECR did not use yellow arm discs!

 

I can see nothing wrong with going for a Barry slip if that is what you fancy and gives you the movements you need but think very carefully about the movements you need t work the area before you make your decision.

 

With your interlocking I would be very strongly inclined to go for the advantages offered by first getting your interlocking between various points sorted before you start adding in the signal locking because getting the points settled first will reduce the amount of locking needed for the signals.

Link to post
Share on other sites

1 hour ago, Lacathedrale said:

Thanks Keith, definitely more to learn - such simple rules feel so bloody complicated when it comes to a real implementation! Thanks for the tip on sorting the locking of points first, I'd been going by signals locking opposing signals first!

 

EDIT: thanks for the diagram, why did you change the lie of 10?

 

Any particular distinction between your recommendation of implementing 21 and 22 as box controlled and Mike's suggestion of hand signaled and hand-levers?

Lie of 10 just personal preference as I said could work either way.

Traps are neccessary to protect the passenger route into platform 1. The Barry slip or double slip cannot provide protection as all moves through them conflict. See Stationmasters latest post agreeing with me.

I suggested yellow ground signal for the principle and as that is what the green books show. If you find what your chosen comany and period used for that function then use that, could be as simple as a point indicator or as complex as using a double arm signal.

 

I would suggest sorting the numbering before doing a locking chart, I just used vacant numbers which are not optimal for a frame.

Link to post
Share on other sites

@LNERGE you're absolutely right - at least from my uneducated eye.

 

Would I need a subsidiary signal under 1,2,3,4 to signal movement from the up line into the ES, or would I assume the pilot would simply move into a platform road and back out onto the down main and use 8?

 

I'm just having so much trouble trying to figure out the order to write this locking down - I feel like I just can't get it :( Some parts seem easy, and others I just get up all in a twist. I'll sort the numbering now.

Link to post
Share on other sites

  • RMweb Gold
2 minutes ago, Lacathedrale said:

@LNERGE you're absolutely right - at least from my uneducated eye.

 

Would I need a subsidiary signal under 1,2,3,4 to signal movement from the up line into the ES, or would I assume the pilot would simply move into a platform road and back out onto the down main and use 8?

 

I'm just having so much trouble trying to figure out the order to write this locking down - I feel like I just can't get it :( Some parts seem easy, and others I just get up all in a twist. I'll sort the numbering now.

I'm no expert on the signalling for that company but I think there would almost certainly have been some sort of 'reduced status' arm to read to the engine shed siding.  Incidentally yoi need another ground signal at the toe of No.9 points in the Down Line.

 

Whay t I would advise you to do before anything else is to think through all the likely movements and check you have provided signals for them

  • Agree 1
Link to post
Share on other sites

Thanks @The Stationmaster . My interpretation of ground signal 8 is that it could read through to any platform road,, and that 9 is used only by P4 to access the down, since the distance between 9 and 6 is too small to hold a train of any sort?

 

These are the routes that I've figured out:

 

image.png.0fa2d824d6241007001e2a6293f3254e.png

 

I have included any turnout which the route traverses, is that about right? What is the next step to translate this into a locking table?

Edited by Lacathedrale
Link to post
Share on other sites

On ‎23‎/‎09‎/‎2020 at 10:15, The Stationmaster said:

One question that came up when Mick and I were talking yesterday was about mechanical sequential (aka rotation) locking and does any still exist anywhere on the national network?  I'm fairly certain that there is none left on any Western frames. (as in WR as it stood at the end of its life).    Although I know, according to drawings, there was still some around in the 1970s but in general full sequential locking seems to have been a later addition at many (most?) Western 'boxes so it had been done electrically.   In a few cases some tappet locking  had been added to older frames to provide sequential locking.

 

Mike,

Just checked the /3's for ex WR boxes, and Droitwich, Lostwithiel, Malvern Wells, St. Erth and Worcester Tunnel Junction still have MSL with sprung nibs.

Its very prevelant in mechanical form on ex LMS type frames, but relatively rate on Eastern frames that I now have.

Hope that helps,

Martin

  • Interesting/Thought-provoking 1
Link to post
Share on other sites

  • RMweb Gold
On 29/09/2020 at 11:16, Lacathedrale said:

Thanks @The Stationmaster . My interpretation of ground signal 8 is that it could read through to any platform road,, and that 9 is used only by P4 to access the down, since the distance between 9 and 6 is too small to hold a train of any sort?

 

These are the routes that I've figured out:

 

image.png.0fa2d824d6241007001e2a6293f3254e.png

 

I have included any turnout which the route traverses, is that about right? What is the next step to translate this into a locking table?

What you have written in effect is the lever leads, i.e. which levers have to be pulled in order to release a particular lever.  So you can use all the ones shown in the pink squares to do the 'Released By' column of a locking chart.  You can use all the ones with the green squares to do part of the 'Locks' chart column.    What you haven't done is sort out the locking between points first.

 

(Using your chart)  Line 20 Up to ES would be Released by 5, 10, [21 an error I think, 21 should be green] and 23.  And it would lock 11.

But if you locked the point levers first then 5 would lock 11 -, so no need for Route 20 to lock 11.  Similarly 13 would lock 21 so, again, no need for Route 20 to lock 21.  And 23 would be released by 13 so there is no need for 13 to release Route 20.

 

Taking another example for Route Line 3, Up to Platform 2 if you do the point locking first Point Lever No.6 would lock Lever No. 5 and Release Lever No.7. (and No.7 could lock No.12). So the lever for your Route No. 3 need only be released by Lever No.7.  Again it shows the advantage of getting the interlocking right between point levers done first using the simple rule that point levers should interlock each other to prevent conflicting routes being set.

 

For your point locking Lever 5 would lock levers 6, 9, & 11 (and conditionally Lock 12 with 10 Normal but taht will be influenced by the distances between signals and any need for operational flexibility)

point lever 6 would lock levers 5, 9, & 12.  And would release Lever 7 ....   and so on

 

PS One thing I would recommend is revising the normal lie of No.10 points to the way it was originally drawn, i.e. towards Platform 1.  The reason for doing this is to avoid a conditional lock because as the revised sketch stands 7 will lock 11 with 10 normal. (and vice versa of course).  If the lie of 10 is altered back to the original 7 will simply lock 10 with no conditional element.   Thus the parallel move facility is preserved with simpler locking

Edited by The Stationmaster
To add PS re normal lie off points 10.
Link to post
Share on other sites

i havent got these myself, but ive noted 3 articles about lever frames and interlocking in my club library. sorry about not knowing the month issue, i only noted by page number

RM 1988 p536 Lever frame interlocking

RM 1996 p311 making a lever frame

 

model railway news 1960 p222

Edited by sir douglas
  • Thanks 1
Link to post
Share on other sites

Quote

PS One thing I would recommend is revising the normal lie of No.10 points to the way it was originally drawn, i.e. towards Platform 1.  The reason for doing this is to avoid a conditional lock because as the revised sketch stands 7 will lock 11 with 10 normal. (and vice versa of course).  If the lie of 10 is altered back to the original 7 will simply lock 10 with no conditional element.   Thus the parallel move facility is preserved with simpler locking

There is no need for conditional locking either way, if 7 can lock 10 it can equally be released by 10. Its a matter of personal preference but I agree putting it back as was saves a lever pull in that case, but will cause an extra one elsewhere, swings and roundabouts.

 

I would still recommend redoing the numbering before getting into the locking, the extra points and signals were just added on the end which is not really the right place.

And I think I would put the double slip back rather than the Barry slip, as it is any access to the goods shed involves shunting via the engine shed which could be very awkward.

 

Edited by Grovenor
Link to post
Share on other sites

  • RMweb Gold
12 hours ago, Grovenor said:

There is no need for conditional locking either way, if 7 can lock 10 it can equally be released by 10. Its a matter of personal preference but I agree putting it back as was saves a lever pull in that case, but will cause an extra one elsewhere, swings and roundabouts.

 

I would still recommend redoing the numbering before getting into the locking, the extra points and signals were just added on the end which is not really the right place.

And I think I would put the double slip back rather than the Barry slip, as it is any access to the goods shed involves shunting via the engine shed which could be very awkward.

 

I'm a bit confused there sorry.  If 10 lies normal towards Platform 2 then either 10 has to release 7 (which seems a bit odd because it isn't in the logical route)  or 7 has to lock 11 with 10 Normal.  If the lie of 10 is changed then 7 locks 10 (and vice versa) so the only levers pulled are those in the direct route  and it is easier to learn the lead because it is all logically in the route being set.   Exactly the reason why some lever frame leads were the subject of work study assessment in the late 1950s although a locking designer would normally hopefully go with a logical series of pulls.

 

And yes, the numbering really should be sorted before anything else happens although what amounts to a route locking table drawn up by Lacathedrale remains valid with - at that stage - only the point lever numbers to revise.

 

 

Link to post
Share on other sites

 

I have reviewed the IRSE green book on lever numbering and have seen these rules:

  • The numbering layout should as far as possible be arranged that signalmen are pulling levers in ascending or descending numerical order
  • All levers for the same routes should be as close together as ppossible.
  • Signals: to be placed in the centre reading outwards, or the ends inwards (rather than interspersed), except for ground signals, which should be next to points through which they read.
  • Point levers should go from low numbers left side (up), High numbers right (down)

THis is what I have come up with.  I've managed to simplify my plan somewhat, so hopefully that should simplify the signalling requirements:

 

image.png.d77b77ffa16eb0c5f9ec309150666865.png

 

Home signals 1-3 read through to P4, 3 and 2 respectively.

Crossover 4 is only used for a) shunting movements and b) departures, but its position means it's either the first lever pulled in an 'up' shunting move, or the last lever pulled in a 'down' departure.

5 and 6 are used for arrivals to P4, 3, and 2 and so are next.

The outer throat lane levers 7-11 are bisected by ground signal 9 on one side, which handles movements from the up line, and capped by ground signal 11 which handles movements from the down line.

I did not think it made sense to insert the  coal stage headshunt or the engine shed trap into the middle of the running line levers, so those are 12-15.

16 is a shunt signal from P4 down to the up main,  no other levers are involved and it's in the same direction of the starters, so I put it here.

Platform starters count down from the other end of the frame so are 20, 19, and 18.

 

Before I get into routing and locking - I just want to clarify, that I think in the era I'm modelling that the starters would have S shunt signals. Whether 17 and 9 would be as such I'm not sure.

 

Link to post
Share on other sites

>>>16 is a shunt signal from P4 down to the up main.....

...but 16 is listed as a spare :-)

 

>>>I think in the era I'm modelling that the starters would have S shunt signals....

 

Not sure what you mean by that? or why?

Link to post
Share on other sites

  • RMweb Gold
21 hours ago, RailWest said:

 

 

>>>I think in the era I'm modelling that the starters would have S shunt signals....

 

Not sure what you mean by that? or why?

The only reasonably indicative box diagram I can find is Cannon Street and there the platform starting signals did have subsidiaries - but they all appear to have been there to read to engine sidings.

 

23 hours ago, Lacathedrale said:

 

I have reviewed the IRSE green book on lever numbering and have seen these rules:

  • The numbering layout should as far as possible be arranged that signalmen are pulling levers in ascending or descending numerical order
  • All levers for the same routes should be as close together as ppossible.
  • Signals: to be placed in the centre reading outwards, or the ends inwards (rather than interspersed), except for ground signals, which should be next to points through which they read.
  • Point levers should go from low numbers left side (up), High numbers right (down)

THis is what I have come up with.  I've managed to simplify my plan somewhat, so hopefully that should simplify the signalling requirements:

 

image.png.d77b77ffa16eb0c5f9ec309150666865.png

 

 

 

The plan reveals a slight problem - where does 17 read to?  Sorry for not mentioning it earlier but It would appear that you need a trailing crossover in the station approach otherwise shunting trains and engines out of Platform 4 is going to be a considerable cause of congestion as it will effectively block the incoming line for an extended period.

 

As it stands the interlocking between point levers (and only between point levers) would be as follows -

4 - Locks 5 .10

5 - Locks 4.  Releases 6

6 - Locks 10 with  7 normal

7 - Locks Nil. Releases 12 (because of the very short distance between 7 and 8).

8 - Locks 14. Releases 12

10 - Locks 4. 6 with 7 Normal

12 - Locks Nil

14 - Locks 8

 

Some of the signal k leads will very straightforward, e.g 5 leads 2;   5.6 leads 3; 10 leads 19  - and so on.  But some will be a bit more complex,  The lead for 9 would NIL OR 4 OR 5 OR 5.6 OR 4.7 OR 4.7.8.12 (which gives two routes to Platform 2 which could be sensibly reduced to only one, via 5.6 ) but there will be an awful lot of unavoidable conditional releases via lever 4.  Some of this could be avoided by using soem sort of shunting subsidiary arms beneath signals 1/2/3  in order to reduce the number of routes from ground signal 4 9 - it really  depends how busy you envisage the terminus being or will you be relying on lots of hand signalling for shunting moves?

 

 

Edited by The Stationmaster
  • Friendly/supportive 1
Link to post
Share on other sites

Quote

And I think I would put the double slip back rather than the Barry slip, as it is any access to the goods shed involves shunting via the engine shed which could be very awkward.

I said this because your plan showed a goods shed, now your latest plan just shows a coal stage, in which case the Barry slip is less of a problem.
 

Quote

 

12 - Locks Nil

14 - Locks 8

 

14 locks 8  I don't think is enough, 14 should lock 7 as well to block a fouling move.

12 released by 7 and 8 prevents a move to the coal stage headshunt. Needs to be 12 released by (7 and 8) or 14.

 

William,

As I tried to say above I now agree with SM that 7 should have its normal the other way as you originally had it. I think losing the runround on platforms 2-3 is a sensible idea but you should re-instate the crossover out of platform 4 for shunting and to allow departures. And it looks good on your photo.

  • Agree 1
Link to post
Share on other sites

Thanks chaps - I was mistaken - the 1896 OS grid shows a building adjacent the engine shed with track running into it - but the 1874 plan issued by the railway company shows it's a lamp hut and  the buffers stop well ahead of it - the gap between that road and the adjacent engine shed road has a square marked 'coal stage' - apologies for any confusion, it wasn't intentional! See below for a picture.

 

@RailWest I wrote '16' in my text but I did mean 17. @The Stationmaster also - platform 4 and shunt signal 17 reads back onto the up main. I guess it should be on the opposite side of the track to read that way, sorry! It absolutely would cause a bunch of congestion, see the real plan below. Please let me review your info re: locking and I will reply - thank you!!! There is no ground signal 4, but I assume you mean 9? three starters and four shunting signals would make a good deal of sense!

 

image.png.5c715cc4b389f1d8481ba9da470b02bd.png

 

This is the original 1874 plan which shows a lack of runarounds and the arrival/departure only platforms - albeit with six roads (three bidirectional) instead of four roads (two bidirectional) of my plan.

HV_plan_1874.png

 

By 1896 the crossover I have chosen to omit between P3/4 on my plan (P5/6 on the real plan) was added, as was extending the outer lane of the throat through a slip to the up main to allow bidirecitonal running of all lines. But in this case, inefficiency is interesting as long as it's not wildly unrealistic so I've chosen to adopt the slightly earlier pattern.

 

 

Edited by Lacathedrale
Link to post
Share on other sites

  • RMweb Gold

If you replaced 4 with a trailing crossover, you could then have all shunt movements using the down main (rather than those involving p4 having to use the up), no need for shunt 9, and (if it were real) two fewer FPLs

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...