Signalling a mixed gauge location

[Pacific231G]

A friend came up with a layout suggestion the other day and I'm intrigued to figure out how it might be signalled under British rules.

 

The idea is that a narrow gauge public railway (so signalled to light railway rules)  shares a section of mixed gauge track with a standard gauge branch. Each railway has its own station at one end and there is a junction at  the other end where the two go their separate ways. The two stations are separate so there are no mixed gauge running lines actually in the station areas. Discuss!

 

 

This situation was fairly common in other countries- the mixed metre/SG line between Noyelles and St. Valery on the Baie de Somme and the former line of the Vivarais at Tournon come immediately to mind and I know how these were signalled. I can't think of any examples on public lines in Great Britain, though there may have been one or two in Ireland where the same BofT signalling rules would presumably have applied.

 

How would you signal such a situation? The assumption is that the NG line was there first but there was no room for a separate SG line in a very narrow location.  I assume that there would be a signalbox at each end of the mixed gauge section (probably with Tyers token machines)

I'm sort of looking at Ft. William- Mallaig Jct. as a model for a single track terminus with a junction between two separate single track lines  some way up the line.   The track is only mixed between the station and the junction so the separating "turnouts" at each end of are fixed. Either railway though may have to use the start of the mixed gauge section as a head shunt. (It would only be the left hand station end that would actually be modelled)  

 

Any ideas ?

Edited March 18, 2019 by Pacific231G

[PaulRhB]

It would probably be signalled as if it was all one gauge although the interlocking would take the split of gauge into account in later periods. If there are some single gauge sidings it would probably have something to detect which gauge track was in use, track circuit or treadles to prevent derailments or it might have a short continuation of the other gauge so they can stop in the braking distance. 

 

[Joseph_Pestell]

How long is the dual-gauge section of track, David? It would need to be quite long (as at Tournon) to justify two boxes. St Raphael, where the dual-gauge was just the length of a bridge, seems a more likely scenario. Quite a few of those shown in the 1904 PLM diagrams book.

 

I'm with Paul that if there are two boxes, it is signalled just as any single track line would be.

 

Interlocking is interesting though because, as Paul says, it needs treadles or track circuits as there are no moving parts on the turnouts that require a lever that can be interlocked with the signalling.

[Oldddudders]

The southern approach to Wolkenstein, in Saxony, had a section of dual gauge track, and where the 750 mm Preßnitztalbahn joined the DR standard gauge there was indeed a signalbox. A further box was located at Wolkenstein station to control the complex layout of standard and narrow gauge sidings and through lines.  

[Pacific231G]

Thanks for the suggestions. 

 

This is a British layout so taking Ft. William station SB to Mallaig jct. SB as an example, that was 1560yds so the block section was a little under a mile.  A shortish bridge section might have been controlled by a single box but I'm assuming something long enough to have warranted a box at each end. 

I agree that the mixed section itself would be signalled like any section of single line track but  there wouldn't be any danger of derailments from trains trying to take the wrong gauge track as the fixed crossings wouild always keep them on their own gauge (That might not be true if you had a set of points for one of the gauges within the mixed section but that doesn't apply here)

 

However, with no moving points there would be a theoretical danger of a wrongly set signal clearing  a train to the wrong  line. For example, if the entrance to the station was controlled by a pair of home signals it would be possible to clear the NG signal and for an SG driver to take that as permission to proceed even if it was the NG station that was actually clear. I guess you'd get round that by some kind of sequential locking so the Junction box couldn't clear a train into  the mixed section and then have the station box clear it for the wrong side.  I don't think the problem would occur in the opposite direction as outbound trains would have to relinguish the token for the mixed section at the junction box and then acquire the token for the next section of their own line before proceeding.

 

I think the difference between a British rules set-up and one in France or Germany would be the need for staff and token or tablet operation with signals preventing conflicting movements. At Noyelles this was and sometimes is taken care of by a Point B ground frame controlling the narrow gauge platforms and the SG access to the mixed gauge section from the Abbeville-Boulogne main line. 

Point B mainly controls the scissors crossover C,D, but I suspect that the signalling (the chequerboard "carrés" are absolute stop signals) was only used when standard gauge operations were in progress and carrés 14,16 &18 were normally kept open. That seems to still be the situation under preservation as I've only seen Point B manned during the busy Fêtes a Vapeur when SG trains also run between there and St. Valery. The Baie de Somme was slightly odd in that the dual gauge section from Noyelles to St. Valery  used to be part of the national network (d'Intêret General) so signalled  but the line on to Cayeux and the line to Le Crotoy were local railways (d'Intêret Local roughly equivalent to light railways) that usually only required physical signals to protect lines of the national network (those distinctions applied irrespective of gauge- there were metre gauge d'Intêret General lines and standard gauge d'Intêret Local railways). 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

It occurs to me that though such a situation may never have occurred in Great Britain with mixed standard and narrow gauge lines it might have been found at mixed standard gauge/ broad gauge stations especially those at the edge of the GW's broad gauge territory.

 

 

 

Edited March 21, 2019 by Pacific231G

[Joseph_Pestell]

39 minutes ago, Pacific231G said:

Thanks for the suggestions. 

 

This is a British layout so taking Ft. William station SB to Mallaig jct. SB as an example, that was 1560yds so the block section was a little under a mile.  A shortish bridge section might have been controlled by a single box but I'm assuming something long enough to have warranted a box at each end. 

I agree that the mixed section itself would be signalled like any section of single line track but  there wouldn't be any danger of derailments from trains trying to take the wrong gauge track as the fixed crossings wouild always keep them on their own gauge (That might not be true if you had a set of points for one of the gauges within the mixed section but that doesn't apply here)

 

However, with no moving points there would be a theoretical danger of a wrongly set signal clearing  a train to the wrong  line. For example, if the entrance to the station was controlled by a pair of home signals it would be possible to clear the NG signal and for an SG driver to take that as permission to proceed even if it was the NG station that was actually clear. I guess you'd get round that by some kind of sequential locking so the Junction box couldn't clear a train into  the mixed section and then have the station box clear it for the wrong side.  I don't think the problem would occur in the opposite direction as outbound trains would have to relinguish the token for the mixed section at the junction box and then acquire the token for the next section of their own line before proceeding.

I think the difference between a British rules set-up and one in France or Germany would be the need for staff and token or tablet operation with signals preventing conflicting movements. At Noyelles this was and sometimes is taken care of by a Point B ground frame controlling the narrow gauge platforms and the SG access to the mixed gauge section from the Abbeville-Boulogne main line.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

It occurs to me that though such a situation may never have occurred in Great Britain with mixed standard and narrow gauge lines it might have been found at mixed standard gauge/ broad gauge stations especially those at the edge of the GW's broad gauge territory.

 

 

 

 

But we come to the usual problem in signalling a model. 1560 yards is about 55' in 4mm scale. Building a layout of "normal" size will put the boxes too close to each other. A problem that could be solved by building the layout as two boxed dioramas linked by a tunnel - more shades of Tournon.

[The Stationmaster]

There were of course mixed gauge stations in Ireland and they were basically signalled in accordance with British practice although i'm not sure what systems were used on various mixed gauge running lines in the sort of situation you describe. 

[Pacific231G]

4 hours ago, Joseph_Pestell said:

 

But we come to the usual problem in signalling a model. 1560 yards is about 55' in 4mm scale. Building a layout of "normal" size will put the boxes too close to each other. A problem that could be solved by building the layout as two boxed dioramas linked by a tunnel - more shades of Tournon.

The layout idea in question would only include the station end of the mixed gauge but I'm curious to know how the whole thing would have been signalled in reality. 

I only travelled on the Vivarais once to Lamastre and back and there was something unsettling  about that dual gauge section between the terminus at Tournon and the point where the metre gauge swung away from the main line. You were very   conscious of what one of the heavy freight trains you saw belting through the (closed to passengers )  SG station  at full speed would have done to the much smaller wooden bodied coaches if the signalling system had failed, especially on the way back to Tournon when the CFV train was effectively running wrong way on the double track main line. 

 

Mike 

Do you know the specifics of any of the Irish examples? I don't think there were any in Grest Britain because there weren't that many public n.g, railways but I could be wrong about that. 

[jim.snowdon]

If the mixed gauge section is long enough to warrant token operation and a signal box at each end, I would suggest that (a) the separating turnout would be outside the distance for the home signal for each box, and (b) standard and narrow gauge trains would have their own bell codes. The homes could be splitting signals, in as much as it allows a confirmation to the driver that his train has been correctly signalled. 

 

Jim

[jim.snowdon]

14 hours ago, The Stationmaster said:

There were of course mixed gauge stations in Ireland and they were basically signalled in accordance with British practice although i'm not sure what systems were used on various mixed gauge running lines in the sort of situation you describe. 

I can't think of any examples of mixed gauge running lines in Ireland, although there were a few examples of mixed gauge sidings. Nothing in mainland Britain either, once you discount the Great Western, and even there, where a standard gauge line diverged from the broad gauge, the BoT rules required that the broad gauge was laid for a sufficient distance up the SG line to allow for an incorrectly routed train to be stopped. (The same, in a different sense, still applies in OLE where a non-electrified line diverges from an electrified one.)

The GW did use fixed points in its mixed gauge track, but as far as I am aware, the only examples on running lines were for transposing the common rail. All other turnouts were provided with moving switches.

 

 There are quite a few examples of mixed gauge running lines in Switzerland and Japan, but all with modern signalling, which may be interesting but not particularly helpful in this context.

 

Jim

[The Stationmaster]

59 minutes ago, jim.snowdon said:

I can't think of any examples of mixed gauge running lines in Ireland, although there were a few examples of mixed gauge sidings. Nothing in mainland Britain either, once you discount the Great Western, and even there, where a standard gauge line diverged from the broad gauge, the BoT rules required that the broad gauge was laid for a sufficient distance up the SG line to allow for an incorrectly routed train to be stopped. (The same, in a different sense, still applies in OLE where a non-electrified line diverges from an electrified one.)

The GW did use fixed points in its mixed gauge track, but as far as I am aware, the only examples on running lines were for transposing the common rail. All other turnouts were provided with moving switches.

 

 There are quite a few examples of mixed gauge running lines in Switzerland and Japan, but all with modern signalling, which may be interesting but not particularly helpful in this context.

 

Jim

I believe there was one in Londonderry but although it was a through line it appears to have been worked as a siding if what I've read about it was correct.

 

Coming back to the OP the answer will in some respects depend on period but is in reality not too difficult to resolve.   The standard gauge and narrow gauge lines could be worked as separate token/tablet sections interlocked with each other so that a token/tablet could only be withdrawn for one of them at any one time - that sorts the separation of the trains for block purposes as required by the 1889 Act.  The signals might or might not be interlinked with the token/tablet instruments - that really depends on the era being modelled although it would fit well with the use of separated token/tablet section.  Obviously signals would be interlocked with each other (at the same signalbox) and with any pointwork involved at that signalbox.  Each end of the section would be signalled as a junction with a splitting Home Signal reading towards either the standard gauge line or the narrow gauge line.

 

Incidentally on the GWR where a narrow gauge (in GW terms) line diverged from a mixed  gauge line there was no need to continue the broad gauge rail over the divergence if b.g trains could not be routed towards the narrow gauge only route (as a well known picture of Slough West Junction and curve shows with the broad gauge rail considerably disarranged after the curve was narrowed.

[Edwin_m]

Surely the simplest arrangement would be one token system (or staff & ticket, train staff or any other variant) with the same tokens used regardless of gauge.  I see little point in devising an interlock between two indepdent token systems. 

[jim.snowdon]

33 minutes ago, Edwin_m said:

Surely the simplest arrangement would be one token system (or staff & ticket, train staff or any other variant) with the same tokens used regardless of gauge.  I see little point in devising an interlock between two indepdent token systems. 

I would agree with that. Signalling is about managing the occupation of a section of railway, regardless of the gauge, or even what the "train" consists of.  What the signalmen at either end need to know before accepting the train is whether the line is clear to the clearing point. After, it helps to know what sort of train it is, even without the niceties of gauge to consider. In this instance, knowing the gauge is material, as although the track will sort which side of the terminus it will be directed into, it does help to have checked that there is space to accommodate it.

 

In the case of departing trains, much the same applies, but with the added complication that unless the receiving signalbox has passing loops on both lines, the token for the next section of the appropriate gauge single line has to be obtained before that for the mixed gauge section, otherwise there is the risk of two trains travelling in opposite directions meeting at the point where the single lines meet end on. (One reason why I believe the Railway Inspectorate was never exactly keen on intermediate block posts without passing loops on single line railways.)

 

Jim

[jim.snowdon]

1 hour ago, The Stationmaster said:

I believe there was one in Londonderry but although it was a through line it appears to have been worked as a siding if what I've read about it was correct.

That is the only example I could think of, and it relates, I believe, only to the the sidings that gave access to the riverside quays, so completely unsignalled.

 

Jim

[eastglosmog]

7 hours ago, jim.snowdon said:

I can't think of any examples of mixed gauge running lines in Ireland, although there were a few examples of mixed gauge sidings. Nothing in mainland Britain either, once you discount the Great Western, and even there, where a standard gauge line diverged from the broad gauge, the BoT rules required that the broad gauge was laid for a sufficient distance up the SG line to allow for an incorrectly routed train to be stopped. (The same, in a different sense, still applies in OLE where a non-electrified line diverges from an electrified one.)

The GW did use fixed points in its mixed gauge track, but as far as I am aware, the only examples on running lines were for transposing the common rail. All other turnouts were provided with moving switches.

 

 There are quite a few examples of mixed gauge running lines in Switzerland and Japan, but all with modern signalling, which may be interesting but not particularly helpful in this context.

 

Jim

The nearest I can think of is the interleaving of the Ravenglass & Eskdale from Ravernglass to Murthwaite mill with the private standard gauge siding between the two places.  Only the R&E would count as a running line, though!  At the Ravenglass end there were interlocked signals on the R&E and siding to prevent incorrect movements.  I would imagine there were some at the Murthwaite end as well, but have not seen any definite evidence.

[Edwin_m]

I don't think you could have a running line sharing with a siding of a different gauge, because there would need to be a means of ensuring that the siding traffic wasn't blocking the running line.  So the standard gauge would have to be considered as a running line and controlled by whatever system the R&ER used for its own traffic at the time. 

Edited March 19, 2019 by Edwin_m

[eastglosmog]

It depends how you define a siding, I suppose.  I have certainly seen it called such and it made an end on connection with the sidings in the LMS station at Ravenglass. I you are happy to call it a branch then so am I!  It then becomes a closer example to what the OP was looking for. 

[Pacific231G]

8 hours ago, jim.snowdon said:

I would agree with that. Signalling is about managing the occupation of a section of railway, regardless of the gauge, or even what the "train" consists of.  What the signalmen at either end need to know before accepting the train is whether the line is clear to the clearing point. After, it helps to know what sort of train it is, even without the niceties of gauge to consider. In this instance, knowing the gauge is material, as although the track will sort which side of the terminus it will be directed into, it does help to have checked that there is space to accommodate it.

 

In the case of departing trains, much the same applies, but with the added complication that unless the receiving signalbox has passing loops on both lines, the token for the next section of the appropriate gauge single line has to be obtained before that for the mixed gauge section, otherwise there is the risk of two trains travelling in opposite directions meeting at the point where the single lines meet end on. (One reason why I believe the Railway Inspectorate was never exactly keen on intermediate block posts without passing loops on single line railways.)

 

Jim

Hi Jim

I wondered about that but, beyond Fort William, the Glasgow line didn't have a loop at Mallaig Junction so a train going that way carrying the token  from Ft. William station box would have exchanged it at Mallaig jct. box for the token for Spean Bridge. The signalman at Mallaig jct. must therefore have had to know that he could pass a Glasgow bound train forward to Spean Bridge - the first passing loop on the Glasgow line- before accepting it from Ft. William Stn. Box.

That wouldn't have applied on the Mallaig extension as there was a passing loop that began at the junction  so a down train for Mallaig could have been accepted from Ft. William and held in the loop to wait for an up train from Mallaig (The more I study the old Ft. William station the more interesting it becomes) 

 

I agree that the two gauges wouldn't have been a reason to treat the mixed gauge section as anything other than one single line block section with a single pair of Tyers token instruments and possibly you'd just have had a notice at either end as you do when non-electrified lines or sidings branch off from electrified lines, Since it apparently never happened in the British Isles I guess we'll never know (though this situation might have arisen elsewhere where British signalling practices prevailed such as India) 

 

[John_Hughes]

Not quite actual mixed gauge, but I understand that the Vale of Rheidol and the Cambrian line in Aberystwyth have their - quite separate -  Llanbadarn level crossings interlocked so that only one of them at a time can be set against road traffic. Presumably this is to prevent traffic which has been held up at one crossing from backing up and blocking the other.

[The Stationmaster]

19 hours ago, Pacific231G said:

Hi Jim

I wondered about that but, beyond Fort William, the Glasgow line didn't have a loop at Mallaig Junction so a train going that way carrying the token  from Ft. William station box would have exchanged it at Mallaig jct. box for the token for Spean Bridge. The signalman at Mallaig jct. must therefore have had to know that he could pass a Glasgow bound train forward to Spean Bridge - the first passing loop on the Glasgow line- before accepting it from Ft. William Stn. Box.

That is the standard procedure in the Single Line Block Regulations for a token station/signal box which is not at a passing place.  

19 hours ago, Pacific231G said:

That wouldn't have applied on the Mallaig extension as there was a passing loop that began at the junction  so a down train for Mallaig could have been accepted from Ft. William and held in the loop to wait for an up train from Mallaig (The more I study the old Ft. William station the more interesting it becomes) 

 

In which case the train from Fort William would first have to be brought to a stand at Mallaig Jcn's Home Signal before being admitted to the loop.  If the train from Mallaig direction was approaching at  more or less the same time both trains would require to be stopped at the respective Home Signals before the first was admitted to its k loop and the other train would not have it Home Signal cleared until the first train to move had come to a stand in its loop.

 

I agree that the two gauges wouldn't have been a reason to treat the mixed gauge section as anything other than one single line block section with a single pair of Tyers token instruments and possibly you'd just have had a notice at either end as you do when non-electrified lines or sidings branch off from electrified lines, Since it apparently never happened in the British Isles I guess we'll never know (though this situation might have arisen elsewhere where British signalling practices prevailed such as India) 

 

No need for any sort of notices as a Driver would have a fixed signal to indicate to him which way the route was set at the exit from the mixed gauge section

[Pacific231G]

16 hours ago, The Stationmaster said:

That is the standard procedure in the Single Line Block Regulations for a token station/signal box which is not at a passing place.  

No need for any sort of notices as a Driver would have a fixed signal to indicate to him which way the route was set at the exit from the mixed gauge section

I've been thinking about this some more and wonder if I've been seeing this all wrong as the route isn't set- it's fixed .  The upper diagram shows my original assumption where the exit from the mixed gauge is equivalent to a junction that therefore requires splitting signals before the "points". However, though entry to the mixed gauge section has to be protected from conflicting movements, by both signals and possession of the token, it isn't actually a junction when exiting.  Trains can only go one way depending on their gauge so could not the home signals protecting their  respective stations or the lines beyond be positioned after the gauges separate as in the lower drawing (Forget that in the OP the station pointwork began immediately after the gauges split) 

 

I've ignored any distant or shunting signals, advanced homes etc.  though they might also be needed but would an arrangement lijke this to exit the mixed  gauge section at each end work? I'm not sure how interlocking with the tablet machines would work but presumably the same as any other single line junction.

Edited March 21, 2019 by Pacific231G

[The Stationmaster]

13 hours ago, Pacific231G said:

I've been thinking about this some more and wonder if I've been seeing this all wrong as the route isn't set- it's fixed .  The upper diagram shows my original assumption where the exit from the mixed gauge is equivalent to a junction that therefore requires splitting signals before the "points". However, though entry to the mixed gauge section has to be protected from conflicting movements, by both signals and possession of the token, it isn't actually a junction when exiting.  Trains can only go one way depending on their gauge so could not the home signals protecting their  respective stations or the lines beyond be positioned after the gauges separate as in the lower drawing (Forget that in the OP the station pointwork began immediately after the gauges split) 

 

I've ignored any distant or shunting signals, advanced homes etc.  though they might also be needed but would an arrangement lijke this to exit the mixed  gauge section at each end work? I'm not sure how interlocking with the tablet machines would work but presumably the same as any other single line junction.

I think it depends on how the third rail is arranged.  If there is any sort of switch rail involved (for either or both gauges) it would logically have to be detected and that means a fixed signal would also be required - especially if the standard gauge is involved.

 

As far as the tablet machines are concerned you might hit a problem if one of the signalboxes is not a passing place and a single section tablet section is used for both the standard and narrow gauges if the machines are interlinked with the machine for the next section in advance (although that interlinking would normally only prevent a token beimng withdrawn from one machine if one has already beeb n withdrawn at the opposite end of the section. However it does - for example - mean that if an n.g. train wishes to leave the station at the left end and the signalbox at the other end has accepted a standard gauge train from the 'box in rear the n.g. train will have to wait at the station until that train has arrived narrow.  Provided the signals provide a Clearing Point  approaching the junction from the right hand direction having a separate machine for the n.g. and s.g lines would be a way around that as the Signalman would not be accepting trains towards each other with no means of crossing them but would then be working as a junction.

Edited March 21, 2019 by The Stationmaster

[eastglosmog]

The mixed gauge section of the Ravenglass and Eskdale did not have switches  - the narrow gauge rails were both within the standard gauge rails.  The "points" each end had moveable crossings, with a pivoting bar where each narrow gauge rail crossed the standard gauge rail.  These were interlocked with the stop signals on each line. 

I suspect the R&E at the time (1929-53) did not bother with such things as tablets or staffs.

Edited March 21, 2019 by eastglosmog

[Pacific231G]

9 hours ago, The Stationmaster said:

I think it depends on how the third rail is arranged.  If there is any sort of switch rail involved (for either or both gauges) it would logically have to be detected and that means a fixed signal would also be required - especially if the standard gauge is involved.

 

As far as the tablet machines are concerned you might hit a problem if one of the signalboxes is not a passing place and a single section tablet section is used for both the standard and narrow gauges if the machines are interlinked with the machine for the next section in advance (although that interlinking would normally only prevent a token beimng withdrawn from one machine if one has already beeb n withdrawn at the opposite end of the section. However it does - for example - mean that if an n.g. train wishes to leave the station at the left end and the signalbox at the other end has accepted a standard gauge train from the 'box in rear the n.g. train will have to wait at the station until that train has arrived narrow.  Provided the signals provide a Clearing Point  approaching the junction from the right hand direction having a separate machine for the n.g. and s.g lines would be a way around that as the Signalman would not be accepting trains towards each other with no means of crossing them but would then be working as a junction.

Thanks Mike

I'm specifically looking at a situation where there are no switch rails involved for either gauge so there is nothing to detect. That's what makes this different from a conventional junction where there would be points that would need to be protected by signals at the facing end . Signals would be needed to enter the mixed section just as they would to enter a single line section from a station with more than one platform  but not to leave it (though they wold be needed to protect whatever came after the split.  So far as the tablet machines are concerned I'm assuming that the situation is equivalent to a single line section with a junction at one end so one pair of machines could cover the mixed gauge section.  The situation actually proposed is that the Standard Gauge station is a terminus (not a busy one)  but the NG line continues to the left past its station to a small terminus a bit further on (I've just realised that this is the situation on the final version of P.D.Hancock's first Craig and Mertonford though I don't think that was ever fully signalled). The assumption is that this final section of the narrow gauge would be single engine in steam so would only need a staff.