Signals

THE SIGNAL BOX


Railway signalling discussion

Tan-Y-Graig (GWR passing)

For railway modellers to discuss and share ideas on realistic signalling arrangements.

Tan-Y-Graig (GWR passing)

Unread postby tynewydd » Sun Jun 8, 2014 3:17 pm

Second last in the series of Port Dinllaen Lines - is Tan-Y-Graig, a GWR passing station on a single line between Port Dinllaen and Pwhelli worked by token. The files are here

The inspiration is Evercreech - the Station rather than the Junction. Thus we have a central road for banking engines with access to the goods yard via a diamond and a level crossing at one end of the station. The bank in question is the roughly 1 in 40 one for a mile up to the summit on the road to Port Dinllaen.

The things that are different to Evercreech are that I
  • Reversed the station direction and removed the dual track in the up direction
  • Added provision for two way working to both platforms for peak demand
  • Borrowed from some other S&D station the idea for use of a backing signal. Briefly it is for occasions when a goods train stalls on the bank out of the station and has to return for "another go" but we don't want to hold up a following passenger train while we get up steam, possibly attach a second banker, back down all the way to get a better run at it, etc.
  • added an industry that its proponents say has an explosive potential for growth :wink:

I'd especially appreciate guidance on the trap/catch point situation. Although the station is basically flat it has a steep bank up to PD and a lesser slope down to Pwhelli at the other to consider as well.

As always there is other stuff to discuss I'm sure. Hopefully at least I got the spelling correct this time! Welsh mutation of the leading end of words is a always a source of confusion (to me).

Thanks for the help,
Adam
tynewydd
Crossing box
Crossing box
 
Posts: 49
Joined: Tue Feb 15, 2011 2:04 pm
Location: Les Etats-Unis, Cote Ouest

Re: Tan-Y-Graig (GWR passing)

Unread postby Mike Hodgson » Sun Jun 8, 2014 4:10 pm

Facing traps on the exits to the loops would be unusual on a passenger line.

If stopping passenger trains require banking, the layout is operationally inconvenient because either they have to draw forward to allow the banker to jojn the rear or the banker needs to run through the opposite platform to the advanced starter to join the train whilst still in the platform. It would probably be easier to attach at the front of the train.

I'm not sure about a running signal leading into an explosive factory. No doubt there are precedents for it, but I'd be happier with a diesel or fireless engine shunting that.

Looks like a very long rodding run to the points at the PD end of the station. Three homes and two starters in one direction, two homes and three starters in the other with pointwork at the outermost homes seems like rather a lot for a one-box station in purely mechanical days. Where are you issuing tokens? I can see a need for auxiliary instruments, or if the box were on the opposite side of the road, I suspect it would be a little easier.
User avatar
Mike Hodgson
Main line box
Main line box
 
Posts: 2464
Joined: Fri Nov 9, 2007 5:30 pm
Location: N Herts

Re: Tan-Y-Graig (GWR passing)

Unread postby Chris Osment » Sun Jun 8, 2014 6:40 pm

I gulped when I saw the amount of signals :-)

From what I can see, there are 3 backing signals, all of which seem to apply to right-direction moves and therefore IMHO are not appropriate here. I suspect the S&D example to which you refer is the one (+ repeater) at Midford, which did control a true 'wrong road' move.
Chris Osment
West Country Railway Archives
http://www.railwest.org.uk
User avatar
Chris Osment
Main line box
Main line box
 
Posts: 2238
Joined: Sun Nov 11, 2007 9:32 am
Location: Somewhere in the West Country

Re: Tan-Y-Graig (GWR passing)

Unread postby tynewydd » Sun Jun 8, 2014 11:57 pm

Chris Osment wrote:I gulped when I saw the amount of signals :-)

From what I can see, there are 3 backing signals, all of which seem to apply to right-direction moves and therefore IMHO are not appropriate here. I suspect the S&D example to which you refer is the one (+ repeater) at Midford, which did control a true 'wrong road' move.


Chris,
Well I had intended that one was a repeater but got carried away and added a second one! I assume that only one is likely needed and no repeater(s)? One thought I had was that if the stalled train wished to back down to the end of the loop beyond the platforms, in order to get a longer run, but on the wrong road (we'll get to that later) how would it be signaled over the LC?

I guess a more basic question is this - what constitutes wrong road when backing? Is it the direction of travel of the train as if it was going in a forward direction (and backing doesn't count) or is it the backing direction that is against the normal flow? I presume I picked the wrong platform and any backing signal should lead into the Down side implying that it is the backing movement that is against the normal flow?

Would the most important point be that by choosing to signal both lines bi-directionally from the PD end, there is no "wrong road" at all?

Back to basics - what I was trying to avoid would be the situation where a stalled goods or other train blocks a fast passenger while it recuperates for another try up the Garn Boduan Incline. And that passenger train might be from ether direction. So backing signal or no, how should that be done?

Adam
tynewydd
Crossing box
Crossing box
 
Posts: 49
Joined: Tue Feb 15, 2011 2:04 pm
Location: Les Etats-Unis, Cote Ouest

Re: Tan-Y-Graig (GWR passing)

Unread postby tynewydd » Mon Jun 9, 2014 12:34 am

Mike Hodgson wrote:Facing traps on the exits to the loops would be unusual on a passenger line.


Aha! - I had read somewhere that in situations where there was not sufficient clearing beyond the signals out of the loop they were required. It's a lot easier if they are not needed. Would there be a requirement to have a trap at the foot of an incline to divert runaway traffic, though?

Mike Hodgson wrote:If stopping passenger trains require banking, the layout is operationally inconvenient because either they have to draw forward to allow the banker to jojn the rear or the banker needs to run through the opposite platform to the advanced starter to join the train whilst still in the platform. It would probably be easier to attach at the front of the train.

I agree and some time playing with the PCRail simulation of Evercreech seems to suggest that the prototype was awkward similarly. If dual heading is preferred the calling-on arms would need to be moved to allow the "header" to back down onto the train, I guess.

I can play about with "just" reversing the cross-overs, but that leads to the problem about the exit being in the middle of the down platform - I'll have to look into moving the platforms to the PD end instead so the central line is beyond the station instead of ahead of it... I'll see what I can do.

Mike Hodgson wrote:I'm not sure about a running signal leading into an explosive factory. No doubt there are precedents for it, but I'd be happier with a diesel or fireless engine shunting that.

OK - actually it is the siding for explosive shipments in closed vans - the closed vans would be loaded and unloaded well away from the actual factory which would be hidden behind protective berms - the prototype I am using is the ICI Nobel Explosives works at Penrhyndeudraeth, which as far as I know had a single siding worked conventionally - but with barrier wagons, of course.

How about if the running signal was removed and we used ground signals only?

Mike Hodgson wrote:Looks like a very long rodding run to the points at the PD end of the station. Three homes and two starters in one direction, two homes and three starters in the other with pointwork at the outermost homes seems like rather a lot for a one-box station in purely mechanical days. Where are you issuing tokens? I can see a need for auxiliary instruments, or if the box were on the opposite side of the road, I suspect it would be a little easier.


OK - I guess that because both sides of the station have tokens (both single lines) the position of the signal box is somewhat arbitrary? There would need to be token exchange machines to gather the incoming tokens and permit use of the new section on the "other side" of the loop anyway? We could move the SB to be on one of the platforms if that would help distances.

Thanks as always,
Adam
tynewydd
Crossing box
Crossing box
 
Posts: 49
Joined: Tue Feb 15, 2011 2:04 pm
Location: Les Etats-Unis, Cote Ouest

Re: Tan-Y-Graig (GWR passing)

Unread postby Mike Hodgson » Mon Jun 9, 2014 9:17 am

tynewydd wrote:I guess a more basic question is this - what constitutes wrong road when backing? Is it the direction of travel of the train as if it was going in a forward direction (and backing doesn't count) or is it the backing direction that is against the normal flow? I presume I picked the wrong platform and any backing signal should lead into the Down side implying that it is the backing movement that is against the normal flow?


I would agree with Chris that "backing" is inappropriate where you have those signals. Backing only makes sense if you are going against the flow of an Up or Down line. Single lines are bidirectional by definition and you have chosen that to apply also to both your platforms, so there isn't a "Wrong" direction. In this situation, although Up and Down trains would usually use the platforms indicated, I don't think it is strictly correct to label them Up Line/Down Line. However I think it unlikely that both platforms would have been bi-directional, although it could be necessary for example to allow an express to overtake a slow passenger. If it was only freights needing to be passed, they would almost certainly be put into a refuge siding or loop outside the station, but I suppose you don't have room for that. It was fairly common for crossing stations on single lines to have one platform (usually the one with the station building) signalled bidirectionally to cater for switching the box out.

The current limit on rodding is 350 yards, but recommended no more that 200 yards for facing points. The IRSE recommend reducing the limits if there are sharp curves or lots of cranks etc. Various even shorter limits have applied in the past. Whilst you could use a ground frame to control a remote siding such as the factory, it would not be acceptable to do this with the main points on the running line. If the overall distance is short enough, to control the points mechanically the box should be roughly mid-way between the two ends of the loops. This would still leave you with the problem of controlling the crossing. Gates could be worked by station staff, but that's obviously undesirable.

In more recent times, of course point motors, colour light signals and CCTV crossings allow you to work remotely, but history would have dictated the position of the original structure, and this would have been used on resignalling if suitable. But these would rather go against the grain of what you are trying to achieve.

Consequently my feeling is that you would therefore probably have had two boxes - one where you have it (convenient for the crossing) and another near the far end. It solves the problem of handling tokens. With that arrangement, I think you also overcome concern about the number of signals, although it does introduce the question of which box controls what, what is slotted, and whether you need distants under any of the homes.

Runaway catch points on steep inclines were usual on double lines, to catch the back end of unfitted freights running wrong direction if the brake van's brake was overpowered after couplings had broken. However they were not usual in the facing direction, and of course both directions can be facing on a single lines. If a train ran away down a single line, it would be usual for it to be directed into the loop intended for its own direction, and if still unable to stop it would run through the trailing points at the far end of the loop onto the single line, and it would be accepted that these points would be damaged. The signalman would of course send emergency messages!

It might be seen as preferable to divert a runway into a bay platform, cattle dock etc rather than have it continue to accelerate if the line the other side was also falling or if there was a passenger train in its likely path. The signalman would have to decide the least worst option taking into account all the circumstances, but one place he would try to avoid would be a siding containing an explosives train! On the subject of which incidentally, five days ago it was the 70th anniversary of the Soham explosion.

I think I'd be happier with a disc into the factory, as I suspect you would probably shunt a few vans into it rather than bring a whole train straight in. Your goods arm on the bracket might be justified though, on the basis of sighting or traffic, and I'm not sure about GWR practice.
User avatar
Mike Hodgson
Main line box
Main line box
 
Posts: 2464
Joined: Fri Nov 9, 2007 5:30 pm
Location: N Herts

Re: Tan-Y-Graig (GWR passing)

Unread postby John Hinson » Mon Jun 9, 2014 9:46 am

Mike Hodgson wrote:I think I'd be happier with a disc into the factory, as I suspect you would probably shunt a few vans into it rather than bring a whole train straight in. Your goods arm on the bracket might be justified though, on the basis of sighting or traffic, and I'm not sure about GWR practice.

The GWR would not have an arm on the bracket signal, or a disc alongside, unless trains arrived off the single line and ran straight into the siding. Otherwise a disc would be provided at the toe of the points coming out of the siding - for setting back into the siding.

In essence, I feel the whole layout is over-signalled. Unless there is very good reason, two bi-directional lines through the station are excessive and unprototypical and the arrangement of siding connections (per Evercreech) are designed for separate Up/Down working. Modeller's licence allows anything of course - this is just my opinion.

Given the apparent arrangement of the yard, I would add a trailing connection (similar to the gunpowder siding at the other end) for ease of access.

And I concur with the removal of the facing traps.

John
Image
‹(•¿•)›
User avatar
John Hinson
Site Admin
Site Admin
 
Posts: 6782
Joined: Thu Nov 8, 2007 1:13 pm
Location: at my computer

Re: Tan-Y-Graig (GWR passing)

Unread postby Chris Osment » Mon Jun 9, 2014 12:03 pm

I would concur with the general feeling that the signalling is rather OTT and surely out-of-step with the basic ethos of your line.

As regards bank/pilot engines, not quite sure what your sim of Evercreech Jcn does, but passenger trains from there were piloted and bank engines only used (when necessary) on goods trains towards Bath. But I agree with John that the whole idea of the middle siding does not really fit in with single-line passing-loops.

As regards traps on the incline.......you could follow the example used at some places whereby the facing point lies to the right rather than the left, with a trailing sprung catch point at the exit from the RH loop. But if that road is going to be bi-directional, then you would need to be able to shut the catch (using a slotted spring joint) and then have a FPL to bolt if for the facing move.

You might want to look at the arrangements at South Molton for an example of a GWR loop signalled for bi-directional on one road, but not equipped for switching-out.
Chris Osment
West Country Railway Archives
http://www.railwest.org.uk
User avatar
Chris Osment
Main line box
Main line box
 
Posts: 2238
Joined: Sun Nov 11, 2007 9:32 am
Location: Somewhere in the West Country

Re: Tan-Y-Graig (GWR passing)

Unread postby tynewydd » Mon Jun 9, 2014 3:01 pm

John Hinson wrote:In essence, I feel the whole layout is over-signalled. Unless there is very good reason, two bi-directional lines through the station are excessive and unprototypical and the arrangement of siding connections (per Evercreech) are designed for separate Up/Down working. Modeller's licence allows anything of course - this is just my opinion.

Given the apparent arrangement of the yard, I would add a trailing connection (similar to the gunpowder siding at the other end) for ease of access.


Ok - chastened by these reviews, I have revised the layout a bit and pruned down the signaling so that only one direction is signaled both ways - with the idea of allowing a faster train to overtake a slower one in the direction of the incline. Apart from banking this looks a bit like South Moulton - but I have yet to absorb the provision of sprung points at that location - I started with 2 "points to nowhere" and would end up with 5!

I have also altered the arrangements for the central road to reverse the direction so that both banking engines leaving the back of trains from PD and bankers to be set back onto trains to PD can directly access from/to the most frequently used lines. This adds to the facing point load, but makes things more practical when operating "normally".

I have clung "dangerously" to the backing signal to allow trains to back into the "wrong road" (i.e, the one that does not normally allow for down traffic). But that could obviously be changed to be a normal home on the bracket (or a miniature).

Let's see if that revision to track plan is any better.

Adam
tynewydd
Crossing box
Crossing box
 
Posts: 49
Joined: Tue Feb 15, 2011 2:04 pm
Location: Les Etats-Unis, Cote Ouest

Re: Tan-Y-Graig (GWR passing)

Unread postby Danny252 » Mon Jun 9, 2014 5:23 pm

tynewydd wrote:Ok - chastened by these reviews, I have revised the layout a bit and pruned down the signaling so that only one direction is signaled both ways - with the idea of allowing a faster train to overtake a slower one in the direction of the incline.


I would have expected the timetable to be designed so that there was never that sort of situation - after all, the faster train still has to fight its way along the Cambrian line to Pwhelli, and any slow freight could be held there until a suitable gap appears.

For comparison with what happened in a similar real-life situation, I've had a look at the 1949 timetable for Neyland and Milford Haven, both of which are single line port branches from Johnston (although shorter than the Pwll-Dinllaen section). Only two trains during were "overtaken" each day, and in both cases these were local freight trains that shunted out of the way whilst also picking up/setting out traffic - from what I've seen in other GWR/BR(WR) timetables, this was a pretty typical maneuver on single lines (or even double tracked lines), with the daily pick-up freight disappearing into each station's yard for an hour whilst a local passenger or two scooted past.

This could be done fairly nicely at Tan-y-graig with the layout you have, with the freight shunting out onto the single line and then back into the goods yard, and then shunting around until the fast train passes and clears the section ahead. I imagine the banker would then head into the goods yard (or perhaps would be sent in before the goods arrived), as shunting the goods out and then putting the banker on would require more of that horrible expensive signalling.

(If you'd be interested in the Milford/Neyland workings, I've tabulated them and could send you a copy - it might prove of interest when working out your own timetable. Alternatively, you can view them yourself on Michael Clemens' site)
Danny252
Rest-day relief
Rest-day relief
 
Posts: 686
Joined: Thu Aug 26, 2010 4:39 pm
Location: As close to the levers as they'll let me be

Re: Tan-Y-Graig (GWR passing)

Unread postby Chris Osment » Mon Jun 9, 2014 6:32 pm

Another place to look at would be Churston, also bi-directional on one road but NOT switching-out. And IIRC in earlier days it had a backing-arm bracketed off the Down Home to take you back onto the Up loop.
Chris Osment
West Country Railway Archives
http://www.railwest.org.uk
User avatar
Chris Osment
Main line box
Main line box
 
Posts: 2238
Joined: Sun Nov 11, 2007 9:32 am
Location: Somewhere in the West Country

Re: Tan-Y-Graig (GWR passing)

Unread postby S&TEngineer » Mon Jun 9, 2014 7:38 pm

Mike Hodgson wrote:Facing traps on the exits to the loops would be unusual on a passenger line.

I don't agree with that statement. I can think of lots of (GWR) places where there were facing traps at the ends of a crossing loop (which I believe were provided to allow simultaneous entry to the crossing loop from both directions). There is an example here not a million miles from the location of the proposed layout: http://www.s-r-s.org.uk/html/gwu/S3244.htm

Also note that there is a wrong direction main running signal with a 3ft arm reading in a facing direction into the siding.
Regards,
S&TEngineer
-----------------------------------------------------
Out of this nettle, Danger, we pluck this flower, Safety.
Henry IV, Part 1, Act 2, Scene 3
User avatar
S&TEngineer
Moderator
Moderator
 
Posts: 1593
Joined: Fri Nov 9, 2007 8:17 pm
Location: Somewhere in the far South West

Re: Tan-Y-Graig (GWR passing)

Unread postby tynewydd » Fri Jun 13, 2014 4:15 am

Danny252 wrote:I would have expected the timetable to be designed so that there was never that sort of situation - after all, the faster train still has to fight its way along the Cambrian line to Pwhelli, and any slow freight could be held there until a suitable gap appears.

Unstated I had thought that selective Cambrian line doubling would have taken place as the demand for services to South Wales increased via Carmathen. Also the likely building of a "proper" bridge over the Dyfi at Ynyslas to remove the need to go around via Dovey Junction. But you are right re: scheduling in any case.

It probably makes most sense to either double the line from TYG to Pwhelli - I'll see if I have the room - need to make sure the clearances are there but might be possible - so that this section would resemble more like the final stretch into Fishguard Harbor (double to single) - for another port for Ireland, incidentally - or make sure that we just have a simple passing situation and remove bi-directional working unless we want to allow for switching out - which however seems unworkable with banking engines involved unless they all double-head run through and from Pwhelli when switched out?

Danny252 wrote:For comparison with what happened in a similar real-life situation, I've had a look at the 1949 timetable for Neyland and Milford Haven, both of which are single line port branches from Johnston (although shorter than the Pwll-Dinllaen section). Only two trains during were "overtaken" each day, and in both cases these were local freight trains that shunted out of the way whilst also picking up/setting out traffic - from what I've seen in other GWR/BR(WR) timetables, this was a pretty typical maneuver on single lines (or even double tracked lines), with the daily pick-up freight disappearing into each station's yard for an hour whilst a local passenger or two scooted past.

This could be done fairly nicely at Tan-y-graig with the layout you have, with the freight shunting out onto the single line and then back into the goods yard, and then shunting around until the fast train passes and clears the section ahead. I imagine the banker would then head into the goods yard (or perhaps would be sent in before the goods arrived), as shunting the goods out and then putting the banker on would require more of that horrible expensive signalling.


Yes - got that - of course Neyland was a former big port at that point with Fishguard being the main one so the working were probably pretty sparse. However - that approach would be a slower pace and still have a person responsible for two sets of block bells, instruments, etc. a good place to learn the ropes, perhaps. If the banking is restricted to the Down, we also don't have that many extra signals for that attachment operation.

Thanks for the thoughts -
Adam
tynewydd
Crossing box
Crossing box
 
Posts: 49
Joined: Tue Feb 15, 2011 2:04 pm
Location: Les Etats-Unis, Cote Ouest

Re: Tan-Y-Graig (GWR passing)

Unread postby tynewydd » Fri Jun 13, 2014 5:00 am

Chris Osment wrote:Another place to look at would be Churston, also bi-directional on one road but NOT switching-out. And IIRC in earlier days it had a backing-arm bracketed off the Down Home to take you back onto the Up loop.


Chris, thanks - yes I can see it - lever 48, I think! Not sure how it would read - presumably just back enough to get to the ground signal that controls the entrance to the bay - or perhaps locking that route out and going all the way - but in that case they are lacking a limit sign to halt progress backwards in the SRS diagram. Anyway - prototypical only as long as it is indicating a reverse move against the flow of normal traffic.

Adam
tynewydd
Crossing box
Crossing box
 
Posts: 49
Joined: Tue Feb 15, 2011 2:04 pm
Location: Les Etats-Unis, Cote Ouest

Re: Tan-Y-Graig (GWR passing)

Unread postby Chris Osment » Fri Jun 13, 2014 7:04 pm

The Churston backing arm was 42 - you must be reading the small thumbnail image!

According to the SRS locking table info, it could be pulled regardless of the position of the up dock siding points 13. It did not require the ingoing disc 42 to be off if the siding points were reverse. My suspicion therefore is that you could go all the way back down the up loop if so required, as with 13 normal it locked the up loop facing points 10 normal and the Up Home, so the 'limit of shunt' presumably would have been the Down Advanced Starting 44.
Chris Osment
West Country Railway Archives
http://www.railwest.org.uk
User avatar
Chris Osment
Main line box
Main line box
 
Posts: 2238
Joined: Sun Nov 11, 2007 9:32 am
Location: Somewhere in the West Country

Next

Return to Signalling - model railways/simulators

Who is online

Users browsing this forum: Yahoo [Bot] and 1 guest