K-crossing check rails

Discuss the prototype and how to model it.
User avatar
Martin Wynne
Posts: 1172
Joined: Mon May 14, 2012 4:27 pm

K-crossing check rails

Postby Martin Wynne » Wed Nov 03, 2021 10:57 am

.
I've mentioned a few times that it can help to avoid mis-tracking at flat-angle fixed K-crossings if you raise the level of the check rails above the level of the running rails. And that it's now being done on the prototype.

This picture has been in the news recently, and clearly shows such raised check rails:

Image

It's easy to see why steam locomotives having flangeless driving wheels have to be banned from such routes.

It's also easy to see why you need a supply of Elastoplast when track cleaning. :)

RAIB press release today on the accident is here:

https://www.gov.uk/government/news/coll ... l-junction

Best wishes to the driver on a full recovery.

Martin.
40+ years developing Templot. Enjoy using Templot? Join Templot Club. Be a Templot supporter.

User avatar
steve howe
Posts: 911
Joined: Sun Feb 01, 2009 10:16 pm

Re: K-crossing check rails

Postby steve howe » Wed Nov 03, 2021 11:57 am

I wasn't aware of that junction until it appeared on the aerial views in the news report. The close proximity of a double junction to a tunnel entrance looked so improbable, a case of prototype following model!

This incident has been much covered on our local TV news, the current theory is 'wheel-slide' on wet or leaf-covered rails. Even so for this to happen the train must have been approaching too fast and/or, as has been suggested, a faulty signal led the driver to believe the road was clear. Either way, thankfully no fatalities.

Steve

User avatar
Martin Wynne
Posts: 1172
Joined: Mon May 14, 2012 4:27 pm

Re: K-crossing check rails

Postby Martin Wynne » Wed Nov 03, 2021 12:24 pm

steve howe wrote:I wasn't aware of that junction until it appeared on the aerial views in the news report. The close proximity of a double junction to a tunnel entrance looked so improbable, a case of prototype following model!

Hi Steve,

Also Worcester Tunnel Junction, with just a loco length between the switch toes and the tunnel:

worc_tunnel_junc_aerial.jpg

cheers,

Martin.
You do not have the required permissions to view the files attached to this post.
40+ years developing Templot. Enjoy using Templot? Join Templot Club. Be a Templot supporter.

User avatar
grovenor-2685
Forum Team
Posts: 3917
Joined: Sun Jun 29, 2008 8:02 pm

Re: K-crossing check rails

Postby grovenor-2685 » Wed Nov 03, 2021 12:41 pm

steve howe wrote:This incident has been much covered on our local TV news, the current theory is 'wheel-slide' on wet or leaf-covered rails. Even so for this to happen the train must have been approaching too fast and/or, as has been suggested, a faulty signal led the driver to believe the road was clear. Either way, thankfully no fatalities.
Steve

The RAIB preliminary report does NOT indicate either faulty signals or overspeeding, best to read it before making assumptions, there was more than enough of that in the early press reports.
Regards
Keith
Grovenor Sidings

John Palmer
Posts: 825
Joined: Fri Jul 15, 2011 11:09 pm

Re: K-crossing check rails

Postby John Palmer » Wed Nov 03, 2021 12:59 pm

The 3rd November statement issued by RAIB is worth reading. The following extract is pertinent to the propriety of the SWR driver's actions:

"Preliminary analysis of data downloaded from the On Train Data Recorder (OTDR) fitted to train 1L53 shows that the driver initially applied service braking to slow the train on approach to the caution signal before signal SY31. Around 12 seconds after service braking started, the driver made an emergency brake demand. As the train approached signal SY31, and with the emergency brake still being demanded by the driver, a second emergency brake demand was made by the train protection and warning system (TPWS). These emergency brake demands did not prevent the train from reaching the junction, where the collision occurred. OTDR analysis indicates that wheel slide was present both when the driver applied service braking and after emergency braking was demanded. This was almost certainly a result of low adhesion between the train’s wheels and the rails."

And from SWR's response to the RAIB statement at https://www.southwesternrailway.com/other/news-and-media/news/2021/november/south-western-railways-response-to-raibs-statement-on-the-salisbury-rail-incident:

"RAIB’s statement on their initial investigation into the train collision near Salisbury on Sunday points to low adhesion between the wheels and the track as the most likely cause. Their early assessment shows the South Western Railway driver reacted correctly to the signals by braking to slow the train down.

"While the investigation has a long way to run, these initial findings are helpful to stop speculation and focus attention on the evidence about the likely cause of the collision."

The series of signal aspects presented to the SWR driver seems likely to have been: SY29R: YY, SY29: Y, SY31: R. Distance from SY29 to point of collision is in the region of 817 metres. Not yet clear whether the service brake application was made on the approach to SY29 or its repeater, but that is probably academic, given historical instances of trains being unable to pull up under wheel slide conditions within distances substantially in excess of 800 metres.

The impression I get is that the combination of lightweight trains, substitution of disc braking for shoe braking (no cleaning of wheel tread) and increased railhead contamination has introduced significant uncertainty into train braking performance, and that so far this has been countered by mitigation measures of dubious efficacy.

Deep respect for the SWR driver who remained at his post, and I echo Martin's best wishes for his full recovery.

User avatar
Noel
Posts: 1972
Joined: Wed Jun 23, 2010 1:04 pm

Re: K-crossing check rails

Postby Noel » Wed Nov 03, 2021 3:45 pm

John Palmer wrote:The series of signal aspects presented to the SWR driver seems likely to have been: SY29R: YY, SY29: Y, SY31: R.


My understanding is that 29R is a YGY- repeater, in which case it would have been showing Y and SY27 YY. I was interested to note that the driver is apparently 70, which in a road accident would probably lead to questions about his reaction speed. Neither SWT or RAIB have yet made any comment about the speeds involved, although the actual collision seems to have been quite low speed. It does, however, leave open for the moment the question of whether the driver made the initial normal brake application early enough.
Regards
Noel

User avatar
grovenor-2685
Forum Team
Posts: 3917
Joined: Sun Jun 29, 2008 8:02 pm

Re: K-crossing check rails

Postby grovenor-2685 » Wed Nov 03, 2021 6:43 pm

Note that the RAIB do not make any mention of SY29 or SY29R. They just say that the driver braked on approach to the caution signal before SY31.
Regards
Keith
Grovenor Sidings

Alan Turner
Posts: 643
Joined: Sun Jul 20, 2008 4:24 pm

Re: K-crossing check rails

Postby Alan Turner » Fri Nov 05, 2021 10:43 am

There is a question that I have that relates to the signalling. If this were Absolute block then there would be stop signals protecting the junction. Each of those stop signals would have their clearing point and that would probably be somewhere in the junction (or tunnel even). So in order to accept a second train into the junction the first train would first have been accepted to their stop signal and stopped at that signal before the road could have been set for the second train and the clearing point established for the second train.

So I am assuming here that modern signalling does not follow the principles of AB and that two trains can occupy the same clearing point? Or am I missing a point here?

regards

Alan

User avatar
Guy Rixon
Posts: 909
Joined: Sun Mar 27, 2011 6:40 pm

Re: K-crossing check rails

Postby Guy Rixon » Fri Nov 05, 2021 12:09 pm

Screenshot 2021-11-05 at 12.03.18.png

That's a screen capture of the satellite image on Google. IIUC, SY31 is near the light-coloured patch in the top right (the main patch looks like an equipment cabinet and I can just see something that looks like the signal post). The safety overlap from the junction is like 100 m.

The curve is quite tight approaching the junction from the north, so is a ~100 m overlap consistent with a cautionary approach on the curve, if the previous signal is at yellow? I.e. is the line speed low enough to justify this?

PS: I measured the signal position on the older OS maps and it's 220 yds from the fouling point, which makes more sense than 100 m. No idea if SY31 is in exactly the same place.
You do not have the required permissions to view the files attached to this post.
Last edited by Guy Rixon on Fri Nov 05, 2021 12:53 pm, edited 2 times in total.

User avatar
grovenor-2685
Forum Team
Posts: 3917
Joined: Sun Jun 29, 2008 8:02 pm

Re: K-crossing check rails

Postby grovenor-2685 » Fri Nov 05, 2021 12:20 pm

Alan Turner wrote:So I am assuming here that modern signalling does not follow the principles of AB and that two trains can occupy the same clearing point? Or am I missing a point here?
regards
Alan

Multiple Aspect Signalling (MAS) with track circuit (or axle counter) block requires a specified overlap to be clear beyond the next signal before a signal can be cleared. Any trailing points in the overlap must also be set and locked. The specified overlaps were 200 yds for 4 aspect signalling and 300 yds for 3 aspect signalling, as the driver in 4 aspect territory gets two warnings so stopping errors should be less. In this case the signal is set back from the junction to allow the 200 yds overlap before the clearing point of the junction. If the signal, for some reason had been closer to the junction then the points, locked for the first train, would have blocked the overlap and the train would have been signalled to stop at the previous signal.
Its the same principle as AB working except that the AB clearing point (equivalent to the overlap) was usually 440yds.
Regards
Keith
Grovenor Sidings

John Palmer
Posts: 825
Joined: Fri Jul 15, 2011 11:09 pm

Re: K-crossing check rails

Postby John Palmer » Fri Nov 05, 2021 2:13 pm

I understand the signalling system over both the Worting Jc-Salisbury and Romsey-Salisbury routes to be Track Circuit Block, under which signal overlaps are the building blocks that take the place of clearing points as found under Absolute Block.

Guy is correct about the location of SY51, and in this case, using the measuring tool supplied with Google maps, I calculate the distance from the diamond to SY51 to be very close to 180 metres.

Permitted line speed from Grateley to Laverstock North Jc (north end of the chord connecting the Worting and Romsey routes) is 90 mph, reducing to 50 mph immediately before Laverstock North Jc.

As Keith quite rightly says, the RAIB statement made no mention of either SY29 or SY29R, but I think it must be the case that the service brake application was made on the approach to one or other of these. SY29 stands about 860 metres in rear of SY51, whilst SY29R stands approximately 2.23 km / 1.39 miles in rear of SY29, according to my Google map measurements.

Mea culpa! I should not have assumed that the 'R' in the SY29R designation necessarily stands for 'repeater'. It is my understanding that it functions not as a repeater of SY29's aspect but as a caution signal for SY29 and that it displays a double yellow aspect when SY51 is exhibiting a red aspect and SY29 a single yellow. In saying that, I'm relying on the concensus of opinion received from a number of apparently well-informed commentators, but I've seen nothing authoritative to that effect.

I can't see a basis for finding fault with the design of the signalling installation here. I can, however, see a significant problem arising from the fact that reliable prediction of where wheel slide conditions are going to occur seems unattainable. The last few miles of the route leading to Fisherton Tunnel are a gentle descent over what is predominantly open downland (clue in the name: “Porton Down”) and as such not an obvious candidate for substantial contamination from leaf fall. Running railhead treatment trains over tracks vulnerable to wheel slide conditions seems to be the current strategy for dealing with the problem, but if it becomes necessary to apply this strategy to tracks apparently having as low a risk of contamination as those of the last few miles into Salisbury then the viability of the strategy itself seems to me to come into question.

User avatar
grovenor-2685
Forum Team
Posts: 3917
Joined: Sun Jun 29, 2008 8:02 pm

Re: K-crossing check rails

Postby grovenor-2685 » Fri Nov 05, 2021 4:09 pm

I'm sure the RAIB report will major on railhead conditions, and maybe sander maintenance/provision with references back to a number of earlier recomendations, but we are going to need to wait months to see it. Average time for producion of a report seems to be about a year.
Regards
Keith
Grovenor Sidings

Alan Turner
Posts: 643
Joined: Sun Jul 20, 2008 4:24 pm

Re: K-crossing check rails

Postby Alan Turner » Fri Nov 05, 2021 8:36 pm

grovenor-2685 wrote:
Alan Turner wrote:So I am assuming here that modern signalling does not follow the principles of AB and that two trains can occupy the same clearing point? Or am I missing a point here?
regards
Alan

Multiple Aspect Signalling (MAS) with track circuit (or axle counter) block requires a specified overlap to be clear beyond the next signal before a signal can be cleared. Any trailing points in the overlap must also be set and locked. The specified overlaps were 200 yds for 4 aspect signalling and 300 yds for 3 aspect signalling, as the driver in 4 aspect territory gets two warnings so stopping errors should be less. In this case the signal is set back from the junction to allow the 200 yds overlap before the clearing point of the junction. If the signal, for some reason had been closer to the junction then the points, locked for the first train, would have blocked the overlap and the train would have been signalled to stop at the previous signal.
Its the same principle as AB working except that the AB clearing point (equivalent to the overlap) was usually 440yds.


Thanks for the explanation Keith. I suppose the bit I had upper most was that in AB the signalman could not release the first clearing point until the signalman had established that the train had come to a stand. It would appear that under present day conditions that two trains can be allowed to move towards each other. In effect occupying the same clearing point.

regards

Alan

User avatar
grovenor-2685
Forum Team
Posts: 3917
Joined: Sun Jun 29, 2008 8:02 pm

Re: K-crossing check rails

Postby grovenor-2685 » Fri Nov 05, 2021 10:58 pm

No, the clearing point is the end of the overlap and is before the junction. If the clearing point/end of overlap extends past the junction then only one train would be allowed to approach, the other would be stopped one signal back, ie at the equivalent of the outer home which has its own clearing point.
Which is what I thought I said. :) (Last sentence of first para). There was a provision for sharing overlaps under certain conditions but a rare situation and I don't remember the required conditions now. But based on the likelyhood of two trains doing a SPAD simultaneously being very low. Even then if one train has the route then the overlap for the other would be blocked and the train held back. ie a shared overlap required both trains to stop before either could be routed through the junction. Facing points in an overlap are less of a problem as the overlap can go either way.
Regards
Keith
Grovenor Sidings

Alan Turner
Posts: 643
Joined: Sun Jul 20, 2008 4:24 pm

Re: K-crossing check rails

Postby Alan Turner » Sat Nov 06, 2021 8:42 am

grovenor-2685 wrote:I'm sure the RAIB report will major on railhead conditions, and maybe sander maintenance/provision with references back to a number of earlier recomendations, but we are going to need to wait months to see it. Average time for producion of a report seems to be about a year.


It appears to me quite astonishing that a train can slip for over 220yds (more in reality if the overlap finishes before the crossing) from a speed commensurate with the approach speed to a stop signal. The rail conditions must have been like ice.

regards

Alan

User avatar
Noel
Posts: 1972
Joined: Wed Jun 23, 2010 1:04 pm

Re: K-crossing check rails

Postby Noel » Sat Nov 06, 2021 12:16 pm

John Palmer wrote: It is my understanding that it functions not as a repeater of SY29's aspect but as a caution signal for SY29 and that it displays a double yellow aspect when SY51 is exhibiting a red aspect and SY29 a single yellow.


The numbers do suggest a connection, but if SY29R is not a repeater for SY29 and showing the same aspect, nor a 'normal' running signal, then the only other option would seem to be that it's restricted to Y-YR, as otherwise it could be a standard signal numbered SY27, and would still produce the same YY aspect if SY31 was R. Or am I missing something?

John Palmer wrote: SY29 stands about 860 metres in rear of SY51, whilst SY29R stands approximately 2.23 km / 1.39 miles in rear of SY29, according to my Google map measurements.


I have no knowledge of this particular area, but generally the signals are about 1,000 yards apart [+/- according to circumstances] in 4 aspect open line areas. This puts a gap of about 950 yards between SY29 and SY31, and 2,400 between SY29R and SY29, which seems a lot?
Regards
Noel

bécasse
Posts: 374
Joined: Sun Jun 17, 2018 8:26 am

Re: K-crossing check rails

Postby bécasse » Sat Nov 06, 2021 1:12 pm

Signals with R or RR suffixes are effectively "distant" or "outer-distant" signals which cannot display red aspects. They are "commonplace" on lines with SIMBIDS bi-directional signalling but can be found elsewhere, as here.

As to sliding on leaf-fall lubricated rails, in the relatively early days of Thameslink (before the problem was properly appreciated), I remember being told by a driver that one evening he had applied the brakes of his 4-CAR class 319 train to make a routine station stop somewhere in Hertfordshire where the linespeed was 100 mph and nothing appeared to happen, the unit sliding for well over two miles (almost to the next station, in fact) before eventually coming to a stand, fortunately with green signal aspects all the way. He told me that the most remarkable thing was that, once he had realised what was happening, he expected to have to take the train out of service with severe wheel-flats, but it transpired that although the train did have minor flats, the leaf-mould "lubrication" was so effective that he was able complete the journey all the way to Brighton.

User avatar
grovenor-2685
Forum Team
Posts: 3917
Joined: Sun Jun 29, 2008 8:02 pm

Re: K-crossing check rails

Postby grovenor-2685 » Sat Nov 06, 2021 1:15 pm

I don't know the current numbering policy for NR Southern but from my past life as a signal engineer the suffix R would imply that the signal does not have a red aspect because there is some reason not to stop there, otherwise it should have a standard aspect sequence, showing Y, YY or G.
Trains have been known to slide very much more than 200 yards in the leaf fall season, distances measured in miles. The RAIB initial report says that it started to slide as soon as the service brake was applied, so well over a Km in this case.
Regards
Keith
Grovenor Sidings

JFS
Posts: 813
Joined: Wed Dec 16, 2009 3:47 pm

Re: K-crossing check rails

Postby JFS » Sat Nov 06, 2021 2:01 pm

Things were different in 1974 and not necessarily worse - Sykes Lock and Block here to Salisbury East!

10 Salisbury Tunnel Jct.jpg


We must not jump to conclusions, but I would like a wager that the phrase "Approach Control" gets a mention in the final report. Anyone take me on?
You do not have the required permissions to view the files attached to this post.


Return to “Track and Turnouts”

Who is online

Users browsing this forum: ClaudeBot and 0 guests