Marcus' Model Railway Journey

9th October 2023

Nene Valley Railway recruits steam loco 45212 as Santa’s helper​

Snowy freight - 45212 hauls a demonstration freight on the Keighley and Worth Valley Railway banked by 45699 Bahamas on March 11, 2023

Locomotive 45212 will arrive at the Nene Valley Railway (NVR) in Peterborough in early November to help with Santa Operations.

The locomotive works part of the year for Riley and Sons Limited and part on the Keighley and Worth Valley Railway.

45212 was built in 1935, and is from a class of locomotive that ran in service until the very end of steam on the national railway network.

It is one of eighteen Black 5s in preservation, having moved straight out of mainline service into preservation.

The loco headed the final steam-hauled, revenue-earning service for British Railways, from Preston to Blackpool on Sunday 4 August 1968.

It will take part in NVR's Driving Experience courses and a one-day running day on Sunday 5 November, before assisting with hauling the First Class Santa services.

Details and tickets are available by clicking these links:

• Driving Experience Course, Saturday 4 November
• The Fenman with 45212, Sunday 5 November

10th October 2023


RAIB
News Story:

Member of staff struck by train at St Philips Marsh depot​

Published: 10th October 2023

Investigation into a member of staff struck by a train at St Philips Marsh depot, Bristol, 26 September 2023.

At around 13:15 hrs on 26 September 2023, a member of staff responsible for shunting trains was struck by a train that had just exited the main shed within St Philips Marsh depot, Bristol.

The member of staff sustained serious lower body injuries and some minor upper body and head injuries as a result of the accident.

Our investigation will seek to identify the sequence of events which led to the accident. It will also consider:

• the actions of those involved and the factors that may have influenced them
• the processes in place for train movements inside the depot, including the training and competence of staff involved with this activity
• the way in which the risks associated with train movements were being managed at the depot
• any underlying management factors.

Our investigation is independent of any investigation by the railway industry or by the industry’s regulator, the Office of Rail and Road.

We will publish our findings, including any recommendations to improve safety, at the conclusion of our investigation.

10th October 2023


RAIB
Press release:

Report 10/2023: Freight train derailment at Petteril Bridge Junction​

Published: 10th October 2023

RAIB has today released its report into a freight train derailment at Petteril Bridge Junction, Carlisle, 19 October 2022.

The accident

Summary of the accident

At 19:53 hrs on Wednesday 19 October 2022, five wagons in a freight train, reporting number 6C00, derailed as it was passing through 679A points, close to Petteril Bridge Junction, south-east of Carlisle station.

The train was approaching from the Settle direction and travelling at around 17 mph (27 km/h)when the derailment occurred.

Two of the derailed wagons fell from the bridge which takes the railway over the River Petteril, with one landing in the river and the other on the riverbank.

No one was injured in the accident, but there was significant damage sustained to the railway infrastructure which resulted in the closure of the routes from Carlisle towards both Newcastle-upon-Tyne and Settle for seven weeks.

Immediately after the derailment, it was identified that the leading wheelset of the ninth wagon of the train (number VTG12455) had a very large wheel flat (a flat spot on the wheel tread) on each wheel.

Extract from Ordnance Survey map showing location of accident at Petteril Bridge Junction

Overview of derailed wagons at Petteril Bridge Junction

Wheel flat observed on the leading wheelset of the ninth wagon immediately after derailment

Context

Location

The derailment occurred at a set of trailing points, in the direction of travel of the train, (numbered 679A) which is located at 307 miles 10 chains from London St Pancras.

This set of points provides access to the sidings for a disused oil depot, on the Down (northbound) Settle to Carlisle line, about50 metres to the east of Petteril Bridge Junction.

The Settle to Carlisle line remains double tracked as it passes through Petteril Bridge Junction.

The permissible speed at the point of derailment, and through Petteril Bridge Junction, is 20 mph (32 km/h).

The permissible speed on the route of the train before the point of derailment is 60 mph (96 km/h), reducing about 140 metres before Petteril Bridge Junction.

At the point of derailment, the gradient is level after a long descent from the north Pennine hills.

The signalling at the derailment location was controlled from the signal box at Carlisle.

However, the rest of the journey of the train over the Settle to Carlisle line was controlled from a series of signal boxes operating under the ‘absolute block’ system.

Aerial view of the derailment site, showing positions of derailed wagons

Organisations involved

GB Railfreight Limited (GBRf) was the operator of train 6C00 and the employer of the train driver.

It was also the employer of the ground staff operator who prepared the train at Clitheroe.

VTG Rail UK Limited is the owner of the wagons involved in the accident.

It is the entity in charge of maintenance for the wagons, with the maintenance work itself being contracted out to DB Cargo Limited at Clitheroe.

Network Rail is the owner and maintainer of the railway infrastructure, and the employer of the signallers along the route.

It is also responsible for the seasonal management of rail adhesion along the route.

Wabtec Faiveley UK3 and Knorr-Bremse Rail Systems UK are the respective manufacturers of the brake cylinders and the brake distributor on the wagons.

GBRf, VTG and Network Rail all freely co-operated with the investigation.

Wabtec Faiveley and Knorr-Bremse Rail both freely provided technical assistance during testing of the components from the wagons.

Train involved

The train consisted of a class 66 diesel-electric locomotive (number 66739)hauling 14 JPA tank wagons that were fully loaded with powdered cement.

Each JPA wagon was carrying a load of approximately 80 tonnes of cement, with the loaded wagons each weighing about 101 tonnes.

The maximum permissible speed of the loaded train was 60 mph (96 km/h).

Train 6C00 was the 17:15 hrs freight service from Clitheroe Castle Cement sidings to Carlisle Network Yard (located to the north of Carlisle station).

The train was scheduled to continue to Mossend, near Glasgow, the following day for unloading.

The same locomotive had brought the empty wagons south from Carlisle to Clitheroe that morning.

The wagons were part of a sub-fleet that was operationally constrained to the Clitheroe to Mossend traffic flow.

Other sub-fleets of similar JPA wagons operated on different traffic flows out of Clitheroe, Hope (Derbyshire), Padeswood(Flintshire) and Dunbar (East Lothian) cement terminals.

66739 working light engine 0E30 1134 Whitemoor Yard L.D.C Gbrf to Doncaster Up Decoy at March West Jct 13th April 22

Equipment involved

Each of the JPA wagons was fitted with two TF25 type bogies.

One bogie on each wagon was fitted with a manually operated handbrake acting on both wheelsets in that bogie.

The train was fitted with a single-pipe air brake system, operating on all the wheels of all the wagons.

A single train brake pipe connects all the wagons along the train, both supplying air to the wagons and controlling braking.

Air pressure in the pipe is generated by a compressor on the locomotive, and the driver regulates the pressure in the pipe to control the brakes on the train.

To release the brakes when running, air pressure is created in the brake pipe.

The pressure in the brake pipe when the brakes are fully released is normally 5 bar.

Reducing the air pressure in the brake pipe proportionally increases the force of the brake application.

Once the brake pipe pressure decreases to 3.3 bar, ‘full service' braking is applied.

Each wagon is fitted with a single brake distributor and a separate auxiliary air reservoir located under the centre of the wagon body.

A reduction in the pressure in the train’s brake pipe causes the distributor to operate and admit air from the auxiliary reservoir to the two brake cylinders on each bogie of the wagon.

Air pressure in the brake cylinder then moves a piston which, in turn, acts through a system of rods and beams to apply the brake blocks to the wheel treads.

As with most freight wagons, JPA wagons are not fitted with a wheel slide protection system.

Such systems, common on passenger trains, detect wheelsets starting to slide when braking and automatically reduce the brake force being applied to the sliding wheelset until the system determines that they are no longer sliding.

A JPA wagon and the location of its distributor (auxiliary reservoir hidden under wagon)

The brake equipment fitted to the bogies of a JPA wagon is known as the block force compact bogie-mounted (BFCB) system.

The BFCB system was developed by Faiveley Transport in Sweden (now part of Wabtec Corporation).

It has been in use since 2001 and is widely used both within the United Kingdom and internationally.

The BFCB system consists of two transverse beams fitted between the axles.

The ends of both beams are attached to brake block holders, suspended from brackets on the bogie by hangers.

The inner primary beam (the one nearest the centre of the wagon) carries a pair of brake cylinders that operate by extending longitudinal rods, which pass through the bogie frame.

These spindles push the primary and secondary beams apart, pressing the brake blocks against the wheels with equal force.

The system is self-adjusting, automatically taking up excessive slack (for example, caused by brake block wear) by means of slack adjusters within the brake cylinders.

BFCB braking system for one bogie (brake blocks not shown)

Side view of BFCB braking system

Staff involved

The driver had worked on the railway for 39 years, with the last 22 of those driving freight services.

The driver regularly drove class 66 locomotives, and often drove this service over the Settle to Carlisle route.

The driver held the relevant competencies for driving class 66 locomotives and had up-to-date route knowledge for the route from Clitheroe to Carlisle.

GBRf stated that there was no record of this driver having been involved in any relevant safety-related incidents.

The ground staff operator who helped prepare the train at Clitheroe had worked in that role for three years and was acting as a mentor for a new trainee.

Before that the operator had worked on the railway in another role for eight years.

GBRf stated that there was no record of this staff member having been involved in any safety-related incidents.

The signaller at Culgaith signal box had been a signaller for 20 years and had operated Culgaith signal box for most of that period.

This signaller held the relevant competencies required to operate Culgaith signal box at the time of the accident.

Network Rail stated that there was no record of this signaller having been involved in any relevant safety-related incidents.

The signaller at Low House signal box had worked as a relief signaller for 23 years, covering several signal boxes on the Settle to Carlisle line, including both Low House and Culgaith.

This signaller held the relevant competencies required to operate these signal boxes.

Network Rail stated that there was no record of this signaller having been involved in any relevant safety-related incidents.

External circumstances

Records from local weather stations along the route of the train, between Settle and Carlisle, recorded no rainfall on the evening of 19 October 2022.

However, witness evidence was that the rails were damp at both Culgaith and Low House signal boxes.

Weather stations at Dent and Garsdale recorded the average wind speeds as being between 15 and 20 mph (24 and 32 km/h).

This, combined with the time of year, meant that there was a high risk of leaves falling from trees and being blown around.

The recorded average air temperatures at Garsdale were around 9°C, very close to the recorded dew point of about 8°C.

Temperatures at lower altitudes on the Settle to Carlisle route were a little warmer at up to 12°C.

It was dark at the time of the derailment, although the train had left Clitheroe during daylight, with dusk falling on the initial climb towards Blea Moor.

Route of train 6C00

Background information

Route of the train

After departing from Clitheroe cement terminal, train 6C00 started its journey from Horrocksford Junction over the (primarily freight-only) route which runs from there to Hellifield.

This route was mostly on rising gradients, with a maximum permissible speed of 30 mph (48 km/h) for freight trains.

At Hellifield, the train joined the route via Settle Junction to Carlisle (the Settle to Carlisle line).

After a gentle climb, the route includes a long, steep ascent into the north Pennine hills, with a 15 mile (24 km) stretch at an average gradient of 1 in107.

There is then a flatter 10 mile (16 km) long stretch along the top of the hills before a 15 mile (24 km) long descent at an average gradient of about 1 in 120,followed by a gentler descent towards Petteril Bridge Junction.

Gradient profile for the route of train 6C00

Pre-existing wheel flats on the train

Eleven of the wagons in train 6C00, including the ninth wagon, had passed a wheel impact load detection site (known as a ‘Gotcha’ site) at Braidwood, near Motherwell, on 13 October 2022, six days before the derailment.

This site recorded small impact forces on the rails from two wagon wheelsets.

These were consistent with wheel flats on the rotating wheels, usually caused by the wheels having previously slid along the railhead without rotating for a short distance.

These forces were small and well below the level that would have required any alarm to be raised or any action to be taken.

These flats were recorded on the fourth (rear) wheelset of wagon VTG12459 (the eighth wagon from the front of train 6C00) and the third wheelset of wagon VTG12450 (the eleventh wagon from the front).

No wheel flats were recorded in the 13 October data for wagon VTG12455 (the ninth wagon from the front).

Examination of the eighth and eleventh wagons after the derailment showed that these pre-existing wheel flats were still visible.

RAIB has no evidence to suggest that the existence of either of these wheel flats contributed to the derailment on 19 October 2022.

Pre-existing wheel flats on wagons in train 6C00

The condition of the ninth wagon before the derailment

The JPA wagons were subject to an annual vehicle inspection and brake test (VIBT), and a four-monthly planned preventative maintenance (PPM).

All the wagons in train 6C00 were found to have been maintained within the required periodicity.

The ninth wagon had undergone its last VIBT five months before the derailment and its last PPM one month before the derailment, with no relevant faults or defects found.

Before departure on the day of the derailment, the brake blocks on the sixth wagon were replaced due to wear, and the whole train was subject to an in-service inspection, with no defects found.

Brake system testing

Following the accident, the ninth wagon was recovered from the River Petteril and transported to a secure storage site.

Examination of the braking system showed that all the main components were still intact, although several air pipes had become dislodged due to the distortion in the wagon’s underframe caused by the accident and the wagon’s recovery.

Reconnection of the dislodged pipework made the brake system airtight, and able to be pressurised and tested.

A series of functional tests was then undertaken on the wagon’s braking system in conjunction with VTG and DB Cargo.

The wagon’s brakes were tested using a portable test rig.

This simulated the brake pipe connection from a locomotive and allowed the wagon’s brakes to be applied and released in line with the requirements of the VIBT procedure.

The brake system was seen to operate as intended and the brakes applied in accordance with the criteria required by the testing.

It was noted that the time the brakes took to release when commanded was longer than required by the VIBT (35 seconds compared to the 20 seconds specified), although the brakes did release fully.

Subsequent examination showed that water contamination had entered the air system, probably during the five- week period between the derailment and reconnection of the dislodged pipework.

38 During the brake tests, the forces applied by the brake blocks on the wheels were measured using instrumented brake blocks.

These showed that the brake block forces on the eight wheels on the ninth wagon were relatively uniform and in line with the expected design values.

This demonstrated that excessive brake force was not being applied to the leading wheelset that had slid and developed the wheel flats.

The ninth wagon (VTG12455) after recovery

Brake distributor testing

The brake distributor was subsequently removed from the wagon and bench tested, with the assistance of its manufacturer, Knorr-Bremse.

These tests showed that it was functioning as designed, with the air pressure being applied to the brake cylinders responding correctly to changes in the brake pipe pressure.

The slow brake release time observed during the brake system testing was also found during the bench test.

Subsequent dismantling and examination of the brake distributor found water inside the air chambers and corrosion in a choke used to control the speed of the brake release.

The last VIBT before the derailment found the brake release timings to be correct, suggesting that the corrosion was not present then.

It is likely that the corrosion was initiated by water getting into the air system of the wagon after the derailment, when the air system pipework was disturbed and open to the atmosphere and the rain.

No other defects were found during the examination of the distributor.

If the brake release timings had been extended from 20 seconds to 35 seconds, the effect would have been to extend the brake application on all the wheelsets of the ninth wagon by a few seconds during each brake application.

RAIB found no evidence of abnormal brake wear on this wagon to support this possibility.

It is likely, therefore, that the slow brake release times were a result of water ingress which occurred after the derailment and that the brakes on the ninth wagon were working within specification at the time when the wheel slide was initiated.

Brake cylinders testing

The brake cylinders were removed from the BFCB beams and tested in conjunction with the manufacturer, Wabtec Faiveley.

All four cylinders from the ninth wagon were found to pass bench tests for force and stroke length, as per their design.

The brake cylinders were also dismantled and inspected to identify any defects.

One of the brake cylinders on the leading bogie was found to have a broken piston spring and a bent guide plate.

This spring was intended to maintain alignment of the brake piston in its cylinder.

Although the piston spring also assists the piston to return to its position after application of the brakes, the main return spring, which was still intact, was configured to do this on its own if necessary.

Wabtec Faiveley stated that failure of a piston spring was very rare, although not entirely unknown.

The piston spring had broken in two locations and RAIB commissioned a metallurgical examination of it to determine the nature of the breaks.

This found that the first break showed signs of fatigue cracking that would have developed over many cycles of operation in normal service.

The second of the breaks was due to the dislodged spring being trapped in the brake cylinder mechanism and being crushed.

The conclusion of the examination was that fatigue had initially led to the spring failing and becoming dislodged, resulting in the spring and its guide plate becoming loose in the mechanism and subsequently becoming crushed during service operations.

The metallurgical report concluded that it was likely that the cracking, and both the resulting breaks, had occurred during normal service, before the derailment.

If the dislodged spring and guide plate had become wedged in the brake cylinder mechanism, it is possible that it could have resulted in the brake force on that bogie not being fully released after a brake application.

The maximum force that could have been applied would have been a normal full service brake application.

However, this would have resulted in the brakes being applied to both wheelsets in the bogie, and not just to the wheelset that slid, which would also cause additional brake wear on the brake blocks for both wheelsets.

Examination of the wagon after the derailment did not show any significant difference in the brake block wear between the two bogies on the ninth wagon, nor any evidence of excess heat having been generated, suggesting that any possible abnormal brakeforce from this brake cylinder had a minimal effect.

Internal arrangement of the brake cylinders

The sequence of events

Events preceding the accident

At 12:27 hrs on 19 October 2022, train 4N00 arrived in the loading sidings at Clitheroe cement terminal from Carlisle, after reversing into the terminal from Horrocksford Junction.

Train 4N00 conveyed twelve of the empty wagons that would later form train 6C00, as well as an additional thirteenth empty wagon that was to be detached at Clitheroe for maintenance.

Train 4N00 on its southbound journey at Appleby

Track diagram of the cement terminal at Clitheroe

Between 12:35 hrs and 16:40 hrs, 12 wagons were loaded one-by-one as the train drew forward through the loading hopper building.

During this process, two already-loaded wagons were attached to the end of the train, and the empty wagon next to the locomotive was detached and taken away by the terminal's shunting locomotive.

Once all the wagons were loaded, the driver overcharged the train’s brake pipe as part of the train’s pre-departure process.

This admits a higher pressure than required in normal operation to allow the control pressures in the wagons’ brake distributors to normalise, thus reducing the risk of the train setting off with a dragging brake.

GBRf procedures required the overcharge to be carried out on the arrival/departure line after the locomotive had run round the train rather than before, but RAIB has no evidence to suggest that this minor non-compliance contributed to the derailment.

From 16:40 hrs to 17:25 hrs, the now‑loaded train of 14 wagons drew forward to the arrival/departure line.

The locomotive was detached and ran round to the opposite end of the train ready for the journey north.

After getting authority from the signaller at Horrocksford Junction, the driver propelled the train out on to the main line, while the ground staff operator carried out the pre‑departure roll- by observational checks.

A public webcam4 at Horrocksford Junction, combined with bodycam footage from the ground staff operator, showed that all the train's wheels were freely rotating at this location.

At 17:25 hrs, train 6C00 departed from Horrocksford Junction towards Hellifieldand Carlisle.

Eight minutes later, at 17:33 hrs, the driver carried out a running brake test, as required by the railway Rule Book, and achieved a speed reduction of 11 mph (18 km/h).

The driver applied the brake for 17 seconds during the running brake test.

Between Horrocksford Junction and Hellifield, the driver made two very short (6 seconds each) and light (<25% of full service) brake applications to manage the train’s speed within the 30 mph (48 km/h) permissible speed limit.

At 18:00 hrs, the driver made another similar brake application to regulate the train's speed to 15 mph (24 km/h) at Hellifield.

From Hellifield to Settle Junction, the driver increased the train’s speed towards the maximum permissible line speed of 60 mph (96 km/h) before starting the climb up towards Blea Moor.

On that climb, the driver kept the locomotive's power handle in its maximum position.

Despite this, the train’s speed slowed to around 20 mph (32 km/h) due to the weight of the train, the rising gradient and low levels of wheel/rail adhesion.

At 18:22 hrs, the train passed a public webcam at Horton-in-Ribblesdale station, and an audible wheel flat can be heard on the recording around the middle of the train.

Due to the absence of significant braking up to this point, RAIB considers it likely that this wheel flat was one of those that were pre-existing on the train before the day of the accident.

Train 6C00 passing Horton-in-Ribblesdale

After Blea Moor, the gradient levelled out and the train was able to increase speed again.

No brake applications were made after Hellifield until the train approached a 30 mph (48 km/h) permanent speed restriction through Dent station.

At this point, the driver made a brake application of approximately 35% of full service, lasting a maximum of 32 seconds, which reduced the train’s speed by 19 mph (31 km/h).

At 18:50 hrs, the train passed another public webcam at Garsdale, by which time it was getting dark.

A review of this footage by RAIB did not identify any visual or audible signs of a problem with the train.

55 At 18:55 hrs, the train passed Ais Gill and the driver made the first of many short brake applications to control the train’s speed on the long, steep descent to Appleby.

Most of these brake applications were considerably less than 20 seconds long.

At 19:03 hrs, the train passed Kirkby Stephen and left marks on the railhead at sets of points there showing that a non-rotating wheelset had passed over them.

Similar marks were left at every subsequent set of points along the route of the train between this location and the point of derailment.

57 At 19:09 hrs, train 6C00 passed a southbound passenger service on the opposite line (train 2H97, the 18:24 Northern service from Carlisle to Leeds).

Rear-facing closed-circuit television (CCTV) footage from this service momentarily showed a small amount of sparking at the wheel/rail interface on one wheelset of train 6C00.

This sparking was only visible once the contrast of these CCTV images had been altered and it is unlikely that the driver, or any other railway staff, on train 2H97 would have been able to see it as train 6C00 passed.

Train 6C00 passing Garsdale

Rear-facing CCTV image from train 2H97showing sparking at a wheelset on train 6C00

At 19:31 hrs, train 6C00 passed a private residence at Little Salkeld.

CCTVfootage from this residence, which was fitted with a night vision system, showed sparking from the leading wheelset on the ninth wagon.

This sparking would not have been as visible to the naked eye as it appears in the night vision images.

Sparking from the ninth wagon in train 6C00 at Little Salkeld

At 19:37 hrs, train 6C00 passed Lazonby & Kirkoswald station.

Here the train left the last of three track sections whose status were displayed in Culgaith signal box(train 6C00 had passed Culgaith signal box at 19:26 hrs).

The status of each track section is determined by axle counters.

These count the number of wheels that enter and then leave a section of track and provide an indication to the signaller to show whether a train is present in that section.

After train 6C00 had left the last track section, all three sections continued to show as occupied (that is to say, a train was indicated to be present).

The signaller at Culgaith contacted the next signaller at Low House signal box to check on the progress of train 6C00.

At 19:42 hrs, train 6C00 passed Armathwaite station, where the station CCTV footage showed a small amount of sparking at the leading wheelset on the ninth wagon.

One minute later it passed Low House signal box.

The signaller watched the train pass to check that it was complete after the contact from Culgaith signal box.

The signaller at Low House did not see or hear anything wrong with the train as it passed by.

Sparking from the ninth wagon in train 6C00 at Armathwaite station

Events during the accident

By 19:53 hrs, train 6C00 had slowed to 20 mph (32 km/h) as it approached Petteril Bridge Junction.

As the train passed over 679A points in the trailing direction, the leading wheelset of the ninth wagon derailed to the left and started to severely damage the track as both of its wheels ran along the sleepers instead of the railhead.

This damage led to subsequent wheels of the following wagons becoming derailed to the left.

The derailed wagons destroyed the set of points at Petteril Bridge Junction, which was located on top of the bridge over the river.

The derailed wagons, which were running to the left of the track, ran against the side of the bridge parapet wall, resulting in the wall and part of the railway formation falling into the river.

This led to the ninth and tenth wagons falling off the left side of the bridge, coming to a rest with the ninth wagon upside down in the river and the tenth wagon on its side on the riverbank.

The rear of the eighth wagon also derailed to the left as it was dragged sideways by the front of the ninth wagon.

The coupling between the fifth and sixth wagons separated as the ninth and tenth wagons fell off the bridge, resulting in the brake pipe becoming separated and the train's brakes applying automatically.

Damage to the track and bridge at Petteril Bridge Junction

Events following the accident

The automatic brake application stopped the train, and the driver contacted the signaller, thinking that the train could have become divided.

Although the adjacent line to the right of the train was still open to traffic, it was unlikely to have been obstructed because the wagons had primarily derailed to the opposite side.

At the same time, the controlling signaller in Carlisle power signal box received indications that some track detection equipment had failed.

This was because of the track damage caused by the derailment.

The driver got out of the locomotive to look back along the train and saw that some of the wagons had become derailed and that the train had become divided.

The driver immediately informed the signaller, who took steps to block all lines to traffic.

Due to difficulties with erecting a suitable crane close enough to the derailed wagons, it took three and a half weeks for all of them to be recovered from the site.

It was seven weeks after the derailment before the lines to both Newcastle and Settle could be reopened to traffic after extensive rebuilding work to the side of the bridge over the River Petteril.

Analysis

Identification of the immediate cause

Train 6C00 derailed, while travelling through 679A points in the trailing direction, due to a large false flange that had developed on the leading wheelset of the ninth wagon during the journey and this had not been detected.

Immediately after the derailment, the leading wheelset of the ninth wagon was found to have a large wheel flat on each of its wheels.

These wheel flats had worn approximately 25 mm into the tread surface of the wheels, resulting in a large false flange on the outside edge.

The wheel flats were approximately 290 mm long.

Wheel profile of the worn wheels, compared to the reference wheel profile

The length of the wheel flats

Marks at the site showed that, during the initial stage of the derailment, the outer edge of the false flange on the leading left wheel of the ninth wagon had become trapped between the converging stock rail and switch rail at 679A points.

In the absence of a false flange, the outer edge of the wheel would have passed over the top of the converging rails, with the contact surface of the wheel transitioning from the switch rail to the stock rail.

Route of the false flange at trailing points

Switch rail at 679A points

The marks revealed on the inside of the stock rail as the outer edge of the trapped false flange on the leading left wheel ran along the inside of it at the tip of the switch rail as the wheel dropped into the four-foot (the gap between the running rails).

This event, and the subsequent stage of the derailment mechanism is illustrated in figure below.

The left wheel dropped inside the left rail after it passed the end of the switch rail, with its outer face then pushing the left rail outwards, while the flange of the right wheel was pushing against the right rail and trying to climb up onto it.

The track fixings holding the left rail in position then gave way and the rail shifted laterally and started to roll over.

A few metres further on, the left wheel collided with a check rail that was mounted inside the left rail as part of the next set of points.

This caused the left wheel to climb over the left rail, and the right wheel to drop into the four-foot.

This derailment mechanism, where a wheel with a false flange on the outside edge derails at a set of trailing points has been seen at a number of previous derailments.

Derailment mechanism

Illustration of locations where false flange marks will be visible at points

False flange marks at Howe & Co’s Siding

False flange marks at Kirkby Stephen