A derailment  occurred at 2.55 p.m. on the 27th June in a deep cutting near the southern portal of the Merstham Quarry Tunnel between Coulsdon and Earlswood, on the London-Brighton Main Line in the Southern Region, British Railways. The 2.25 p.m. electric passenger train, Victoria to Littlehampton, comprising 10 coaches, was travelling on the Down Main at a speed of about 55 m.p.h. when three out of the last four coaches became derailed on a curve in a deep cutting and obstructed both lines. It was a very hot day and the track distorted under the train. Fortunately none of the 150 passengers was injured, though three subsequently complained of shock. They were detrained quickly and conducted to Merstham Station, mile away on the Redhill line, where they left for their destinations by local trains.

The derailment was observed by Signalman King of Merstham box and he took commendably prompt steps to warn the Earlswood Station signalman, who was thus able to stop the 2.25 p.m. fast electric train Brighton to Victoria, which was approaching at high speed on the Up Main. Electric current on the Down line was cut off immediately by the opening on short circuit of the breakers at the Three Bridges Electrical Control Centre ; current on the Up line was cut off at 2.59 p m . on request from the Earlswood signalman. The Main line services were diverted and there was comparatively little delay to traffic.
The train consisted of a six and a four coach set, each with a motor-coach at front and rear. The total length over buffers was 665 feet and the weight 426 tons. The coaches were standard Central Section main line electric service corridor stock, except the third which was a Pullman car. They were all of composite construction with timber framing and steel panelling on steel underframes except two motor-coaches and the Pullman car, which were all steel. Heavy duty self-contained buffers and screw couplings were fitted throughout. Westinghouse brakes operated on all wheels and the overall brake power was 61.4% of the 
total weight. Cast iron brake blocks were fitted to the wheels of the motor-coaches and non-ferrous blocks to all others.

The train stopped with the leading coach 675 yards beyond the initial point of derailment. The first six coaches remained on the rails and the next three were buffer locked, with the leading bogie of the seventh vehicle, a motor-coach, derailed to the right, and the leading bogie of the ninth coach to the left. The motor-coach at the rear broke away, owing to the coupling lifting from the drawbar hook, and came to rest with its leading end 200 yards beyond the point of derailment ; both bogies were derailed to the right, with the right-hand wheels pressing against the Up track which was slewed over and obstructed.

Damage to rolling stock was slight and was confined to the last four vehicles ; the sole bars of three of them were bent, one door light was broken, and gangways were torn and twisted. The Down line was badly distorted at the point of derailment ; it was destroyed for 120 yards and sleepers were damaged for a further 440 yards. All of this track was relaid during the night with 113 lb. F.B. rail, and both lines were reopened for trafficby 7.0 a.m. the following morning.

The weather had been warm and dry for several days and on the day of the accident it was very hot.

1. SITE: It will be seen from the accompanying drawing that the London-Brighton four-track line, which runs almost due south, bifurcates at Coulsdon North. The Through lines run direct to Earlswood passing under the North Downs in Merstham Quarry Tunnel, 2,113 yards long. The Local Redhill lines pass under the Through on the north side of the Downs and then enter a separate tunnel, which emerges some 175 yards to the north-west of the Quarry tunnel. Merstham Station is about 4 mile further on, after which this route passes through Redhill Station and then rejoins the other at Earlswood.

The Down Through rises from Coulsdon to a summit at the north end of the tunnel, and then falls at a gradient of 1 in 206, which changes to 1 in 230 at the southern exit, where it enters a deep cutting 450 yards in length on a right-hand curve of 75 chains radius, before running onto a 20 ft. high embankment. Two bridges cross the line as indicated ; Rockshaw Road is at a high level, hut the clearance under Peters Bridge, near the tunnel mouth, is only 13 ft. 7 inches and head room is so restricted that it is not possible to lift the tracks at this point. 

Both Through lines are continuously track circuited and equipped with automatic colour light signals ; the relevant Up and Down Signals, Nos. C.A. 43 and C.A. 42 are indicated on the drawing. A telephone at each is connected through an omnibus circuit to the adjoining signal boxes.

2. TRACK: In May 1941, the Down line in the cutting was re-laid with new 95 lb. R.B.S. rails on timber sleepers. Rails were 60 ft. long, except for two pairs of closures at the tunnel mouth. There were 22 standard and two joint sleepers per 60 ft. length with standard chairs and through bolt fastenings. Keys were mixed steel and hardwood and rail joints in the cutting were held by two-bolt fishplates.

1 examined the track at about 5.30 p.m. on the day of the accident, when it was still very warm. The first mark of derailment was clearly seen on the low rail some 40 yards from the tunnel mouth. The second mark was 11 ft. further ahead and, from subsequent marks on the c h a h and sleepers, it is clear that the first bogie to be derailed left the track at this point, where the rails were distorted 20 inches towards the cess. Thereafter the distortion increased and the rails took up an S curve, with marks of other bogie wheels visible on the sleepers.

There was a shortage of ballast from Peters Bridge to the point of derailment ; there was none on the shoulders of the sleepers on the cess side, the boxing was not up to standard and on the 6 ft. side there was also a deficiency, particularly near the bridge, where there was no ballast between the ends of the sleepers. The 6 ft. space under the bridge had been opened and filled with rubble covered with timber boards, in an attempt to overcome drainage difficulties caused by clay oozing up in wet weather.

Three or four weeks before the accident the maintenance gang had riddled the ballast under the bridge and for two lengths on the south side of it ; this had lowered the height of the boxing and, by loosening it, had reduced its lateral resistance. This gang had also been lifting joints just before the arrival of the train and the places where the sleepers had been opened out for fettling and not refilled are shown on the drawing.

Cant, gauge and alignment of the Down track in the tunnel were well maintained. High and low rails weighed 87 and 89 lbs. per yard respectively ; there was some sidecut on the high rail but it was not excessive. Fastenings were holding well and sleepers were sound ; some of them under Peters Bridge had been "pumping”.

There were no signs of creep, though many of the rails were butting or showing very small expansion gaps. These were measured later in the evening and the results are recorded on the drawing. From this it can be concluded that most, if not all, of the joints in the cutting were tight during the extreme heat of the afternoon and many of the rails were under compressive stress.

I had a number of fishplates removed and examined their condition. I was informed that they had been greased in April but there were few signs of grease on either rails or plates ; there were indications of binding which might have restricted movement. I formed the impression that the work bad not been well done and that the grease had dried quickly. 

WEATHER

3. The weather during June had been warm and dry and on the day of the accident maximum shade temperature at Croydon (7 miles from Merstham Quarry Tunnel) was 87° F., and the maximum solar radiation temperature at Greenwich Observatory was 142°F.

On 1st July, temperature tests were made in the cutting and the hottest place was near No. 7 joint in the Down track ; o n this day the sun temperature rose to 124°F. a t 1.15 p m . compared with a maximum solar radiation at Greenwich of 138°F. and maximum shade temperature at Croydon of 79°F. Solar radiation temperature is usually about 5°F. higher than sun temperatures, so it can be assumed that the maximum sun temperatures in the cutting during the period ranged from about 110°F. on 19th June to 135°F. on 27th June when the rise in rail temperature from the minimum during the previous night was probably about 75°F. 

In tunnels, normal gaps are allowed for the first four joints at each end, but as the average range of temperature is small, gaps for the remaining joints are reduced to 1/16th inch, or in cold weather a little more, sufficient to permit the rail being removed and replaced in case of breakage.

The theoretical expansion of a 60 ft. rail is Bin. per 25°F., and this produces a compressive stress of approximately 2 1/4 tons per square inch or 21 tons in a new 95 lb. B.H. rail when fully restrained. Thus, the expansion allowances are sufficient to cover all normal conditions, though it is not practicable to cater for extremes of cold and heat, which may range from 10°F. (22°F of frost) to 140°F., producing an expansion of over 5/8in. compared with maximum rail gap of from 1/2in. to 5/8in. Consequently in very hot weather rails may be in compression, e.g., if they were laid with tin. gaps at a temperature of 60°F., and were able to expand freely, not only would joints be closed but in addition a compressive stress of 2 3/4 tons per square inch would be set up when the temperature rose to 135°F.

5. FISHPLATES: On 21st February, 1949, the Divisional Engineer sent the following letter to all his Permanent Way Inspectors :-
''In accordance with Instruction No. 8 (d) of the Green Book of Instructions to Engineer's Department staff the work of greasing fish plates should be completed by the 3lst March. Will you please therefore put this work in hand, commencing the 28th February, if you have not already commenced.

Each pair offish plates is to be taken off and thejish plates turned end for q d so thaf the outside plate becomes the inside plate and vice versa, but the plates must on no account be turned upside down.

When the fish plates are taken off to be turned, care must be taken to see that all rust is removed from the fishing angles of both the fish plates and the rails before the grease is applied, and at the same time the expansion of the joints must be attended to.

As in previous years, the material to be used for greasing of fish plates is cleaned axle oil and petroleum residuum, the proportion for mixing being 3094 oil to 70% petroleum residuum, approximately two gallons of the mixture being required per single line mile of track.

You have received from the Stores Department the material for carrying out of the work and while it is in hand please let me have a report by not later than each Monday morning of the number of pairs of  fish plates greased during the week ending the previous Saturday, so that I may be aware of the progress being made, and on completion of the whole of your district please indicate on the last return that the work has been finished".

6. GENERAL HEAT PRECAUTIONS On 26th April, 1949, the Divisional Engineer issued the under- mentioned instructions, which were a repeat of those issued in 1948 :-

"Will you please draw the attention of all your Gangers to the possibility of track buckling during thejirst spell of hot weather after the winter. Although the winter set taken up by the track is eased to some extent during the greasing offish plates, the track is more likely to buckle under the first heat than it will do when it has adjusted itself to warmer conditions.
The attention of Gangers should also be drawn to the fact that track which has been relaid or reballasted for twelve months or less is much more liable to give trouble than older track.
Buckling should not take place if proper precautions have been taken, i.e., by adjustment of expansions, pulling back of rails, ballast kept to top of sleepers, keys driven tight, etc.
As a rule a Ganger should know those sections of his length which are susceptible to buckling, and under the conditions set out above, such stretches of track should be kept under special observations.

Gangers should also be njarned that buckling can well be caused by using a jack during the hottest period of the day.

Please draw the attention of all Gangers to these points and acknowledge receipt”.

EVIDENCE

7. Motorman H. H. Howell, in charge of the Littlehampton train, said that from Victoria to Merstham Quarry Tunnel everything was working satisfactorily. The brakes had been tested before he left and he made the normal stop at East Croydon. After passing the summit at the entrance to the tunnel he shut off power and was coasting at a speed of about 55 m.p.h. when he emerged from the southern portal. Almost immediately afterwards he noticed some rough riding and applied the brake so as to reduce speed. The lurching became severe near Signal C.A. 42 but before he could make a second application he saw the air pressure in the train pipe drop to zero ; the train came to rest some 400 yards further on.

8. Passenger-Guard H. V. Fish said he worked the Littlehampton train from Victoria, where he tested the brakes before leaving. 

He travelled in the brakevan at the rear of the sixth coach and had an uneventful journey until his train had passed through the tunnel, when he felt a jerk and the carriage seemed to sway. At first he was not particularly alarmed, but on looking back through the door window he saw a coach swinging about and throwing up a cloud of dust.

9. Other Railway servants confirmed the trainmen's evidence. Motorman G . N. Cummings, who was in the cab learning the road, felt a shuddering movement and saw the driver apply the brake. Divisional Inspector W. J. Holmes in the fifth coach felt a sharp jerk as if, something was dragging at the rear, Goods Guard H. E. Hayler in the front of the seventh coach felt the bogie make "a leaping movement", and Motorman C. J. Hay, in the front of the ninth coach, said "it gave a terrific lurch up in the air and down again; it swayed from side to side and for just one moment it seemed to ride on an even keel and then started to sway again". He realised it had left the rails and immediately pulled the communication cord.

10. Ganger A. Stone, who was in charge of the Main line length of 2 1/4 miles from the south end of Quarry Tunnel to the north end of Redhill Sand Tunnel, said his gang had just finished work near the Quarry Tunnel mouth when the Littlehampton train approached. He stood at the south end of Peters Bridge clear of the Down cess and watched the track as the train passed over it. 

After the first four or five coaches had gone by, he saw the line suddenly bulge out towards the cess ahead of him, he heard a bang and the train was enveloped in a cloud of dust. Immediately it had passed he telephoned from Signal C.A. 43 to the Earlswood Signalman and told him to block the road. Stone also sent men with flags and detonators to protect the line in each direction. Sub-Ganger J. Packham, who was standing on the Up side, also saw the track move outwards after the first six coaches had passed. Both men were confident that the alignment was correct before the train arrived and that the distortion took place under it.

11. At 7.30 a.m. on the morning of the accident Stone hegan walking his length towards Earlswood, returning along the Down line to the Quarry Tunnel. He found a few loose keys, hut nothing unusual, and he was quite sure expansion gaps were normal ; most of them were open about t inch. In the mean- while, Packham with the rest of the gang had started work on the Down line between Rockshaw RoadBridge and Signal No. C.A. 42, where they opened out sleepers preparatory to packing them. Stone returned to his men at about 10.0 a.m., and told them there were some joints which required lifting near Peters Bridge. He then left again to attend a joint at the Redhill end of his length and did not come hack till after lunch.

He said that by this time (1.0 pm.) it was very hot in the cutting and rail joints were very nearly closed, hut it was still cool near the tunnel mouth. His men had already removed the ballast a t several joints and Stonedecided to complete the work. He lifted the track at each place in turn using a jack under the rail. He said he only raised it about an inch and spread two canisters of granite chippings evenly over each sleeper bed ; then he removed the jack and allowed the work to he consolidated by a train or two before the ballast was replaced. He packed both the high and low rail at several joints ; two were lifted before the 2.0 p.m. 
Victoria-Brighton train passed at 2.20 p.m. and two more just before the Littlehampton train arrived. After packing the sleepers he verified that gauge was correct and he checked the alignment and cant from the monuments.

Stone said that there had been no creep hetween the tunnel and Rockshaw Bridge for many years and he had never had to pull hack the rails. The length of track under Peters Bridge, however, frequently gave trouble in wet weather when the clay oozed up through the sleepers which could not he lifted properly at this point owing to the restricted headroom under the bridge. He had cleaned the ballast for several rail lengths some three or four weeks before in order to remove the dirt and consequently the boxing was not up to the level of the sleepers. He realised also that there was a shortage on the shoulders, but he had not reported this nor had he appreciated that the lateral strength was seriously weakened by the work he was doing.

He admitted that he received written instructions from his Permanent Way Inspector regarding heat precautions, but he thought that they applied mainly to the first onset of hot weather, though he realised be should not use a jack during any hot spell unless it was absolutely necessary. He was not, however, at all apprehensive when be was working in the cutting because he was near the tunnel mouth which was comparatively cool and he was only lifting individual sleepers and was not disturbing a long length of line. He said that he had oiled the fish plates in April and that his routine was to remove and clean the plates, scrape any rust off the rail, apply oil to the top and bottom of the rail, reverse the plates and screw them up again. He was present when plates were removed, at my request, on the evening of the accident and he agreed that the oiling had not been very well carried out.

12. Sub-Ganger Packham said that he hegan work in the cutting on Stone's instructions and first of all opened out two groups of sleepers near Rockshaw Bridge. He then worked towards the tunnel and opened out several joints before reaching Peters Bridge, where there were five requiring attention. He did not start lifting and spreading chippings under any of them until Stone arrived back after lunch. By this time it was getting very hot and the rail joints which had been open in the morning were now nearly all closed. Packham stated that be commented on this to his ganger and said "These joints are getting pretty tight ; look at them”. 

They continued with their work, however, and had just lifted two joints near Peters Bridge before the Littlehampton train arrived.

Packham agreed that he had read the instructions regarding work in hot weather, he knew that a jack should only he used in exceptional circumstances and he thought that the worst time for buckling was about mid-day. He also confirmed that the fish plates were oiled early in April. He said that some of the plates were taken off, cleaned and turned before oiling, hut in the majority of cases he only loosened the bolts and ran in oil with a brush, because his ganger said that this would he sufficient. 

Stone, when questioned about this, explained that these instructions referred only to those cases when it was necessary to oil a plate a second time.

13. Permanent Way Inspector W. E. Sinden said that the track between the tunnel and Rockshaw Bridge was generally satisfactory and well supplied with ballast, except for the short length already referred to under Peters Bridge, which gave trouble in wet weather. He did not think it was much use keeping large supplies of ballast there because it soon got clogged up when the wet clay came through the formation, but he had instructed Ganger Stone on several occasions to watch this place particularly.

Although Hallade recordings taken on 18th May showed that the track was riding well, with little signs of lurching or bad joints, Sinden did not consider that Stone's length was one of the best and when he last inspected it on 15th June he had to speak about the bad state of the weeds. Stone, however, had been short of a lengthman for six weeks and the only recruits so far had been unsuitable. Usually there was not much trouble over labour in this area though there was sometimes difficulty in replacing men on the Tonbridge line, Sinden agreed that he had issued written instructions to all his gangers about heat precautions but had never questioned Stone to find out whether he understood them properly. He had also written about greasing fish plates but he had not had any removed for inspection, and had always taken the ganger's word that the work was properly done:

14. Mr. Parker, the Divisional Engineer, said that negotiations were in progress regarding Peters Bridge and he hoped it would be removed by the end of the year. If this were done, the line could be lifted and additional ballast placed under the sleepers.

15. I accept the evidence of Ganger Stone and Sub-Ganger Packham that before the Littlehampton train arrived the alignment and gauge of the track was correct ; it was, however, in an unstable condition, as proved by their own statements and those of the motormen and other railway servants all of whose evidence goes to show that the derailment was due to the track distorting under the train.

The day was extremely hot and there was a deficiency of ballast. This in itself should not have been enough to account for distortion, but the lateral resistance had been weakened by cleaning some weeks earlier and it was still further reduced by lifting some of the sleepers a few minutes before the derailment. Thus, when the train passed over it, the thrust of the wheel flanges, coupled with the vibration set up by the train's movement, was quite sufficient to start the rails moving outwards. Once the initial resistance was overcome, buckling became progressively worse.

It is probable that the short distance between the trailing motor bogie of the sixth and the leading motor bogie of the seventh vehicle prevented the latter from negotiating the distorted track and so the right-hand wheels mounted the low rail and became derailed. The lurching and rolling of the last coach was sufficiently severe to cause the screw coupling to ride over the drawbar hook and thus cause the break- away. It is fortunate that the track was not burst open nor more vehicles derailed, otherwise the results might have been much more serious.

16. Primary responsibility for the derailment must rest with Ganger Stone, who failed to appreciate the danger of disturbing track, already short of ballast, at the hottest time of a very hot day. He had received written instructions about heat precautions but it was clear that he had not understood them properly. He is 60 years old with 38 years service, all on the same length, of which he has been ganger for the last four years. He has a clear record.

17. Permanent Way Inspector Sinden must also share responsibility for failing to ensure that Stone carried out his work in a proper and efficient manner and was fully alive to the precautions to be taken in hot weather. Although Sinden was prompt in sending out copies of his Divisional Engineer's instructions, he did not follow them up by explaining personally to his gangers nor did he check by personal inspection that the greasing of fish plates was properly carried out.

Sinden realised that the track under Peters Bridge was not well supplied with ballast yet he took no steps to increase it and I do not accept his suggestion that it was not worth while doing so because of the difficulty of dealing with the clay in the formation. 

It is probable that he was unduly influenced by the hope that the low bridge would be removed soon, and that he would then be able to lift the line and place a substantial layer of ballast under it. He is 59 years old with 41 years service, of which he has been an Inspector for seven years. He has a very good record.

18. During very hot weather, when sun temperatures rise as high as 135"F., it is unavoidable that some rails are under stress due to restriction of expansion, but dangerous conditions are not set up on properly maintained and ballasted track, because its lateral strength is ample to resist buckling. If, however, this stress becomes high due to irregular expansion gaps or tight joints, and at the same time the holding power of ballast is lowered by loosening or lifting, danger may arise, especially in those places where there is a deficiency of ballast. It will be noted that in this case, although maximum temperatures in the cutting probably ranged between 110°F. to 135"F., for ten days prior to the derailment, buckling did not occur until after the joints were lifted.

19 Records show that the number of accidents due o heat distortion have risen during the last two years and this aspect of track maintenance is receiving speed attention. The whole problem is, unfortunately, still affected by a shortage of manpower in many play and difficulty in obtaining and retaining men of the right type, often due to housing problems and competition from other industries. Courses of instruction have been arranged and a standard system of periodical assessment of every permanent way length has been introduced in all Regions. Investigation into the causes of heat distortion has been made during the last 18 months and special instructions regarding heat precautions have been issued but, as already explained, they must be reinforced by verbal explanation and proper supervision if they are going to be fully effective.

Research into the strength of different types of ballast was instituted as a result of the Wath Road derailment on 18th May, 1948, and the method of oiling fish plates is now being re-examined, with a view to improving the quality of the lubricant. Flat bottomed rails have been standardised and their gradual introduction in place of hull headed rails will increase materially both the lateral as well as the vertical strength of track. In connection with the research, it might be desirable to measure the actual stresses set up in rails in hot weather, so as to supplement the information previously obtained regarding the maximum buckling forces which have to be resisted ; also to investigate whether the smoother sleeper beds produced by modem methods of measured shovel packing have affected lateral resistance to any marked degree. Finally, I hope that the results of this work will he published in due course so that engineers may be further informed on a subject which does not seem to have received in the past all the scientific study which its importance merits.