Read an Excerpt
By Jim Pike
The History PressCopyright © 2013 Jim Pike
All rights reserved.
The origins of trackwork are, obviously, the origins of railways. Just what constitutes a railway needs to be addressed: it is here taken as a prepared way of rails to accommodate specially adapted rolling stock. The rails can be of any section (most sections, likely and unlikely, have been tried at one time or another) and the wheels can be flanged or plain. The invention of the wheel is shrouded in antiquity, but it does not take much imagination to devise a way of making a wheeled vehicle follow a pre-set path. The main application of a self-steering system was in conditions of total darkness where a man could not see to steer a truck. Such conditions existed underground, in mines.
The first railways of which there are definite records were the mining tramways of the sixteenth century, as described and illustrated by Agricola. It was quickly realised that a truck of coal running on flat wooden planks was much easier to push than one running on the rocky floor of a mine gallery. It was, however, soon found that if the truck were to be pushed along in total darkness, then some means was needed to hold it on course. The system devised consisted of arranging planks on the floor with a gap of about 6in between them, and fixing a vertical pin on the front of the trucks to engage in this gap. The truck was called a hund by the German miners, and the pin was termed a leitnagel. This system is shown in Fig. 2.
Above ground, wooden rails were popular. The name 'rail' comes from rail-and-post fencing, which can still be seen in places. Wooden rails were cheap and easy to lay, repair and renew; they were soon very popular. They did, however, have one drawback: they wore out quickly. So the custom soon became widespread of laying a second, renewable, strip on the top of the 'permanent' rail. Where these strips were made of iron, they were known as 'plates' and the men who looked after them became known as 'platelayers' – a term still in use.
Wooden rails were often used in conjunction with wooden wheels; cast-iron wheels wore the rails out even faster. But wooden wheels tended to slip on the rails in wet weather, which made braking difficult. Slipping on wet rails is a problem also faced by modern railways.
There was once an extensive system of wooden railways in the North- East, all conveying coal from the mines to the rivers Tyne, Tees and Wear. Here the coal was loaded onto ships, either for export or for the coastal trade, principally to London. Traction on these railways was supplied by horses. All these routes were eventually relaid with iron rails, either as plateways or as edge railways, and many have been repeatedly upgraded to become main line railways.
The distance between the rails, or the track gauge, differed. In the North-East it varied between 4ft 6in and 5ft, while in Shropshire at Ironbridge in the Severn Gorge it was around 3ft. The two areas developed independently of each other, and the reasons for the two sizes are largely historical: a wooden 'chaldron' wagon in the North-East made a full load for a single horse, while in Shropshire it was usual for a horse to pull two or three trucks. In the North-East, the general lie of the railways was downhill, from the pits to the staithes on the river banks where the coal was loaded into ships for export. Wagons were fitted with good brakes, but these were sometimes insufficient when wooden or cast-iron wheels slipped on wet wooden rails.
Some means had to be found of keeping wheels on the plain, rectangular rails (see Fig. 3). A flange was added to the wheel. Projecting below the level of the rail top, it stopped the wheel from slipping off in one direction, and a flange on the opposite side kept it from slipping off in the other. This pre-supposed, of course, that the wheels could not slide along their axles.
Generally flanges are on the inside of the wheels, but there have been one or two isolated lines where they were on the outside. Only slightly more common was the line with flanges on both sides of the wheels. This was to accommodate simply appalling track which kept to gauge plus or minus several inches! The Nantlle Railway in North Wales was one such. Here, the track gauge was nominally 3ft 6in, but the wheels could slide along their axles to take up variations in gauge. The Nantlle was horse-operated, and lasted long enough to be closed by British Railways in the 1950s.
So far, we have considered plain rails carrying flanged wheels. A rival system developed, whereby plain wheels ran on L-shaped rails. The perceived advantage was that wagons with plain wheels could also run on the public roads. This was not really true, because the wheels were necessarily of narrow tread – about 1in – and they soon bogged down in the unmetalled roads of the times. Turnpike road operators soon laid down a minimum tread width to avoid damage to their roads' surfaces. The narrow wheel rims are clearly shown in Fig. 6.
Pointwork presented no problem, apart from arranging that the rails did not trip up the horses pulling the wagons. The drawback to plateways was that the L-shaped rails, or plates, soon became clogged with dirt and rubbish. This did not fall off, but collected in the angle of the plates and resulted in wagons getting a very rough passage. It also reduced the hauling capacity of the horses that supplied the motive power. However, if a handful of dirt is dropped onto a modern rail, most of it will fall off. What remains will move when the rail vibrates at the approach of a train.
Curves were a potential problem. It is possible to bend a I-section rail to a continuous curve, but not so L-section ones. Curves were built up by a series of tangents, which meant that rails were limited to 3ft to 6ft lengths. The rough, jarring motion when traversing a curve was no problem when traffic moved at the speed of a horse led by a man, namely walking pace, but anything faster would have been unacceptable. This, and the inability of cast iron to support a heavy load, goes a long way to explain the general lack of success of steam locomotives on plateways. Indeed, several plateways that tried locomotives found that breakages were so frequent that they reverted to horse traction.
Plateways had a good innings. The last one, in the Forest of Dean, closed in 1944, but a demonstration plateway has been assembled at Blists Hill Open Air Museum, Coalbrookdale. This not only displays pointwork, but also includes some rolling stock – including that rarest of items, a plateway tank wagon.CHAPTER 2
EDGE RAILS OF IRON AND STEEL
Cast-iron rails were cheap to make. Usually in 6ft lengths, they were laid on separate stone sleeper blocks to leave a clear path down the centre of the track for the horses that provided traction. The Silkstone tramway in South Yorkshire was unusual in that its sleeper blocks were laid in a diamond form, rather than the usual rectangular layout. The cross-section of the rails could be almost anything – T-section, T-section, I-section, plain rectangular were all tried.
The Vale of Belvoir Railway track, illustrated in Fig. 8, shows an unusual form of interlocked joint between rails. This was ingenious, but it was soon found that odd gaps had to be filled with short stretches of rail. The interlocked system was simply not suitable for joining to plain-cut pieces of rail and so did not become commtonplace.
Cast iron is strong in compression but brittle in tension. Between sleeper blocks, where it was unsupported, it was strong enough to support the horse-drawn wagons but not the pioneer steam locomotives, which weighed 3 or 4 tons and had an axle-load of between 1 and2 tons. The more expensive wrought iron was tried and found to be far less prone to breaking. In 1820 John Birkenshaw of the Bedlington Ironworks in Northumberland produced a T-section rail with a 'fish-bellied' cross-section in malleable iron. The fish-bellied section, first used in the cast-iron rails of William Jessop, was produced by a system of cams operating rolls. The first type produced was of 17lb/yard, but the Stockton & Darlington Railway, opened in 1825, used 28lb/yard, later increased to 35lb/yard. These rails were secured in the chairs with iron pins, rather after the fashion of Fig. 9. It is said that by 1837 parallel rails of 50lb/yard were in use, and 64lb/yard by 1842. These were used with iron chairs and secured by wedges. By 1841 Bolckow & Vaughan's Middlesbrough Iron Works was rolling rails of 73lb/yard for the Great North of England Railway. Stone sleeper blocks were used which were drilled with two holes, plugged with hammered in oak pins; the chairs were spiked to the oak. Wrought-iron rails with a square section had been used elsewhere for several years before Birkenshaw developed his fish- bellied section, but they were inferior to the Birkenshaw product.
Another attempt to combine a steam locomotive with brittle cast-iron rails involved putting the locomotive on more than four wheels. This was tried by William Hedley (1779–1843) at Wylam, and produced the world's first eight-wheeled locomotive – also the first articulated locomotive. It was mounted on two four-wheeled trucks or bogies and all wheels were driven by gearing. But it was ahead of its time: keeping the complicated machinery in order was quite an assignment, and the locomotive spent more time under repair than in service. It was converted back to a four-wheeler.
Stone sleeper blocks, while adequate for horse-drawn traffic, soon proved unsuitable for locomotives. They were used on the Stockton & Darlington Railway, but as soon as horse traction for passenger services was abandoned, the line was relaid with wooden sleepers. The reason for this is not hard to deduce: the sideways thrusts of a locomotive's wheels tended to force the blocks out of gauge, and one rail tended to subside more than the other. So wooden sleepers were adopted, and the rails could be fastened down to something that could be relied upon to hold the gauge.
The next step was the adoption of a dumbbell-section rail, with the two heads identical. This 'double-headed' rail was mounted in cast-iron supports, termed 'chairs', and wedged in place with wooden wedges or 'keys'. When one head became worn, the whole rail could be turned upside-down and the other head used for traffic. It was an ingenious idea, but it failed because the chairs left definite imprints on the underside of the rail, thus making it useless for running. It was replaced by a rail, still dumbbell-shaped in section, but with the top half appreciably larger than the bottom, which was secured in chairs as before. This was termed a 'bull-head' rail, and it can still be seen in service today.
Steel-making had been practised since at least medieval times, always on a small scale and with variable, not to say unpredictable, results. It was only with the development of the Bessemer converter in 1856, which allowed steel-making on a large scale, that steel rails became possible. The first were introduced in 1857 on the Midland Railway and the London & North Western Railway. They lasted much longer, but were appreciably more expensive. The railway companies were cautious about the new material: the cost was high, and little value as scrap metal was perceived for worn rails. Advantage was then taken of the rolling mill process to produce longer rails. The London & North Western Railway led the way by rolling 60ft lengths at Crewe in the 1880s, a length that had become common by 1914. From 1979 120ft lengths were introduced. Each extension in length reduced the need for fished joints. This not only gave a smoother ride but also reduced the cost of joints and their maintenance.
Steel of controlled quality in abundant quantities was assured with the introduction of the Siemens-Martin open-hearth method in 1864, and the metal produced was put to many other uses besides rails. In fact, most steel manufacturers produced rails more or less as a side-line. With the nationalisation of the steel industry in 1967 the new British Steel Corporation concentrated rail production at Workington, home of the Workington Iron & Steel Company (latterly a subsidiary of the United Steel Co. Ltd), where rails had been produced since 1876. Today, steel 'blooms', or heavy rectangular bars, are produced at Teesside and taken to Workington by rail to be rolled into – rails! Typical steel compositions for rails are shown in Table 1.
Flat-bottomed rails are manufactured in 60, 70, 75, 80, 90, 95, 100, 110 and 113 lbs/yard, plus light rails, conductor rails, and special sections for switches and crossings. The specification for conductor rails is quite different and is shown in Table 2. Again, the balance, up to 100%, is iron. The conductor rail has low residuals to maximise electrical conductivity.
Rails are manufactured by heating a steel bloom until it is bright red hot and has a degree of plasticity. It is then passed between the rolls of a rolling mill, which looks rather like a giant kitchen mangle. In much the same way that pastry becomes longer when it is rolled out thin in the kitchen, steel becomes longer when the blooms are rolled. Several passes are made, each time bringing the material closer to the desired size. The rolls have grooves cut into them, so that the steel, as it emerges, has the profile of a rail. The rolls are not quite symmetrical and as a result the emerging rail is 6 degrees off the horizontal. The purpose of this is to take account of the fact that the rolls wear, and to facilitate their reprofiling after the production of between 1,000 and 1,500 tons. Since the mid-1980s British Steel at Workington has developed an electric induction treatment which produces a rail with a hardened head. This has since been replaced by an on-line heat treatment process, utilising the heat that remains in the rail immediately after rolling. The head of the rail is cooled by water sprays from 850°C to 500°C, thereby producing a refined microstructure which resists wear in a similar manner to an electric induction hardened rail. Consequently the electrical induction process has now been discontinued. More expensive than the conventional product, most of this rail is exported and it finds only occasional use in Britain.
A further development in bloom manufacture at Teesside is continuous casting. Instead of ingots being produced, then reheated and rolled, the molten steel is poured (or teemed, as it is called), straight into a casting machine to produce blooms in a single operation. Costs are saved on a massive scale and there is less wastage. The continuously cast bloom is then cut into convenient lengths, ready for transport to Workington.
Wear is important, as steel fatigues with use. If the rate of wear exceeds that of fatigue, then cracks will wear away instead of growing in the rail. If very wear resistant rails are used in locations of low wear, then cracks may develop and if they remain undetected the rail will eventually break. To avoid this the affected rail must be replaced. The alternative is to introduce 'artificial wear' by grinding the rail surface, a solution that finds favour in some countries. British Rail and Railtrack have largely defeated the problem by choosing a softer grade of steel than most European railways for most applications, resulting in higher wear and lower crack formation. But the problem has not been beaten absolutely. The occasional broken rail still occurs, but is so rare as to be regarded by the media as highly newsworthy.
The reason why rails break is worth discussing, especially in the light of the Hatfield accident on 17 October 2000. Defects can arise within the rail, either as a result of welding or a flaw in the manufacturing process. A minute pinhead of gas, trapped in the hot metal, results in a weakness invisible to the eye and only detectable by ultrasonic equipment. There is only one cure – replacement. This flaw, if unchecked, will develop until the rail snaps under a train, and this may have been the cause of the Hatfield accident. Defects arising outside the rail include weaknesses caused by faulty welds, drilling holes in a rail for signalling, track circuiting and other purposes, and wheelburns, where a braked wheel has skidded along the rail. Gauge corner cracking is a phenomenon which has only appeared within the last ten years and is not yet fully understood. It is thought to be caused by the stresses of wheel against rail at very high speeds, and results in a series of cracks, each running from the top centre of the rail to the inside of the head. Again, the only remedy is replacement. This book will be with the printers before the report into the Hatfield accident becomes available, and so further comment would be inappropriate at this time.
The action of wheels on the rail tops causes the phenomenon termed 'cold rolling'. That is to say, it causes the top of the rail to become longer than the bottom, producing a vertical curve in the rail. When rails have been removed for renewal and are lying by the trackside, this condition can often be clearly seen.
Rail is sold by the ton, so, provided the cost of the material is constant, the lighter the rail section specified, the more rail one gets for one's money. There was once a strong temptation, especially in countries such as the USA and Russia, to get the track laid with the lightest rail that would conceivably do the job, and upgrade the permanent way once sufficient revenue had been earned to meet the cost.
Excerpted from Track by Jim Pike. Copyright © 2013 Jim Pike. Excerpted by permission of The History Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.