Track (rail transport)

Notwithstanding modern technical developments, the overwhelmingly dominant track form worldwide consists of flat-bottom steel rails supported on timber or pre-stressed concrete sleepers, which are themselves laid on crushed stone ballast.

Most railroads with heavy traffic utilize continuously welded rails supported by sleepers attached via base plates that spread the load. A plastic or rubber pad is usually placed between the rail and the tie plate where concrete sleepers are used. The rail is usually held down to the sleeper with resilient fastenings, although cut spikes are widely used in North American practice. For much of the 20th century, rail track used softwood timber sleepers and jointed rails, and a considerable extent of this track type remains on secondary and tertiary routes. The rails were typically of flat bottom section fastened to the sleepers with dog spikes through a flat tie plate in North America and Australia, and typically of bullhead section carried in cast iron chairs in British and Irish practice. The London, Midland and Scottish Railway pioneered the conversion to flat-bottomed rail and the supposed advantage of bullhead rail - that the rail could be turned over and re-used when the top surface had become worn - turned out to be unworkable in practice because the underside was usually ruined by fretting from the chairs.

Jointed rails were used at first because contemporary technology did not offer any alternative. However, the intrinsic weakness in resisting vertical loading results in the ballast becoming depressed and a heavy maintenance workload is imposed to prevent unacceptable geometrical defects at the joints. The joints also needed to be lubricated, and wear at the fishplate (joint bar) mating surfaces needed to be rectified by shimming. For this reason jointed track is not financially appropriate for heavily operated railroads.

Timber sleepers are of many available timbers, and are often treated with creosote, Chromated copper arsenate, or other wood preservatives. Pre-stressed concrete sleepers are often used where timber is scarce and where tonnage or speeds are high. Steel is used in some applications.

The track ballast is customarily crushed stone, and the purpose of this is to support the sleepers and allow some adjustment of their position, while allowing free drainage.

• track structure images
• 
  traditional railway track (showing ballast, part of sleeper and fixing mechanisms)
• 
  Track of Singapore LRT
• 
  Ballastless high-speed track in China

A disadvantage of traditional track structures is the heavy demand for maintenance, particularly surfacing (tamping) and lining to restore the desired track geometry and smoothness of vehicle running. Weakness of the subgrade and drainage deficiencies also lead to heavy maintenance costs. This can be overcome by using ballastless track. In its simplest form this consists of a continuous slab of concrete (like a highway structure) with the rails supported directly on its upper surface (using a resilient pad).

There are a number of proprietary systems, and variations include a continuous reinforced concrete slab, or alternatively the use of pre-cast pre-stressed concrete units laid on a base layer. Many permutations of design have been put forward.

However, ballastless track has a high initial cost, and in the case of existing railroads the upgrade to such requires closure of the route for a long period. Its whole-life cost can be lower because of the reduction in maintenance. Ballastless track is usually considered for new very high speed or very high loading routes, in short extensions that require additional strength (e.g. railway stations), or for localised replacement where there are exceptional maintenance difficulties, for example in tunnels. Most rapid transit lines and rubber-tyred metro systems use ballastless track.[1]

Diagram of cross section of 1830s ladder type track used on the Leeds and Selby Railway

Ladder track at Shinagawa Station, Tokyo, Japan

Early railways (c. 1840s) experimented with continuous bearing railtrack, in which the rail was supported along its length, with examples including Brunel's baulk road on the Great Western Railway, as well as use on the Newcastle and North Shields Railway,[2] on the Lancashire and Yorkshire Railway to a design by John Hawkshaw, and elsewhere.[3] Continuous-bearing designs were also promoted by other engineers.[4] The system was tested on the Baltimore and Ohio railway in the 1840s, but was found to be more expensive to maintain than rail with cross sleepers.[5]

This type of track still exists on some bridges on Network Rail where the timber baulks are called waybeams or longitudinal timbers. Generally the speed over such structures is low.[6]

Later applications of continuously supported track include Balfour Beatty's 'embedded slab track', which uses a rounded rectangular rail profile (BB14072) embedded in a slipformed (or pre-cast) concrete base (development 2000s).[7][8] The 'embedded rail structure', used in the Netherlands since 1976, initially used a conventional UIC 54 rail embedded in concrete, and later developed (late 1990s) to use a 'mushroom' shaped SA42 rail profile; a version for light rail using a rail supported in an asphalt concrete–filled steel trough has also been developed (2002).[9]

Modern ladder track can be considered a development of baulk road. Ladder track utilizes sleepers aligned along the same direction as the rails with rung-like gauge restraining cross members. Both ballasted and ballastless types exist.

The earliest rails were made of wood, which wore out quickly. Hardwood such as jarrah and karri were better than softwoods such as fir. Longitudinal sleepers such as Brunel's baulk road are topped with iron or steel rails that are lighter than they might otherwise be because of the support of the sleepers.

Early North American railroads used iron on top of wooden rails as an economy measure but gave up this method of construction after the iron came loose, began to curl and went into the floors of the coaches. The iron strap rail coming through the floors of the coaches came to be referred to as "snake heads" by early railroaders.[12][13]

Rail is graded by weight over a standard length. Heavier rail can support greater axle loads and higher train speeds without sustaining damage than lighter rail, but at a greater cost. In North America and the United Kingdom, rail is graded by its linear density in pounds per yard (usually shown as pound or lb), so 130-pound rail would weigh 130 lb/yd (64 kg/m). The usual range is 115 to 141 lb/yd (57 to 70 kg/m). In Europe, rail is graded in kilograms per metre and the usual range is 40 to 60 kg/m (81 to 121 lb/yd). The heaviest rail mass-produced was 155 pounds per yard (77 kg/m) and was rolled for the Pennsylvania Railroad. The United Kingdom is in the process of transition from the imperial to metric rating of rail.[14]

The following are lengths of single sections produced by steel mills, without any thermite welding. Shorter rails may be welded with flashbutt welding, but the following rail lengths are unwelded.

• (1762) 3 feet (0.91 m) various tramways by Reynolds [15]
• (1767) 6 feet (1.83 m) various tramways by Jessop and Outram [16]
• (1825) 15 feet (4.57 m) Stockton and Darlington Railway 5.6 lb/yd (2.78 kg/m)
• (1830) 15 feet (4.57 m) Liverpool and Manchester Railway
  • fish-belly rails at 35 lb/yd (17.4 kg/m), laid mostly on stone blocks.
• (1831) 15 feet (4.6 m) long and weighing 36 pounds per yard (17.9 kg/m), reached Philadelphia the first use of the flanged T rail in the United States.

• (1880) 39 feet (11.89 m) United States to suit 40-foot (12.19 m) long gondola waggons
• (1950) 60 feet (18.29 m) British Rail
• (1900) 71 ft (21.6 m) - steel works weighing machine for rails (steelyard balance)[17]
• (1940s) 78 feet (23.77 m) - double 39ft US[18]
• (1953) 45 feet (13.72 m) Australia[19]

Welding of rails into longer lengths was first introduced around 1893, making train rides quieter and safer. With the introduction of thermite welding after 1899, the process became less labor-intensive and ubiquitous.[20]

• (1895) Hans Goldschmidt developed exothermic welding
• (1899) The Essen Tramway becomes first railway to use thermite welding; also suit track circuits.
• (1904) George Pellissier welded the Holyoke Street Railway, first to use the process in the Americas
• (1935) Charles Cadwell develops non-ferrous exothermic welding
• (1950) 240 ft (73.2 m) welded - (4 x 60 feet or 18.3 metres)[21]

Modern production techniques allowed the production of longer unwelded segments.

• (2011) 120 metres (393.7 ft) Voestalpine, [22]
• (2011) 121 metres (397.0 ft) Jindal[23]
• (2013) 108 metres (354.3 ft) Tata Steel Europe[24]

Newer longer rails tend to be made as simple multiples of older shorter rails, so that old rails can be replaced without cutting. Some cutting would be needed as slightly longer rails are needed on the outside of sharp curves compared to the rails on the inside.

Rails can be supplied pre-drilled with boltholes for fishplates or without where they will be welded into place. There are usually two or three boltholes at each end.

Bonded main line 6-bolt rail joint on a segment of 155 lb/yd (76.9 kg/m) rail. Note the alternating bolt head orientation to prevent complete separation of the joint in the event of being struck by a wheel during a derailment.

Jointed track is made using lengths of rail, usually around 20 m (66 ft) long (in the UK) and 39 or 78 ft (12 or 24 m) long (in North America), bolted together using perforated steel plates known as fishplates (UK) or joint bars (North America).

Fishplates are usually 600 mm (2 ft) long, used in pairs either side of the rail ends and bolted together (usually four, but sometimes six bolts per joint). The bolts have alternating orientations so that in the event of a derailment and a wheel flange striking the joint, only some of the bolts will be sheared, reducing the likelihood of the rails misaligning with each other and exacerbating the derailment. This technique is not applied universally; European practice being to have all the bolt heads on the same side of the rail.

Small gaps which function as expansion joints are deliberately left between the rail ends to allow for expansion of the rails in hot weather. European practice was to have the rail joints on both rails adjacent to each other, while North American practice is to stagger them. Because of these small gaps, when trains pass over jointed tracks they make a "clickety-clack" sound. Unless it is well-maintained, jointed track does not have the ride quality of welded rail and is less desirable for high speed trains. However, jointed track is still used in many countries on lower speed lines and sidings, and is used extensively in poorer countries due to the lower construction cost and the simpler equipment required for its installation and maintenance.

A major problem of jointed track is cracking around the bolt holes, which can lead to breaking of the rail head (the running surface). This was the cause of the Hither Green rail crash which caused British Railways to begin converting much of its track to continuous welded rail.

Where track circuits exist for signalling purposes, insulated block joints are required. These compound the weaknesses of ordinary joints. Specially-made glued joints, where all the gaps are filled with epoxy resin, increase the strength again.

As an alternative to the insulated joint, audio frequency track circuits can be employed using a tuned loop formed in approximately 20 m (66 ft) of the rail as part of the blocking circuit. Some insulated joints are unavoidable within turnouts.

Another alternative is the axle counter, which can reduce the number of track circuits and thus the number of insulated rail joints required.

Welded rail joint

A pull-apart on the Long Island Rail Road Babylon Branch being repaired by using flaming rope to expand the rail back to a point where it can be joined together

Most modern railways use continuous welded rail (CWR), sometimes referred to as ribbon rails. In this form of track, the rails are welded together by utilising flash butt welding to form one continuous rail that may be several kilometres long. Because there are few joints, this form of track is very strong, gives a smooth ride, and needs less maintenance; trains can travel on it at higher speeds and with less friction. Welded rails are more expensive to lay than jointed tracks, but have much lower maintenance costs. The first welded track was used in Germany in 1924.[25] and has become common on main lines since the 1950s.

The preferred process of flash butt welding involves an automated track-laying machine running a strong electric current through the touching ends of two unjoined rails. The ends become white hot due to electrical resistance and are then pressed together forming a strong weld. Thermite welding is used to repair or splice together existing CWR segments. This is a manual process requiring a reaction crucible and form to contain the molten iron.

North American practice is to weld ¹⁄₄ mile (400 m) long segments of rail at a rail facility and load it on a special train to carry it to the job site. This train is designed to carry many segments of rail which are placed so they can slide off their racks to the rear of the train and be attached to the ties (sleepers) in a continuous operation.[26]

If not restrained, rails would lengthen in hot weather and shrink in cold weather. To provide this restraint, the rail is prevented from moving in relation to the sleeper by use of clips or anchors. Attention needs to be paid to compacting the ballast effectively, including under, between, and at the ends of the sleepers, to prevent the sleepers from moving. Anchors are more common for wooden sleepers, whereas most concrete or steel sleepers are fastened to the rail by special clips that resist longitudinal movement of the rail. There is no theoretical limit to how long a welded rail can be. However, if longitudinal and lateral restraint are insufficient, the track could become distorted in hot weather and cause a derailment. Distortion due to heat expansion is known in North America as sun kink, and elsewhere as buckling. In extreme hot weather special inspections are required to monitor sections of track known to be problematic. In North American practice extreme temperature conditions will trigger slow orders to allow for crews to react to buckling or "sun kinks" if encountered.[27]

After new segments of rail are laid, or defective rails replaced (welded-in), the rails can be artificially stressed if the temperature of the rail during laying is cooler than what is desired. The stressing process involves either heating the rails, causing them to expand,[28] or stretching the rails with hydraulic equipment. They are then fastened (clipped) to the sleepers in their expanded form. This process ensures that the rail will not expand much further in subsequent hot weather. In cold weather the rails try to contract, but because they are firmly fastened, cannot do so. In effect, stressed rails are a bit like a piece of stretched elastic firmly fastened down. In extremely cold weather, rails are heated to prevent "pull aparts".[29]

CWR is laid (including fastening) at a temperature roughly midway between the extremes experienced at that location. (This is known as the "rail neutral temperature".) This installation procedure is intended to prevent tracks from buckling in summer heat or pulling apart in the winter cold. In North America, because broken rails (known as a pull-apart) are typically detected by interruption of the current in the signaling system, they are seen as less of a potential hazard than undetected heat kinks.

An expansion joint on the Cornish Main Line, England

Joints are used in the continuous welded rail when necessary, usually for signal circuit gaps. Instead of a joint that passes straight across the rail, the two rail ends are sometimes cut at an angle to give a smoother transition. In extreme cases, such as at the end of long bridges, a breather switch (referred to in North America and Britain as an expansion joint) gives a smooth path for the wheels while allowing the end of one rail to expand relative to the next rail.

Various methods exist for fixing the rail to the sleeper. Historically spikes gave way to cast iron chairs fixed to the sleeper, more recently springs (such as Pandrol clips) are used to fix the rail to the sleeper chair.

Measuring rail gauge

During the early days of rail, there was considerable variation in the gauge used by different systems. Today, 54.8% of the world's railways use a gauge of 1,435 mm (4 ft 8 ¹⁄₂ in), known as standard or international gauge.[34][35] Gauges wider than standard gauge are called broad gauge; narrower, narrow gauge. Some stretches of track are dual gauge, with three (or sometimes four) parallel rails in place of the usual two, to allow trains of two different gauges to use the same track.[36]

Gauge can safely vary over a range. For example, U.S. federal safety standards allow standard gauge to vary from 4 ft 8 in (1,420 mm) to 4 ft 9 ¹⁄₂ in (1,460 mm) for operation up to 60 mph (97 km/h).