Articles
This was a portable seismometer placed in a coach above the bogie to record vertical, transverse and rolling movements on a paper roll.
In the late 1950s recording trollies produced mechanically driven track parameter plots and used red-paint nozzles to highlight severe faults. Rail flaw detection cars that used induction were pioneered in the US in the 1920s. A broken rail caused the 49 fatalities of the 1967 Hither Green derailment. This led to the introduction of ultrasonic test unit (UTU) trains. Before then such testing was done by hand.
The main track geometry parameters that are monitored are gauge (ensuring that rails are the standard 1,435mm apart), twist (variations in cross level) and cyclic top (regular vertical irregularities). Although these were measured by the recording coaches developed by British Rail, track inspection was, until recently, a largely manual affair.
Flying Banana
The 2000 Hatfield train crash was a catalyst for a new approach to track inspection which led, in 2003, to the introduction of the yellow-painted New Measurement Train (NMT), affectionately known as the Flying Banana. This is a converted Intercity High-Speed train that can monitor track geometry and other infrastructure data, such as overhead line parameters and distance to adjacent line at 125mph as well as producing high-definition driver’s-eye videos of each route.
The NMT covers 110,000 miles each year as it monitors all Network Rail’s main lines in an eight-week cycle. Vertical profile measurements are taken from an inertial box with a triaxial arrangement of accelerometers and gyroscopes with optical scanners locating the position of each rail. Linear variable differential transformers are used for other systems.
For each 1/8th-mile section of track, track geometry parameter standard deviations are used to identify trends and deterioration. If safe limits are exceeded the track is blocked by stopping the train and reporting the defect. Network Rail’s inspection fleet records exact positions using differential GPS with inertial measurement, high-resolution axle-end tachometers with reference to a track centreline model.
The NMT is supplemented by a loco-hauled track-recording coach, a self-powered track-recording diesel multiple unit, a specialised vehicle to monitor switch and crossing units and four Plain Line Pattern Recognition (PLPR) coaches. PLPR is also fitted to the NMT. This uses an array of high-speed video cameras to detect track defects such as missing fastenings. At 125mph these cameras record an image every 0.08mm.
Ultrasonic testing
UTU trains have ultrasonic probes inside to detect various types of rail faults. These transmit an ultrasonic signal and detect its return. UTUs can only operate at a maximum speed of 30mph and so can only be used when there are no service trains running. UTU vehicles also determine rail wear using laser rail-head profile measurement units and have ground-penetrating radar to identify problems such as differential track-bed stiffness.
The vast amount of data collected by all these trains is turned into useful information by Network Rail’s data collection team of 30, which has to advise local managers of faults within 72 hours.
The present generation of track-monitoring trains supports the move from reactive maintenance to fault prediction and prevention and have improved track quality and safety performance. For example, the number of broken rails has been reduced from 444 to 80 over the past 16 years and the requirement for hazardous manual track inspections has been significantly reduced.
Much of this is due to increasing miniaturisation and computing power which offers the potential for unattended measurement systems to be fitted to service trains which may eventually largely replace the requirement for a dedicated infrastructure-monitoring fleet.
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.