Institute of Rail Welding’s (IoRW) 18th technical seminar – Rail welding… nothing stays the same!

Listen to this article

Rail inspection technology and its application develops at a remarkable pace. So there was much to hear about when delegates gathered recently for the Institute of Rail Welding’s (IoRW) 18th technical seminar – new Network Rail specifications, enhanced training and competence requirements, as well as changes in inspection and NDT (non-destructive testing) techniques as a result of European projects and other initiatives.

All the key issues that affect rail inspection were addressed during the event – preventative strategies, Network Rail’s requirements, training obligations, optimisation of inspection and NDT, and advances in NDT technology. The speakers were all well-known figures in the industry who not only provided a clear picture of the state of the art but also a vision of future improvement priorities.

Predict and prevent

Brian Whitney, IoRW chairman, opened proceedings with a presentation entitled Track integrity – preventative strategies. The aim is to drive downwards both the frequency of rail defects and their criticality, and a number of initiatives are in place to achieve that. The defects that cause broken rails were described including those associated with welds and weld repairs. Particular attention is being paid to rolling contact fatigue (RCF) defects which are currently analysed on the basis of ‘prediction and prevention’, involving the use of modelling techniques.

The adoption of new high performance (HP) rail steels is also helping to combat RCF, as well as providing better wear resistance. Other approaches include the use of eddy current testing, optimisation of rail grinding and special training of ultrasonic testing operators. Attention is also being paid to rail-end failures, such as those caused by bolt hole cracks. Monitoring of the dip angle – abrupt variations in vertical alignment – at the rail-end has proved to be an effective technique for deciding on which areas need remedial action.

50% of rail breaks result from problems in the rail foot and even small defects like pitting corrosion can eventually cause failure. These are sometimes associated with a dipped (concave) weld which causes a dramatic increase in the dynamic loading of adjacent track, deterioration of pads and fasteners and, in turn, fatigue cracking in the foot. Again the monitoring of geometric features has proved a valuable tool in highlighting critical areas.

The good news is that the number of rail breaks has dropped steadily from 952 in 1999 to 152 this year. But Network Rail is not satisfied with that and the quest for further reductions continues. Effort is currently being directed at the ‘precursors’ – identifying the combinations of factors that may exist before failure occurs.

Inspection levels

Rail welds were the focus of attention in a second presentation from Network Rail, this time given by Bill Mosley. There are currently about 50 rail weld failures per year compared with about 500 in 1980. Most of these arise from ‘old’ aluminothermic welds; new welds contribute only about five per year, out of about 50,000. Around 1,000-1,500 welds are rejected annually as a result of the inspection regime.

Three levels of welding inspection were outlined –

• immediately after welding, the welder inspects his/her own weld and decides on its suitability, including surface NDT for arc welding

• within one month, a qualified welding inspector carries out the final inspection of the weld and decides on its suitability, again including surface NDT for arc welds

• NDT is carried out by a qualified operator if the contract requires it although this level is rarely applied.

The above activities are all performed in accordance with Network Rail specifications and these are constantly reviewed and improved. There has been some recent relaxation in specification 0032 which deals with aluminothermic welds and in 0132 that addresses arc welding in order to avoid repairs to defects which are not detrimental.

There is a suite of three Network Rail specifications for the inspection of flash butt welds covering depot-welded and site-welded strings as well as switches, crossings and transition rails. A Letter of Instruction (163) has been issued to provide acceptance criteria for site-welded strings. Complete new versions of the specification will be issued this month.

The presentation concluded with a description of the training and competence requirements for rail weld inspectors. A workplace assessment takes place every 12 months and if there is any doubt as to a welding inspector’s ongoing competence, it is withdrawn and a reassessment carried out.

Magnetic approach

Kevin Winchester of MRX Technologies presented a new approach to track inspection based on the principle of detecting the leakage of magnetic flux arising from discontinuities in a magnetised component. In terms of rail inspection, the method focuses on discontinuities in the rail head down to a depth of 8mm. The rail has to be magnetised prior to testing and then demagnetised afterwards. The technique has been used successfully on track in Australia for four years and is currently being evaluated for use on the UK’s permanent way.

Trials have been conducted using a multi-sensor test head which shows that the technique is capable of giving an indication of the vertical depths of cracks in the rail head down to 8mm. The equipment can also detect flux leakage from subsurface cavities and areas of the rail head that have been heated – for example due to wheel burns – because of changes in permeability. Wheel inspection is also possible.

The trials, involving both train-mounted equipment at 45kph and a hand trolley at 5kph, found that the technique was also capable of weld detection and classification, and the detection of surface artefacts within welds. There were also suggestions that the detection of pre-stress regions was possible. The equipment specifies how much of the rail head surface needs to be ground or milled in order to remove defective areas.

Sobering thoughts

Network Rail’s John Harris began his presentation on the non-destructive testing of rails with some statistics. There are 42,000 miles of rail and 19,500 switch and crossing units on the network. Since privatisation, passenger traffic has increased by 35% and freight by 45%, with more growth expected. This data, and the fact that previous rail breaks have sometimes had catastrophic consequences, indicate the nature of the challenge facing Network Rail.

John reviewed the types and causes of defects that can lead to rail breaks and followed this with a summary of the NDT methods that are available to find and evaluate them. The aims of NDT are to test often, find everything and replace crisis recovery with a system of ‘managing’ the defects found. The main method adopted to achieve these objectives is ultrasonic testing using a combination of train-based, hand-held ‘walking stick’ and manual scanning techniques.

Since 1991, extensive use had been made of roller search units (RSUs) in which an array of ultrasonic transducers is housed. In the train-mounted system, the principles are –

• the train tests the track at 30 mph

• RSUs transmit and receive ultrasonic signals

• on-board computers store the data, plotting it against the location (using differential GPS)

• the data is transmitted to the Engineering Data Centre

• analysis is performed using graphical user interface (GUI) software producing B Scan (longitudinal cross section) displays

• suspect areas are prioritised based on risk and repaired as necessary.

Eddy current testing is currently being developed, using both train-based and manual systems, to detect rolling contact fatigue and other rail head defects. Thermal imaging and automatic video inspection are also being investigated. Although some integration has been achieved in terms of collecting and analysing data from the various NDT methods, further developments in this area are also targeted.

Every picture tells a story

The inspection of tramway rails was the subject of a presentation by Dr Rob Carroll of Stagecoach Supertram Maintenance Ltd. He described progress with an ongoing European collaborative project called PMnIDEA. This is aimed at improving the quality of inspections and to develop a predictive maintenance regime.

The first phase involves the fitting of digital cameras and lighting to service vehicles, as well as the application of automated intelligent image analysis. This will enable the system to be evaluated in terms of its ability to accurately detect, locate, classify and reference all the key deteriorations on the track for structural integrity assessment. It is hoped that this will reduce the need for track inspections on foot, helping to minimise costs and associated health and safety issues.

The criteria for the selection of camera, lighting and data acquisition equipment were described, as were the different image analysis techniques which included the use of neural networks. To improve effectiveness, each available technique was supplemented with bespoke filtering and analysis software. Algorithms to negate the effects of interference from bogie vibration, ballast and camera position – as well as other artefacts – were developed. The results so far are positive, showing that the image processing techniques are efficient in extracting relevant features for analysis and classification.

New NDT technique

Progress on another European collaborative project, RAILECT, was summarised by Tamara Colombier of TWI Ltd. With this, attention is focused on the detection and evaluation of volumetric defects in aluminothermic welds in rails. The work involves –

• the production of test welds, some defect-free and others containing commonly encountered defects

• NDT using conventional techniques

• mechanical testing to establish defect acceptance criteria via ECA (engineering critical assessment)

• the development and evaluation of a prototype for a new NDT technique.

The latter is based on the phased array (PA) ultrasonic principle. A PA probe consists of many small ultrasonic elements, each of which can be pulsed individually. By varying the timing – for instance, by pulsing the elements one-by-one along a row – a pattern of constructive interference is set up which results in a beam at a set angle. In other words, the beam can be steered electronically. It is swept like a searchlight through the object being examined and the data from multiple beams is brought together to make an image showing a cross section.

Laboratory trials were firstly conducted to evaluate the performance of the prototype, with field trials then carried out to assess its performance in real-world conditions. The developed system is a semi-automated one for the inspection of rail welds, enabling full volumetric inspection in 15 minutes. Efficient and operator-friendly, it saves time and resources as well as being unique to the marketplace.

The next stage of the project will involve the commercialisation of the RAILECT system and further funding will be applied for, to turn the prototype into a production system.

End note

Throughout the day the event was punctuated by questions and discussions – sometimes lengthy – about the topics presented. All the speakers and delegates expressed satisfaction with the event, indicating that is was well worth the effort of getting there despite appalling weather conditions which made travel a real challenge.

Tim Jessop is Executive Officer of the Institute of Rail Welding.

Thanks to Dr. Rob Carroll for photos.

Article courtesy of the rail engineer magazine.

1 COMMENT

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Latest Rail News

ORR review leads to 50% reduction in maximum fees for ticket refunds

New rules will mean that from 2 April the maximum fee that train operators and ticket retailers can charge...

More like this...