The Institute of Rail Welding’s seminar at the BOC site in Wolverhampton on 15 November 2011 had a slightly less technical flavour than usual and, as the title suggested, focussed on strategic matters instead.

However, there was a special session in the afternoon on developments in ultrasonic testing of rails and rail welds.

The day, chaired by Mick Downing of Sky Blue Welding, commenced with a presentation about the National Skills Academy for Rail Engineering (NSARE), delivered by Chief Executive, Gil Howarth.

National Skills Academy

Gil described the current setting for NSARE, referring to an initial industry plan drawn up last October by Network Rail, ATOC, RFOA and RIA.

It emphasises key issues relevant to the training requirements of the rail industry, including the continually increasing demand for rail services of all kinds, the very large planned programmes of investment and renewal, and the significant efficiency savings that are being demanded.

It concludes that the existing workforce needs to be significantly re-skilled – 30% more professionals will be needed over the next 5 years and the number of apprenticeships offered annually will need to double.


Roger Griffiths of Network Rail spoke about his company’s Innovation Strategy. This is driven by the demands of the ORR and the McNulty report, and the general industry resource shortfall.

Network Rail recognised a need to clarify and simplify its approach to innovation to maximise the benefits to be gained from its own people and its suppliers.

The innovation process focuses on the needs of the company’s customers, following the sequence “Think-Explore-Prove-Do” before checking “Did we deliver?”.

Setting the strategic agenda are four Innovation Portfolio Groups. Each covers a specific area of the business and controls the innovation process.

The company has embarked upon a process of targeted engagement with its own staff through “iStorm”, a special area of the in-house intranet which links together people and groups with common interests and ideas.

In addition, a dedicated section of the company website is designed to engage with suppliers and steer their innovative activities by offering guidance about the company’s needs and allowing people to submit innovatory proposals.

The website currently holds 19 open challenges to innovation, and has received over 250 supplier submissions. Proposals adopted to date, such as the weld repair of tri-metal zones of cast crossings, have shown high rates of return.

Mobile flash butt welding (MFBW)

Third up was another Network Rail speaker, Sean Heslop. He spoke about “Network Rail’s Deployment Strategy for Flash-butt Welding” which aims to make MFBW the preferred site welding methodology for Network Rail’s infrastructure.

Sean described the history of MFBW, including how early machines had limited mobility and could not stress the rail, how the choice of weld method was based solely upon welding cost and how major concerns about possible damage to the infrastructure, in particular electrical damage to S&T systems, inhibited full application.

The requirement to reduce significantly the time taken for rail defect replacement drove the need to re-examine the situation. Recently introduced methods, e.g. repair welding of certain defects, were good, but more needed to be achieved.

A big change was the introduction of MFBW machines that could stress the rail. An acceptance procedure was developed for stressed flash-butt welds and trials were instigated for welding heads.

Work was undertaken to make best use of MFBW, to improve mobility of the machines and to reduce the costs of welds.

One area of concern was the cost of moving the machines between sites, which could amount to 25% of the shift cost.

“Network Rail MFBW” was developed – a package including working practices, performance specification, and crew specification and competences.

The objective was to be able to go into a site, replace a section of defective rail, stress the rail, and get off site again within 2 hours.

Sean described the challenges that had to be met by the project, including obtaining the funds needed and convincing the sponsors that the project was a winner.

The outcome was an order for four machines with the option for six more later. Those ten machines will cover 30% of Network Rail’s welding needs.

The specification is for the welding head to pull up to 400mm, with external tensors to take this to 1000mm, while welding rails with a depth differential up to 3mm.

Provision of MFBW services will be managed by a dedicated in-house organisation. Welding machines will be based within a given geographic area, rather than on a route or asset basis.

This was shown to be the best way to reduce time lost in travel between sites.

One machine is to be brought into service at a time, with two dedicated teams for each. The first MFBW is due in service in March 2012, with the other three following sequentially. If these are a success, others may be purchased under the option in the contract.

Tram rail resurfacing

The fourth speaker was Tamas Sandor of ESAB with “Resurfacing Tram Rails by Arc Welding without Pre-heat”.

Tamas explained that tram rails need to be welded without pre-heating because the rails are normally embedded in materials which cannot be exposed safely to pre-heating temperatures. Possession times available on tramways are too short to permit the digging out and restoration of the embedding material.

However, rails are almost always basic grade, susceptible to the development of brittle martensite if welded without pre-heating and without controlled cooling.

Recently developed techniques use the heat generated by the weld deposition process itself to control the martensite formation.

When a sidewear scar is repaired by applying several passes of weld deposition, the heat of the second and subsequent passes of the welder anneal the metal effectively enough to prevent significant risk of brittle fracture in service.

It is crucial to use the correct number of passes and the right cycle times, and these parameters vary depending upon where on the rail the repair is to be made; for example gauge corner repairs need different treatment from repairs on the top of the head.

Europe 1 – RAILECT

Christopher Spree, of Spree Engineering, spoke about the European Commission RAILECT project.

Involving eight partners from across Europe, including TWI and Network Rail, the objective is to fill the gap in the market for a quick, full volumetric assessment of aluminothermic rail welds using a simple, clamp-on assembly of phased array ultrasonic transducers to produce an image and analysis for comparison with a specified standard.

Full coverage and sensitivity for the whole weld should be achieved without needing to move the probe.

Modelling using ES Beam and CIVA software enabled the design to be fine-tuned. The resultant device has eight phased-array transducers, weighs less than 10kg, and takes 20 minutes to scan a full weld.

The data produced is easily interpreted (there is also a proposal to incorporate automatic defect recognition software and UK firm and project partner KCC is working on this).

The device does need to be moved once to give full coverage of the weld, having to be turned 180˚ to provide dual sided inspection.

The full rail section is then covered including all of the foot. The system has been verified against other detection methods, but more work is required to develop and agree critical defect sizes and acceptance criteria.

Consultations have been carried out with industry stakeholders and the next steps for the project have been identified and agreed.

Full volumetric analysis

Geismar UK also have ideas about the testing of rail welds. Dr Neil MacCuaig included these in a presentation about the new GS70 tester, which was also displayed to seminar attendees in the BOC laboratory during the lunch interval.

The GS70 was developed to address the gap in the market already described by other presentations. Geismar appreciated that rail infrastructure operators had concerns about the difficulty and expense of the manual ultrasonic examination methods that were the only non-destructive means of checking welds.

It was not practicable to check all welds by these methods, it was difficult to comprehensively and consistently document the tests and it was hard to differentiate between “good” and “bad” when defects were detected.

The quality of testing was heavily operator dependent and current systems only recorded evidence of defects detected, being incapable of recording evidence of their absence.

As a result it was impossible to be certain of the quality of the tests that were done, there was no possibility of an “audit trail” and there was always the concern that amongst the welds not tested might be some that had hazardous defects.

The new tester gives full volumetric analysis of a weld, flash-butt or aluminothermic. It has 70 probes in five banks which conduct a complete examination of the rail section.

The process is fully automated, and involves moving the unit only once, a short distance along the rail between the two scans (forward and reverse) of the examination cycle.

An automatic check for full acoustic coupling before the examination cycle begins inhibits the examination from proceeding if a problem is detected.

All data and alarms are displayed to the operators in real time, and are fully digitally recorded. Results may be analysed at once if a qualified operator is present, or undertaken off-site using the digitised record.

The equipment incorporates facilities to carry out an immediate manual verification of any detected defects if this is desired.

Sub-critical indications are recorded, allowing the comparison of results with earlier findings and permitting the detailed monitoring of the development of any sub-critical defects over time. This should permit pre-planned removal before they reach a size where emergency action is necessary.

Balfour Beatty Rail

Ultrasonic testing was also the subject of the talk by Bob Sawdon and Sam Broujeni of Balfour Beatty Rail entitled “Recent developments in the ultrasonic testing of rails and welds”.

Their vision for the future embraces an RRV-based ultrasonic system and adoption of a “stop and verify” approach which will get away from the hazards and costs incurred in sending staff to manually verify defect reports from train borne UT systems.

It bridges the gap between pedestrian and train borne systems, reduces costs and leads to higher productivity on lower category lines.

The Balfour Beatty Rail 8000SX ultrasonic rail flaw detection system, produced by RTI, is the fourth generation of a system that was developed for Australian heavy haul railways.

Critical defect sizes there are very much smaller than those applicable on European networks, driven by higher axle loads and annual traffic tonnages.

The system shows the operator the whole of any defect in one picture. Its “Flawview” system automatically recognises and displays defects and will identify them at sub-millimetre sizes if required.

It incorporates “SmartCal”, an automatic calibration system that deals with temperature changes and variations in rail depth without manual input, as well as GPS location. The outputs of different probes can be displayed in different colours and results are digitally recorded.

The system in the UK is mounted on an RRV based upon a Land Rover Defender 130 and can inspect rails at up to 45km/h (though in the UK at present, RRV operational rules reduce this to 32km/h). Train borne variants are also available.

The 8000SX has detected porosity in aluminothermic welds which were verified by destructive examination, so there is the possibility that, in future, it may be used for the examination of aluminothermic welds.

Europe 2 – MonitoRail

A speaker from TWI Ltd introduced the seminar to another European Commission project, MonitoRail. Carmen Campos Castellanos, the project leader, described how the EC FP7 programme has funded this work on the possibility of looking along the rail ultrasonically to detect defects from a distance.

It was recognised that the rail industry needs to be able to non-destructively examine the whole of the rail profile, including all of the rail foot. Current techniques at best reach only the area of the foot directly beneath the web.

As was shown later in the day, rail breaks initiated by foot defects are becoming significant and a cause of concern to railways.

There are clear safety and economic benefits to be gained from an ability to look at long lengths of rail from a single examination point.

The MonitoRail project concentrated on achieving full volumetric examination of significant lengths of a rail from a single location, obtaining full defect detection in all areas where defects are likely.

Challenges to this investigation are the environment of operation, the restricted access available to the rail, the interfaces with engineering and operating staff and the way in which features of the rail and track (such as pads, clips etc) attenuate the ultrasonic signal.

For this reason, guided waves are used which are lower in frequency than those used in conventional ultrasonic testing, a technology which is already commercially used to examine pipelines. However, the geometry of rails makes their examination more complex.

The project was preceded by theoretical modelling and experimental verification, considering the rail in 3 separate sections, foot, web and head. The foot has been given priority for the reasons already mentioned.

The optimum wave mode to use to examine the rail foot has been determined as well as the best location for the transducer. Experimental trials were carried out with good results at TWI Cambridge on a section of plain rail.

Further trials have been made on a Network Rail sample complete with pads and clips. Results are encouraging.

The next steps include improving the quality of the propagated wave, determining the full effects of items such as clips and how to filter these out, improving signal analysis in other ways, carrying out further empirical validation and then repeating these processes for the rail head and web to complete the means to examine the whole rail.

Network Rail’s ultrasonic strategy

The day’s final speaker was Brian Whitney of Network Rail who is also the current chairman of the Institute. He outlined his company’s strategy for the ultrasonic testing of rails.

Having considered why the company should test rails ultrasonically at all, and described what needed to be detected, Brian considered in some detail Network Rail’s needs and priorities for the present and the future.

Inspection technologies need to give accurate defect location data so reported defects can be easily found when the time comes to repair or replace them.

Test outputs must be consistent and repeatable, often this is more important than absolute definition, and the applicability and limitations of any method must be fully understood.

Any inspection strategy must take account of many factors including route criticality, the position of the defect in the rail, the probability of detection, the track access and the test method.

Network Rail’s strategy demands early and reliable detection of common defects, improved management of high risk defects, moving from “find and fix” to “predict and prevent”, earlier detection to permit the action to be taken to be less onerous, and improved management of factors such as poor track conditions and wheel flats.

Conventional ultrasonic testing alone may not be enough; high track forces need to be understood and controlled and data must be gathered from many sources and combined to give the full picture.

Track support conditions are a significant driver of the rail failures that remain now that the “easy” successes in rail failure management have been achieved over the last 5 – 10 years.

Steep dip angles, voiding and abrupt track stiffness changes are examples of significant contributors to failure. Changes such as the introduction of a dip angle limit are already resulting in a reduction in failure rates.

Technological developments Brian mentioned included the Sperry RSU, which has similar capabilities to the Balfour Beatty Rail 8000SX described earlier, work being done on the detection of rail foot defects, and the removal of poor track conditions based around the company’s Route Asset Management Plans.

Brian discussed a significant list of planned ultrasonic test unit (UTU) developments, particularly the purchase of a fourth example.

This will be used to permit each of the existing UTUs to be withdrawn in turn for refurbishment without diminishing the programme of inspections. Afterwards, it will be used to allow coverage by UTU of the lower category lines now dependent upon pedestrian testing, a process which Network Rail wishes to eliminate wherever possible.

A significant addition to the UTU fleet will be the Sperry Railfix system, which automatically takes high definition photographs of any detected defect and despite all the testing, management of rolling contact fatigue continues to be important and increasingly involves preventative action as well as detection, monitoring and removal.

Brian’s talk brought the day’s proceedings to a close. The Chairman thanked the hosts, BOC, for their excellent hospitality and for making workshop space available for the demonstrations.

This blend of formal presentations and informal workshop demonstrations has become an attractive feature of IoRW seminars held at BOC and delegates were pleased with the variety of subjects that had been covered during the day.