Taking the temperature
As ever, the Institute of Rail Welding (IoRW) offered some fascinating insights when its latest technical seminar was held at the old Railway School of Engineering on London Road, Derby.
Richard Johnson of Thermit Welding GB kicked things off, explaining his firm’s SmartWeld concept which aims to reduce the reliability gap between aluminothermic welding (ATW) and flash butt welding (FBW).
Two years ago Richard described to another IoRW seminar the porosity problems discovered in some welds made using his company’s products. SmartWeld is part of Thermit’s ongoing response to those issues.
Immediate improvements were made to ensure that welding materials reach site in acceptable condition and a revised preheating procedure was introduced. These early changes brought significant improvements in the failure rate through porosity.
SmartWeld aims to automate site welding as much as possible. Preheating has been the first priority as it is sensitive to small changes. Gas pressure variation is particularly critical and current mechanical gauges are not sensitive enough for welders to control pressures with sufficient accuracy.
The project has introduced the SmartWeld Control v1 to assist in the setting up and formation of welds, to monitor the process and record it – logging the weld portion used (via a barcode reader), gas pressures, preheating time and the time when the portion is tapped (via an infrared detector).
It will also record mould details when Thermit completes arrangements to barcode them. Audible and visual alarms alert the user to problems like low gas pressure and remind them of critical process timings.
Users have to log in as there are different access levels available from ‘welder’ up to ‘administrator’. The former may only use the system to set up and make a weld, record the process parameters and check results.
But administrators can use Control to review all the welds of a given welder or put the device into training mode. Welders will be given magnetic tags for log-in purposes which will automatically record their unique ID and Sentinel details against the welds they make.
The unit has a real-time clock and GPS (accurate to 5m), enabling it to log the location and time of each weld made. The GPS can be supplemented manually to show on which line and rail the weld was placed.
Gas cylinder and ambient temperatures are recorded throughout welding. Control can monitor two welds at a time, allowing welders to achieve good productivity. Operation is via a remote handset or the built-in screen.
Weld records can be in standard Network Rail or other user-selected formats. Work is already well advanced with v2 and a suite of SmartWeld devices have been developed to work with it, delivering the ability to automate much more of the process including preheating, igniting the preheating flame then automatically controlling gas pressures, and preheating time. An automatic portion igniter is also being developed.
Other proposed devices include SmartWeld Setter, which will introduce fully automated weld alignment, and SmartWeld Cutter. This will trim the aligned rail ends to leave the correct weld gap. Self-sealing and adjustable moulds, together with a universal mould shoe, are all in preparation.
The second talk was presented by Bernhard Lichtberger of Plasser & Theurer. The firm has been supplying FBW machines for decades but the recent introduction of European Standard EN 14587-2 has demanded the development of an improved welding head, which has already been certified to the standard for both UIC54 and UIC60 rail sections.
The use of a DC head avoided the issues of head impedance associated with AC operation. Positioning, alignment and welding are all fully automated. The 200mm pulling stroke ensures that up to 500m of rail each side of the weld can be stressed as the weld is made. Optionally the stroke can be increased to 400mm.
Weld shearing is entirely independent of the fixing system so can be done whilst keeping the rails clamped in the correct alignment and maintaining pressure on the weld. The shearing blade is plasma-coated and very accurate, minimising subsequent weld grinding. A pyrometer gives integrated temperature measurement.
Air cooling is incorporated to allow accelerated cooling of the running surface if desired, as might be the case when welding head-hardened rail. Conversely current can be pulsed through the rail during cooling if necessary to avoid the weld cooling too quickly. The formation of brittle martensite must be avoided and this depends upon correctly controlling the temperature and cooling rate.
Control, monitoring and recording systems are fully automated and alarms warn of any important exceedances. Process parameters and GPS location are recorded for each weld. Again, reports can be delivered in standard Network Rail format or others meeting client specifications.
Welders will be available in containerised form, or track/truck-mounted with road-rail running gear: a four-axle 28t device with a road speed of 85kph.
Next Jay Jaiswal of Tata Steel described development of a low pre-heat method of arc repair for rail wear and defects on light rail systems where requirements arise due to tight curves. To minimise costs, emissions and disruption, operators want the ability to repeatedly repair worn/damaged rails in situ. Repairs that would resist spalling were essential.
Conventional arc weld repair methods are inappropriate because rails are usually embedded in polymer materials, meaning preheat temperatures less than 170C are needed, normally considered too low for repair welding. Furthermore, the wear-resistant steels common on tight curves are usually of poor weldability.
Jay described how he designed a weld process to avoid martensite formation and hydrogen inclusion in the repaired rail without exceeding acceptable temperatures. This led to a preheat temperature of 60-80C – just above the critical figure below which martensite forms.
The fact that hydrogen tends to migrate preferentially into the martensite rather than the austenite is used to an advantage. The welding repair process is designed so that each pass will convert any martensite left earlier into austenite. Each time, hydrogen will tend to be left in the martensite of the last completed pass.
Completion with an extra sacrificial pass that is ground off completely means the final areas of martensite are removed by the grinding, taking any hydrogen away with them. This theoretical solution was validated by quite complex laboratory tests on repair samples.
The process is capable of repairing 10-12m of wear up to 12mm deep on in-situ rail in a 4½- hour possession. Working in Sheffield with Stagecoach Supertram Maintenance Ltd, Tata has repaired about 3,000m of rail over six years without problems. Re-repair after rail has worn again has been entirely successful and life extension of as much as seven years has been estimated. A patented weld repair process based upon this is now available from the firm.
Jay next looked at how applicable it might be to main line railways. 10,000 squats and wheel burns per year are removed from Network Rail’s tracks so there is a possible market for a more efficient repair process. Tata therefore asked whether an adaptation of the light rail repair process would be possible and beneficial, and tests have confirmed that it is.
A machine has been developed to carry out such repairs through a largely automated process, employing two interchangeable heads. One carries a milling machine to remove a defect; the second is a welding head which performs the repair.
The milling head is returned to use later, to machine away the sacrificial final weld deposit. Manual intervention is required at only two stages: a pin gun must be used to clear away the slag after each welding pass and the rail profile has to be restored by manual grinding after final milling. Track trials of Tata’s Discrete Defect Repair Process are required to prove the process times but Jay suggested that a repair 100x72x10mm could be completed in an hour.
Stress and weld
A second account of current mobile FBW developments was presented by Jurg Wahrenberger of Schlatter Industries AG. He provided a brief summary of four welders – the static GAAS60 (for joining plain rail) and GAA100 (for S&C), and the mobile AMS60 and AMS100.
The two mobile types differ in their alignment systems, with the AMS60 aligning the running edge of the rails whilst the AMS100 aligns the centres of the webs. The former has an upsetting force of 600kN and the latter 1,000kN. Both can be produced in containerised form or on a road-rail chassis.
The main topic of the presentation was the newly developed AMS200. This incorporates the ability to stress and weld the rail at the same, in a similar fashion to the Plasser machine. It also has an independent weld stripping system that allows for the weld being kept under compression as it is stripped.
The AMS200 shares many features of the AMS60 but has an improved rail alignment system and better means of applying stressing and welding forces. Like the earlier mobile welders, this one is available in a containerised form or on a road-rail vehicle.
Finally, Jurg described the improvements made to the Schlatter Weld Analyser. This now allows the creation and recording of log files for the welds made, either individually or by job/project/time.
Put a coat on
The fifth presentation, by Sean Gleeson of Tata Steel, explained how Network Rail asked the company to develop a coating to overcome rail corrosion being experienced in certain tunnels, at some level crossings and in coastal areas.
One instance, at a crossing in Barnes, was so severe that new rail lasted only three months. Stray third-rail currents were significant there. The coating needed to significantly extend rail life, be resistant to stray current damage, easily repaired and simple to prepare for welding.
The outcome is Railcote – a one-pack coating applicable by spray, brush or roller at temperatures between -15C and +40C. Three coats provide the required film thickness; each one can be over-coated after 15 minutes.
Once cured, it is resistant to abrasion and mechanical damage, and hardens after the first exposure to moisture. If damaged, the surrounding coating will continue to protect the damaged area and prevent outward spread of corrosion. Damage is easily repaired on site.
Tata can supply pre-coated rails up to 18.3m long and, if there is demand to justify the investment in the facilities, longer rails could be made available. 300mm is normally left bare at each end to allow welding, but the coating can easily be stripped back if needs be and retouched after welding with brush or roller.
Results so far look promising. The coated rail installed at Barnes is still in place two years later with no visible corrosion, even though nothing has been done to improve conditions or stop the stray currents.
John Hempshall of Network Rail provided a summary of the company’s approved procedures for welding premium grade rail. The relevant types of steel in use on its infrastructure are MHH, HP (hypereutectoid) and Compact1400 – the latter being a specialised heat-treated alloy steel only used currently for the manufacture of crossing nose blocks.
For each type, John described the processes and weld materials to be used for joining rails (by ATW and FBW methods respectively) and repair welding. The repair of Compact1400 is problematic, with Network Rail finding that use of the manufacturer’s recommended consumables results in a weld softer than the parent metal. Further work is therefore being done to determine an acceptable means of repair.
John said authorisation must be obtained from Network Rail before carrying out welding on any of the premium rails mentioned and come from either the Senior Technical Engineer (Welding) or one of the Senior Maintenance Support Engineers (Welding).
The next speaker was a representative of RailTech International, Frederick Delcroix. Like Thermit, Frederick’s firm recognised the need to improve the quality of ATW welds. Two systems have been produced to this end – Gasbox being the first. Available now, it comprises a system designed to improve preheating.
Gasbox itself fits by snap connectors into the gas hoses close to the torch. It incorporates accurate digital gas pressure gauges and automatically controls the gas flow and flame cone length. It integrates with a new, lighter torch assembly, designed along with a slightly modified mould to ensure that the torch self-centres within it. Gasbox automatically adjusts to ensure correct delivery of gases to the torch at all times.
The other development, Startwell, should be available in September and is a system for automatic electric ignition of the portion. Because conventional igniters are not easily exported for security/customs issues, Railtech has developed a thermic starter which is fired electrically using a rechargeable device.
The final presentation of the day came from Tim Jessop of TWI Ltd. He summarised the outcomes of RAILSAFE 2 which was described in Issue 63 (January 2010) of the rail engineer. The whole RAILSAFE project concerned the development of harmonised standards and procedures for the training of rail welders in Europe, allowing them to work anywhere within the EU. RAILSAFE 1, dealing with aluminothermic welding, concluded some time ago; RAILSAFE 2 – also now finished – covers arc welding.
Tim described the structure, funding and objectives of the project as well as the participating countries and organisations.
Guidelines have been produced for the training, testing and development of welders. There are two parts, covering arc welding for joining rails and rail repair. Rules for implementation were also produced. Pilot training courses were run in 2010, including one at Network Rail’s centre at Larbert which was well received.
The European Welding Federation (EWF) is the European body for oversight of all this. Presently, TWI is acting as the UK national body. All of the documents are available free from the EWF website.
Prorail in the Netherlands intends to fully implement RAILSAFE 2 and Network Rail is very close to compliance, with only a module on basic rail metallurgy needing to be added to its existing training course. All other EU countries intend to implement in due course.
Once again this event was of great value to all who attended and our thanks go to Tim Jessop and the IoRW.