Institute of Rail Welding 15th Technical Seminar:

- Rail welding underground and related developments

This meeting took place at the premises of The Welding Institute (TWI), on Granta Park at Great Abington, just outside Cambridge. The venue consists of an impressive group of buildings on a very attractive business park, and has a restaurant and overnight accommodation available as well as the TWI's business and training facilities and the conference centre in which the seminar took place. The day was chaired by Joe Small of Jarvis Rail.

The event commenced with a paper by Angharad Rees of TubeLines and Andy Kent of Metronet Rail: 'Rail welding underground - problems & solutions'. The two speakers came from quite different backgrounds it appeared. Andy is a former welder ex BR's Walsall Depot; Angharad clearly did not come from a front line welding past. This worked well though, as the latter's section of the talk covered LUL standards, the verification and approval of welders to work on the Underground and the procedures and paperwork required to be in place before work may commence. This dovetailed nicely with Andy's discussion of the practical issues peculiar to welding underground. LUL's standards have been having to change rapidly with the private sector's increasing involvement in work on their railway. Many standards have gone from LUL into the hands of the contractors. Prescription of method has had to be eliminated and be replaced by defined outcome specifications. Angharad showed examples of such changes and then went on to describe how standards are also being developed and updated to reflect changes in technology and good practice. An example of this is the rail head defect repair standard, where originally only isolated wheel-burns were permitted to be weld repaired. A concession has been granted now, pending a standards change, which allows weld repair of squats, in a process similar to that in Network Rail's (NWR) equivalent standard. Other similar improvements to the standards are being developed.

Andy's part of the presentation explained how much more demanding work on the underground can be compared with that in the open. Access is generally through stations via escalators or lifts. Not only does this restrict the size and weight of loads that can be moved, it also increases the number of staff needed and necessitates great care to avoid damage to the stations (eg sand has to be very carefully sealed in double bags; sand in an escalator's mechanism does it no good at all!). Additionally, extremely high standards affect fire protection arrangements. The dust present in the tubes and tunnels is extremely dry, as are the timber sleepers that remain in many locations. Sparks from cutting or welding will easily start these materials smouldering imperceptibly, only to break out into serious fire when 2 or 3 trains have fanned things nicely after resumption of the train service. It takes at least an hour to deal with such a fire, so this is not a good outcome from a work shift. LUL require every worksite to be overseen by a qualified fire-watch person, who has to book on and off site separately from the work force. Very painstaking arrangements have to be made to keep potential sources of ignition from reaching dust and other flammable materials. Space in the tubes is extremely tight, though cut and cover tunnels are a bit better. As a result of the confined working space the rejection rate of welds in the tubes is about three times higher than for open lines or on NWR. Power supply is another issue. Ventilation being restricted underground, there are tight controls on the number and type of machines and generators that may be employed. There are 110V supplies in all tube tunnels, but the quality of the supply is not reliable everywhere. Sometimes machines will not work at all because of phase problems with the supply, and reduced outputs due to low voltage are quite common.

Another TubeLines representative, Max Gidney, presented the second paper of the day. His talk: 'Maintenance of rails underground - grind, mill or replace?' Starting by explaining why rails need to be re-profiled at all, Max went on to discuss how the newly available Schweerbau rail milling machine has considerable potential benefits compared with rail grinding, especially in underground working. Re-profiling is done to remove surface defects like squats, corrugations and wheel burns or to manage conicity at the wheel/rail interface. The benefits are many, including the reduction or elimination of defect growth, reduced incidence of rail failures and a reduction in noise and vibration. Significant reductions in the requirement to re-rail are achievable through re-profiling, saving appreciable costs. Generally grinding has been the procedure used to re-profile, and many administrations have shown that the expected benefits do follow if the work is done appropriately. However, grinding causes heat, sparks and fumes, and the swarf created is difficult to contain and collect. These are particularly important disadvantages underground. It also typically takes 6-12 passes on LUL sites to achieve the desired steel removal. Grinding is a proven technology, however, and the design profile can be adjusted readily and achieved accurately. Milling the rail is an exciting alternative to grinding because the 1320mm diameter wheels on the Schweerbau machine, each containing 540 cutters, can remove all the required steel from both rails in one pass. The speed of operation is very much higher than that of a rail grinder, and the accuracy of the profile is comparable or even slightly better (DB uses the machine on their high-speed lines). Importantly, temperatures are much lower, meaning no risk of fires or heat damage to the rails; and the type of swarf created is more easily collected and managed than that of a grinder. One significant disadvantage appears to be that the profile of the cut is not readily changed, as it is determined by the arrangement of the cutters and the profile of the milling wheels themselves. As a pair of wheels with their cutters cost over £200k, changing them is not likely to be a regular occurrence! A second issue is that the machine does not work well on LUL's older concrete embedded timber sleeper track, though it is fine where concrete sleepered trackforms exist. In the end, whether to mill or grind depends upon the site. Trackform, rail condition, whether in tube or not, risk of fire and lead time are all relevant factors in the decision. Per shift the milling machine costs significantly more than a grinder, and because only one exists at present there is a long lead time for bookings. When available, it can generally achieve far more finished rail in a shift than a grinder, and has other clear advantages for underground sites. Rail depth is rarely an issue for LUL, defects are the driver for action. Whichever re-profiling method is chosen, re-railing is almost always the last resort because it is so much more expensive.

Dr. Jay Jaiswal of Corus Group appeared next, presenting 'Discrete defect repair - a novel process' a paper authored by Dr. Vijay Jerath and Dr. Howard Smith together with Jay himself. NWR estimates that some 10,000 squat type defects are removed annually from their infrastructure, and that about 60% of these could have been weld repaired. As they estimate that re-railing costs £2,500/defect, considerable savings might be made if an economic repair was applied instead. Typically wide-gap welds or manual metal arc (MMA) weld repairs are used instead of re-railing, also there is the recently introduced technique of aluminothermic weld repair. However, each of the two established methods has its drawbacks, and the aluminothermic alternative, though innovative, is still very novel. Corus decided to look for a better method still, and went right back to the metallurgical basics in their search for it. Essentially, they wanted a technique that eliminated all risk of the repair causing undesirable metallurgical changes in the repaired rail. The thermal history of the rail throughout the repair was thus the key. The new method ensures that temperatures and temperature changes during the repair never create the right conditions for permanent formation in the steel of dangerous structures like martensite. Other objectives included reductions in the time and costs of repair, simplification of the skills and equipment required, greater integrity under cyclic loading, and improved consistency both within each repair (eg reduced variations in hardness) and from one repair to another. For speed, consistency and to allow automation, the new process uses a portable milling machine to remove the defect. The machine is based upon a readily available commercial product. After temperature conditioning (essentially pre-heating, but to only 80°C as compared with the usual 300°C or so) the repair is made by semi-automatic open arc welding with flux cored wire. Particular attention has been given to the welding process to ensure the necessary control of maximum and minimum temperatures and of rates of temperature change. The finishing is currently by profile grinding, but consideration is being given to the use of the milling machine for this. Further development of the process would see the milling, welding and finishing automated, to further simplify the skills required and ensure greater consistency of results. Metallurgical examinations of trial repairs and fatigue failure tests have shown no cracks or adverse metallurgy in any sample, and fatigue resistance vastly higher than that achievable by conventional MMA repair. Corus has patented the details of the thermal control process, and is looking for a partner with welding delivery resources to join with them to complete the development and to exploit the process commercially. They do not have their own welding organisation to participate in this and do not intend to move into that field of work. This looks like a good opportunity for a company with welding resources and expertise!

Bob Sawdon's paper 'Applying remote flash butt welding to the construction of the Terminal 5 railway' made a contrast with the earlier ones, as it looked back at a major rail welding project executed by his company, Balfour Beatty Rail. Bob was involved with his company's contract to build the extensions to the Heathrow Express and the Piccadilly Lines of LUL so that these railways could serve the new terminal. Bob described the vast T5 project before going into details about Balfour Beatty Rail Technologies' technical management of the flash butt welding. The welding referred to in the paper's title was the welding up of 60ft rails into long strings, not in itself that unusual. The difference was that the work was to be done in the new rail tunnels at the bottom of access shafts some 25 to 40m deep. To do this a mobile flash butt welder (MFBW) was lowered into the shaft and set up for work. The rails were lowered down to the MFBW, welded onto the rail string, ground and finished. Completed strings were fed into the tunnel behind for track installation. Bob described the checks and tests that were done to prove quality, and some of the challenges that were met during the job. The key issue seems to have been the client's requirement that because the trackform meant that the FB welds might be seated on the rail pad, the base of each weld was to be ground off flush with the rail foot. In essence it appears that the extra heat input to the rail by this grinding, done while the weld was still hot, led to microstructures in the steel which caused some substandard results in bend tests. Once it was agreed that the base grinding was not needed, the problem stopped. In all, over 2,000 welds were made on site, and there has been no problem despite the un-ground bases to the majority of the welds.

The next paper, 'The single-use crucible - service experience and development' was presented by Richard Johnson of Thermit GB. Richard changed the theme of his paper significantly because of problems which had emerged with two batches of weld portions. There have been unusually high rates of rejection due to weld porosity (NWR rejecting 8% of welds made from batch 13216 and 11% from 13272 so far). The highest incidences of rejections in both cases have been in welds joining worn rails or new to worn. Richard was clearly keen to show that his company has taken this matter extremely seriously and is determined to establish its cause. Thermit GB is looking at both rejected welds and at unused portions which have been returned. The gas pores in the welds appear to originate from small slag inclusions. The pores are believed to be caused by hydrogen, which seems to come from the slag. This suggests water contamination having entered the process at some point, but despite careful investigation of all possibilities so far conceived, no culprit has been definitely identified. Richard described the wide range of possibilities considered so far, eg faulty storage, faulty packaging etc allowing water ingress to portions; contaminated or substandard ingredients supplied and mixed into portions; contamination from packaging, plugs or other similar and many more ideas. Thermit GB would like customers to return any other portions from the affected batches as it would like to examine as many as possible. It has made test welds from some of the returned portions in non-NWR, non-LUL track and is monitoring these.

It seemed to many of the audience that the company has taken this issue almost too seriously, and the good intent of their actions on this problem was very evident in Richard's presentation. There appears to be no evidence that any other batches of portions have a problem. Further, we heard that a large proportion of one of the affected batches had, for various reasons, spent an inordinate amount of time in transit on lorries, itself a cause for problems perhaps.

Future developments of the weld repair process will still be made, although the current emphasis on researching the problem batches has meant some delay. These developments should include: application using acetylene; application to bull-head rail; wide gap application and rail head repair applications. On the last, Richard said that 250 trial welds are currently being installed and monitored. Isolated squats, taches ovales and small visual surface defects have been the subjects of these repairs.

'Aluminothermic welding developments for underground rail and metro systems' was the theme of Nicholas Penverne of Railtech UK. He described again the particular difficulties faced on such rail systems: access, limited space/clearance, 3rd/4th rails and general track condition. He described the benefits of the Railtech process and the contribution these make in alleviating such problems. Lighter weight, consistency of process, versatility of the three piece moulds and the 'pate a lute'/paste in a tube combination and the significant reduction in fumes (especially with the filter fitted to the one shot crucible) are all great advantages in tube or tunnel locations. New developments such as the 'gas box', which digitally controls pre-heating to give greater efficiency and security of the process, are continually being introduced. He described the specialist applications made available to LUL and DLR and to international users. Finally he described Railtech's 'Matweld' hydraulically powered range of welding equipment, where the power pack can be installed up to 100 feet from the site of work, with the advantages that gives in respect of fumes, vibration and noise at the worksite. The range includes a disc cutter, a weld stripper, rail grinders and more.

The last two papers were both about European developments. First came Hans van den Brug of RI, The Netherlands. He described the RAILSAFE2 project. The overall RAILSAFE project concerns the introduction of a harmonised system across the EU for the education, qualification and certification of rail track welders. The lack of common standards was considered a barrier to the mobility of workers within the EU, and was also seen to be a hindrance to interoperability and safety on the railways of Europe. RAILSAFE1, which was completed 18 months ago, dealt with aluminothermic welders. The current phase concerns all forms of arc welding of rails. RAILSAFE2 began in October 2008 and is on a two year programme. It will take the development of common standards right through to the training and certification of welders through pilot training courses. It is currently at the stage of holding workshops of interested parties to obtain the opinions of stakeholders and to seek consensus on what is required. In fact, one of these workshops had taken place at TWI on the day before the IoRW meeting. Hans described the currently proposed qualification process and the administrative structure and organisation (including the involvement of the European Welding Federation as the overseeing European body). A key date for interested parties is 21st October 2009, when a European Seminar on RAILSAFE2 is to take place at TWI in Cambridge. Discussion will centre on the proposed guidelines and rules. The pilot training courses are scheduled for 2010, and will include one in the UK.

The second European paper was presented by Angelique Raude of TWI, and was entitled: 'RAILECT'. The RAILECT project is intended to develop a non-destructive examination system capable of reliably, comprehensively and consistently examining aluminothermic welds. Begun in September 2008, and with eight project partners from five member states, it has a 2 year programme. UK participants include Jarvis, Spree Engineering Ltd (UK), NewRail (UK) and TWI. As well as developing the equipment and technology, the project will agree the acceptance criteria for the different types of defect it will detect. So far the work has prioritised the types of defect so as to concentrate on the most important, and has identified project objectives and work packages.

The day was rounded off by a lively question and answer session, when all of the speakers formed an expert panel to deal with queries from the floor.

The above is an extract of an article by Chris Parker in 'the rail engineer', May 09 issue. 'the rail engineer' can be obtained free for engineers who work in rail.

Email tre.subs@railstaff.co.uk
www.therailengineer.co.uk