Return to Archive -by date - by topic - Archive 2002.
As we suspected, the increase in rolling contact fatigue is the fault of John Major and his Government
Discussion of what is colloquially termed Gauge Corner Cracking (GCC), but should really be called Rolling Contact Fatigue (RCF), tends to focus on what happened at Hatfield, which can be misleading. Hatfield was down to bad management which allowed RCF on a short section of rail to get out of control. What is more important for the railway is the pandemic of RCF which emerged subsequently and brought the railway to its knees.
RCF has been with us as long as steel wheels have rolled on steel rails. It is not dangerous provided it is managed professionally. Why then was it ‘discovered' to be prevalent all over the Railtrack network after 19 October 2000 ?
This was the question put to the consultants Ove Arup & Partners International Limited and Transportation Technology Center, Inc (Arup/TTCI) who were brought in to look at the problem. They have now reported on the results of their 10 month investigation.
Numerous aspects of the problem have been covered by the investigation which, say Arup/TTCI ‘all lead to a clear set of conclusions'. Significantly, the resulting Control Strategy to control and manage RCF on the UK railway network covers both infrastructure and rolling stock.
Arup/TTCI warn that while the results of their work may ‘unfairly' seem to focus on infrastructure aspects this is because the information on the effects of rolling stock on the track is so poor. As a result, the Control Strategy calls for systems to measure and monitor vehicle condition and performance. This mirrors the requirement for augmented and enhanced infrastructure measurement and monitoring systems.
But, and read this carefully, Arup/TTCI warns that until such systems are fully implemented ‘on both sides of the wheel-rail interface and effectively utilised to correct identified problems, RCF will continue to cause significant disruptions to the railway system'.
Another fundamental truth from Arup/TTCI, which was first adumbrated in this column after Hatfield, is that the industry must recognise that RCF is a product of running steel wheels on steel rails. And a year ago, Arup/TTCIs preliminary report said that there was no ‘clear and single change in recent operating practices' which led to RCF.
Instead, a number of interacting factors had changed simultaneously over the past 10 years without appropriate controls having been put in place to manage the effects of the aggregated changes. This shortcoming was ‘partly, but not wholly, the result of responsibility splits introduced as part of the privatisation process.'
A year of research and practical investigation work has validated this preliminary conclusion. Better knowledge of the wheel-rail interface ‘clearly shows that the initiation and growth of RCF is a systemic failure, and, consequently, dependent on systemic solutions.'
For example, only around 20% of RCF has been found to occur at the gauge corner of the rail. This is significant because the mechanisms causing RCF on the rail head (head checking) and GCC are often different and require different controls.
Even more important is the demise of the “steady-state” theory. Along with others, I assumed that RCF was a function of wheels rolling normally on rails. Not so, according to Arup/TTCI. Many of the mechanisms leading to RCF are induced by small track geometry irregularities, vehicle performance deterioration or both. These result in vehicle transient responses that can cause the high contact stresses and wheel-rail creepage forces which then cause RCF.
From which the consultants conclude that the mechanisms generating RCF are specific to location and traffic and are engineered into the wheel/rail interface.
For example, in curves of 1000m radius and less, the immutable conflict between curving and stability, first identified by Alan Wickens and his team at BR research in the 1960s and 1970s, still applies. You need a stiff suspension for high speed stability and a compliant suspension for good curving. Remember how the prototype IC125 power car encountered just such a conflict until BR Research came up with the yaw damper which was ‘hard' at high frequencies and ‘soft' at low frequencies, allowing a stable bogie to steer?
Despite this, on curves under 1000m you can get hard flange contact against the gauge face of the outside rail resulting in continuous RCF leading to GCC. This is nothing new, merely old knowledge lost in the fragmented railway and rediscovered.
More important for Railtrack is the role of short wavelength track geometry anomalies in generating RFC. These deviations induce transient responses in vehicles which then generate excessively high contact forces. To simplify, the vehicle hits a high spot, the suspension absorbs the shock, but then rebounds, momentarily increase the forces in the contact patch.
Such track anomalies can be caused by normal wear, poor maintenance workmanship, or a combination of the two. Arup/TTCI observe that even where anomalies may be within Railway Group Standards and Railtrack Line Standards, they are not good permanent way maintenance practice.
Equally, vehicles with poorly performing bogies or hollow worn wheels generate excessively high contact forces. Once again the condition of these vehicles may be within the Railway Group Standards and train operators' maintenance practices. But their performance is detrimental to the railway.
And, of course, in addition to ‘good' vehicles on poor track and vice versa, poor vehicles on poor track almost guarantee the high track force which lead to RCF. Both sides of the wheel/rail interface may be operating within their respective safety – based Standards, but the combined effect of barely acceptable wheel on barely acceptable rail is unacceptable.
Analysis of the East and West Coast main lines, attributed the sources of RCF as follows:
1)Inherent design, including curves below 1000m radius on which RCF has been found – 5%.
2)Poor geometry on straight track, switches and crossings and very light curves (above 5000m radius) where RCF has been found - 40%
3)Poorly performing vehicles or the combined effect of vehicles and track at the lower ends of their performance curves. This includes track with curvature between 1000m and 5000m radius on which RCF has been found - 55%.
Do not forget that we are considering a complex interaction. In the case of (3), for example, there is no prime suspect. If there were virtually all curves in this range would incur RCF which clearly is not the case.
So, no one is innocent when it comes to RCF.
Which is why the Wheel Rail Interface System Authority (WRISA) is so important. As Arup/TTCI observe, It is through WRISA that the railway industry will be able to constructively manage the interface issues that have not been integral to the management of the railway network since privatisation'.
As part of the Arup/TTCI study, instructions for controlling RCF have been developed. These are blindingly obvious and include: compliance with current Standards and principles of good workmanship, correction of track geometry faults, the implementation of a new rail grinding profile and grinding programme, rail lubrication, and the establishment of performance-based standards for track geometry and wheel wear condition.
This prescription is described as ‘a simplification of the ideal technical solutions, recognising the practical difficulties of implementation on the operating railway'.
For those who believe that the fragmented railway would work fine if only everyone worked to their contracts, Arup/TTCI offer little comfort. RCF is a system problem ‘resulting from the unintentional consequences of the individual actions of a fragmented set of Duty Holders, operating with competing commercial agendas in a very high capacity railway, with very little access time for maintenance.
Every system in the railway has been pushed to its limit of capability and, at times, beyond. In retrospect, it is not surprising that the system has fallen prey to vulnerabilities inherent, but controlled, in most other operating railways. Arup/TTCI report |
Arup/TTCI press the need for new research and development work on the mechanics of the wheel rail interface, based on the railway as a system. Shall I remind you that before privatization BR Research at Derby was a world leader in this arcane science? No, it will make you as cross as I was on reading the report.
Yes, yes, I know I am an old railway revanchist, but, Arup/TTCI will keep encouraging me. Referring to challenges ahead they comment, ‘Because of the competing commercial agendas of Duty Holders, implementation of system–based solutions is likely to be impeded. These agendas and the expectation to operate the railway at increasingly high levels of utilisation are challenging to the long-term efficient and economic survivability of the GB railway system'.
They add, that without a cooperative system-based culture. the railway will become increasingly confrontational while continuing to deteriorate.
There is the impression that this is solely a Railtrack problem. It is not. There are issues of performance improvement throughout the GB rail industry that must be resolved. Arup/TTCI |
This is a warning for those who demand a ‘24/7 railway'· Arup/TTCI spells out what we all know, that maintaining a reliable high capacity railway demands adequate access for p/way maintenance and renewal.
There is also a small bombshell for those responsible for implementing Cullen 2 on the management of safety. Arup/TTCI reject the assumption that safety-based Railway Group Standards and Railtrack Line Standards will result in good performance ‘because they do not incentivise the economically optimal operation of the railway as a system'. Sorry, a bit of raw jargon slipped in there, but I thought you ought to see the seamy side of my job this once.
Instead, and this is the controversial bit, there should be performance-based standards owned and maintained by the Duty Holders, separate from Railway Safety and within a systems framework developed by WRISA. There should also be a review of the technical basis for some vehicle performance Group Standards, particularly those for wheel impact load limits.
And the contractors get some stick too. Work carried out on the railway infrastructure is often ‘below accepted standards of good workmanship'. This can be seen in ‘the poor quality of track component installations, the overall poor quality of track geometry and sub-grade drainage, the number of visibly misaligned joints and welds on plain track and at switches and crossings, and the observable lack of ballast and sub-grade quality at switches and crossings'. Ouch.
The fundamentals of sound and proven railway engineering are often being neglected. This is evident in the many infrastructure maintenance shortcomings for which standard best practices can be found in the British Railway Track Design Construction and Maintenance handbook. Arup/TTCI |
My experienced civil engineering chums, who have seen the infrastructure they nurtured go to pot, would not forgive me if I did not include this observation: ‘There is a definite need to train current and future maintenance engineers in the fundamentals of proper Permanent Way Engineering, and providing them with the tools and resources to perform their work'.
And so it goes on. There are ‘significant' short wavelength track anomalies that are not seen through filtered recorded track geometry data and have not been identified for repair by visual inspection. There are inadequate vehicle monitoring systems in place to identify and remove poorly performing vehicles from the railway.
In any case, Vehicle Acceptance gives ‘inadequate consideration' to the effects of the vehicle on the infrastructure. This does not incentivise the rolling stock provider to reduce infrastructure damage at the design stage. That should wind up Powell Duffryn who saw their Track Friendly bogies rejected in favour of cheap and cheerful American kit. So cheap that the manufacturer is now in receivership.
Sadly, Arup/TTCI fails to name the guilty men. All those they approached were helpful, except for two unnamed infrastructure maintenance companies. One was ‘repeatedly unavailable' and did not return numerous phone calls. The other company refused to take part, citing required internal approval mechanisms not being in place.