While a huge amount of progress has been made, vehicle/track interaction is still not fully understood some 200 years after the first public railway opened. For over 20 years, RSSB has sponsored a Vehicle/Track System Interface Committee (V/T SIC) to sponsor research in this area. Adhesion is part of the interface, and an offshoot of V/T SIC is the ADHEsion REsearch challenge programme (ADHERE), sponsored by the Adhesion Research Group (ARG). Each year it reports at an annual seminar which your writer has covered many times. This article presents the work of V/T SIC and ADHERE is covered in a companion piece.
This year’s seminar took place in March at the IET Austin Court conference centre in Birmingham. It was attended by attended by over 130 people – a record for the event. As ever, there was new work presented, updates on work in progress, and generational reminders about issues that may have been forgotten. A welcome trend over the years has been increasing cooperation between train operators and track engineers, and greater use of data.
Data confidence
Network Rail’s Brian Whitney, engineering expert for track and S&C, is a familiar face at the event. His long-term ambition has been to build enough confidence in data gathered by specialist measurement plant that repairs can be ordered immediately, without first needing ‘boots on the ground’ to verify defects identified by measurement trains.
As an example, the ultrasonic inspection units typically identify over 3,000 issues per month (more than 38,000 annually) that require action by pedestrian testers to verify the suspect. Based on typically two to two-and-a-half ‘suspects’ inspected per shift, this equated to 15,000 to 19,000 shifts per year which could be saved if repairs can confidently be ordered without verification. This requires: (i) confidence that the fault has been correctly identified which involves linking the Sperry ultrasonic test results with the One Big Circle Automated Intelligent Video Review (AIVR); and (ii) that it can be accurately located, ideally with +/- 30mm.
The latter has been achieved with a system developed by Machines With Vision (MWV) whose RailLoc Fault Navigator is a Neuromorphic Vision system which provides a single centralised map of track features, with automatically positioned condition data, from inspection and in-service trains, and meets the required accuracy of 30mm. It works everywhere on the railway, including tunnels and stations, in all conditions: light/dark, rain, wind. It works at speeds up to 125mph and provides geospatial and linear reporting. The MWV approach provides a mobile application that can be used by an operative on the track to navigate to the precise location identified previously by an inspection vehicle.
Brian referred to the tragic derailment and collision at Adamuz near Córdoba in Spain where a rail break at a weld led to a derailment where derailed vehicles were directed into the path of an approaching train at a crossover. This has led him to add weld breaks to his routinely presented broken rail chart below. Brian added that he is in close contact with colleagues in Spain so that we can learn lessons from the Spanish tragedy. While derailments from a broken rail are very rare, they continue to present a serious risk to railway safety.
Enhanced designs
At the V/T SIC seminar in 2024 (Issue 208, May-June 2024), Professor Jason Zheng Jiang from the University of Bristol & FOSTER Technology talked about developing an enhanced trailing arm bush design based on inerter technology and associated network synthesis methods. This year, he presented with Sharon Odetunde from RSSB who introduced the opportunity/benefits of improved bushes.
Jason outlined the difference between conventional hydraulic damping and inerters. The former provides a force proportional to the relative velocity at the damper’s mounting points. Modern HALL or hydro bushes are this type. An inerter’s restraining force is proportional to the acceleration between the mounting points. He introduced the ETAB bush design which will be tested on a Mark 4 coach. This has a significantly lower longitudinal static stiffness than either the new GMT supplied hydro bush and even lower than the Mark 4’s current bush. It delivers all this and achieves the performance parameters for curving, stability, passenger comfort, and a nine-year life.
More modern carriages use smaller bushes, and a mini-ETAB design has been developed which has an even lower longitudinal static stiffness. Jason outlined an even better NEXATAB bush which is expected to: (i) reduce primary yaw stiffness when curving to deliver a better trade-off between wheel/rail surface damage and passenger comfort; and (ii) improve the performance of existing bush designs.
Sharon outlined the benefits of reduced longitudinal stiffness bushes for several use cases based on Variable Usage Charges designed to encourage track friendly bogies. This work suggested that more rural and suburban routes which typically have more curves will see more benefits than inter-city main line routes. The planned Cardiff-Manchester/Holyhead test route is forecast to deliver moderate savings from reduced rail surface damage and a 25% reduction in wheel maintenance cost. The plan is to fit eight ETAB bushes to one Mark 4 vehicle and retain the existing HALL bushes on the remaining carriages in the rake.
Project group
Professor Mark Burstow, Network Rail’s principal vehicle track dynamics engineer, another regular, gave an update from the V/T SIC Permanent Project Group covering two topics. The first was new methods for inspections at switches and crossings. For many including your writer, Mark’s description about current measurement standards, processes, and gauges felt something like a lecture about the dark arts in Harry Potter. Network Rail’s standard clearly outlines the key hazards, and each one has its own inspection method and tools. The challenge is applying them on site, in the dark and in bad weather. One of the gauges requires the use of a 200µm feeler gauge to assess clearances. Mark explained that because of the manual nature of the gauge, it’s possible to get different results on the same switch.
New switches can also appear to fail the inspection and when grinding teams arrive to make adjustments, they sometimes cannot find a problem they are supposed to fix. Clearly there is variability between inspectors. Work has been carried out to develop software to allow use of Greenwood Engineering’s MiniProf device. This provides a detailed profile including angles, and assesses the actual flange contact point rather than a pass/fail ‘tick box’ exercise of the manual gauge. MiniProf data show that because of the manual nature of the standard feeler gauge process, the results are highly conservative as well as providing inconsistent results. There were big differences between actual contact angles and those recorded using the feeler gauge. Although safe, many switches were failing leading to facing moves being banned and unnecessary grinding repair or even replacement taking place.
While manual gauging continues, Mark explained that using a 100µm feeler gauge gave a more consistent result and this has now been mandated. In addition, there is now a laser scanning gauge becoming available from Abtus to supplement the MiniProf. For more complex junctions where its use can be justified, there is also an automated measurement trolley called FELIX.
Mark’s second topic was vehicle track interaction at lubricators, or rather, lack of it. Lubricators are fitted at suitable locations so that wheel flanges contact the lubricant and distribute it around curves helping to reduce rail side cut and flange wear. Mark was alerted to an issue because of reports of signalling faults in locations contaminated by iron filings. This was on a route with a number of tight radius curves, running a new fleet with better curving performance than the old fleet. All this meant that the new trains’ flanges were contacting the high rail later than the old trains and thus were missing the lubricator. Although the better curving performance was generally successful this was an unintended consequence causing increased wear in tight radius curves, but it was comparatively easy to assess the right location at each curve and move the lubricators.
On-train monitoring
Next were two presentations showing the benefits of on-train monitoring equipment.
First, Tom Lendhill from Hitachi Rail and Nick Swift standing in for Peter Dixon from train leasing company Beacon presented ‘Sine waves to sound decisions’. Rail Engineer has previously presented results from the Perpetuum – now Hitachi Rail – self-powered triaxial accelerometers fitted to axle bearings and other locations on trains which, following analysis can provide information about wheel wear/damage, poor ride, and track condition (Issue 211, Nov-Dec 2024). These articles have highlighted how using data has helped optimise wheel turning frequency to optimise ride quality and to detect wheel anomalies not visible to the naked eye. This meant turning the wheels more often but increasing wheel life as less metal is removed at each turn.
Central to this presentation was dealing with RCF on the wheels of Driving Pantograph Trailer coaches (DPT) of the Class 80X units. The routine wheel reprofiling periodicity was found to be inappropriate on these vehicles. Turning when RCF was present led, only after the first cut, to the discovery of defects under the surface, necessitating a second cut, and reducing wheel life. The vibration monitoring system was used to determine an optimum turning periodicity. The outcome was a periodicity of about 90,000 miles for DPTs to prevent RCF cracks, and around 200,000 miles for intermediate vehicles because ride quality tended to deteriorate at higher mileage.
The benefits were summarised as follows:
- Operations/Planning: fewer wheel defects, with less out of course turnings, leading to planned lathe use and hence improved lathe capacity. All this provides better fleet availability and provides planners some flexibility to briefly postpone some maintenance for events where extra trains are required.
- Technical: better understanding of component life leading to better defined maintenance periodicity and improved vehicle ride. Also, every train provides some track monitoring capability.
- Asset Management: improved asset/component life as a result of reduced depth of cut leading to less frequent wheel replacement. The benefit is reduced material consumption leading to reducing carbon emissions.
Dr. Mani Entezami from the University of Birmingham and MoniRail, Edwin Cornish from Eurostar, and Professor Yann Bezin from University of Huddersfield discussed in-service monitoring: passenger comfort, rough rides, and enhancing track/train interaction. It covered monitoring kit from MoniRail, applied in cooperation with the operator Eurostar, and analysis from academia. MoniRail’s kit (Rail Engineer Jan-Feb 2026) has been applied to many fleets and has demonstrated how the equipment fitted to a regular passenger train would show real world issues that could be missed by a dedicated measuring train whose dynamics are different from the regular fleet.
Examples included:
- Long wavelength track features up to 140 metres (which can disturb ride comfort on high speed trains).
- Effect of rail grinding on axlebox vibrations.
- Axlebox vibrations as a proxy for noise levels – good correlation was demonstrated.
- Short wavelength features such as squats.
- Rail head corrugation (another short wavelength issue), linked to cant deficiency.
- Vehicle instability such as hunting linked to evaluation of equivalent conicity.
Many of these examples were illustrated by a case study of MoniRail equipment fitted to a Eurostar Class 374. Body, bogie, and axlebox sensors allowed monitoring and assessment of ride comfort against EN 12299, body and bogie performance against EN 14363, as well as track geometry, wheelset, and suspension performance monitoring.
It was found that long wavelength track input at 300km/h excited a secondary suspension mode which led to vertical displacement exceeding the suspension travel limit, resulting in bump stop contact. Sensors on the axlebox, bogie, and car body captured the full dynamic propagation through the vehicle which enabled quantification of severity and identification of critical speeds. The results supported the diagnosis that the cause was track induced excitation leading to corrective action.
The system was also able to help rule out a possible cause of an issue which was reported by drivers as a rough ride at speeds over 265km/h. The sensors showed conclusively that although the driver could feel vibrations at this speed, the vibrations did not appear on the bogie, suggesting that the vibrations could be from some other source such as aerodynamic forces.
Once again, the V/T SIC seminar illustrated that there is much to learn or, in a generational sense, to be reminded about. Long may it continue.
Image credit: iStockphoto.com/ClaireLickman