Rail Engineer has reported on poor ride from some fleets on the main line several times, as have other commentators. Unfortunately for the people who might investigate, reports of “rough ride” is rather unspecific. That said, at a recent seminar featuring the work carried out under the auspices of the Vehicle/Track System Interface Committee, Dr Mark Burstow, Network Rail’s Vehicle Track Dynamics Engineer, rose to the challenge. He discussed an investigation about poor ride on several fleets at specific locations.
These included:
- Midland Main Line (South of Bedford).
- West Coast Main Line (Winsford/Acton Bridge).
- Great Western (Didcot-Swindon area).
- East Coast Main Line (Currently under investigation).
Mark said the reports were initially described as ‘rough rides’, but he explained that, in reality, they were ‘hunting’. Technically, hunting is a well understood problem with known causes, but ‘rough ride’ is more vague description which can refer to a number of different issues. It seems that, across the industry, hunting is not as well understood as it used to be and many see it as a ‘track’ problem because it occurs at specific locations, although it tends to affect only some vehicles on a route. In reality it is a true vehicle/track interface problem.
Although the cause of hunting – high equivalent conicity – is known, investigating the phenomenon is complex. It requires co-operation between the infrastructure manager, the vehicle operator and, where different, the vehicle maintainer. It is necessary to carry out site visits to take detailed measurements of wheels, rails, and other track features such as gauge and cant. The automated systems for measuring wheel profiles from trackside and rail profiles from trains are not yet proven to be accurate enough for this process.
“After measuring a train’s worth of wheel profiles or 100 metres of rail profiles, anyone can become an expert in using a Miniprof,” said Mark.
Whilst measurement train results for gauge and cant are generally accurate, these are also measured by hand and correlated well with the automated measurements. Once the data has been acquired it is processed using specialist software.
Analysis
Analysis showed that hunting was occurring on straight sections of track. In these locations the track gauge was tighter than 1,435mm but well within specified maintenance limits although it was noted that for many years the nominal gauge was 1,432mm and some track installed to this value may still be in service. Measured rail profiles showed no obvious issues and were well within allowable wear limits.
Equally, wheel profiles were well within limits set out in GMRT2466 ‘Railway Wheelsets’ and there had been no flange wear even on wheels that had run for over 300,000km.
But when wheel and rail were put together, there was a general trend that higher mileage wheels took equivalent conicity above the usual limit of 0.5. tight gauge and the absence of flange wear leading to slight tread hollowing were significant factors.
How to manage this
The National Technical Specification Notice (NTSN), the UK version of EU Technical Specification for Interoperability for Rolling Stock – Locomotive and Passenger is clear that (summarising):
- The combined (i.e., vehicle and track) equivalent conicity for which the vehicle is designed shall be specified for in service conditions in maintenance documentation taking account of the contributions of wheel and rail profiles.
- If instability is reported the operator and the infrastructure manager shall identify the locations affected in a joint investigation.
- The operator shall measure the wheel profiles of the wheelsets in question and calculate the equivalent conicity using a reference rail profile to check if this value is withing the set limits. If not, the wheel profile has to be corrected.
- If the wheelset conicity is within limits (despite hunting), then a joint investigation by the operator and infrastructure manager shall be undertaken to understand the reasons for the instability.
A similar procedure is specified in the Infrastructure NTSN, using measured rail profiles and reference wheel profiles.
As is often the case when dealing with wheel rail interface issues, delivering the NTSN requirements is easier said than done for a number of reasons. Firstly, calculating equivalent conicity using techniques in EN15302 ‘Railway applications – Wheel-rail contact geometry parameters – Definitions and methods for evaluation’ (referenced in the NTSN) is not simple and unsuitable for day-to-day management and monitoring.
Secondly, the requirement is to check wheel profiles against a reference rail profile (and vice versa for rail profiles). Thirdly, when making those comparisons, the results often show that all is well. It is the combination that is troublesome.
Mark referred to previous research into a technique for estimating equivalent conicity using measurements that are relatively easy to collect. This work sought to identify features in profiles that increase equivalent conicity. The researchers had proposed to measure wheel flange thickness at 3mm above the tread datum (in addition to the current 13mm) and track gauge 3mm below the top of rail rather than side of rail (in addition to the normal gauge point 14mm below the rail top).
The work shows that this is not yet a perfect measure as it showed that serviceable rails were already below an acceptable limit and that wheels would need to be reprofiled at too low a mileage.
Mark said that strategies to reduce wheel rolling contact fatigue have been successful and design solutions such as lower primary yaw stiffness have reduced flange wear. All this has led to much higher mileages being achieved between reprofiling if the only wear criteria are used, but there is no monitoring of factors that lead to high equivalent conicity.
This is not helped by the absence of these factors in wheelset and infrastructure standards. It is reasonably clear that tight gauge, but within maintenance limits, can lead to high equivalent conicity and that the absence of flange wear leads to hollow treads which can also lead to high equivalent conicity.
What next?
Clearly those involved in train and infrastructure maintenance need to be better informed; that rough riding may well be a system property not a sole train or infrastructure issue. The industry needs to focus on delivering acceptable equivalent conicity, a wheel and rail management parameter. Automated wheel measurement in depots can help, but it is understood, automated rail profile measurement has not, so far, proved to be accurate enough.
The good news is that a university project is carrying out further work to develop a ‘quick conicity’ method to help wheelset and infrastructure maintainers to determine when to intervene.
Hunting – small changes are important
‘Hunting’ is the dynamic instability of wheelsets, a higher frequency lateral wheelset oscillation. It is uncomfortable for passengers, feeling like lateral vibration. It also imposes high forces on both track and vehicle.
It usually develops when the ‘equivalent conicity’ is high and causes forces that are higher than can be damped by the suspension. Equivalent conicity is a measure of the combination of wheel shape, rail shape and track gauge. The frequency of oscillation increases as conicity rises and train speed increases.
Equivalent conicity
A rolling radius difference graph describes how the wheel rolling radius changes as the wheel/rail contact point changes, as shown in the diagram below.
Equivalent conicity is defined from the gradient of that relationship. Equivalent conicity is mostly good; it is required to allow wheels to steer around curves and higher conicity allows vehicles to be guided round sharper curves without flange contact reduced wear. But equivalent conicity can be bad; high conicity leads to instability on straight track and on shallow curves.
A standard exists (EN15302:2021) that specifies how to calculate equivalent conicity, but it is not an easy read. The calculations are not straightforward and are usually undertaken using specialist computer software.
Lead image credit: iStockphoto.com
