Each year, the UK hosts numerous railway industry events. Most of these concern domestic issues and few have complex technical content throughout. In contrast, more than half the 150 or so participants at the biennial Railway Engineering Conference held in Edinburgh were from outside the UK (33 per cent from Europe, 14 per cent from Asia, eight per cent from USA, plus individuals from Brazil and Australia). Hence, most of those present had travelled a long way to present their in-depth technical papers.

There were a hundred such papers, covering all aspects of railway engineering, although track, structures and civil engineering accounted for more than three quarters of them.

The conference is chaired by Professor Mike Forde of the University of Edinburgh and organised by ECS Publications, part of the multi-award-winning Edinburgh Railway Group. The first Railway Engineering Conference in the series was held at Brunel University in 1998, thereafter the event was held in London until it moved to Edinburgh in 2011.

This year’s two-day conference was held on 3-4 July. Each day started with keynote presentations, after which papers were presented in three parallel sessions. Two of the keynote presentations were the only ones without detailed engineering content.

Chris Jackson, editor of Railway Gazette International, gave his review of specific railway developments in each continent from which he saw globalisation, increasing urbanisation and decarbonisation to be common worldwide issues. He felt that the data-driven ‘fourth industrial revolution’ had huge implications for asset monitoring, maintenance, train control and automation for which the critical challenge is attracting and developing new skills.

Professor Rod Smith asked whether railways in rural areas were a financial drain. His key point was that high fixed infrastructure costs don’t change much with use. Therefore, proposals for lightweight rail vehicles showed an inability to learn from history as such vehicles do not satisfy the requirement to make the best use of the infrastructure. In this respect, the UK has one of the best records in Europe with 11,200 passenger-km/route-km/day, although this compares poorly with Japan’s 40,900.

High speed track

Niall Fagan, HS2’s head of track engineering, explained the thinking behind the design of HS2’s track, which will carry 18 trains per hour and over 60 million gross tonnes per annum, and for which there will be a five-hour overnight maintenance window, with eight hours on Sunday. Phase one consists of 486 linear kilometres of track and 153 S&C units, for which there is a 20-month construction window.

HS2’s survey grid will be a snake projection, which has been developed to provide a unified coordinate system for long, linear projects as the Ordnance Survey grid does not take account of the curvature of the earth. The difference between the two projections is 50 metres over the 170 kilometres between London and Birmingham and Niall illustrated the importance of this grid by explaining why millimetres matter at the highly constrained Euston approaches.

He also explained the issues that had to be considered to determine HS2’s trackform. These included the predicted tamping of ballasted track, which is largely a function of tonnage carried. It also determines the renewals requirement, as ballast life is a function of the number of tamps. If, as seems possible, HS2 is to consist of largely slab track, measures will be required to prevent ground-borne sound and vibration reaching buildings above tunnels.

By the end of 2018, China has built 29,000 km of high-speed lines, of which more than 80 per cent is slab track. Professor Xuecheng Bian, of Zhejiang University in China, described his research into the mechanisms that trigger mud pumping under slab track, for which polyurethane injection is an effective remedial measure.

He also described how the University had used a full-scale ballasted-track test rig to observe the dynamic responses of ballasted track at speeds up to 360km/h. This showed that high-speed wheel loading increases ballast particle rearrangement, due to greater particle rolling and sliding, and that dynamic ballast settlement was 75 per cent more than that caused by stationary cyclic loading.

As HS1’s head of track engineering, Dr Sin Sin Hsu is responsible for 109 route kilometres and 143 sets of S&C, of which 62 are high-speed swing-nose turnouts. Her presentation considered how high-speed line maintenance, especially S&C, must consider higher dynamic forces, for example HS1’s track maintenance tolerances are essentially the same as Network Rail’s construction tolerances for 200km/h track.

Dr Hsu described the complex geometry of high-speed swing nose crossings. On HS1 these are produced by Vossloh Cogifer and are 1 in 65, 230km/h turnouts with 152 metres from toe to nose.

She also gave an example of the problems of maintaining high-speed S&C – a badly vibrating point machine which had to be changed every three months. After trying various solutions, the cause was eventually found to be a one-millimetre rail dip, for which the solution was a 0.5mm rail grind. She stressed that this showed the importance of obtaining the correct data to understand the root cause of any problem.

Other high-speed rail papers included an assessment of critical speed by Pedro Alves Costa of the University of Leeds, which concluded that this was governed by soil properties up to a depth of eight metres, and a presentation from the Austrian PORR on the Slab Track Austria (STA) system.

A team from the SNCF also presented a paper on a holistic approach for high-speed lines maintenance and renewal.

Subgrade including asphalt

One of the keynote presentations was given by Professor Carlton Ho of the University of Massachusetts, Amherst, on substructure track design principles and how these differ between the USA and China. He noted that, in the USA, where heavy freight has axle loads of between 33 and 39 tons, standards are based on the American Railway Engineering Maintenance-of-Way Association’s (AREMA) Manual of Railway Engineering (MRE). These are based on geometrics and absence of defects and so allow railroads the flexibility to use the most appropriate design practice.

In contrast, in China, track design is more prescriptive as it must meet the various codes for different aspects of railway engineering.

Professor Ho’s presentation featured probably the longest equation presented to the conference, for the amplitude of elastic displacement of the subgrade bed.

Two presentations considered transitions at bridges. Giacomo Ognibene of the University of Southampton has studied ballasted railway bridge transition using a finite element model to assess the effects of train speed, sub-base soil and under sleeper pads and found that both the train speed and the sub-base material affect transition performance. In particular, it was found that a stiffer, wedge-shaped backfill mitigated the support stiffness variation at the bridge approach.

A paper by Stark and Wynn of the University of Illinois, Urbana, considered ballast-based reinforcement, mechanically stabilized earth reinforced walls, and geosynthetic reinforced and pile-supported embankments (GRPE). This concluded that segmental retaining walls with geosynthetic reinforced soil is a cost-effective solution to mitigate differential movement at railway/bridge transitions and that GRPEs are a more cost-effective method than unreinforced pile-supported embankments for the reduction of soil deformation.

Professor Jerry Rose of the University of Kentucky is clearly a fan of asphalt. His presentation described the benefits from the US railroad industry’s selective use of a 25-37.5mm hot-mix asphalt layer in the track substructure since the 1980s. It described how testing such trackbeds, from 12 to 29 years old, had shown that the asphalt had no brittleness, weathering, or deterioration due to the insulating effects of the overlying ballast. The benefit of its load bearing properties was evident from the asphalt mat being subject to typical dynamic pressures of 13-17psi from the heaviest freight trains whilst the layer below it is subject to 5-7psi.

The use of asphalt outside the USA was considered by Dr Diego Cardona of Eiffage Infrastructure in France. His presentation showed that Italy first used it in the 1970s, for the country’s first high-speed line between Rome and Florence, and now has 1,200 kilometres of asphalt track. In France, a short trial section of the Paris to Strasbourg high-speed line was provided with an asphalt mat in 2004.

After this was shown to require much less tamping than the rest of the line, a further 283 kilometres of French high-speed lines have been built with an asphalt base. Short lengths of asphalt track are in use in Spain, Germany and Austria, where the first asphalt trackbed laid in 1967 had not required any maintenance by 2011, 44 years later. There is also widespread use of asphalt trackbeds in Japan for high-speed and conventional lines.

Dr Cardona noted that this experience highlighted the reduction in both maintenance and line closures from the use of asphalt, which had justified its higher initial cost. However, the use of asphalt required careful consideration of drainage requirement, due to its higher run off, and the need for tamping and ballast cleaning to take account of the asphalt layer.

Drainage and Flooding

With 300,000 hours of delay recorded each year in the UK due to flooding issues, papers considering how potential drainage problems could be better analysed and predicted were well received. A joint paper produced by Network Rail, the University of Birmingham and the University of Lampung in Indonesia proposed a better method to understand underlying problems and failure mechanisms associated with drainage failures. This used expert input to produce a fault tree with 22 casual factors basic events, eight casual factors mid events and three failure modes leading to one top event.

When this method was used to investigate drainage failures at Ardsley tunnel, it was concluded that the underlying problems were change in land use, resulting in increased surface runoff, changes to drainage upstream and damage caused by others or third-party assets.

Yiqi Wu of the University of Sheffield and Raja Jamie of Network Rail presented a Markov chain model for predicting the degradation of various classes of railway drainage assets. This approach, a widely used probabilistic model for simulating infrastructure deterioration, considered the influence of various factors, such as construction material, size, shape and location, to quantify the rate of the degradation on all 329,781drainage assets on Network Rail’s Ellipse database.

Cowley Bridge Junction between Tiverton and Exeter St Davids has been subject to frequent flooding and washouts as described in issue 169 (November 2018). Here, the depth and velocities of flows overtopping the railway have exceeded 0.5m and 0.5m/s respectively. This is due to the complex hydrology and character of the River Exe system, which has a sinuous channel that meanders severely back and forth beneath the mainline.

In their presentation, Sinead Lynch and Thomas Mymors of Arup described the complex hydraulic modelling process used to determine the best flood mitigation option, which was the selective lowering of the flood plain on the approach to the embankment into which twin concrete box-culvert sections, 3.5 metres wide x 2 metres high, were inserted.

Earthworks and Bridges

Although the closure of the railway at Dawlish highlighted its vulnerability to the sea, the stability of the 50-metre-high cliff above it poses an equally serious problem. As Tim Laverye of Network Rail described in his presentation, there have been 50 recorded cliff failures in the vicinity. He described the current mitigation for such failures, including numerous sensors in the cliff and its drape netting, and outlined plans to ensure the long-term resilience of the railway. The many issues to consider include the complex groundwater regime and the nature of the dominant Teignmouth Breccia strata.

The problem addressed by the paper produced by Raynor and Bennett of Ove Arup is the design of OLE structures. In a wide-ranging presentation, this addressed ground investigations, selection of foundation type, constraints of construction plant and the need for cost effective design. For example, it showed how pile depth could be reduced, resulting in only a slight increase in permissible contact wire movement.

The fatigue life of riveted railway bridges was the subject of the paper presented by John Mander of Texas A&M University. He noted that, whilst appropriate for new bridges, current conservative design codes are not helpful in assessing the remaining life of older structures. His paper outlined a systematic process that considered both initial fatigue-life and post-crack fracture propagation life through to fracture. This gives a 20 per cent life extension beyond crack initiation, providing a grace period for remedial repairs.

The longevity of masonry bridges was considered by Manicka Dhanasekar of Queensland University of Technology in Australia. He described how digital-image correlation had been used to determine the deformation of masonry arches and a flat jack method was used to measure the elastic properties of aged masonry. This showed that the maximum deformation at the crown of a 150-year old bridge was 0.5 mm for freight trains and that the absolute maximum strain was well within the limit of the masonry arch barrel.

Safety and Environment

The paper “Are Hydrogen trains the answer?” was one of the few about rolling stock. This was presented by your writer and considered the environmental benefits and limitations of hydrogen trains. It concluded that they are not the answer to “life, the universe and everything”.

Loss of refrigerant contributes to greenhouse gas emissions and air conditioning failures. In his presentation, Andrea Stanio of Alstom described how a virtual twin of each train’s HVAC system, coupled with sensor data acquired from the associated physical counterpart, can provide accurate assessment of the actual amount of refrigerant in the system. This reduces both the cost of maintenance of the air conditioning system and the risk of its failure.

Comparing different optimisation algorithms to analyse metro eco-driving was the subject of a paper presented by the Universitat Politècnica de València. This was intended to take advantage of the advanced communications between train and track, which now make it possible to define multiple speed-profiles for ATO (automatic train operation) systems.

This study compared genetic algorithms with the particle swarm optimisation algorithm inspired by the collective behaviour of insect colonies and concluded that, in terms of spread, the swarm algorithm performed better.

Two papers considered train derailments. Shinya Fukagai of Tokyo’s Railway Technical Research Institute considered how the size of machining marks after tyre turning can increase risk of wheel-climb derailment. Richard Bullet of Arup referred to historic accidents as he considered mitigation for the risk of bridge collapse after derailment.

Using an intelligent vision system to improve platform safety was the subject of a presentation by Howard Parkinson of Lancaster University. In it, he identified the potential for such systems to detect potentially dangerous situations, including automatic indication on the driver’s monitor, and reduce platform dwell time. He also identified the issues that a pilot scheme would need to address.

Prize winning presentations

The presentations mentioned in this review are about a quarter of those presented at the conference. They are, of necessity, an arbitrary selection of the 98 presented to the conference, but they give an indication of the breadth, the intellectual rigour and complexity of the issues covered. Four of the papers were given special prizes. These were:

• Best paper by a university researcher: “Analysis of a bridge approach: Long-term behaviour from short-term response” by G. Ognibene, W. Powrie, L. Le Pen, J. Harkness of the University of Southampton;
• Best engineering application paper: “A holistic assessment approach for high-speed lines maintenance and renewal” by A Dhemaied, G Saussine, S El Janyani, Q A Ta, J M Cornet, J Lossignol, M Koscielny, A Schwager Guillemenet, A Hily C Renaud of SNCF;
• Railway Gazette International innovation award: “Proposal of track renewal method using prepared concrete method” by S Matsuo, T Fujioka, S Watanabe, I Arai,Y Yonehara, S Kubota of the Tokyo Metro;
• Best Paper demonstrating use of Geophysics and NDT: “Autonomous vehicle-track interaction monitoring to improve infrastructure maintenance” by S Jovanovic, P Tešić, University of Novi Sad, Serbia and M Dick, Ensco Inc, Springfield, USA.

The award for the best exhibition at the conference was jointly awarded to edilon)(sedra and Staytite.

A lifetime achievement award for distinguished international service in the field of railway track engineering was also awarded to Dr Jerry Rose of the University of Kentucky. A surprise award was that given to Edna Forde for her contribution to the technical development of the PWI by the Institution’s technical director, Dr Brian Counter.

With many of those present travelling half-way around the world to present their papers, the conference demonstrated that railway engineering is an international community from which there is much to learn. It would be good to know if some of this international practice is adopted in the UK as a result of this conference.