Over 200 delegates attended the recent 2017 Rail Research UK Association (RRUKA) conference and exhibition, held at King’s Place London. Luisa Moisio, RSSB R&D programme director and industry co-chair of RRUKA, reminded delegates that RRUKA exists to provide a link between the rail industry and academia through which problems can be shared and the results of work reported.
Trenitalia chairman and managing director Ernesto Sicilia, and the company’s engineering director Marco Caposciutti, delivered the first keynote address. Ernesto highlighted the opportunities in the UK – the most open rail market in the whole of Europe. He is delighted by the success of the c2c operation Trenitalia has bought from National Express and is looking forward to collaboration with First Group on the West Coast Partnership, bringing the experience of being the only high speed operator exposed to competition. It was also stressed that Trenitalia can draw on the experience of its infrastructure provider, also part of the FS group (Ferrovie dello Stato Italiane).
With the twin aims of improving customer satisfaction and reducing costs, Marco quoted three examples that typify Trenitalia’s approach to innovation. Firstly, it has specified new high-speed trains with a capability to run at up to 360km/h, even though the current infrastructure only supports 300km/h, as it is confident that it will identify innovative ways of delivering the higher speed without building new infrastructure.
Secondly, Trenitalia is investing heavily in sensor and data solutions to enable maintenance based on condition/cycles of operation rather than time. Thirdly, solar panels are being installed on depot and workshop roofs to reduce energy costs.
Partnerships
Sharon Odetunde of RSSB and the University of Birmingham’s Clive Roberts talked about academic partnerships, past, present and future. RRUKA now has a membership of some 55 academic institutions. Sharon highlighted current research opportunities including a £500,000 research fund to identify data-driven solutions to some of the performance challenges, such as dwell time variations, reactionary delays and lack of berth availability, and so help increase capacity, improve punctuality, and reduce delays.
Clive introduced the UK Railway Research Network – UKRRIN. This has been developed in response to the Rail Technical Strategy, which called for innovations to be introduced faster with fewer of them “getting lost” between academia and commercialisation. A number of the UK’s best academic research organisations have collaborated to propose centres to bring academics together to solve industry problems.
It would take a whole article to describe the process required to set them up, but there are to be four centres, supported by government funding of £28 million for capital investment, and £64 million pledged by 17 industrial partners, with support from organisations in the public sector such as Transport for London and Network Rail. The centres will be:
- Digital Systems, led by the University of Birmingham;
- Rolling Stock, led by the University of Huddersfield in collaboration with Newcastle University and Loughborough University;
- Infrastructure, led by the University of Southampton in conjunction with the University of Nottingham, the University of Sheffield, Loughborough University and Heriot Watt University, Edinburgh;
- Existing testing and trialling facilities.
A formal launch will take place early in 2018 and they are due to start operation on 1 April 2018.
It is planned that there will be a consolidation of RRUKA and UKRRIN academic institutions to deliver five core functions – the link to industry strategies, promotional activities, bridging services, capability mapping and the route to market.
The proposed structure for academic engagement will start with the UKRRIN centres stated above with partner universities supporting delivery in core areas and affiliate universities providing cross-discipline skills that are of value to the network as a whole, in areas such as human factors, economics and education. Clive pointed out a number of risks in delivering this plan, which will need careful mitigation.
Robotic servicing
Following a conference programme full of presentations and debate, it is impossible to report on everything. However, certain topics stood out as being relevant to today’s (or even tomorrow’s) railway.
Neil Drury from South Western Trains introduced three presentations on the theme of robotics and autonomous systems for rolling stock maintenance. All were funded through RSSB while the first two of which were also supported by Chiltern Railways.
The results of a feasibility study on the robust automated servicing of passenger train fluids were presented by Mark Atherton, Brunel University. His team has worked closely with Chiltern Railways to understand the nature of fluids that need routine attention, including windscreen washers, toilet effluent, toilet clean fluid, washbasin & catering water, and sanders together with the locations of the various fillers. Two concepts have been developed which suggested very significant savings in the cost of staff required to replenish fluids – a nasty, smelly and potentially risky job often carried out in the open.
To safely and reliably attach the robotic devices to the train, and afterwards reseal the train fillers, Brunel has used rapid prototyping to develop a number of potential devices. The concepts have attracted much interest and next steps include developing proof-of-concept for connectors and port interfaces including lab tests. They will shortly commence an InnovateUK project with TBG Solutions and Chiltern Railways to develop a working prototype for field tests. Mark said that industry involvement is key to getting the technology to market and that “we welcome additional partners”.
The Cab-Front Cleaning Robot was introduced by Tetsuo Tomiyama of Cranfield University. The problem to be addressed is that cab front cleaning is often a manual activity, is hazardous for the staff (awkward positions, working at height, electrical supplies in the vicinity) and the result is often not very good! The latter point is also true of the automatic cab front cleaning machines on the market today. Cleaning between cars is quite rare.
Cranfield aimed to develop a proof-of-concept prototype of a semi-autonomous robotic cab front cleaner, with versions for in-between the carriages, which would be low cost, robust, and easy to use. The project included system design, mechanical arm design, cleaning head design, designing control algorithms for the arm position, and building a one-eighth scale model to demonstrate the concepts.
Tetsuo explained that the brush design was critical. The force on the brush must be high enough to provide the scrubbing action but not so high that it scratches the surface. The brush also must be controlled to remain parallel to the surface being cleaned.
The other key issue was control of the cleaning arm movement. For comparatively simple cab fronts such as that in class 168/170 Turbostar trains, the control strategy needs no pre-knowledge of the design – the system works it out from the starting point (set manually or from a visual sensor) and then scans the cab front whilst both maintaining constant force and avoiding windscreen wipers and handles.
For more complex front ends, for example trains with corridor connections, visual information must be supplemented by a CAD model of the cab front.
Tetsuo concluded that a cab front cleaning robot is realistic, is likely to be cost effective and overcome all the problems seen with today’s systems. It is also capable of being expanded to clean both coupled cabs and gaps between cars. His parting remark was “on to the full-scale version!”
Reworking wheelsets
Stephen Kent, from the University of Birmingham, talked about the feasibility of the use of autonomous robotic systems for wheelset reworking, especially using laser welding to restore worn wheels.
Normally, after four or five visits to the wheel lathe, wheelsets have to be replaced as all of the machining allowance has now been used up. Whilst this can usually be planned to coincide with bogie overhauls, it is very costly – wheelset maintenance typically accounts for around 20-30 per cent of total vehicle maintenance costs for passenger vehicles. It would therefore be good if material could be added back on to avoid changing wheels.
The work was split into three phases:
- Industry consultation: wheel users and maintainers explained issues, constraints and value to the industry which indicated that a significant saving is available if wheels could last multiple bogie overhaul intervals.
- Robotic Control Development: involved developing the algorithms appropriate to the railway environment rather than the factory production line environment.
- Inspection Technology: the results indicated the best wheelset inspection approaches would be phased-array ultrasonic, and magnetic flux leakage.
- Repair Technology: the most promising additive process is laser deposition as it is highly controllable and minimises heat input.
Whilst no fundamental technical barriers to laser deposition had been identified, Stephen emphasised that repairing wheels is challenging as it is difficult reliably to obtain the correct microstructure and it is tricky to achieve the required wheel rim compression with additive processes. Even assuming these difficulties could be overcome, wheels are single point failure/safety critical items and thus the railway’s approvals process would be difficult and time-consuming.
Stephen said that there are many other ways to extend wheel life and maximise fleet availability before investing in additive repair. One attractive possibility, however, is to apply a low friction material to the flange for routes with sharp curves, or look at additive rail repair, something that would be easier to achieve and arguably of higher value.
In a nutshell
Roger Goodall of Universities of Loughborough and Huddersfield introduced a number of elevator pitches – four to five minute presentations. The University of Huddersfield has been working on a low adhesion braking dynamic optimisation for rolling stock (LABRADOR) simulation model. Hamid Alturbeh briefly described the objective as being to develop a computer simulation tool to predict the braking performance and behaviour of a range of multiple unit passenger train types in normal and low adhesion conditions. This will provide a means of investigating:
- Root causes of occasional ‘very long’ train slides;
- Different brake blending strategies;
- Adhesion gains that different sander placements and/or delivery rates might provide;
- The benefits of a true train speed value to assist low adhesion braking;
- Comparison of current train brake control strategies to identify best practice;
- Improved brake control strategies to improve existing trains cost-effectively.
Hamid said that a model had been built in Matlab/Simulink, integrating all the subsystems of the train and track that affect braking. LABRADOR simulates braking system configurations under varying track gradient and adhesion profiles. So far, it is developed for one to four car trains but the model can be extended to represent longer trains and model brake systems for any rolling stock.
Tingyu Xin described the University of Birmingham’s research into railway pantograph dynamic behaviour measurement and fault diagnosis. Pantograph faults may cause poor current collection, loss of energy supply, reduction of system life, lower reliability, as well as increased maintenance cost and downtime.
The project proposed that it is possible to develop a pantograph dynamic behaviour measurement device that will enable condition based fault diagnosis and maintenance. Dynamic tests have allowed characterisation of critical pantograph parameters and have shown that key faults can be detected. The next phase of the research will work towards transferring the laboratory system to a depot environment.
At the University of Huddersfield, Gareth Tucker has been studying how deployment of available rail steels can reduce life cycle costs. Previous research has focused on investigating vehicle-track characteristics to reduce wheel-rail forces and less effort has been spent on increasing rail steel’s resistance to the imposed forces. Rail manufacturers have recently developed new steels, such as British Steel’s HP335, which provide improved resistance to wear and RCF.
Experimental testing was undertaken to understand the performance of current rail steels. A laboratory twin-disc facility has been developed for future testing of rail steels under more realistic contact conditions and further work controlled testing and microstructural assessment is proposed to cover more rail steels.
Case studies
Stuart Hillmansen of the University of Birmingham and Trevor Dowens from Ricardo Rail presented the SmartDrive Project. A variety of technologies, such as automatic train operation (ATO) and driver advisory systems (DAS), have been proposed or introduced to minimise energy consumption for a given run time. The SmartDrive project is proposing a driver-centric system which would deliver savings by driver education.
Working with Edinburgh Trams, the team calculated the optimum speed-distance curves for all the inter-stop runs to deliver the run-time required and assessed where cues for drivers might be required – for example by introducing coast boards. During night-time trials, trams that were driven by drivers briefed on the new process achieved an average energy saving of around 15 per cent.
Iain Roche, head of innovation at HS2, delivered a second keynote address. Iain said, for him, innovation is about coming up with solutions to HS2’s key challenges:
- Passenger experience;
- Design and aesthetics;
- Safety, health and wellbeing of staff, passengers and neighbours;
- Delivery/construction excellence;
- Logistics, with 130 million tonnes of spoil to move and, mostly, re-use;
- Robust operations;
- Sustainability.
Iain went on to describe examples of how innovation is already helping:
- BIM – to be able to demonstrate in models that HS2 can be built;
- Condition monitoring – with trains running at 300km/h, one needs to be confident the infrastructure will be working and intact;
- Sustainable materials to reduce the embedded carbon footprint;
- Remote survey using drones and satellites to identify materials and monitor progress;
- Nearly at market is an automated method for changing tunnel boring machine cutting teeth to eliminate a hazardous and unpleasant manual task;
- A new chemical technique for removing asbestos and turning it into an inert material that can be used in aggregate;
- Copied from Crossrail is the provision of extra bores to provide sources of heat free-of-charge whilst undertaking grouting operations – at little extra cost to the project but of great future benefit to neighbours.
Iain concluded by describing some of the processes being used to encourage innovation. These include hacks – use some real data, indicate desired outcomes and encourage analysts/software writers to come up with solutions – innovation forums and an innovation hub where people might post ideas. All of these are intended to break down barriers and encourage working together.
Analysing collaboration
The final presentation was from Alexander Knight-Percival, of Unipart Rail and Manchester Metropolitan University, who had just received the RRUKA Best Young Researcher Award for his work on mapping of the electromagnetic environment on the railway and condition monitoring of signalling assets. It involved using trains to monitor the performance of track circuits, which can be unreliable, but rarely fail without giving some sign of impending failure.
Alexander cited the successful installation of infrastructure-based monitoring of London Underground Victoria line track circuits, which has eliminated service-affecting failures.
Alexander had decided to measure the magnetic fields created by track circuits using a very sensitive sensor on board a train. This was tested in practice with the sensor built into a housing mounted on the bogie and connected to a portable module containing a GPS system, analogue to digital converter and hard disk storage.
A vast quantity of data, some 10GB per hour, was collected and analysis showed that the concept works. Alexander showed how track circuit signal strength can be colour coded and plotted on a map of the railway, allowing easy comparisons to be made. Once areas of concern are identified, it is possible to use the base data to see exactly what is happening.
Alexander concluded that his work opens the way to bring track circuit condition monitoring to much of the network at low cost.
This was the final example of the collaboration between industry and academia.
Summing up what had been a very interesting day, the conference was closed by Stuart Hillmansen, senior lecturer in electrical energy systems at the University of Birmingham and academic co-chair of the RRUKA. He said that the day’s discussions had illustrated the need for continued research and innovation in rail, and demonstrated the high level of interest and investment in the sector at present.
This article was written by Malcolm Dobell.
Read more: Cornwall’s Capacity Challenge
