Optimising running gear

Railway vehicles are complex machines. So their running gear – the bogies and wheelsets – need regular maintenance if they are to perform at their best, as well as being safe, reliable and comfortable for passengers. In recent years, railways have increasingly focused on Condition-Based Maintenance (CBM) for the rolling stock. CBM, an alternative to maintenance at pre-fixed times, determines when and what of maintenance should be done, by monitoring physical assets. This reduces needless labour and optimises servicing when it is most necessary.

CBM is enabled by sophisticated equipment, by installing sensors, actuators, simulation tools, and diagnostics in the running gear of railway vehicles. However, this is easier said than done, for several reasons:

• A harsh environment for high-tech electronic components, due to vibration and shock, dust and dirt, ice and snow, and a wide temperature range.
• The relative movement of mechanical components, plus the rotation of wheels and axles: this complicates integrating a power supply or wiring sensors and actuators.

Under the Shift2Rail programme, taken over up by the Europe’s Rail Joint Undertaking, a consortium developed a series of demonstrations to explore the huge potential of sensors, actuators, computer modelling, and diagnostics of running gear. Even if the scope of work within the running gear activities under Shift2Rail have worked on developing innovative combinations of architectural concepts, lighter materials and significantly improved performance levels, in relation to CBM activities six demos have been specifically addressed:

Demo: Hardware architecture and main components

• Solution: A suitable architecture to collect sensor data in the bogie and transmit them to the TCMS (Train Control and Monitoring System) of the vehicle, as well as to the railway’s maintenance management system.
• TRL: 4 (technology validated in a lab).

Discussion

In this demo, two core hardware components were developed: a sensor gateway to provide the power supply to the sensors and collect their measured data; and a triaxial sensor prototype, connected to the sensor gateway. Both prototypes were successfully tested in line with the standard EN 50155, which defines the requirements for tests of electronic equipment installed on rolling stock.

Key findings

If fully developed, this CBM concept by Alstom could reduce maintenance costs for running gear of up to 12% and increase reliability by up to 15% (mean distance between failures) – by comparison with the ways that maintenance activities were tasked/performed without the use of the technology developed and the different sensor types used. However, these benefits would mean increasing capital costs by approximately 15%.

Demo: Hardware and software architecture, components and algorithms

• Solution: Collection and analysis of wired and wireless sensor data onboard the train and on the trackside.
• TRL: 7 (system prototype demonstration in an operational environment) for the wayside measurement prototype.

Discussion

Data acquisition equipment developed in this demo included hardware, a measurement architecture, machine learning techniques, and web-based CBM software. Their prototypes were tested and installed on a train and a wheelset, and tested under very harsh conditions on a real track. A wayside measurement site was installed on a metro line in regular service, to validate the concept and the CBM algorithms for detecting wheel flats and other defects of the running gear in an operational environment.

Key findings

Thanks to this developed/tested solution, the expected impact of the proposed system is a 2% increase in reliability (2% longer mean distance between failures), and a 5% reduction in running gear maintenance costs, if compared with no data acquisition equipment that helps detecting anomalies in the operation. Capital costs would only rise by 1%.

Demo: Bogie component diagnostic system

• Solution: Monitoring of health status of all critical bogie components of a train and transmission to the cloud/dashboard.
• TRL: 7 (system prototype demonstration in an operational environment).

Discussion

All 85 regional trains of the Rhein-Ruhr-Express (RRX) in Germany were fitted with sensors in each car and in each bogie, to continuously measure rotational speed and accelerations at various locations. These parameters, recorded in real time and stored on the train in a diagnostic unit, were analysed with a tool trained with machine learning techniques. The health status of all critical bogie components of the train is transmitted to the cloud, for use by a land-side dashboard. With this dashboard, maintenance staff get an overview of the status of all trains, down to the bogie components. So they can detect anomalies that require immediate attention or those that can be inspected or exchanged at the next planned maintenance. The long-term field test, with the whole fleet in passenger service, has proven this system’s capability under normal operating conditions.

Key findings

Thanks to this developed/tested solution, the maintenance cost of the bogie components could be reduced by 10%, and the mean distance between failures of the same components could be increased by 20%, if compared with non-existing sensors equipped that helped on the analysis of the operating conditions. There is clearly a benefit for the safety against derailment, since severe suspension failures can be detected early. The estimated additional capital cost is 10%.

Demo: Health Monitoring of High-Speed trains

• Solution: Cloud-based data analytics of continuously measured parameters for a fleet of trains and selected train components, viewable on a dashboard.
• TRL: 7/8 (system prototype demonstration in an operational environment/ system completed and qualified).

Discussion

In this demo, the health monitoring system developed by Talgo was tested in several fleets of high-speed trains in Spain. The system’s six main features are 1) cloud-based automatic data collection and processing of continuously recorded sensor data of all trains; 2) data analytics to compare the measured data with estimated values and alarm thresholds; 3) dashboard presentation of the health status in a hierarchical structure, from the whole fleet to trains, individual wheelsets, and down to running gear components; 4) recorded data are correlated with the geographical locations, allowing collateral health monitoring of the track quality, mainly through the accelerations measured on the trains; 5) maintenance activities carried out on the equipment are considered in the dashboard tool, e.g. changes of components or adjustments of wheel steering; 6) and statistical analysis of the measured data shows trends.

Key findings

This health monitoring system, as developed and tested here, shows the running gear maintenance costs can be reduced by 8%, and the mean distance between running gear failures is increased by 10%, if compared with not such features made available to help on the health monitoring of the trains. Although the additional capital cost could be 10%, this stems mainly from the costs of data transmission and processing in the cloud.

Demo: Train-borne Ultrasonic Detection System

• Solution: Monitoring of the condition of gauge-face lubrication in curves.
• TRL: 4 (technology validated in a lab).

Discussion

Some railways use trackside equipment to lubricate the gauge face in narrow curves, in order to reduce the friction between the wheelset (the flange) and the rail. This in turn reduces the wear of wheels and rails, as well as the running resistance. In this demo by Network Rail, a train-borne ultrasonic detection system was mounted on the inner side of the wheel. Sophisticated algorithms enabled the system to distinguish between dry contact, wet contact, and a film of grease between wheel flange and rail face. A prototype wheelset was equipped with a suitable ultrasonic transducer for static and dynamic tests in the lab, including tests on a roller rig simulating speeds up to 160 km/h.

Key findings

The concept worked well, with a success rate of up to 90% for detecting insufficiently lubricated gauge faces. The rate could be further increased through machine learning and by analysing results from several runs on the same line.

Conclusion

Condition-Based Maintenance can be increasingly implemented on European railways, by using a combination of sensors, actuators, simulations tools and diagnostics. Despite the difficulties of installing on them on running gear, it is clear that CBM has a lot of potential to reduce rail wear and track settlement. It also enables early detection of potential failures in running gear, which can result in derailment.

Radial steering

Focused on reducing the wear of wheels and rails, as well as avoiding noise in curves, four demos specifically addressed radial steering on the bogie:

Demo: Active hydraulic actuator

• Solution: A hydraulic actuator, in parallel to the passive hydro bush between the axle box and the bogie frame, to guide the wheelset in curves.
• TRL: 4 (technology validated in a lab).

Discussion

The demo by Siemens included a control unit, power unit and four actuators, forming a self-contained system per bogie. The active steering allows an optimum alignment of the wheelset in curves, while the passive hydro bush provides the necessary stiffness and stability for smooth running on straight track. A prototype was built and tested to proof the concept in the lab.

Key findings

The test results confirmed the concept’s suitability and demonstrated the benefits of combining a service-proven passive hydro bush with an innovative active steering system. The concept can improve the curvature performance of the existing passive wheelset guidance. When compared with no hydraulic actuator installed which contributes to improve the radial steering, this developed/tested solution could lead to an estimated reduction of maintenance costs of up to 10%, while the capital costs would increase by 10%. Reliability is hardly affected by the added complexity, with 1% lower mean distance between failures.

Demo: Hydro Elastic Bushing

• Solution: Replacement of conventional rubber bushings for the guidance of the axle boxes on regional trains by innovative hydro-elastic bushings.
• TRL: 7 (system prototype demonstration in an operational environment).

Discussion

In this demo, a specification was developed for the new hydro-elastic bushings, based on simulation results and in line with applicable standards. After virtual testing and optimisation of the dynamic behaviour of the new hydro-elastic bushings, prototypes were tested in the lab against relevant standards for fire safety, harsh environments, and long-term endurance against fatigue. After successful completion of these tests, three regional trains of SNCF were partly equipped with the newly developed hydro-elastic bushings.

Key findings

After 9 months testing in regular passenger service, the benefits of the hydro-elastic bushings were clear: the wheelsets equipped with the new bushings showed significantly less wear than the wheelsets with conventional bushings. This suggests that passive steering with innovative hydro-elastic bushings can reduce running gear maintenance costs by 3%, while additional capital costs are low (1%), if compared with no hydro-elastic bushing being installed in the axle boxes.

Demo: Electromechanical actuators

• Solution: Continuously adjust and optimise the angle of attack of the single rotating wheels of high-speed trains.
• TRL: 6 (technology demonstrated in a relevant environment).

Discussion

Prototype equipment was developed by Talgo and installed on one axle of the prototype AVRIL G3 high-speed train. This concept called on an electromechanical actuator, to adjust the length of the mechanical guiding bar if necessary. There followed a series of tests on track, without passengers, on a section of the high-speed line between Madrid and Barcelona. The prototype components were tested in line with all relevant standards.

Key findings

Field tests on track proved successful. Optimisation of the guiding system for the trains’ wheels resulted in the significant reductions of lateral forces, when running on straight track. Thanks to this developed/tested solution and by comparison with not having the electromechanical actuators installed, running gear maintenance costs could be reduced by c. 10%, with a marginal increase of capital costs (1%) and a small reduction of payload due to the extra weight of the equipment (+1%).

Technology validated in lab: Active steering system

• Solution: Two single axle running gears replace two conventional bogies.
• TRL: 4 (technology validated in a lab),

Discussion

This demo, involving the companies TRV and KTH, called on a two-axle metro car with single axle running gear, instead of conventional bogies. With the long distance between the two axles, steering of the axles is essential for a smooth ride in curves. So several concepts of passive and active steering were investigated and simulated in a complex computer model, a virtual twin of the metro car. Active steering proved to be the most effective solution.

Key findings

Modelling with the virtual twin led to three results: 1) the optimum concept for the active and passive suspension; 2) conclusions about the impact of possible failure modes on stability and safety against derailment; and 3) predictions of the impact of different concepts on wheel and rail wear. When compared with no active steering systems in place, the active steering applied to the two-axle metro car, should reduce the running gear’s maintenance costs by 10%. This would only increase the capital cost by 1%, but there is a penalty of 10% on reliability, in relation to the potential for improving the hydraulic system’s performance, due to the added complexity of an active steering system.

Conclusion

Tests have highlighted the potential for better radial steering of wheelsets. Promising new solutions include the use of active hydraulic actuators to improve the curvature performance of current passive wheelsets; hydro-elastic bushing to provide stiff steering on straight and curved track; electromechanical actuators for more active steering; and a system to actively steer single-axle running gear.

SUSPENSION

Two demos were developed to assess the benefits of active or semi-active suspension systems, to boost passenger comfort:

Demo: Computer Model (Virtual Twin) of a two-axle metro car

• Solution: Simulation model, a virtual twin, for investigating active suspension and active steering.
• TRL: 3/4 (experimental proof of concept / technology validated in a lab)

Discussion

The demo uses the same two-axle metro car (TRV and KTH) as in the TD for active steering and the same simulation models (virtual twin) used for investigating active suspension and active steering. Here, the focus was on selecting actuators and control for the suspension. Passive, semi-active and fully active concepts were studied and simulated. This included assessing accelerations experienced by passengers at different locations in the car, against the limits for passenger comfort as defined in European standards.

Key findings

The studies highlighted the potential of active suspension: a two-axle Metro car can meet the recommendations for passenger comfort for speeds up to 100 km/h or more. This is not possible with purely passive suspension. The virtual vehicle model allowed the simulation of conditions with new or worn wheels, plus the simulation of different load conditions, and of failure modes.

Demo: Design of Experiment to select the most suitable active dampers

• Solution: Virtual testing of the influence of passive, semi-active and active dampers in the suspension of high-speed trains.
• TRL: 3 (experimental proof of concept)

Discussion

This demo by Talgo evaluated three different concepts for suspension (passive, semi-active and active), through virtual tests of a multi-body model of a complete nine-coach high-speed train. It focused on passive, semi-active and active vertical dampers for the primary and secondary suspension, looking at their influence on passenger comfort at medium to high speeds. Passenger comfort is not solely linked to the characteristics of rolling stock, so the demo also simulated the quality (irregularities) of the track and different track qualities, as defined in European Standards.

Key findings

The result was clear: suspension with active vertical dampers, particularly for the primary suspension, is best for passenger comfort, in comparison with semi-active and purely passive solutions. With the same virtual vehicle model, Talgo also studied the suitability of active suspension systems (e.g. the actuators used) for adjusting the floor level to different platform heights, as well as compensating for the variable passenger load.

Conclusion

Future trains will run faster, sometimes over track that is irregular. Virtual tests and simulation show that active suspension will likely be the best way to ensuring passenger comfort.

(START BOXED TEXT)

Universal Cost Model UCM2.0

The Universal Cost Model was introduced in an earlier Shift2Rail rail project, to counter the lack of evidence of the economic benefits from the solutions deriving from the Shift2Rail runing gear activities . It aims to incentivise the application of innovative running gear solutions, by developing an assessment methodology that can quantify the impacts of the running gear performance on the whole rail system economics – capital costs, operating costs, and maintenance costs. One of the main improvements is the damage to switches and crossings together with the track settlement or ballast module, that can influence the maintenance cost of the track. Project PIVOT2 enhanced the UCM, to persuade train operators to implement new technologies, through three use cases:

• Predicting the influence of running gear suspension on the settlement of track (track irregularities at high speed).
• Calculating rail wear in curves (e.g. simulating the effect of rail face lubrication).
• Correctly predicting damage on the outer rail (rolling contact fatigue).

(END BOXED TEXT)

Next steps

• Gauge face lubrication: the prototype equipment will be installed on a rail vehicle, for proving with on-track testing.
• Talgo active steering: the concept developed for the Talgo AVRIL platform will be implemented on part of the AVE S-106 fleet of Renfe. The aim is to study the long-term behaviour, mainly on wheel wear, based on the positive data obtained during the tests. These field tests are for a period of two years, to compare the evolution of wheel wear.
• Calibration of the Universal Cost Model (UCM2.0) to make the prediction of absolute cost more accurate.