The dynamic interaction between a magnetically levitated vehicle and a flexible track
The only commercially operating magnetically levitated (maglev) transport system in the world is the link between Birmingham International Airport and the National Exhibition Centre. Comparative financial analysis for this route showed that the construction costs for both wheeled and maglev systems were similar and that the cost of the guideway accounted for over 70% of the total. In part this was because the guideway was elevated; a likely requirement for any future urban system. A substantial reduction in installation costs for a system of this nature can only be achieved by the use of cheap, lightweight and flexible guideways. The British Rail Research maglev vehicle was designed for use on a rigid guideway and it was known that excessive flexibility would make the suspension control system unstable. The aim of the study was to develop a maglev suspension control strategy that was insensitive to guideway flexibility. Vibration measurements were carried out on the Birmingham guideway to establish its modal properties. It was found to be sufficiently rigid to allow the existing controller to work without problems .Measurements were also made on the guideway of a Swiss cablecar transit system. This was felt to represent the extremes of both lightness and flexibility and established the range of guideway dynamics that were likely to be encountered. For the initial experimental work, a section of the British Rail maglev test track was modified to incorporate three sections of flexible track. A personal computer was installed on board the vehicle and software was written to aid frequency response testing and dynamic system modelling. Tests were carried out to establish the dynamic parameters of the new sections of guideway. The existing rigid guideway controller separated magnet control from suspension control. Guideway flexibility destroys this separation and induces additional feedback terms that degrade system stability. Theoretical studies of an improved controller took advantage of the fact that that the suspension magnets act directly onto the guideway and affect the position of both vehicle and guideway. As the guideway is lightly damped it is only flexible over a narrow bandwidth and the new suspension controller is able to use vehicle inertia to react forces that control the guideway at its natural frequency. Theory suggested that this would restore the separation of magnet and suspension control even with a flexible guideway. For a variety of reasons, experimental implementation of the new controller proved to be difficult. Suspension performance on the flexible portions of the guideway was never adequately demonstrated. The work did however enable a very accurate theoretical model of the system to be developed. This model contrasted with earlier predictions because, on rigid guideways, it predicted substantially smaller phase margins than the earlier models had suggested. It showed that the new controller had only modest benefits relative to the original rigid guideway suspension controller. This led to the development of an improved controller, a "lumped" controller where magnet and suspension control are not separated. Modelling for a single degree of freedom vehicle on a single mode guideway showed that large improvements in suspension performance could be made. Further modelling of a three degree of freedom vehicle and a five mode three degree of freedom flexible guideway used parameters that represent the production vehicles at Birmingham. This work defined limits for guideway flexibility and vehicle dynamic performance and showed that maglev guideways for production scale vehicles, with the "lumped" controller, can be very flexible indeed. The major aim of the project was achieved. A suspension controller was developed that will allow a maglev vehicle to work on a guideway that is far lighter, more flexible and far cheaper than the guideway required for a conventional wheeled vehicle.