Related resources
Search for item elsewhere
University researcher(s)
Automotive electric actuator modelling and design methodologies
[Thesis]. Manchester, UK: The University of Manchester; 2014.
Access to files
-
FULL-TEXT.PDF (pdf)
Abstract
Electromechanical position actuation systems typically consist of an electric motor, driven by a set of power electronics, effecting output through a mechanical transmission. Whilst an optimal fully integrated actuator design from first principles could be considered, this is often not a cost-effective option. It is common to construct designs utilising commercially available subcomponents – the Cummins variable geometry turbocharging application detailed in this thesis provides a typical example. The design problem studied in this work is therefore one of meeting requirements through careful subcomponent selection.Electromagnetic, mechanical and thermal equations are developed to model actuator performance. These may be parameterised based on datasheet values or sample component test data. A set of tests is proposed to extract the required information from example motors; this is demonstrated using five different sample motors. Validation is performed to assess the accuracy of the parameterised models for the sample motors. A process is then developed to use the validated models to assess actuator design performance against a set of requirements.A key contribution of this work is the derivation of a computationally efficient motor model, which may be used with an integrated low-order lumped-parameter thermal model to investigate actuator performance at elevated temperatures – since this is often the limiting factor in machine rating. This allows a user to select the appropriate modelling fidelity, allowing accuracy to be traded against simulation performance. The overall process is demonstrated through the assessment of a full actuator design.The models and design process developed in this work allow a candidate actuator design to be appraised through calculations and simulations at a range of different fidelities, and using only a minimal set of subcomponent parameters. This allows designs that cannot meet the performance requirements to be quickly identified and excluded. Satisfactory designs may then be modelled and evaluated in detail to optimise other requirements, such as cost or volume.