Preliminary Design Rules for Electromechanical Actuation Systems – the Effects of Saturation and Compliances
Electromechanical actuator (EMA) is a type of power-by-wire (PBW) actuator that is becoming widely implemented in aerospace industry. Given the application area, designing an EMA is highly constrained in weight, integration space, maintenance costs, dynamic performances, reliability, etc. In order to reduce the EMA’s design time, cost and effort, it is of utmost interest to account for these constraints during the preliminary design stages. This requires simple and parametric models that can predict the main impact of the sizing variables on performance. In the proposed communication, control of a linear EMA is the main regarded in terms of stability, settling time and overshoot. Since cascade control architecture is most often used for position control, linear control theory permits any desired dynamics for the system to be imposed. However, owing to the other constraints (e.g. mass, integration, etc.), the EMA operates in conditions that are not captured by the linear model used for controller design. For instance, since the airframe is designed under mass constraints, anchorage stiffness between the actuator and the load/airframe is not negligible. Since the driven load inertia and the force disturbances are important, the stiffness of the EMA’s nut screw may become a critical parameter. In these conditions, the coupling between the structural and the actuator dynamics alters the performance of the whole actuation function that can even become unstable. As the actuator is also subjected to mass and volume constraints, the motor rated torque has to be minimized. Consequently, this limit introduces a saturation effect between the demanded and the produced electromagnetic torque. Thus, these technological limitations/imperfections make the real performance far different from the performance expected when a liner model is used for control design. Moreover, since the control becomes a constrained problem, some performance requirements may be even unreachable. In these conditions the controller will have to “live” with these limitations and still ensure the required performance. In order to reduce the round trips between the mechanical designer and control engineer, this communication aims to give simple mathematical relations linking stiffness and torque limitation to the achievable dynamic performance by modelling and simulation in MATLAB/Simulink environment. Thus, the mechanical engineer can use these relations as “best practice rules” for the specification and preliminary design of the mechanical part, enabling the dynamic performance requirements to be met in practice. The control engineer may use these design rules to reduce the number of design iterations through rapid verification of consistency between closed-loop performance requirements and early choices and definition of the mechanical components. The proposed approach is illustrated in the case of an aileron actuator.