Optimization of a dual acting, magnetically driven, linear actuator

Abstract

In this study the geometry of a dual acting, magnetically driven, linear motion actuator will be optimized. This will be accomplished by modeling the system through a set of differential equations to be solved in Matlab. An ANSYS finite element program will be used to model the thermal response of the electrical coils so a maximum temperature can be found. These simulations will be part of a Matlab optimization routine. This routine will optimize the actuator's geometry by minimizing a cost function comprised of the floater displacement, the actuator weight, and the maximum coil temperature. Different simulations will be run, each with a unique set of operating parameters. These tests will be analyzed and an optimal configuration will be found. Once complete, a different approach in meeting the design objectives of a stroke length of at least 10 mils, a low maximum coil temperature and a low actuator weight, will be taken. Instead of optimizing the geometry, the rubber pads that are found between the ECOREs and the floater will be replaced by a classical PID type controller. The effects that this PID controller has on the actuator response in reference to the design objectives will be noted and discussed. During the geometrical optimization, an optimal configuration was found that increases the stroke length of the actuator from approximately 1.01 mils to 2.24 mils. However, by implementing a PID control scheme and keeping the rubber pads, with a stiffness of 2854 lb/in, in the system, this stroke length can be increased to 30 mils. It is recommended that the geometry be changed to geometry found in iteration 53 of optimization trial 4 in conjunction with a PID control scheme. This best satisfies the design objectives.