Optimal Design of Magnetic Springs; Enabling High Life Cycle Elastic Actuators

Magnetic springs are an energy storage alternative to mechanical springs in applications that require long lifetime with no fatigue failure. Although this concept exists in academia for more than 30 years, industrial applications have been sparse. The goal of this article is twofold: (i) to position magnetic springs sideto- side with mechanical springs and (ii) develop a design methodology that will allow us to improve dynamic performance and/or reduce average and peak power consumption in highly dynamic industrial motion systems. Additionally, an extensive exploration of industrially feasible magnetic spring design space is performed using 2D finite element models combined with multiobjective genetic algorithm, resulting in Pareto-optimal fronts and parameter sets for each of the studied topology. Aside from geometry optimization, different magnetizations and permanent magnet materials are studied. Modelling efforts are validated on a physical prototype using both static and dynamic measurements of a two-pole magnetic spring within a dedicated setup. The validated results will be used within this paper to position magnetic spring assisted actuators back-to-back with classical industrial solutions.