Design and Optimization of a Permanent Magnet-Based Spring–Damper System

Abstract

This paper reports the design and optimization of a permanent magnet-based spring. The aim of the optimization, performed using a particular form of the self-organizing map (SOM) algorithm, was to determine the dimensions of a ring PM-based spring with a force–displacement curve similar to a desired one. For each step in the optimization process, a spring composed of different ring-shaped magnets was analyzed using a semi-analytical model. Its characteristic was compared with the desired one to search for a minimum cost function obtained by subtracting the evaluated and the desired force–displacement curve. The resulting algorithm was efficient in the design of a spring with a desired characteristic. The geometry obtained was used to study an electrodynamic damper based on the exploitation of the interaction between the moving magnet of the spring and a conductive cylinder. A parametric analysis was performed: the damping effect grows when the cylinder thickness increases and decreases with the gap between the cylinder and the magnets. Also, the cylinder thickness needed to reduce to one the number of overshoots in the moving magnet’s position decreases with the gap increase. Computations were performed using the research code EN4EM (Electric Network 4 ElectroMagnetics) developed by the authors.

Keywords

• self-organizing maps
• permanent magnet spring
• damper
• optimization
• eddy currents