Design, analysis, and optimization of a magnetoelectric actuator using regression modeling, numerical simulation and metaheuristics algorithm

Abstract

In this study, a new method was proposed for the design of composite magnetoelectric actuator. Design of experiment (DOE) was utilized to investigate the mutual effect of geometric parameters. Moreover, the effect of impedance phase angle, magnetic field, and bias field were studied through finite element (FE) modeling. Resonance frequency, displacement value, magnetoelectric coefficient, and mode shape were considered as response variables. Analysis of variance (ANOVA), regression modeling and response surface method (RSM) were used to investigate the pair-wise effect of input parameters on the response variables. ANOVA results showed that the magnetoelectric length and piezoelectric thickness are the most important parameters affecting the magnetoelectric performance. The optimization process was performed using Metaheuristics algorithm. Optimum results were verified using magnetoelectric measurement setup and Laser Doppler Vibrometry device. The accuracy of the FE model in resonance frequency prediction was estimated at 97%. The prediction error of the FE model for the magnetoelectric voltage parameter was 14.6%, which was about 12.9% better than the regression model. The confirmation test showed that the regression modeling can only predict magnetoelectric behavior and for determining magnetoelectric performance, a precise FE model would be more reliable. Such proposed optimization technique can be used in the design of magnetoelectric composites.