Enhancing Actuation in Magnetic Soft Actuators: An Inverse Design Framework Incorporating Internal Magnetic Coupling

Abstract

This paper investigates the role of internal magnetic interactions in magnetic soft actuators and introduces a novel framework for their design and optimization. Magnetic soft actuators, composed of soft composites embedded with magnetic particles, enable untethered deformation and actuation under external magnetic fields and hold great potential for robotic applications such as minimally invasive surgery and intervention. Existing design approaches primarily address external field effects and mechanical resistance, often neglecting the contributions of internal magnetic interactions, which are critical to actuator performance. To address this gap, we develop an inverse design methodology that integrates material properties, actuator geometry, and internal magnetic dynamics. By analyzing the interplay between particle volumetric fractions, magnetic moment strengths, and intrinsic properties such as saturation magnetization and coercivity, we propose that internal magnetic coupling can significantly enhance deformation. Rotating-square magnetic metastructures are employed to evaluate actuation behaviors, comparing performance with and without internal coupling. Our findings demonstrate that incorporating internal magnetic interactions into the design process can lead to up to a 36-49% increase in deformation, with simulations showing an increase from 7.7mm to 12.4mm at an external field of 6mT. In addition, the inverse design framework accurately predicted actuator performance, with experimental measurements of relaxed and actuated state deformations matching simulation results within 10%. This study contributes to the field by 1) quantifying the role of internal magnetic interactions in enhancing actuator performance, 2) developing a fabrication process that reliably produces a magnetic soft composite with targeted remanence and Young's modulus, and 3) introducing an optimization framework for predictive actuator design.