Continuum and Lumped-Parameter Modeling of a Dielectric Elastomer Array Actuator System

This paper presents the modeling of a soft array of interconnected dielectric elastomer actuators (DEAs). The array exploits selective actuation of dielectric patches through the application of a high voltage to achieve localized deformations. The soft nature of the system induces electromechanical coupling between neighboring DEAs, whose interplay with the elastomer’s large deformations and viscoelasticity results in a highly complex and nonlinear dynamic response. Motivated by the need to model those effects in an accurate and computationally-efficient way, so as to enable both design optimization and real-time control, we present a set of models with different levels of complexity and accuracy. First, we introduce a continuum model formulation able to predict the dynamic electro-mechanical response of the overall array, including coupling effects and local deformation patterns, in a completely physics-based fashion. Second, we develop a control-oriented dynamic model that predicts the average response with a significantly reduced number of variables (and, in turn, lower computation burden), while still accounting for several key physical mechanisms of the underlying system. While the former accurately infers the coupled system response solely from material and geometry data, the latter requires a dedicated parameter identification procedure that is here cast as a numericallyefficient optimization problem. Both approaches are extensively validated on an experimental DEA array prototype, demonstrating high accuracy under both static and dynamic actuation.