Model-based design of multi-stable 3D soft manipulators: a dielectric elastomer case study

Multi-stable design principles represent an effective means to enhance the performance (e.g., stroke, energy efficiency) of several engineering systems. Systematic integration of multi-stability in soft robotic manipulators, on the other hand, is far less explored in the literature. In this paper, we investigate how multi-stable design principles can be used to enhance the performance of soft robotic manipulators, using as case study a dielectric elastomer actuator (DEA) driven spatial manipulator with a flexible backbone. Leveraging a model-based approach, we aim at expanding the robot workspace by exploiting the buckling instability induced by the DEAs into the flexible backbone. First, we investigate the ability of established soft robot models, namely Polynomial Curvature and Piecewise Constant Curvature, in describing buckling load and post-buckling shape of soft beams. Building upon those results, we develop a quasistatic physics-based model for a 3D DEA-driven soft bending system. It is proved that a suitable choice of the geometric parameters allows this model to simulate both mono-stable and multi-stable actuation. Numerical simulations based on realistic parameter values reveal that if the system is designed to be multi-stable the static workspace is significantly larger in comparison to mono-stable configurations. Furthermore, multi-stable designs give rise to two possible large-bending actuation modes, which will be used in future 3D soft robot designs.