FEA-Based Inverse Kinematic Control: Hyperelastic Material Characterization of Self-Healing Soft Robots

Recent advances in soft continuum robots revealed the need for accurate models, required to develop advanced control strategies. In this paper a general methodology is presented that allows to create an accurate inverse kinematic control based on hyperelastic models, fitted on mechanical material properties. This methodology is based on finite element analysis (FEA) and links the mechanical properties of the material to the simulated behaviour of a Pneunet actuator. This procedure is valid for any sort of hyperelastic material used in soft robotics, however, as a case study, a specific material with a self-healing ability is considered. This elastomeric polymer can recover from macroscopic damages without losing its mechanical performances, after undergoing a heat treatment. In this article, an accurate characterisation of the mechanical behaviour of this material is provided, involving mechanical testing, both in tensile and compression. On this experimental characterisation data constitutive laws are fitted, using a FEA simulator. The robustness of this material modelling is shown, re-fitting the curve for material samples that were exposed to multiple damage-healing cycles. With this obtained constitutive model, a FEA simulation of a bending soft pneumatic actuator is developed. The simulation results are experimentally validated with a dedicated test bench consisting of a pressure control unit and a motion tracking camera. Using this validated model, an inverse kinematic control is developed, including the FEA simulation in the control scheme.