Investigation of self-healing actuators for robotics

The need for robots that can safely interact with humans has led to the development of the novel field of “soft robotics”. In soft robots, compliance is integrated through flexible elements, which are in many cases elastomeric membranes. Because of their intrinsic flexibility these robots are suitable for applications in uncertain, dynamic task environments, including safe human-robot interactions. However, the soft polymers used are highly susceptible to damage, such as cuts and perforations caused by sharp objects present in the uncontrolled and unpredictable environments these soft robots operate in. In contrast with stiff robots, in soft robotics a large part of the robot’s body will experience dynamic strains. As a result, fatigue will occur throughout the entire soft robotic body. These two damaging conditions lead to a limited lifetime of soft robotic components. Most flexible polymers currently used in soft robots are irreversible elastomeric networks, which cannot be recycled. Therefore, damaged parts are disposed after a limited life cycle as not recyclable waste. In this research I propose to increase the lifetime of soft robotic components by constructing them out of self-healing polymers, more specifically out of reversible Diels-Alder (DA) networks. Based on healing capacities found in nature, these polymers are given the ability to heal damage. As an additional benefit, these polymers are completely recyclable and can pave the way towards sustainable, ecological robotics. A variety of DA-networks was synthesized and characterized that vary in concentration, functionality and (non)stoichiometry ratio of the maleimide and furan reactive components. The knowledge of the direct effect of these three network design parameters on the material properties allow the design and preparation of DA-networks with customized thermomechanical and thermo-responsive properties for dedicated applications. Through modeling of the thermodynamics and kinetics of the DA-reaction and experimental characterization of various synthesized DA-networks, the relation between these three design parameters and the thermomechanical properties are obtained. A new manufacturing technique “folding & covalently bonding” that exploits the healing ability was invented. In addition, the processing parameters to extrude and 3D print DA-networks using fused filament fabrication were determined through modeling of the material behavior and practical optimization. These novel manufacturing techniques were used to develop the first healable soft robotic components including; soft grippers, soft robotic hands, artificial muscles and mechanical fuses. These components, of which some consist of multiple DA-materials, were designed through finite element modeling and their mechanical performances were characterized using customized dedicated test benches. It was experimentally validated that the healing ability of these components allows healing microscopic and macroscopic damages with near complete recovery of initial characteristics after being subjected to a healing process that, depending on the DA-networks used, is completely autonomous or requires mild heating.