Applicability of models to predict the bending behavior of soft pneumatic grippers

Abstract Conventional robots are of a rigid and unadaptable nature. They provide highly automated solutions for specific technical handling operations. In contrast, soft robotics have been proposed for tasks, which require more adaptable and compliant properties of a robotic system or its gripper. To describe the behavior of soft robotics, hyperelastic material descriptions need to be incorporated in geometric deformation models which may be—for instance—analytical or numerical. This paper studies different models to describe the bending behavior of a soft robotic actuator. Mechanical material tests are applied to parameterize a hyperelastic material model to describe additively manufactured thermoplastic polyurethane. A numerical simulation based on the finite element method is parameterized to describe the actuator’s deformation and compared to a recently published analytical model for such an actuator. Both model predictions are compared with the experimental results of the soft robotic actuator. While larger deviations are achieved by the analytical model, the numerical simulation predicts the bending angle with average deviations of 9°, although the numerical simulations take significantly longer for the calculation. By not only fitting the material model to tensile tests but also compression tests, sufficient correlation between experiments and numerical model was achieved. The findings depict the limitations of current modeling approaches and emphasize the importance of refining analytical and numerical models for novel materials and configurations.