Reconfigurable three-dimensional structures for robotic applications

Abstract

Soft robotics, a growing field in robotics, can overcome conventional robots’ constraints such as rigidity and Degree of Freedom (DOF). Unlike conventional robots, soft robots are compliant, giving them a higher DOF. However, current shape changing materials are based on soft materials, which lead to limited stiffness that is undesirable in real soft robotic applications. To solve this dilemma, some scholars are using rigid discrete particles in soft robots. These particles can form a target shape when they are assembled. Their mechanical properties can be tuned with various confining boundary constraints. However, these particles are hard to assemble in regular topology because of their discrete characteristics. This project aims to combine the flexibility of artificial muscles with the rigidity of architectured particles. The large deformation of artificial muscles could assemble the discrete particles while the high force generated by the artificial muscles could provide enough boundary constraints around the particles for a higher mechanical performance. Two kinds of artificial muscles were designed, and their mechanical performances evaluated. After comparing and optimization, one kind of artificial muscle was chosen as the actuation method to assemble the particles. The mechanical properties of the morphed structure were then tested based on particles actuated by the artificial muscles. This design will provide a feasible way to improve the performance of soft robots.