Chamber dimension optimization of a bellow-type soft actuator for food material handling

Soft robots demonstrate interesting possibilities of handling fragile and highly deformable objects without complex and accurate control. So far, the design of soft robots mostly rely on designer's intuition and trail-and-error method which is not efficient and therefore hinders the wide applications of soft robots. In this paper, we presented a way to investigate the optimal chamber design of a bellow-type soft actuator using Abaqus and Isight software. The finite element (FE) model of the actuator was developed in Abaqus with two design variables. The model was then imported into Isight and two objective functions of maximizing bending deformation and contact force were implemented. The optimal parameters were found to be at the boundaries of the predetermined parameter sets. Four kinds of actuators, having parameters of the initial guesses, optimal values, and two other optional sets, were fabricated and experimentally tested. Good agreements were achieved. Two-fingered grippers were constructed using different actuators and grasping tests were performed on defrozen broccolis. Results showed that using the optimized actuators required less air pressure to handle the same targets.