Development of Topology Optimized Bending-Twisting Soft Finger

Abstract

This paper proposed a systematic framework to automatically design and fabricate optimized soft robotic fingers. The soft finger is composed of a soft silicone structure with inner air chambers and a harder outer layer, which are fabricated by molding process and 3D printing, respectively. The softer layer is utilized for actuation while the supportive hard structure is used to impose constraints. The framework applies a topology optimization approach based on rational approximation of material properties (RAMP) method to obtain an optimal design of the outer layer of the soft fingers. Two basic motion primitives (bending and twisting) of the soft finger were explored. A multi-segmented soft bending finger and a soft twisting finger were designed and fabricated through the proposed framework. This work also explored the combination of bending and twisting primitives by developing a combined bending-twisting soft finger. The soft fingers were characterized by free and blocked movement tests. The experiments showed that the triple-segmented soft finger can achieve a maximum of 50.5 deg no-load bending under the actuation pressure of 53 kPa. The blocked movement test on the multi-segmented soft actuating finger showed that this finger could generate up to a maximum of 0.63 N force under 57 kPa actuation pressure in 7 s of inflating time. The developed twisting soft finger was shown to achieve tip rotation of up to 219 deg under 29 kPa actuation pressure. Finally, the potential capability of the bending-twisting soft fingers was verified through applications like screwing and object grasping.