A Human-Inspired Soft Finger with Dual-Mode Morphing Enabled by Variable Stiffness Mechanism

Soft Robot. 2022 Apr;9(2):399-411. doi: 10.1089/soro.2020.0153. Epub 2021 Jun 7.

Abstract

Elevating stiffness without compromising compliance and agility is a key problem for soft finger applications, especially for articulate ones. Inspired by human finger, a multijoint soft finger with dual morphing through active/passive variable rigidity is proposed. The fabricated soft finger weighs 27.4 g. Conductive thermoplastic starch polymers (CTPSs) are embedded in a U-shape-joint pneumatic soft actuator segmentally like biological phalanges. Their stiffness can be independently adjusted utilizing customized thermomechanical property. Yoshimura origami imitating ligaments can passively match deformation and stiffness of the joints. Through electrothemal activation of CTPSs, the finger can realize dual independent articulate morphing: stiffened phalanges (mode 1) for dexterous manipulation and heavy load, softened phalanges (mode 2) for large deformation contact and light load. Comparative experiments of bending angle, output force, and stiffness are carried out between the active and passive stiffness adjustment of mode 1 and mode 2. The results show that the output force and stiffness of the finger adopting mode 1 can be improved more than two times and five times, and its compliance using mode 2 is almost similar, compared with the pure soft one. To further demonstrate performances of dual-mode morphing, a three-fingered gripper is assembled for grasping and manipulating targets with different shapes, sizes, rigidity, and weight, including playing card, unshelled raw egg, grapes, and unscrewing the bottle cap. It can successfully lift a dumbbell weighing 1460 g with a 7.6 load/weight ratio through a two-mode switch.

Keywords: Yoshimura origami; conductive TPS; dual-mode morphing; human-inspired soft finger; variable stiffness.

MeSH terms

  • Equipment Design
  • Fingers
  • Hand Strength
  • Humans
  • Mechanical Phenomena
  • Robotics* / methods