Search

Most existing approaches to developing robotic manipulators or artificial hands have used rigid components, with joints, linkages, gears, and motors. Rigid robotic systems can perform tasks with precise and articulated motion, but require complex integrated feedback-based control systems. Soft robotics is an emerging research field that uses deformable materials to build systems that are compliant and adaptable via simple integrated mechanisms, enabling biomimetic behavior with compact systems. Here, we report a novel tendon-driven bending actuator using smart soft composite (SSC) and shape memory alloy (SMA). First, an artificial finger was designed based on a SMA wire and a sliding mechanism, which mimics flexion of the human hand. This artificial finger has a soft hinge structure to enable the bending motion of the actuator. Experiments were conducted to evaluate the bending and load resistance of the artificial finger, and an optimal material composition was identified. The bending performance of the actuator was measured with various numbers of glass fiber sheets, and two-layered actuator showed the best performance in terms of the trade-off relationship between the bending capacity and the load holding capacity – bending angle of 305° with weight of 20 g and bending angle of 61° with weight of 60 g. Finally, a prototype robotic hand was then developed using four tendon-driven SSC fingers and a thumb, and grasping capabilities were demonstrated with various objects with diverse shapes.

This work introduces the design of a smart soft composite (SSC) hinge actuator capable of a pure bending motion concentrated on specific sections of the actuator. This actuator makes use of a shape memory alloy (SMA) wire in a polydimethylsiloxane (PDMS) matrix embedded with segmented rigid components. The bending deformation was accomplished by actuating the SMA wire embedded along the length of the matrix eccentrically from neutral surface. This study introduces the design and manufacturing method of the actuator and experimental results for the maximum bending curvature for different configurations of the actuator. A model was presented to calculate the maximum bending curvature and validated against the experimental data. Results show that a smaller flexible length ratio is better for obtaining a larger maximum bending curvature, and that for an actuator containing multiple hinges, the total length of the hinges is more important than the number of hinges in the actuator. Finally, a gripper consisting of three fingers was developed where each finger consists of a hinge actuator with two hinges. The fingers of the gripper have a working length of 90 mm and the gripper was evaluated by grabbing cylindrical objects with diameters ranging between 1 mm and 80 mm. Results show that the gripper has a good passively adaptive capability for grasping a large range of objects with different sizes by virtue of the properties of the hinge actuator. Afterwards, the pulling force through different grasps was tested and it was shown that the gripper is capable from transitioning directly from a power grasping to a fingertip grasp and that power grasping is better than the fingertip grasp.

This paper introduces the design of a smart soft composite (SSC) actuator capable of multiple modes of actuation. This actuator combines four shape memory alloy (SMA) wires embedded in a soft matrix where one or two SMA wires can be activated to induce the actuator into either the bending mode, the twisting mode or the combined bending and twisting mode of actuation. Experimental results for actuators of different lengths were obtained for all modes of actuation and the actuator is capable of large deformations in all modes and directions of actuation. Then a simple FEA model was used to predict the range of deformation for different lengths in the different modes of actuation. This model is able to predict accurately the bending and twisting angles of the actuator for the different modes of actuation. The 120 mm actuator is capable of deformations up to approximately 160° in both the pure bending and pure twisting modes and of approximately 80° for both twisting and bending in the combined twisting and bending mode of actuation.

Natural creatures morph their soft skins to adapt to the surrounding environments. Such unique morphing could be of great utility when creating bio-inspired soft robots. Here, we design shape memory textiles that may be used to create soft morphing shells woven using various kinds of fibers, including shape memory alloy wires. We explore the effects of fiber configurations, textile orientations, and lamination numbers in terms of actuation behaviors (deflections and blocking forces). Finally, soft morphing shells prepared from multiple laminates of shape memory textiles were created to allow three-dimensional surface morphing using two-dimensional laminated composites.