Latest Papers

ASME Journal of Mechanisms and Robotics

  • Investigation on a Class of 2D Profile Amplified Stroke Dielectric Elastomer Actuators
    on September 24, 2024 at 12:00 am

    AbstractDielectric elastomer actuators (DEAs) have been widely studied in soft robotics due to their muscle-like movements. Linear DEAs are typically tensioned using compression springs with positive stiffness or weights directly attached to the flexible film of the DEA. In this paper, a novel class of 2D profile linear DEAs (butterfly- and X-shaped linear DEAs) with compact structure is introduced, which, employing negative-stiffness mechanisms, can largely increase the stroke of the actuators. Then, a dynamic model of the proposed amplified-stroke linear DEAs (ASL-DEAs) is developed and used to predict the actuator stroke. The fabrication process of linear DEAs is presented. This, using compliant joints, 3D-printed links, and dielectric elastomer, allows for rapid and affordable production. The experimental validation of the butterfly- and X-shaped linear DEAs proved capable of increasing the stroke up to 32.7% and 24.0%, respectively, compared with the conventional design employing springs and constant weights. Finally, the dynamic model is validated against the experimental data of stroke amplitude and output force; errors smaller than 10.5% for a large stroke amplitude (60% of maximum stroke) and 10.5% on the output force are observed.

Novel and Robust Forward Kinematic Algorithm for Real-Time Control of General Six-Degree-of-Freedom Parallel Robot for Tele-Manipulation and Tele-Navigation

Abstract

The forward kinematics (FK) of a 6-6 universal-prismatic-spherical (UPS) structure of a parallel robot is highly nonlinear, coupled, and has a one-to-many nature of mapping. There exists no close form solution to a forward kinematic problem (FKP), and real-time kinematic control is extremely difficult. This paper presents the implementation of time efficient and robust solution for FKP using a trajectory modifier algorithm along with a Newton Raphson (NR) method. One micrometer in translation and 0.001 deg in orientation accuracy with an average pose computation time of 2.3 ms are achieved. The novel algorithm is elaborated and the detailed performance parameters are tabulated. The paper presents trajectory following experiments to show robust, real-time FK solution and efficient kinematic control on both standalone and master–slave modes to be used for robot-assisted neurosurgery. The neuro-registration using the FK solutions in real time in a tele-manipulation mode is demonstrated.

Read More

Journal of Mechanisms and Robotics Open Issues