Optimal grasp synthesis to apply normal and shear stresses of failure in beams

Abstract

This paper investigates the less-studied problem of failing/yielding an object purposefully by a robotic hand. A grasp synthesis capable of using the whole limb surface of the robotic hand is designed based on internal force decomposition. The introduced approach is based on quasistatic assumption and optimization of active internal forces in order to counterbalance the formulated task wrench/load of yielding. As different geometrical constraints are dictated by the manipulation circumstances (e.g. metallic sheet shaping or robotic harvesting), the yielding wrench optimization is developed to be not only sufficient for yielding the object but also effective in meeting all motion restrictions on manipulator. Maximum-shear-stress theory is used for yielding analysis of a grasped object. Finite Element Modeling (FEM) simulation results are provided as a validation of our proposed approach.