Ability to hold grasped objects by underactuated hands: Performance prediction and experiments

To evaluate and optimize the mechanical design of underactuated hands, a benchmark test is defined that quantifies the ability to hold grasped objects subject to force disturbances as occurs during for instance pick and place operations. The ability to hold is quantified by the magnitude of the maximum permitted static force on the center of a cylindrical object of known radius at which it can still be hold within the underactuated hand. A static grasp model is developed to efficiently calculate the effect of design parameters of the hand on this maximum force. This model is applied to a planar, underactuated hand with six degrees of freedom, calculating the maximum force while the rotational joint stiffnesses and distance between the fingers was varied. We demonstrated that the rotational joint stiffness has no effect on this force, while decreasing the distance between the fingers has a negative effect. These results were validated by measurements in an experimental setup. These preliminary results show that the developed benchmark test can be effective to evaluate and optimize the performance of underactuated hands to hold objects in pick and place operations.