Contact Force Estimation and Control in Soft Fin-Ray Grippers Using FEM and Experimental Approach

To optimize the performance of soft robotics systems for perfect handling of the objects, contact forces should be estimated and controlled. This study employs finite element method with experimental validation to predict the forces between cylindrical objects and the gripper in contact for non-feedback control. The thermoplastic polyurethane material is selected to fabricate the gripper and two uniaxial tensile tests are performed to characterize it. Based on the test data, the material hyper-elastic behaviour is then approximated with a three- parameter Mooney - Rivlin model within the strain range. Finite element analysis is subsequently conducted to understand the distribution of stress and strain within the gripper leading to precise contact force estimation. Simulation results are next compared with the sensor measurements obtained from the gripper setup. The setup included a stepper motor-driven slider mechanism which allows further contact force analysis under different conditions to validate the finite element method predictions. With this finite element method, the open-loop control of this system is achievable. Moreover, system identification approaches are employed using the least squares method to model the dynamic force responses of the gripper. Finally, the closed-loop control of the contact forces and the desired dynamical performance are achieved by implementing the proportional-integral-derivative controller. Our findings demonstrate that our proposed methodologies can be effectively used to refine the design and control of the soft robotic grippers.