Multi-Objective Topology Optimization and PID-Controlled Grasping of Compliant Fingers for Parallel Robotic Grippers
Year: 2026
Type: book-chapter
Abstract: Adaptive grasping of objects with varying geometry remains an important problem in robotic manipulation.Although compliant Fin-Ray-inspired fingers provide better passive adaptation than rigid grippers, the combined design of lightweight finger topology and controlled force interaction is still limited in many practical parallel gripper systems.This paper presents a topology-optimized compliant finger design together with a PID-based force control framework for a parallel robotic gripper.The finger was modeled with an inner triangular design region and an outer non-design frame, and density-based structural optimization was performed in ANSYS Workbench using the Sequential Convex Programming solver.The optimization objective combined maximization of directional deformation in the desired wrapping direction with minimization of equivalent von Mises stress, while a mass constraint of 15 g was imposed to match the reference Fin Ray finger.The structural model was analyzed under a fixed support at the mounting face and a distributed load of 10 N at the selected contact region.The optimized density result was subsequently reconstructed into a printable three-dimensional finger geometry using SpaceClaim with manufacturing-aware post-processing.In addition, a PID-based force control strategy was formulated using a second-order servo approximation and real-time force feedback from a SingleTact sensor.A dedicated experimental test bench was developed for comparative grasping tests.Grasping experiments were conducted using an orange at 10 N and 3Dprinted cylinder and sphere objects at 5 N, with ten repeated trials for each condition.The optimized finger improved the wrapping angle by 9.5%, 12.3%, and 28.7% for the orange, cylinder, and sphere, respectively, compared with the reference Fin-Ray finger.The study provides a practical design, control, and validation workflow for compliant robotic fingers and demonstrates improved object-contact coverage for the optimized design under the tested grasping condition.
Institution: FH Kärnten
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7204987688