SOFT PNEUMATIC ACTUATOR DESIGN FOR HAND REHABILITATION ROBOT USING MODELING BASED ON FINITE ELEMENT METHOD

• Main Author: Fidinillah, Tsabita
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/80000
• Institution: Institut Teknologi Bandung

In the conduct of daily life activities, the movement of fingers and hands is deemed essential. The ramifications of the loss of these motor capabilities are substantial, resulting in a discernible decline in an individual's overall quality of life. Although physical therapy has demonstrated efficacy in restoring hand function through repetitive tasks, its success is contingent upon consistent and extended engagement under the guidance of a skilled therapist. To address the significance of repetitive flexion-extension exercises, the integration of rehabilitation robotics has emerged as a necessity, facilitating the development of a system conducive to independent patient exercises. Traditional rigid robotic rehabilitation modalities, grounded in an electro-mechanical paradigm utilizing DC/stepper motors as actuators, face challenges in achieving designs characterized by personalization and ergonomic accommodation tailored to individual patient needs and distinct physiological conditions. The contemporary paradigm shift toward rehabilitation with soft robotics, supported by an electro-pneumatic framework, presents an adaptive solution distinguished by flexibility, relative lightweight attributes, and prospects for personalized interventions. Soft pneumatic actuator (SPA) assumes a pivotal role in soft robotics, particularly in facilitating jlexion-extension exercises on fingers. SPA design has to conform to specific size requirements and can be fabricated through additive manufacturing (AM), also known as 3D printing. This manufacturing technique, characterized by layer-by-layer additive construction, has been harnessed for SPAfabrication through fused deposition modeling (FDM) 3D printers deploying elastic filaments. Notably, FDM technology accommodates the creation of geometrically intricate objects. To optimize the design,fabrication, and evaluation of SPA, this research endeavors to conceptualize and scrutinize SPA with a pneumatic network (Pneu-Net) structure using the finite element method (FEM). FEM modeling, executed through ABAQUS CAE software, is based on a mathematical model representing hyperelastic material properties. The chosen SPA material, thermoplastic polyurethane (TPU), used as the 3D printing filament, is validated through uniaxial tensile tests conducted on a "dogbone" specimen. The results indicate that the Ogden order 3 hyperelastic model is the most appropriate representation of TPU's mechanical characteristics. The identified material parameters, gleaned from the Ogden model's congruence with the stress-strain curve derived from tensile tests, subsequently inform simulations/or nine distinct Pneu-Net design variations. These variations are delineated by disparities in SPA wall thickness and the quantity of constraints at the SPA channel base. All Pneu-Net designs are physically instantiated through 3D printing, facilitating a comparative analysis between experimental and simulated outcomes. The comparative evaluation centers on the bending angle characteristics of the SPA model under predefined pressure inputs. Simulation-derived bending angles are ascertained through the deployment of the angle tool within the ABAQUS CAE software, while experimental bending angles are deduced through meticulous image processing a/post-pressure application snapshots. Discrepancies between simulation and experimental data are discernible, influenced by inherent Pneu-Net design factors, the use of uniaxial tensile tests, material complexity, and experimental conditions that are difficult to model precisely in FEM simulations. Consequently, the research endeavors to extrapolate actual bending angle characteristics through a meticulous analysis of simulation-derived data. This methodological approach yields a commendable degree of accuracy, quantified at 96.93%, 94.55%, and 94.48% for three distinct SPA model variations. This scholarly inquiry furnishes comprehensive insights into the authentic bending angle characteristics shaped by Pneu-Net design considerations, thereby fostering an enriched understanding and supporting the progressive evolution of rehabilitation technology within the realm of soft robotics.