Towards a Modular Framework for Visco-Hyperelastic Simulations of Soft Material Manipulators with Well-Parameterised Material

verfasst von

Max Bartholdt, Rebecca Berthold, Moritz Schappler

Abstract

Controller design for continuum robots maintains to be a difficult task. Testing controllers requires dedicated work in manufacturing and investment into hardware as well as software, to acquire a test bench capable of performing dynamic control tasks. Typically, proprietary software for practical controller design such as Matlab/simulink is used but lacks specific implementations of soft material robots. This intermediate work presents the results of a toolchain to derive well-identified rod simulations. State-of-the-art methods to simulate the dynamics of continuum robots are integrated into an object-oriented implementation and wrapped into the Simulink framework. The generated S-function is capable of handling arbitrary, user-defined input such as pressure actuation or external tip forces as demonstrated in numerical examples. With application to a soft pneumatic actuator, stiffness parameters of a nonlinear hyperelastic material law are identified via finite element simulation and paired with heuristically identified damping parameters to perform dynamic simulation. To prove the general functionality of the simulation, a numerical example as well as a benchmark from literature is implemented and shown. A soft pneumatic actuator is used to generate validation data, which is in good accordance with the respective simulation output. The tool is provided as an open-source project∗∗∗∗Code available under gitlab.com/soft_material_robotics/cosserat-rod-simulink-sfunction.

Organisationseinheit(en)

Robotik & autonome Systeme
Institut für Mechatronische Systeme
Institut für Dynamik und Schwingungen

Typ

Aufsatz in Konferenzband

Publikationsdatum

2023

Publikationsstatus

Veröffentlicht

Peer-reviewed

ASJC Scopus Sachgebiete

Artificial intelligence, Maschinenbau, Steuerung und Optimierung, Steuerungs- und Systemtechnik, Maschinelles Sehen und Mustererkennung

Elektronische Version(en)

https://doi.org/10.1109/robosoft55895.2023.10122047 (Zugang: Geschlossen)