Multi-physics based simulation of hyper elastic soft robot motion with hydraulic actuation

A robust method is needed to model and simulate the complex material properties of soft robots. The simplicity of traditional robot materials and actuation methods make them easy to model and simulate. However, this simplicity prevents traditional robot modeling and simulation software from extending their capabilities to the robust physics of soft robots. The goal of this project is to demonstrate the validity of utilizing the open-source Multi-physics Object Oriented Simulation Environment (MOOSE) based Kraken software to accurately model the physics and simulate the movement of a hydraulically actuated soft robotic arm. To achieve this goal, a 3-D CAD model of a soft robot is coupled with the MOOSE-based Kraken software to simulate expansion which is commonly used for reaching tasks. The intellectual merit of this thesis is the accurate modeling and simulation of the complex material properties and interactions between hyperelastic materials and hydraulic actuation. The MOOSE framework allows both accuracy and convergence control via direct user control of simulation parameters. The object- oriented nature of MOOSE allows for simulations to mix different physics properties depending on the nature of the simulation problem and build out additional modules which can prove useful to soft robotic applications. The Kraken simulation framework recently developed specific modules on top of MOOSE to allow for the simulation of different hyperelastic materials and contact mechanics. MOOSE-based Kraken opens the door for soft robotic simulations with various actuation methods, material properties, and task spaces. [1]–[3] The broader impact of this work includes validating MOOSE-based Kraken, an open-source method, to provide increased accessibility of information to the public where code and project improvements can be made by the community. Simulation results can be used as a proof of concept of soft robotic operating parameters before conducting expensive experimental tests. Simulation results can also serve as inputs to machine learning models for improving soft robotic control.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Nuclear Plasma and Radiological Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2024

Author dc:creator

• Luo, Kelsey

Contributors dc:contributor

• Zhang, Yang
• Kozlowski, Tomasz

• Soft Robotics
• Physics-based Simulation
• Moose

Statement dc:rights

• Copyright 2024 Kelsey Luo

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/124605

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F124605