Computational Optimization of Collinear Printed Pneumatic Soft Actuators

Student’s name: Theresa Franklin Home Institution: Cornell University NNCI Site: CNS @ Harvard University REU Principal Investigator: Jennifer A. Lewis – Wyss Professor, Harvard University REU Mentor: Jackson Wilt – School of Engineering and Applied Sciences, Harvard University Abstract: Pneumatically actuated soft robots are often used for their flexible range of applications and similarity to natural biological forms of motion. On the centimeter scale, fluidic soft robots act at a high strength-to-weight ratio and compatibility with natural structures, but require high pressure inputs to operate. However, micron-scale applications can achieve similar morphing motifs at lower pressures and while demonstrating less plastic strain, thus enabling the fabrication of complex microarchitectures and precise actuations. The ability to pattern and process such microactuators is simplified by the coaxial 3D printing of elastomeric tubes on a flexible substrate. This fabrication approach can produce actuators that closely mimic natural structures, such as the furling and unfurling of leaves or vascular networks in the human body. To first explore this computationally, we utilize Finite Element Modeling (ABAQUS) python scripting compatibility and multi-physical analysis. We propose a method to create and run a full static analysis of the pneumatic expansion using an empirically derived hyperelastic model. This method is optimized by implementing the use of a Newton Raphson root-finding algorithm, catering to the highest actuation curvature while minimizing overall strain for improved repeatability and robustness. Thus, the optimization expands the range of applications, enabling the actuators to be used for precise and repeated use environments due to their enhanced durability and fabrication success rate.