Harnessing Elastic Instabilities in Liquid Crystalline Networks for Adaptive and Autonomous Actuation

Soft robotics is a rapidly evolving field that leverages compliant and stimuli-responsive materials to develop adaptable and multifunctional systems. Unlike rigid robots, soft robotic systems can deform, adapt, and delicately interact with their environment, making them ideal for biomedical devices, autonomous exploration, and biomimetic engineering. However, achieving rapid motion remains a challenge. Liquid crystalline networks (LCNs) provide an excellent platform for harnessing such instabilities due to their ability to undergo programmable, reversible deformations in response to stimuli such as heat, light, and solvents. By integrating mechanical buckling and snap-through instabilities into LCN-based actuators, spontaneous, self-sustained, and even synchronized motion can be achieved, paving the way for adaptive and autonomous actuation. This thesis explores how elastic instabilities in LCNs can be utilized to design actuators that exhibit high-speed transformations and embodied intelligence. Chapter 2 presents a method for embedding elastic instability into LCNs using two-step masked photopolymerization, creating a structure with splay-bend and isotropic regions. This design allows for multiple instability modes due to strong buckling, which were studied experimentally and validated through numerical simulations. The observed instability behaviors are material-independent, meaning they can be adapted to different materials with varying stiffness or strain-temperature dependencies. By controlling geometry, different modes can be accessed to release large amounts of energy efficiently, even in soft materials. The strongest mode results in snap-through events as fast as 6 ms, with peak speeds reaching 60 cm/s in millimeter-scale systems. This design has the potential to develop high-performance soft robotic actuators. Stimuli-responsive materials are critical for soft robotics, sensors, and biomimetic devices. The most common fabrication method involves bilayer structures of active and passive layers, which suffer from stress points at connection areas. Chapter 3 presents an alternative approach: a thin film with a controlled cross-link gradient achieved through polymerization-induced diffusion of dithiol molecules in a thiol-ene network. This method allows for bending deformation under light or chemical exposure, enabling dual responsiveness. Light actuation is achieved through photothermal effects via embedded dyes, while solvent-induced actuation occurs due to anisotropic swelling. This simple yet effective fabrication method enables complex deformations using photomask patterning, making it ideal for soft robotic and biomimetic applications. Chapter 4 introduces an LCN snapper that integrates sensing and actuation, enabling adaptive responses to environmental stimuli. Under continuous light illumination, the snapper undergoes spontaneous snap-through transformations driven by elastic instabilities. Through continuous energy transfer with the environment, the snapper sustains dynamic, reversible snapping cycles without external control. These snappers can detect temperature shifts, surface roughness, and color, demonstrating embodied intelligence. Operating at 0.5 Hz to 3.5 Hz and below 55°C, their compact design makes them suitable for soft robotics applications, enhancing environmental interaction. This work introduces a novel strategy for designing biomimetic devices that merge autonomous motion with adaptive intelligence. Chapter 5 explores collective motion and synchronization in snapping LCNs, drawing inspiration from biological systems like fish schools and firefly synchronization. By leveraging light-driven interactions, multiple snappers communicate and synchronize autonomously, demonstrating coordinated collective motion. The study investigates the influence of network design and oscillator coupling on synchronized snapping behaviour. These findings provide insight into distributed soft actuation, offering potential applications in biomimetic swarms, neuromorphic computing, and self-organizing robotic networks. Collectively, the findings presented in this thesis establish a new pathway for the development of autonomous and adaptive soft actuators. By harnessing elastic instabilities in LCNs, we demonstrate how embodied intelligence and high-speed, collective behaviour can be achieved.