Ultrasound Wave-Matter Interactions in Soft Matter

Abstract

Soft materials that undergo controlled shape change in response to external stimuli are foundational to emerging technologies in soft robotics, minimally invasive medicine, and adaptive biomedical devices. However, most existing actuation strategies rely on embedded electronics, optical fibers, or magnetic fillers, which complicate fabrication, limit miniaturization, and restrict operation in confined or deep-tissue environments. This dissertation establishes focused ultrasound (FUS) as a unified, non-invasive mechanism for remotely and programmably actuating soft matter, advancing ultrasound-responsive polymers, architected locally resonant metamaterials, and hierarchical structures while developing predictive multi-physics and multi-fidelity modeling frameworks. The first part investigates 4D-printed shape memory polymers (SMPs) actuated by FUS. A systematic spatiotemporal characterization links printing parameters, thermomechanical properties, and ultrasound exposure conditions to shape recovery response. Experiments show that FUS generates localized heating, accesses intermediate configurations, and achieves full recovery in functional structures, with optimized conditions yielding >95% recovery while delineating degradation thresholds governed by thickness and acoustic intensity. The second part introduces liquid crystal elastomers (LCEs) as intrinsically anisotropic, ultrasound-responsive materials. Pre-programmed LCE strips exhibit rapid, reversible bending under FUS, driven solely by viscoelastic dissipation and nematic–isotropic transitions, without the need for embedded optical or magnetic components. The dependence of actuation kinetics on geometry, crosslinking density, and ultrasound parameters is quantified through coupled thermal and mechanical measurements. Building on these material-scale insights, the third part develops a locally resonant metamaterial (LRM) comprising a PDMS matrix with polyurethane-filled Nylon 6,6 fibers housing central tungsten microspheres. Finite-element simulations that couple pressure acoustics, viscoelastic solid mechanics, and transient heat transfer reveal deeply subwavelength dipolar resonances and hybrid fiber–cavity modes that concentrate shear, enhance dissipation, and increase local heat generation by an order of magnitude. The fourth part introduces multi-fidelity machine learning framework that predicts ultrasound-driven deformation and shape morphing by integrating geometric, material, and acoustic parameters. Finally, the fifth part extends the framework to biological soft matter, developing a multiphysics multiscale k-space pseudospectral model of tendon to elucidate how focused ultrasound interacts with anisotropic, hierarchical, tissue architectures. The model captures aberrated focusing, harmonic generation, micro-mechanics, and shear-dominant effects, deformations under impulsive acoustic loading, providing mechanistic insight relevant to therapeutic ultrasound. Together, these studies establish a cohesive experimental, computational, and data-driven foundation for ultrasound-responsive soft matter laying the groundwork for remotely powered soft robots, actuators, adaptive implants, and programmable therapeutic systems.