No abstract

No abstract

Piezoelectric-based semi-active vibration reduction techniques typically rely on rapid changes in the electrical boundary conditions or corresponding stiffness state. Approaches such as state switching and synchronized switch damping on a resistor or an inductor require four switching events per vibration cycle, with switch timing associated with displacement extrema. Any deviation from this switch timing affects the performance of these techniques. Typical harmonic forcing analyses focus on the energy dissipation and only evaluate the performance at resonance. This study evaluates displacement reduction for harmonic excitation, both at resonance and for frequencies near resonance. Furthermore, it examines the effect of sub-optimal switch timings. Numerical simulations of a non-dimensional model are performed, and an analytical solution is derived for any switch time. This analysis shows that the optimal switch timing depends on the forcing frequency relative to the natural frequency of the structure. Thus, the classical switch time at peak displacement is only optimal when the excitation is exactly at resonance. Even when the optimal switch timing is known, uncertainties in vibration sensing cannot guarantee that switches will occur at the desired moment. Therefore, this work characterizes the degradation in vibration reduction performance when switching away from the optimal switch time based on global, non-dimensional parameters.

Active wearable tremor suppression devices apply actuators to the human body to produce joint torques that reduce tremor motion. This potential alternative to medications and surgery has the advantage of providing robust tremor treatment that is non-invasive, but the bulkiness of typical engineering actuators currently prohibits clinical implementations. Dielectric elastomer stack actuators (DESAs) offer a potential pathway towards achieving soft, low-profile tremor suppression devices: DESAs have similar mechanical properties as human muscles and can conform to the human limb. However, low actuation levels and a lack of commercial availability limit the development of DESA-based orthoses. Employing a control approach that only suppresses tremor while allowing the actuators to follow voluntary motion passively significantly decreases actuation requirements to improve potential for clinical devices. Still, DESAs that may offer the necessary actuation characteristics require specialized equipment and techniques. This research advances DESA-based tremor suppression by experimentally demonstrating DESA-based suppression of tremor-like signals on a scaled system using easily manufactured DESAs. Further discussion quantifies the DESA parameters that will enable physical implementations of human-scale tremor suppression and identifies pathways towards achieving those parameters.

Synchronized switch damping (SSD) techniques exploit low-power switching between passive circuits connected to piezoelectric material to reduce structural vibration. In the classical implementation of SSD, the piezoelectric material remains in an open circuit for the majority of the vibration cycle and switches briefly to a shunt circuit at every displacement extremum. Recent research indicates that this switch timing is only optimal for excitation exactly at resonance and points to more general optimal switch criteria based on the phase of the displacement and the system parameters. This work proposes a self-tuning approach that implements the more general optimal switch timing for synchronized switch damping on an inductor (SSDI) without needing any knowledge of the system parameters. The law involves a gradient-based search optimization that is robust to noise and uncertainties in the system. Testing of a physical implementation confirms this law successfully adapts to the frequency and parameters of the system. Overall, the adaptive SSDI controller provides better off-resonance steady-state vibration reduction than classical SSDI while matching performance at resonance.