Piezoelectric Impedance-Based Structural Health Monitoring using Bistable and Adaptive Piezoelectric Circuitr

Structural health monitoring (SHM) has been extensively explored for various aerospace, civil, and mechanical systems due to its significant importance in enhancing life-safety and economic benefits. Among various SHM approaches, the piezoelectric impedance-based method has shown excellent potential in identifying small-sized structural defects, while maintaining simplicity in implementation. This method utilizes high-frequency interrogation to detect small damages based on the electromechanical coupling effect of piezoelectric transducers. This coupling effect enables self-sensing, i.e., the transducer serves as sensor and actuator simultaneously, which facilitates simple implementation with reduced number of transducers and associated electrical wirings while consuming relatively low electric power. Furthermore, the damage characteristics such as the location and severity can be identified by employing baseline models.;Despite the promising potentials, important limitations exist to achieve reliable SHM implementations. For example, the number of available independent impedance data set is generally far smaller than the number of required system parameters. As a result, the inverse problems for damage identification are often underdetermined, which severely undermines the reliability of damage prediction since the inverse solutions become extremely sensitive to even small measurement errors, especially in practical implementations where the response anomaly induced by small-sized damages may be easily suppressed by damping and buried in signal noise.;To address the limitations and advance the state of the art, this thesis presents a novel methodology that fundamentally improves the underdetermined inverse problem and accurately measures the damage-induced impedance variations to reliably identify small damages under noise influences. This is achieved by strategically integrating bistable and adaptive piezoelectric circuitry with the monitored structure. First, adaptive piezoelectric circuitry with tunable inductor is integrated with the monitored structure, which introduces additional degrees of freedom into the system. By systematically tuning the inductance values, the dynamic characteristics of the electromechanically coupled system can be altered; thereby significantly increased number of different independent impedance variations can be obtained with respect to same damage profile. The enriched data set is then utilized to fundamentally improve the underdetermined inverse problem for damage identification. Next, new bifurcation-based sensing approaches are developed, capitalizing on the strongly nonlinear bifurcation in bistable electrical circuits that exhibit dramatic changes in the response due to small input variations. By utilizing the voltage measured from the piezoelectric transducer as an input to the bistable circuit, the enriched damage-induced piezoelectric impedance changes can be assessed by tracking the circuitry bifurcation points. Considering the stochastic and non-stationary influences on the bifurcation points that are theoretically explored in this thesis, a novel bifurcation-based sensing methodology is developed to provide accurate and robust measurements of the damage-induced impedance changes against unavoidable noise influences. Lastly, the impedance enrichment technique utilizing adaptive piezoelectric circuitry and the advanced bifurcation-based sensing approaches employing bistable circuits are integrated to significantly enhance the reliability of piezoelectric impedance-based damage identification.;The important scholarly contributions of this thesis include: (a) newly developed impedance-based SHM method that fundamentally improves the underdetermined inverse problem, (b) novel integration of the monitored structure with bistable circuits for bifurcation-based sensing, and (c) fundamental understanding of the stochastic and non-stationary influences on the saddle-node bifurcation in non-smooth dynamical systems. The bifurcation-based sensing and identification approaches not only enhances the impedance-based SHM, but has the potential of providing high impact to a broad range of sensing and identification systems that are exposed to noise problem.