A novel integrated framework for damage detection in steel structures using magnetic piezoelectric sensors and hybrid damage index

Electro-mechanical impedance (EMI) and the vibration-based monitoring techniques deploy piezoelectric sensors that are permanently bonded to the structural surface. This increases the risk of sensor degradation over time and raises monitoring costs due to the need for multiple sensors. This study explores the development and performance of reusable single magnet-bonded piezoelectric sensor (MBPS) for damage monitoring in steel structures to address these limitations. In this study, MBPS is evaluated for capturing bolt looseness by measuring root mean square deviation (RMSD) using the EMI technique and curvature change during modal testing. The MBPS configuration demonstrates high repeatability (Correlation Coefficient > 99.5%), ensuring reliability for SHM applications. Piezo disc buzzer MBPS, noted for having the highest sensor-to-magnetic base area ratio, displays the highest RMSD index in the miniature specimen. It also effectively identifies damage in a thin-walled steel plate using the RMSD damage index (RDI) and the curvature damage index (CDI) across multiple damage states. The damage indices are found to be highest at the actual damage locations but are at the same time significantly spread over the adjacent undamaged panels, exhibiting an umbrella-like formation. A novel hybrid damage index (HDI) is proposed that fuses the local and the global monitoring techniques, integrating higher-order RDI and CDI, thereby effectively reducing noise over the undamaged panels adjacent to the damaged ones. MBPS paired with HDI offers a highly effective and robust monitoring solution as an alternative to bonded piezoelectric sensors, providing a blend of high sensitivity, flexibility, and resilient damage monitoring for various SHM applications in steel structures.