Extraction of damage-sensitive eigen-parameters for supervised SHM

Abstract

These last decades have seen an exponential increase in the amount of research related to structural health monitoring (SHM) due to its potential for significant life-safety and economic benefits. However, the success of this powerful tool strongly depends on the implemented damage identification strategy. Reliable and efficient damage identification algorithms enable to detect faults that lie beneath the surface of the structure and to spot system’s vulnerabilities at a very early-stage. This allows to adopt appropriate remedial measures in a timely fashion thereby minimizing the risk of unexpected collapses. The present paper describes a spectrum-driven damage identification method that investigates three levels of damage, i.e. detection, localisation and assessment. Peculiarity of the method is the use of spectral frequency-dependent Eigen-parameters estimated from the response Power Spectral Density (PSD) matrix, which is demonstrated to be very sensitive to damage-induced changes. The approach is detailed, including initial assumptions, scientific formulation of the problem and derivation of the algorithm. Finally, the effectiveness of the method is validated through a numerical simulation and verified on a case-study structure.