Combination of a Time Reversal Process and a Consecutive Outlier Analysis for Baseline-free Damage Diagnosis

ABSTRACT: A structural health monitoring system is developed for continuous online monitoring of delamination initiation and growth in composite structures. Structural health monitoring problems are often cast in the context of a statistical pattern recognition paradigm, in which a damage state of a structure is inferred by comparing test data with baseline data. However, subtle signal changes due to damage can often be masked by larger ambient variation of operational and environmental conditions of an in-service structure. Therefore, it is critical for the development of a robust monitoring system to minimize false-positive indications of damage caused by the undesired operational and environmental variation of the structure. The issue of minimizing damage misclassification has been addressed in this article by developing an instantaneous damage detection scheme that does not rely on past baseline data. The proposed instantaneous damage diagnosis is based on the concepts of time reversal acoustics and consecutive outlier analysis, and the proposed damage diagnosis has been tested for detecting delamination in composite plates. Key Words: damage detection, time reversal acoustics, consecutive outlier analysis, delamination, composites, generalized extreme value distribution.