Trend analysis of acousto-ultrasonic signals for integrity evaluation of metallic aerostructures

By attaching a network of piezoelectric transducers over a metallic aircraft fuselage panel, the paper shows the changes of multi-path acousto-ultrasonic (AU) signals which pass through the aerostructure under fatigue loading. Based on the whole life-cycle of the multi-path AU signals collected from a metallic aerostructure until complete fracture, two different trend analysis approaches are presented in detail for evaluation of structural deterioration. One is based on damage indices computed from the changes of the AU signals with respect to their baseline signals in terms of extracted waveform envelopes and phases, and the other is based on principal component analysis (PCA) to reveal the hidden variation patterns of the AU signals. Consistent trends are shown to be produced by both approaches. Whilst the relationship between the complete set of the multi-path AU signals acquired and the gradual deterioration in structural integrity is found to be represented by the statistical trends of the damage indices in the first approach, it is found to generally follow an inverse U-shape trajectory in the feature space formed by the two most significant principal bases in the second approach. Furthermore, the PCA results are shown to provide a good basis to develop an AU-based traffic light system for aerostructure integrity.