Substructure Damage Detection by Principal Component Analysis : Application to Environmental Vibration Testing

This paper deals with the application of statistical process control techniques based on principal component analysis to vibration-based damage diagnosis of structures. It is well known that localized structural damages with relative small amplitude may not much affect the global modal response of the structure, at least at low frequencies. Nevertheless, it can be expected that the local dynamic behavior of the damaged structural component is significantly affected. By applying principal component analysis on the sensor time responses, it may be shown that any change of a particular sensor will affect the subspace spanned by the response of the complete sensor set. The subspace corresponding to the damaged structure can then be compared with the subspace of an initial state in order to diagnose possible damages. The problem of structural damage detection is addressed in the case of a fatigue test by means of an electro-dynamic shaker. In this example, monitoring of the structural responses is performed during a qualification test in order to detect any structural damage.