State space modelling and representation for vibration-based damage assessment

This paper discusses the application of state-space representation of the structural vibration response for the purposes of structural health monitoring and damage detection. The offered methodology considers the changes in the state space and especially in the distribution of points on the attractor that are introduced by the initiation of damage and its accumulation. The probability density of this distribution represented using a basis of orthogonal functions is employed and the coefficients in this distribution are used to characterize the damage state of the structure. The method is applied to a simulated case study of an aluminium beam. The development of the coefficients is tracked with the initiation and the accumulation of damage in the structure. A damage feature based on the introduced coefficients is suggested to detect and quantify damage in the structure. The proposed damage feature is compared to similar ones, which however make use of the modal characteristics of the structure and its frequency response functions.