Structural response to stochastic agencies: damage detection

For civil engineering structures it becomes of primary importance to detect and quantify damage at the early stage, much before structural collapse. One of the greatest difficulties in damage detection lays in the low damage sensitivity of structural response. When a very low level of damage occurs, for instance a little variation in structural stiffness, this phenomenon cannot be detected either in time domain or in frequency domain. Different structural response characteristics can be used for this purpose including modal data (Adams et al., 1991; Cerri and Vestroni, 2000), curvature measures (Luo and Hanagud, 1997; Ratcliffe, 1999), frequency response functions (Sampaio et al, 1999; Lee and Shin, 2002) and strain energy (Shi et al., 1998; Cornwell et al., 1999). A robust technique should be able to detect damage at very low level; engineers need tools to recognize when dangerous situations may be initiated and well hidden behind an apparent structural integrity condition, avoiding the use of any complex algorithm. Recently it has been developed a damage identification procedure very useful to detect low variation of structural stiffness (Cottone et al., 2006; Di Paola et al., 2006; Barone et al, 2006). This procedure is based on applying Hilbert Transform, to obtain the analytical representation of the system response to an impulse. In fact by considering some characteristics of the analytical signal, such as the phase, since the latter is very sensitive in detecting damage at early stage, it has been possible to develop a damage identification procedure based on a functional of the square of the difference between the characteristics of the analytical theoretical and measured signal. So far this investigation has been performed on the impulse response function, now we wonder: what about a damage detection procedure based on the analysis of the structural response to the environmental noise. The aim of this paper is just to provide a damage detection procedure analyzing the response of a shear-building type to a stochastic process that may proper simulate environmental noise.