Structural monitoring for post-earthquake damage

Accurate and quick location of seismic damage is critical to establish a hierarchy of both immediate rescue or evacuation operations and of future repairs, strengthening and rehabilitation interventions. The Interpolation Damage Detection Method (IDDM) has proved efficient in localized seismic damage detection. This method is based on the analysis of the significant variation of a properly defined error function occurring as a consequence of a local reduction of stiffness in the structure. In order to distinguish significant variations from random fluctuations of the error function a statistical approach based on a binary hypothesis test is applied. The design of the hypothesis test is based paper upon a probability of false alarm defined basing on data relevant to the undamaged configuration. In this paper the joint influence on the accuracy of the method of the noise in recorded data and of the value selected for the probability of false alarm is investigated.