Ricerc@Sapienza

In this work uncertain vibrations of beam-like structures are considered. Attention is paid to sensors placement and a statistically based method for the selection of sensors networks is proposed. The objective is to characterize the SHM process in terms of identifiability and maximum accuracy that a given network of sensors can provide, regardless of the damage detection techniques to be developed downstream.

Modal curvatures have been claimed to contain local information on damage and to be less sensitive to environmental variables than natural frequencies. However, simply using the difference between modal curvatures in the undamaged and damaged states can result into localization errors, due to the complex pattern that this quantity presents when considering broad damages or higher order modes.