Damage detection techniques based on data acquired using permanent and/or temporary monitoring systems directly installed on structures, and/or infrastructures, have received a significant attention in the recent scientific literature. The recourse to experimental methods it is necessary also with the aim to characterize the seismic linear and nonlinear behaviour of real structures excited by earthquakes. Structural Health Monitoring (SHM) systems provide also the possibility to better understand the effects of the dynamic soil-structure interaction, together with the role played by the non-structural components on both linear and nonlinear behaviour of the monitored structure. A new methodology for damage detection and localization on framed structures, based on the maximum modal curvature variation related to the fundamental mode of the monitored structure, is proposed in this paper. Particularly, the main outcomes retrieved from several numerical nonlinear dynamic models, and from several shaking table tests, performed at the University of Basilicata using a scaled framed model, have been discussed.
