Abstract

Nowadays analytical models of seismic isolators can fairly reproduce the force response of such devices, when implemented in a large variety of structural systems, such as buildings and bridges. Consequently, realistic hysteretic rules are available for the definition of the dynamic system for Non-Linear Time History Analyses, and earthquake simulations of the considered isolated structural systems can be computed. Such models are generally defined, according to mean values of mechanical properties of isolation devices, even though a certain variability has been experimentally assessed: precisely, statistical analyses of the outcomes of test database have outlined that the main response parameters of isolators should be considered as random variables, rather than as deterministic values. On the other hand, in the common practice both design and assessment procedures are mainly based on deterministic approaches, and bound analyses are ruled in just few standard codes. The present endeavor presents a wide parametric study on a case study structure, in order to assess the variability of the main response parameters, by accounting for random mechanical properties of isolation devices. Precisely, a combination of Lead Rubber Bearings and Flat Slider devices have been considered, and the spatial layout of isolators has been defined, according to a given performance point. The structural response of the case study building has been computed through Non-Linear Time History Analyses, by extracting 10'000 individual values of mechanical properties of devices. Presented results are related to the mean response of a spectrum-compatible set of natural records, in terms of displacement and force of both superstructure and isolation system.