Vibration characteristics and higher mode coupling in intermediate isolation systems (IIS): a parametric analysis

Abstract

Intermediate isolation system (IIS) is currently spreading and gaining significant popularity, mainly in Japan. However, its potentials are not so well-known in European countries, and in USA only one application to building retrofit is registered. The dynamic behaviour of intermediate isolation systems, more complex than the two-degree-of-freedom behaviour of base isolation systems, gives rise to a twofold control mode, which combines isolation and mass damping strategies. However, the research contributions provided in the scientific literature usually concentrate on one single control mode, either isolation or mass damping, and the relevant design methods and criteria. This paper addresses the IIS design problem from a wider perspective and presents an explorative study on the vibration characteristics and dynamic behaviour of IIS, in order to identify the range of different behavioural modes and to propose relevant design guidelines. For these aims, a parametric analysis is carried out, varying the main design parameters, namely: isolation period and ratio, location of isolation layer and mass ratio, distributions of stiffness and mass in the upper and lower structures. A classical modal approach is initially assumed for assessing the contributions of each vibration mode on the global dynamic behaviour of IIS, with a particular focus on the effect of coupling of higher modes. However, since IIS is a non-proportionally damped system, a state space formulation is subsequently adopted for establishing the cases for which the simplified classical approach, only considering two damping values for the isolation and structural parts, can be adopted in a preliminary design stage. Finally, frequency response analysis is carried out for identifying the ranges of predominant isolation and mass damping behaviour and the effect of mode coupling both in terms of local and global response of the isolated models. Design implications are finally derived from the analysis results.