PROCEDURES FOR THE PREDICTION OF FLOOR RESPONSE SPECTRA IN BASE ISOLATED BUILDINGS

This study examines horizontal seismic demands imposed on non-structural elements within base isolated structures and proposes a procedure for their prediction using floor response spectra. Although several previous studies have focussed on the ability of base isolation to limit building damage, very few have closely examined the effect of base isolation on floor response spectra, which represent the peak demands on acceleration-sensitive non-structural elements. Hence, this study contributes towards an improved design philosophy for non-structural elements located within base isolated structures. Key parameters that influence the peak demands described by floor response spectra are identified using results from numerical modelling and full-scale shake table testing conducted elsewhere. It is found that the floor response spectra are influenced by ground motion intensity, the relative masses and stiffnesses of the isolators and superstructure, the base isolated system’s modal characteristics and the degree of nonlinear response of the isolation level. These findings align with prior work in literature, such as work by (Kelly, 1990). From the observations made in this study, a modal floor response spectrum prediction method recently developed by the authors for fixedbase structures is extended to consider the effects of nonlinear base isolation response. The approach is found to perform well for base isolated buildings configured with lead-rubber bearings, outperforming current international code approaches. The floor spectrum approach does not perform as well for structures isolated with friction sliding devices, which undergo significant changes in vibrational characteristics once the isolation system is activated. Accordingly, this study further identifies how the recently developed provisions in the technical seismic design standard for non-structural parts and elements in New Zealand may be adapted to consider the observed behaviours in any base isolated structure. Results obtained from application of the modal and technical standard approaches are compared with those obtained from shake table testing of several base isolated buildings. It is concluded that the methods show significant potential and should be developed and tested further as part of future research.