Development of improved intensity measures for probabilistic seismic demand analysis

Abstract

Performance-Based Earthquake Engineering (PBEE) is a multidisciplinary procedure for seismic assessment of existing structures and design of new structures. One of the phases of PBEE is the determination of the mean annual frequencies of exceeding specified values of a structural response parameter (e.g., maximum interstorey drift) due to future earthquake motions. This is done using probabilistic seismic demand analysis (PSDA), which combines the seismic hazard at the location of the structure and the structural response obtained from nonlinear dynamic analysis for a selected set of earthquake records. Past research work has shown that the PSDA results depend greatly on the intensity measure used for scaling the records for the computation of the structural response. This thesis is focused on the development of new intensity measures for use in PSDA. Three reinforced concrete frame buildings (4-storey, 10-storey, and 16-storey) designed for Vancouver were used in the study. Eighty ground motion records representative of seismic motions in the Vancouver region were selected for use in the analyses. Based on comprehensive analyses of the frames of the buildings, two intensity measures designated SN1 and SN2 are proposed. The intensity measure SN1 takes into account the first mode response and the period elongation of that mode during nonlinear response. This intensity measure is intended for short-period (i.e., first mode dominated) building structures. The intensity measure S N2 takes into account the contributions of the first and second modes to the response, and is suitable for long-period buildings. It is demonstrated in the thesis that both SN1 and S N2 are superior in the prediction of structural responses relative to the intensity measure represented by the spectral acceleration at the first mode period, Sa(T1), which is currently the most used intensity measure. They are easy for use, provide reliable results, and are suitable for probabilistic seismic demand analysis of structures.