Probabilistic estimation of residual drift demands for seismic assessment of multi-story framed buildings

This paper presents the implementation of a probabilistic approach to estimate residual drift demands (e.g. residual roof, residual drift at specific stories, and maximum residual drift over all stories) during the seismic performance-based assessment of existing multi-story buildings. The approach combines residual drift demand fragility curves obtained from an inelastic intensity measure, incorporates explicitly the aleatory uncertainty (i.e. record-to-record variability) inherent in the estimation of residual drift demands at the end of the seismic excitation, with maximum inelastic displacement seismic hazard curves to obtain site-building-specific residual drift demand hazard curves which express the mean annual frequency of exceeding residual drift demands. Recognizing the evolution of central tendency and dispersion of residual drift demands with changes in the ground motion intensity, this procedure makes use of functional models that capture that variation. It is shown that the relationship between transient (maximum) and residual (permanent) drift demands depends on the mean annual frequency of exceedance and the building’s number of stories for a similar lateral load resisting system.