A probabilistic framework for post-earthquake safety assessment of mainshock-damaged buildings

Evaluating the safety of buildings for reoccupation after a mainshock, while accounting for possible aftershocks, remains a major challenge in post-earthquake decision-making. Conventional tagging methods rely on deterministic safety thresholds that assume fixed residual collapse capacity and neglect the variability inherent in nonlinear structural response. In addition, the selection of such safety thresholds is often subjective and based on expert judgment, while decisions are typically derived from a single deformation-based damage measure. This study introduces a probabilistic framework for post-mainshock safety evaluation that addresses these limitations by explicitly quantifying uncertainty and incorporating multiple engineering demand parameters (EDPs), including maximum drift, residual drift, and the Park-Ang damage index. The reduction in collapse capacity is obtained from sequential incremental dynamic analyses, and the probability that a damaged building becomes unsafe under aftershocks is estimated from statistically fitted EDP distributions with bootstrap-based uncertainty quantification. Application to a four-story steel moment-resisting frame demonstrates the implementation of the proposed methodology and shows its ability to provide an objective probabilistic basis for safety classification and post-earthquake reoccupation assessments.