The seismic vulnerability functions for portfolio-level loss estimation are typically developed for general classes of buildings which may not be suitable to assess building-specific risks. Performance-based earthquake engineering (PBEE) provides the means to conduct building-specific seismic risk assessments. However, such assessments often rely on computationally-intensive analytical frameworks such as incremental dynamics analysis (IDA) which poses a challenge for many types of risk assessment projects. To expand its accessibility, FEMA P-58 outlines a simplified method to predict the nonlinear responses of buildings in which the scope is limited to lower levels of inter-story drifts (less than 4%). This limitation restricts its application to ductile structures, particularly when predicting the vulnerability of modern special moment frame systems. To overcome this shortcoming, this paper proposes an enhanced methodology by which the nonlinear responses of some common structural systems can be predicted by interpolating from a structural response database, itself developed by IDA. The database adopted in the current study consists of structural responses of 61 distinct modern buildings with variety of heights (number of stories), construction material, and lateral load resisting systems. Two building reference models, light-wood frame and special reinforced concrete moment frame with varying heights, are selected to validate the performance of the proposed statistical method. The predicted structural responses for these buildings are benchmarked against the corresponding IDA results. The estimated vulnerability of buildings based on the enhanced simplified method is in good agreement with IDA results. The proposed framework can be used in expedited seismic risk evaluations to estimate the losses of buildings in a large portfolio of diverse structures.
Ziaei, Ershad; Safiey, Amir; Pang, Weichiang; Rokneddin, Keivan; Javanbarg, Mohammad; and Gu, Mengzhe, "Seismic Vulnerability Assessment of Buildings Using a Statistical Method of Response Prediction" (2018). Publications. 24.