A method for automated development of model and fragility inventories of nonductile reinforced concrete buildings

Abstract

The nonductile reinforced concrete building (NDRCB) stock—typically, pre-1974 structures in the U.S.—is a well-known high-risk group for seismic hazards. Prior studies indicate that there are approximately 1500 NDRCBs in Los Angeles. Through various ordinances, the owners are currently required to choose between demolition and, when appropriate, seismic retrofitting. Because fulfilling these ordinances will take decades, the potential risk of major losses will persist. In this study, a method for automated development of structural analysis models and damage fragilities for non-ductile moment-resisting frames is established. This capability enables seismic risk assessment at a regional scale using relatively limited building metadata and the era-specific seismic design code. The approach is used first to develop archetypal models in OpenSees, verified through static pushover and nonlinear time-history analyses against prior detailed studies. Fragility curves for discrete damage states are developed through a probabilistic seismic demand model. Additional investigations are carried out to consider the influence of soil-structural interaction effects and to determine the most suitable seismic intensity measures to quantify the seismic damage levels of NDRCB frames. The sensitivity of the proposed modeling method to variations/uncertainties in building configurations and properties is also examined through parametric studies. The method is limited to a particular subcategory of NDRBCs—namely, moment-resisting frames—but extensions to other types appear straightforward.