Various sources of uncertainties including material, modeling, and loading uncertainties can have a significant influence on prediction of earthquake-induced damage and estimates of the collapse risk of a building obtained through the performance-based earthquake engineering framework. As a first- step toward quantifying the influences of these uncertainties on the collapse risk of buildings, this paper systematically quantifies the influence of material, modeling, and loading uncertainties on the performance of structural steel beam-columns. In this study, material uncertainty is characterized by probability distribution functions (PDFs) developed using available material coupon data. These PDFs are employed in stochastic simulations to quantify the influence of material uncertainty on the calculated maximum drift ratio of beam-columns. The influence of employing different modeling techniques and software packages, i.e., modeling uncertainty, on the predicted response of steel beam-columns is also quantified, by development of multiple nonlinear models for a beam-column tested experimentally. The dynamic analysis results are used to develop a set of PDFs representing the variability in the structural response due to modeling uncertainties. The uncertainty in the applied ground motions, i.e., loading uncertainty, is quantified by performing Incremental Dynamic Analysis on a nonlinear beam-column model, which is calibrated to and validated against the experimental results. The variation in the response of beam-column is quantified in terms of a set of PDFs and means, medians, and standard deviations for each source of uncertainty at multiple intensity measures (Sa(T1)). The results show that record-to-record and material variability as having the largest and least impact on the uncertainty in the predicted maximum drift ratio, respectively.
June 25-29, 2018
Los Angeles, CA
11 National Conference on Earthquake Engineering