Bayesian-based probabilistic models for the ultimate drift capacity of rectangular reinforced concrete columns failed in flexure mode

Abstract

To accurately predict the ultimate drift capacity of reinforced concrete (RC) columns failed in flexure mode under seismic loading, a probabilistic methodology is proposed to correct the biases in deterministic models and establish probabilistic models. Probabilistic correction models are constructed based on Bayesian updating, which can consider potential critical influences and also yield probability distribution associated with the model parameters and predictions. The probabilistic models are simplified to identify the significant informative terms by Bayesian updating. Then, the influences of the physical properties and size of the sample on the probabilistic models are discussed. The results show that the Bayesian-based correction method can increase the accuracy of predictions and quantify uncertainties. Additionally, adding new samples with different physical properties in Bayesian updating can expand the scope of application of probabilistic models, and the sample size should be at least two times the number of variables involved in Bayesian updating.