Dynamic Modified Model for RC Columns Based on Experimental Observations and Bayesian Updating Method

Abstract

In this paper, a probabilistic framework for constructing a dynamic modified model for RC columns was developed based on dynamic loading test results and the Bayesian updating method. The influences of strain rate on the mechanical behaviors of RC columns were investigated by analyzing the experimental data obtained from dynamic loading tests. The basic equations of dynamic modified models for RC columns in terms of yielding strength, ultimate strength, pre-yielding stiffness, and ductility coefficient were established based on the Bayesian updating theory. Posterior distributions of the unknown model parameters were estimated using the Markov chain Monte Carlo (MCMC) algorithm, which provided insight into the mechanism of the dynamic effect on RC structural members. By comparing the predicted results with available experimental data, the accuracy and effectiveness of the proposed models were verified. It was found that the Bayesian-based models can obtain more-accurate predictions of the skeleton curve of RC columns under both uniaxial and biaxial dynamic loadings. Moreover, the superiorities of the Bayesian-based models over the non-Bayesian models were presented through quantitative comparison. The presented approach provides a valuable assessment of the dynamic behaviors of RC structural members, which can be used to improve the accuracy of the seismic response predictions of RC structures.