Reliability Estimation for Deteriorating Reinforced Concrete Structures using Bayesian Updating

Abstract

The deterioration mechanisms of reinforced concrete (RC) structures have many kinds of uncertainties, and it is impossible to completely predict the condition of RC structures throughout their lifetime at the initial design stage. Therefore, observation or inspection data, which reflect the actual condition of the existing structures, should be used to reduce the uncertainties of the reliability prediction model of deteriorating RC structures. In this paper, a novel reliability estimation method for deteriorating RC structures using observation data is proposed. Bayesian updating method is used to combine corrosion models with observation data and to update prior probabilistic models, while the structural reliability is calculated with a time-dependent three-dimensional finite element (FE) analysis. Chloride-induced corrosion of reinforcing steels, which is one of the most significant deterioration mechanisms of RC structures, is considered. With Bayesian updating method, the uncertainties in the chloride-ion diffusion model can be reduced, and the probability of corrosion initiation is updated. Finally, as an illustrative case study of the proposed method, the time-dependent structural safety of a box-girder bridge is calculated over its lifetime of 50 yrs. Based on the results of this paper, the structural reliability of deteriorating RC structures can be quantitatively updated, and it can be useful for the appropriate and reasonable decision-making of maintenance.