Failure probability analysis of corroded RC structures considering the effect of spatial variability

Abstract

A reasonable assessment of the probability of structural failure is essential to ensure the safe performance of corroded reinforced concrete (RC) structures. In this paper, a novel cross-sectional area loss model of corroded steel bars is proposed and the scale of fluctuation of some key parameters is estimated using the semi-variogram function method. A stochastic resistance deterioration model considering the effect of unbalanced corrosion and spatial variability and a time-dependent reliability assessment method of corroded concrete structures (CCSs) are then presented. A simple case analysis of a CCS is used to illustrate the application of the proposed method. The case analysis results show that the cumulative failure probability (CFP) is overestimated using the popular pit area model, while it is underestimated without considering the spatial variability of calculation parameters. Random field parameter analysis shows that an appropriate choice of the element size is very important for the safety evaluation of RC structures. Additionally, compared with a marine atmospheric environment, the CFP of a case study structure in tidal and splash environments was found to be increased by 20.15% and 70.40%, respectively, due to the gradual increase of corrosion current density. Of the three service environments, the splash environment thus has the greatest impact on structural failure probability.