Accuracy in prediction of reinforcement corrosion initiation: effects of time, chloride binding, and multidirectional exposures on concrete diffusivity

This study presents a comprehensive investigation into the effects of time-dependent diffusion, chloride binding, and multi-dimensional exposure on chloride transport in concrete, and their combined impact on the accuracy of predicting reinforcement corrosion initiation. Unlike conventional Fickian-based approaches, the proposed framework integrates these critical mechanisms. Concrete specimens were exposed to one- and two-dimensional (1D and 2D) chloride exposure for three durations. Four numerical models were employed to estimate the time to corrosion initiation based on the experimental chloride profiles. Model 2, incorporating time-dependent diffusion, predicted more than double the corrosion initiation time compared to the baseline Fickian model (Model 1). Model 3, which combines time-dependency and chloride binding, produced similarly extended predictions, confirming the importance of these mechanisms. Model 4, which includes chloride binding alone, also showed improvements over the baseline but to a lesser extent. The novelty of this work lies in the comparative application of these models to both 1D- and 2D-exposure conditions, using free and total chloride data, and adapting boundary conditions to simulate multi-directional ingress. The results indicate that concrete with higher maturity and chloride binding capacity exhibits significantly lower diffusivity and longer service life. In addition, 2D-exposure accelerated chloride ingress, resulting in shorter predicted corrosion initiation times compared with 1D-exposure. Overall, the findings highlight that integrating time-dependent diffusivity, chloride binding, and multi-dimensional transport leads to more accurate and realistic predictions of corrosion initiation in reinforced concrete structures.