Models for chloride ingress into concrete – from Collepardi to today

Abstract

The first model to predict chloride ingress into concrete was presented by Collepardi in 1970 in Italian. That model was used for some 20 years for the design of concrete structures before it was realised that the main parameter, the apparent chloride diffusion coefficient, is not a material property. Since then, the Collepardi model has been significantly improved and developed into more or less sophisticated probabilistic models. Today, we have numerous models for predicting chloride ingress, but all of them have serious limitations that restrict the present use for long-term predictions. The paper describes the history of developing chloride ingress models. An overview is given of the fundamental differences between various models, from those based on Fick's 2nd law and constant or time-dependent diffusion coefficients and surface chloride contents to those based on chloride transport equations with or without a multi-species approach. The three main limitations are shown to be the lack of understanding the time-dependency of the apparent chloride diffusion coefficients and the lack of good long-term data, the chloride content increase with time close to the exposed surface and the difficulties in quantifying the boundary conditions for sophisticated ingress models.