Prediction of chloride binding isotherms of cementitious materials by analytical model or numerical inverse analysis

In this paper, various methods of prediction of chloride-binding isotherms (CBIs) of cementitious materials at equilibrium and in saturated conditions are proposed. The first group of methods consists in numerical inverse analysis of an experimental total-chloride-concentration profile. In this group, a first method requires as input data the effective-chloride-diffusion coefficient (DCl−), while a second method predicts DCl−. Once DCl− and the CBI are assessed, profiles after other exposure times can be predicted. The second group of methods is an analytical model based on the composition of the material, which includes physical and chemical components. The “numerical” and “analytical” CBIs respectively obtained, as well as DCl− and the profiles predicted, have been compared to experimental data for various mixtures (with and without SCM) and ages. The excellent agreement observed between predicted and experimental results, as well as between “analytical” and “numerical” CBIs, demonstrates the reliability and accuracy of the proposed methods.