ABSTRACT
Whereas the SL prediction according to the performance-based approach (PBA, see BaroghelBouny’s works) seems in agreement with ours simulations if one considers C1 and C3, PBA leads to an important underestimation of the SL for C2. This con-firms that the analysis of a single parameter, such as the porosity, to assess the performance of a concrete regarding its durability against carbonation, is not relevant. Indeed a defect on the porosity due to a high content of paste or the presence of supplementary ce-mentitious materials like fly ash or slag, which can lead to a more porous microstructure if the measurement of porosity is performed too early, can be counterbalanced by good performances of other DI such as transfer or moisture properties.
