A Methodology for Optimizing the Calibration and Validation of Reactive Transport Models for Cement-Based Materials

Reactive transport models are useful tools in the development of cement-based materials. The output of cement-related reactive transport models is primarily regarded as qualitative and not quantitative, mainly due to limited or missing experimental validation. This paper presents an approach to optimize the calibration process of reactive transport models for cement-based materials, using the results of several short-term experiments. A quantitative comparison of changes in the hydrate phases (measured using TGA and XRD) and exposure solution (measured using ICP-OES) was used to (1) establish a representative chemical model, limiting the number of hydrate phases and dissolved species, and (2) calibrate the transport processes by only modeling the initial tortuosity. A case study comprising the early age carbonation of cement is presented to demonstrate the approach. The results demonstrate that the inclusion of a microstructure model in our framework minimizes the impact of the initial tortuosity factor as a fitting parameter for the transport processes. The proposed approach increases the accuracy of reactive transport models and, thus, allowing for more realistic modeling of long-term exposure.