Evaluation of predictive models for water absorption and saturation in cementitious materials

This study presents a comparative analysis of the performance and accuracy of predictive models for capillary absorption and water saturation in cementitious materials. Water absorption data from the literature are evaluated using formulations based on the power law, stretched exponential function, Weibull distribution, and the enhanced Terzaghi model to describe the temporal dynamics of water penetration and characterize the transport mechanisms involved. The ability of each model to represent the distinct phases of the absorption process is evaluated, with emphasis on anomalous water absorption, characterized by an initial rapid flow in capillary pores followed by a slower flow in gel pores. The distinction between these flows is essential for describing the interaction between the microstructure of cementitious materials and the evolution of absorption over time. The results provide a quantitative basis for assessing the suitability of the models in predicting the transition between absorption and saturation stages, contributing to the improvement of concrete durability modeling under real-world conditions and to the revision of technical standards. The results of the modeling and statistical analyses show that Weibull-type and stretched exponential models more accurately describe water absorption in concrete, while the square-root-of-time method produces significant errors.