Meta-Analysis and Machine Learning Models to Optimize the Efficiency of Self-Healing Capacity of Cementitious Material

Abstract

Concrete and cement-based materials inherently possess an autogenous self-healing capacity. Despite the huge amount of literature on the topic, self-healing concepts still fail to consistently enter design strategies able to effectively quantify their benefits on structural performance. This study aims to develop quantitative relationships through statistical models and artificial neural network (ANN) by establishing a correlation between the mix proportions, exposure type and time, and width of the initial crack against suitably defined self-healing indices (SHI), quantifying the recovery of material performance. Furthermore, it is intended to pave the way towards consistent incorporation of self-healing concepts into durability-based design approaches for reinforced concrete structures, aimed at quantifying, with reliable confidence, the benefits in terms of slower degradation of the structural performance and extension of the service lifespan. It has been observed that the exposure type, crack width and presence of healing stimulators such as crystalline admixtures has the most significant effect on enhancing SHI and hence self-healing efficiency. However, other parameters, such as the amount of fibers and Supplementary Cementitious Materials have less impact on the autogenous self-healing. The study proposes, through suitably built design charts and ANN analysis, a straightforward input–output model to quickly predict and evaluate, and hence “design”, the self-healing efficiency of cement-based materials.