Optimizing Sustainable Self-Compacting Mortar Properties Using Natural Pozzolan and Glass Powder as Cementitious Materials: An Artificial Neural Network Approach

This study aimed to investigate the use of natural pozzolana (PZ) and glass powder (GP) as binary and ternary cementitious materials to improve the sustainability of self-compacting mortar (SCM) for construction purposes. The research assessed PZ and GP's impact on SCM's flowability, mechanical strength, and durability. To do this, artificial neural networks (ANNs) were utilised to model and forecast the behaviour of the materials, offering a comprehensive understanding of how these novel composite materials affect the properties of the mortars. The findings indicated that the inclusion of GP boosted the workability and filling capacity of the mortars. In contrast, PZ improved compressive strength, with both materials exhibiting a synergistic impact when combined in a ternary blend. This combination also resulted in decreased water absorption and porosity; hence, it improved the durability of the SCM. The ANN models accurately forecasted the behaviour of the mortars and their influence on the parameters of the mixtures. The study found that using sustainable cementitious additives instead of PZ and GP improves the quality and performance of mortars while also helping the environment by lowering energy use and CO ₂ emissions.