Structure, morphology and compressive strength of Alkali-activated mortars containing waste bottle glass nanoparticles

• Main Authors: Huseien, Ghasan Fahim, Iman Faridmehr, Faridmehr, Nehdi, Moncef L., Abadel, Aref A., Aiken, Timothy A., Ghoshal, S. K.
• Format: Article
• Published: Elsevier Ltd 2022
• Subjects: QC Physics
• Online Access: http://eprints.utm.my/103459/; http://dx.doi.org/10.1016/j.conbuildmat.2022.128005

Alkali-activated materials (AAMs) have emerged as a sustainable, clinker free binder alternative to portland cement, with reduced consumption of raw materials and very low CO2 emissions. They are produced via the activation of aluminosilicates with alkaline solutions. Millions of tons of bottle glass waste are generated annually worldwide and only a small amount are re-utilized and recycled. Various recycling processes need to be innovated to reduce the volume of waste glasses, which is often landfilled, causing environmental concerns. Waste glass has been utilized as fine aggregate in concrete or as partial replacement for cement. In this study, waste bottle glass nanoparticles (WBGNPs) were incorporated as a precursor combined with ground blast furnace slag (GBFS) and fly ash (FA) to achieve high-strength alkali-activated mortars (AAMs). The effect of WBGNPs, sodium hydroxide molarity, solution modulus (SiO2:Na2O), alkaline solution content, and binder-to-aggregate ratio on the compressive strength development of AAMs were investigated. The strength characteristics of the studied AAMs were assessed by developing an informational model united with a metaheuristic shuffled frog-leaping algorithm (SFLA). The experimental results indicated that after 28 days of curing, AAM prepared with 5% of WBGNPs (precursor) as GBFS/FA replacement attained an improved compressive strength in the range of 16 to 18.4%. Additionally, the optimum mixture containing 5% of WBGNPs as GBFS replacement achieved a strength of 68.6 MPa after 28 days of curing. It was shown that the novel informational SFLA model attained accurate predictions and could appreciably simplify generative design in future computational intelligence of a construction materials platform in civil engineering.