Exploring an intelligent model to predict the compressive strength of UGGBFS based geopolymer concrete

Geopolymer cement (GC) is made by reacting aluminosilicate by-products with an alkali activator such as fly ash. As the production of GC significantly reduces CO2 when compared to ordinary Portland cement (OPC), it can be used as a potential sustainable alternative for OPC. Geopolymer concrete (GPC) uses the GC as the main bonding agent together with aggregates and water. This study aimed to investigate the compressive strength of GPC with varying NaOH concentrations from 5 to 8 mol. Ultrafine ground granulated blast furnace slag (UFGGBFS) was used as the binding element and Crushed stone sand (CSS) was used as fine aggregates. The investigation shows that an increase in NaOH concentration and the addition of UFGGBFS with CSS improve the compressive strength of GPC samples. The GPC containing 10% UFGGBFS with a concentration of 8 molar exhibited the highest compressive strength, achieving 35.7 MPa, 47 MPa, and 60.2 MPa at curing ages of 7, 28, and 90 days, respectively. In addition to the experimental study, this paper also presents an Artificial Neural Network (ANN) model developed based on numerical analysis of literature data to predict the compressive strength of GPC samples. The proposed model predicts the compressive strength of experimental results obtained and discussed extensively.