Artificial Intelligence Approach For Predicting The Strength Properties of Fly Ash And Ggbs Based Geopolymer Concrete Based on Experimental Data

Abstract

The development of low carbon construction materials has gained significant attention as the environmental impacts associated with conventional Ordinary Portland Cement (OPC) concrete continue to increase. Geopolymer concrete (GPC), synthesized from aluminosilicate rich industrial by-products, offers a sustainable alternative while simultaneously promoting the beneficial utilization of industrial waste. In this study, geopolymer concrete was developed using fly ash and Ground Granulated Blast Furnace Slag (GGBS) as precursor materials, with four different Fly Ash–GGBS proportions investigated to evaluate their influence on mechanical performance. A combination of 12M Sodium Hydroxide (NaOH) and Sodium Silicate (Na₂SiO₃) solutions was employed as the alkaline activating system. The developed mixtures were experimentally evaluated through compressive strength, split tensile strength and flexural strength tests to characterize their load-bearing, tensile and bending behaviour.

To complement the experimental investigation, an Artificial Neural Network (ANN) model was developed to predict the mechanical strength characteristics of the geopolymer concrete using the experimental dataset. The model was trained and validated against the experimentally measured responses and its predictive capability was assessed using Root Mean Squared Error (RMSE) and prediction-error analysis. The ANN predictions exhibited close agreement with the experimental results, with relatively low prediction errors, demonstrating the model's ability to capture the nonlinear relationship between mixture parameters and strength characteristics. The findings indicate that the integration of experimental investigation with ANN-based prediction provides an efficient and reliable approach for evaluating and optimizing geopolymer concrete mixtures. The proposed methodology can reduce dependence on extensive experimental trials and facilitate rapid, economical and data-driven mix design optimization. The study highlights the potential of fly ash and GGBS based geopolymer concrete as a sustainable construction material and demonstrates the applicability of artificial intelligence techniques for predicting its engineering performance.