PREDICTION OF COMPRESSIVE STRENGTH OF BINARY BLENDED CEMENT- SNAIL SHELL ASH CONCRETE USING ARTIFICIAL NEURAL NETWORK METHOD

Abstract

The increasing price and environmental impact of Portland cement have heightened the search for green alternatives in concrete production. Snail Shell Ash (SSA), an agricultural waste rich in calcium oxide content, was explored in this study as a partial substitute for cement. Concrete specimens were produced using 0–30% SSA as a replacement for cement by weight and their compressive strength after 7, 14, and 28 days of curing. At 28 days, mixtures with 10–20% SSA achieved compressive strengths of 23.8–24.0 MPa, which are very comparable to the control mixture (25.1 MPa), confirming the feasibility of SSA at moderate replacement ratios. For compressive strength modeling and prediction, a feedforward backpropagation-based Artificial Neural Network (ANN) was trained using experimental data. The model had a very good predictive power with a coefficient of determination (R²) of 0.981, Root Mean Square Error (RMSE) of 0.92 MPa, and Mean Absolute Error (MAE) of 0.66 MPa. Compared with Scheffe's optimization method, which gave an average prediction error of 2.3 MPa, the ANN method improved accuracy by approximately 60%. The findings confirm that SSA can be effectively used as a cement substitute up to 20% without significant strength loss, toward economic and sustainable construction. Further, the ANN model provides an efficient and reliable tool for concrete mix design optimization through the reduction of the need for extensive laboratory testing and enabling sustainable material development.

KEYWORDS Binary Blended Cement; snail shell ash concrete; Compressive Strength Prediction; Artificial Neural Network (ANN); Sustainable Concrete Materials

-