Toward greener construction: Compressive strength prediction of rice husk ash concrete using soft computing models

Abstract

Manufacturing Portland cement, the second most widely used material after water, is a highly energy-intensive process that contributes to 8–10% of global CO2 emissions. With the rising demand for construction materials, the search for sustainable alternatives has become imperative. This study examines rice husk ash (RHA)-based concrete as a promising alternative to Portland cement, highlighting its significantly lower carbon footprint and improved mechanical properties. Utilizing agricultural by-products such as rice husk, this research investigates the effects of various factors, including concrete age, superplasticizer dosage (ranging from 6.2 to 7.36 kg/m³), fine aggregate content (1819 to 1859 kg/m³), and RHA (55 to 100 kg/m³), on the compressive strength of RHA-based concrete across 186 different mix designs. Five modeling techniques Linear Regression, Non-Linear Regression, Multi-Linear Regression, Artificial Neural Network (ANN), and M5P-Tree were employed to predict compressive strength, ranging from 16 to 104.1 MPa. Model performance was evaluated using metrics including correlation coefficient, Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Scatter Index (SI), and Objective Function (OBJ). The results indicated that the ANN model outperformed all other techniques, exhibiting superior predictive accuracy and minimal residual error. Sensitivity analysis revealed that age, superplasticizer, fine aggregate, and RHA content were the most influential factors on compressive strength. This research underscores the significant potential of RHA-based sustainable concrete as an eco-friendly alternative to traditional Portland cement, paving the way for more sustainable construction practices.