High-strength fiber reinforced concrete production with incorporating volcanic pumice powder and steel fiber: sustainability, strength and machine learning technique

This study examines the properties of high-performance fiber-reinforced concrete (HPFRC) mixes fabricated with five different replacements (0%, 5%,15%,20%, and 25%) of cement with volcanic pumice powder (VPP)and 0.5% and 1% of steel fiber. The outcomes reveal that the VPP and steel fiber blends exhibited significantly higher compressive and splitting tensile strength than the control mix, where a decline in workability and enhancement in density was registered. The HPFRC fabricated with 10% VPP and 1% steel fiber produced the best mechanical performance results among all the combinations. Furthermore, to predict the natural and mechanical properties of the HPFRC as a result of the influencing factors, extensive comparative modeling was performed, and various predictive models were proposed using regressions and machine learning (ML) techniques, i.e., artificial neural network (ANN), random forest (RF). Root-mean-squared error, mean absolute percentage error, and coefficient of determination were just a few of the metrics used to assess the quality of the models. RF was shown to have the highest R² and the lowest Root Mean Squared Error (RMSE), considering it the most effective model. Considering a strategy for environmental sustainability, this study highlights the importance of minimizing carbon footprint by lowering cement consumption.