An artificial neural network model prediction of mortar strengths reinforced with henequen and sisal fibres for sustainable construction

The increasing demand for sustainable building materials has spurred interest in using natural fibers for construction. This study evaluates the performance of henequen and sisal fibers in mortar to enhance their tensile properties and provide alternatives to synthetic fibers. Mortar samples were prepared using a 1:3 mix ratio, with fibers introduced at 0%, 0.5%, 1%, 1.5%, and 2% by weight. The particle size distribution, setting times, consistency, and water ab-sorption and tensile strength of mortar constituents and fibers were determined. Compressive and splitting tensile strength tests were conducted on samples of 75 mm cubes and 50 × 100 mm cylinders at 7, 14, 21, and 28 days, respectively. Data were analyzed using ANOVA at α = 0.05 and a validated Artificial Neural Network (ANN) model. Henequen fiber exhibited lower water absorption and higher tensile strength. The 1% mixture recorded the highest com-pressive strength of 15.08 MPa and 15.41 MPa, and the split tensile strength at 0.5% mixture was 3.43 MPa and 3.51 MPa at 28 days, respectively. ANOVA confirmed significant dif-ferences in tensile strength due to fiber inclusion, and compressive strength results indicated no statistically significant differences across samples. The ANN demonstrated high predic-tive accuracy, with R² values of 0.8785 and 0.7885, and low mean squared errors of 0.3855 and 0.4855, for compressive and split tensile strengths, supporting its use in performance forecasting. The optimum compressive strength was observed at a 1% fiber content, while the split tensile strength improved up to 0.5%. The experimental results support its application in sustainable construction.