Experimental characterisation and machine learning (ML)-based tensile strength prediction of sugarcane bagasse, hemp and flax fibre hybrid composites for potential biomedical applications

This study presents a novel tri-fibre hybrid composite consisting of flax fibre, hemp fibre and sugarcane bagasse embedded in a vinyl ester matrix. Six composite formulations (H1–H6) were fabricated by varying the fibre composition at a constant resin content. The mechanical properties were evaluated through tensile, flexural, impact, and Shore D hardness tests. Sample H6 (SB 20%, HF 15%, FF 15%) exhibited superior performance, achieving improvements of 40.7% in tensile strength, 44.1% in flexural strength, 44.4% in impact strength, and 10.2% in hardness compared to H1. Thermal and structural stability were confirmed using TGA, XRD, and FTIR analyses. One-way ANOVA indicated highly significant improvements (p < 0.001). ML models were applied to predict the tensile strength, with the random forest model achieving the highest accuracy (95.4%). The results demonstrate the potential of these hybrid composites for sustainable biomedical applications.