Experimental and explainable AI-based investigation of bamboo fiber reinforced concrete

Bamboo Fibre Reinforced Concrete (BFRC), a more sustainable alternative to traditional concrete, is limited in terms of structural applications because of things such as poor fibre-matrix bonding, high water-absorbing capabilities and varying durability. By combining research with artificial intelligence that could demonstrate how to achieve optimal performance in mechanical and durability testing of BFRC, this study evaluated the effects of modifying certain variables on workability, water absorption, compressive strength and residual strength using experimental testing and finite element analysis to simulate stress distribution and crack resistance. In addition to response surface methodology and an integrated KNN-LRP-SSO model to make predictions, interpret and optimise BFRC performance using feature relevance analysis, it also demonstrated that treated fibre improves mechanical and durability properties at a 1.5% inclusion rate, with fibre length and concentration being the most impactful; this successfully created high predictive values on the tested variables with R2 values over 0.97. Thus, this research provides an effective, data-driven process for creating sustainable BFRC for future structural engineering applications.