Assessment of fiber loading effects on mechanical properties and prediction of mechanical properties via artificial neural networks of hybrid jute /sheep wool epoxy composites

Abstract The primary purpose of this study is to realize the effect of fibre loading on mechanical properties like tensile, bending, impact, interlaminar shear strength and fracture toughness of hybrid composites made of alkaline treated jute and sheep wool hybrid fibres and epoxy. A series of composite laminates were prepared with varying fibre (30%, 40%, 50% and 60%) loadings using hand layup technique. The findings indicated that hybrid composite with equal amount of fiber and resin (JW50) showed a higher strength of 40 MPa against tensile load, maximum strength of 99 MPa against bending load, superior interlaminar shear strength (ILSS) of 3.09 MPa, and a maximum fracture toughness of 72 MPa.m 1/2 compared to other fibre loadings. In terms of impact test, JW50 displayed the highest impact strength of 85.98 J/m, underscoring its exceptional capacity to absorb energy compared to other composites. The mechanical strengths of these composites were also predicted using an Artificial Neural Network (ANN) model, which generated accuracy of 87.46% for tensile strength, 86.31% for flexural strength, and 83.52% for impact strength. The absolute percentage errors were found to be 12.53%, 13.68%, and 16.47%, respectively, demonstrating the model’s reliability in estimating composite properties based on fibre content and composition. SEM images of fractured samples under tensile load reveal that higher resin content reduces fibre-matrix bonding, while a balanced fibre-to-resin ratio improves load transfer, but too much fibre leads to debonding due to inadequate resin justifying the experimental results.