Exploring tensile properties of bio composites reinforced date palm fibers using experimental and Modelling Approaches

The objective of this study was to assess the tensile strength of epoxy bio-composites reinforced with palm fibers, both untreated and treated with sodium carbonate NaHCO3 at a concentration of 10 % (w/v) for 24 and 96 h, with varying weight percentages of fibers (15 %, 20 %, 25 %, and 30 %). To predict the mechanical performance of the composites, two methods were employed: artificial neural network (ANN) and response surface methodology (RSM). A Box-Behnken RSM design was used to conduct experiments and establish a mathematical model of the bio-composite behavior as a function of the fiber percentage in the samples, specimen cross-section, and treatment time. The ANN forecasts showed consistent expected values for the bio-composite sample behavior, with a correlation coefficient (R²) greater than 0.98 for Young's modulus and 0.97 for stress. Similarly, the correlation coefficients obtained by RSM for the mechanical properties were also highly satisfactory, with an R² of 0.89 for Young's modulus and 0.87 for stress. Finally, the errors generated by each method (Box-Behnken and ANN) were compared to the experimental results.