Physico-chemical/mechanical properties of treated groundnut shell fibre; Response surface methodology and artificial neural network performance evaluation and optimisation

Abstract

The difficulties associated with forecasting mechanical behaviours of fibres based on the theoretical method is a big threat to precise design when dealing with complex variables; hence, the need to evaluate the fibres' properties using artificial intelligence techniques to minimise error. This study predicted the effect of acetylation process parameters (treatment time and concentration) on the mechanical properties (elongation at break and tensile strength) of groundnut shell fibre (GSF) using Response surface methodology (RSM) and Artificial neural network (ANN). The GSF was processed, dried and treated with NaOH varying the time and concentration of the treatment and then acetylated. The characterisation results from the physicochemical analysis, SEM and FTIR indicated that acetylation treatment removed excess lignin, hemicellulose and cellulose from GSF and improve its bonding characteristics. Assessment of the models through the coefficient of determination (R2) and mean square error (MSE) indicates that ANN for elongation at break (R2 =0.9999, MSE=2.769E-13) and tensile strength (R2 =0.99997, MSE=3.5146E-18) is superior to the RSM model for elongation at break (R2 =0.9264, MSE 1.59) and for tensile strength (R2 =0.9326, MSE=7.71) in predicting mechanical properties of GSF. Numerical optimisation revealed that optimal values of 3 h and treatment concentration of 3% gave a maximum percentage elongation at break of 10.583% and 135.045 MPa for tensile strength. This result shows that acetylation treatment enhanced the physicochemical/mechanical properties of GSF thereby making it suitable for composite processing.