Integrated optimization and prediction of mechanical properties in FDM printed polyamide/carbon fiber composites using PSI–VIKOR and ANN

The present study intends to find the best possible combination of 3D printing control factors to achieve better impact and tensile strength for the carbon fiber reinforced nylon (ePA-CF) composites. The effect of Fused Deposition Modelling (FDM) parameters namely nozzle temperature, infill percentage and printing speed on the impact and tensile strength of samples are investigated by following the Taguchi method. Tensile and impact assessments were carried out for the FDM printed samples according to the designed L9 orthogonal array in line with ASTM standards. The experimental results of tensile and impact testing were analyzed by converting it into Signal-to-Noise (SN) ratios. Results indicated that infill percentage has the maximum effect on the assessed properties and the same is confirmed by analysis of variance (ANOVA). Linear regression and Artificial Neural Network (ANN) predictive models were developed among which the ANN showed higher prediction accuracy. Multi-objective optimization performed using the PSI-VIKOR techniques suggested the optimal parameter configurations for improved mechanical performance. The study demonstrates the optimal combination of 280 °C nozzle temperature, 60 mm/s printing speed and 70 % infill percentage which yields better tensile and impact properties for ePA-CF composites.