Comparative Analysis of Response Surface Methodology and Artificial Neural Network on the Wear Properties of Surface Composite Fabricated by Friction Stir Processing

Aluminium surface composite having ceramic reinforcement is successfully developed using friction stir processing at different tool rpm. Pin-on-disc test was performed at different sliding distances (300 m, 600 m, 900 m) and at different applied loads (20 N, 30 N, 40 N), to analyse wear behaviour of the fabricated composites. Response surface methodology (RSM) and Artificial neural network (ANN) are used to successfully develop two different models and a comparative study was done of the predictive capacity of both the developed models. The comparative study shows that the predictive capacity of the ANN model is more efficient than the RSM model. RSM is also utilized to optimize the process parameter. Optimum condition predicted by the model is for the composite developed at 1200 tool rotational speed, applied with a load of 20 N for a sliding distance of 300 m. Scanning electron microscopy (SEM) and Energy dispersive spectroscopy (EDS) analysis of wear surface were done, revealing that adhesive wear is the major wear mechanism and oxide layer formation is present on the wear surface. © 2021, The Author(s), under exclusive licence to Springer Nature Switzerland AG part of Springer Nature.