Soft computing model for predicting the wear resistance of friction stir processed aluminum alloy AA5083

Abstract

In this study, AA5083 plates were friction stir processed to improve the hardness by refining the grains and dispersing the β phase in the matrix. Tool rotation speed, tool traverse speed, and tool shoulder diameter were varied as per the face-centered central composite design and FSP trials were performed. The wear resistance of the friction stir processed specimens was estimated by conducting a wear test using a pin-on-disc tribometer. It was observed that the wear resistance was higher than the base specimens and TRS, TTS and SD had a significant effect on the wear resistance. An artificial neural network model and a fuzzy logic model were developed to predict the wear resistance of the friction stir processed specimens. As the Sugeno Fuzzy model had greater prediction efficiency than the ANN model, it was used to study the effect of FSP process parameters on the wear resistance of the specimens. The results indicate that a TRS of 1000 rpm, TTS of 30 mm/min, and SD of 18 mm maximized the wear resistance of the processed plate to 8820 N-m/kg.