Enhancing wear resistance of AZ61 alloy through friction stir processing: experimental study and prediction model

Abstract

In this study, friction stir processing (FSP) is proposed for the treatment of AZ61 alloy, and an artificial neural network is built to predict and compare the experimental wear results. The effects of different processing parameters, including spindle speed (800–1200 rpm), traveling speed (5–15 mm min ^−1 ), and depth of press (0.8–1.2 mm) on the microstructural evolution, mechanical properties, and wear behavior are investigated. Microstructural analysis reveals a grain size of 14 ± 2 μ m for the FSP1 sample, with observed shifting of x-ray diffraction (XRD) peaks, indicative of texture development. Increasing spindle and traveling speeds increase the surface roughness, as observed by average roughness (Ra) values of 68.4 nm for a rotational speed of 800 rpm, traveling speed of 5 mm min ^−1 , and shoulder depth of 0.8 mm (FSP1) and 116.3 nm for rotational speed of 1200 rpm, traveling speed of 15 mm min ^−1 , and shoulder depth of 1 mm (FSP9). Microhardness values increase to 113.36 Hv for FSP1 and 79. 51 Hv for FSP9 compared to 65.92 Hv for the base material (BM) sample. The decrement in hardness from FSP1 to FSP9 can be attributed to increased heat input, resulting in coarse microstructure. Wear results show that FSP1 exhibits the lowest weight loss (0.003 g) and coefficient of friction (COF) (0.28) compared to other FSP conditions and BM samples (weight loss of 0.022 g and COF of 0.68). This work demonstrates the efficacy of friction stir processing in enhancing the wear resistance of magnesium alloys.

Keywords

• AZ61
• friction stir process
• artificial neural network
• microstructural analysis
• wear