High-temperature tribological performance of stir-cast and heat-treated EV31A magnesium alloy: Experiments and predictions

M Somasundaram; U NarendraKumar; A Raja Annamalai; A Muthuchamy

doi:10.1016/j.heliyon.2023.e19055

High-temperature tribological performance of stir-cast and heat-treated EV31A magnesium alloy: Experiments and predictions

Heliyon. 2023 Aug 12;9(8):e19055. doi: 10.1016/j.heliyon.2023.e19055. eCollection 2023 Aug.

Authors

M Somasundaram¹, U NarendraKumar¹, A Raja Annamalai², A Muthuchamy³

Affiliations

¹ Department of Manufacturing Engineering, School of Mechanical Engineering, Vellore Institute of Technology, Vellore, 632 014, Tamilnadu, India.
² Centre for Innovative Manufacturing Research, Vellore Institute of Technology, Vellore, 632 014, Tamil Nadu, India.
³ Department of Metallurgical and Materials Engineering, National Institute of Technology, Tiruchirappalli, Tamil Nadu, 620 015, India.

Abstract

The temperature effect on the wear behaviour of EV31A Mg alloy during dry sliding wear was investigated. Wear tests were carried out at 50, 100, 150, 200, and 250 °C using a standard load of 10 N and a sliding distance of 1000 m. Weight loss method was used to calculate the wear rate. Optical microscopy was used to examine the microstructure of the EV31A alloy. FE-SEM with EDS analysis was used to investigate the wear morphology, and XRD analysis was performed both before and after the wear test. A high wear coefficient (K) value (more than 10^-4) indicates extreme wear for EV31A in all the scenarios. T4 EV31A had a maximum wear rate of 20.2 mg at 150 °C. The as-cast EV31A alloy exhibits an excellent wear rate at the price of mechanical properties under all test scenarios. Wear resistance is improved by Nd and Zr oxides, although Mg and Gd oxides have little effect. Zn has no effect on the wear behaviour of the EV31A. In as-cast, T4, and T6 heat-treated conditions, the EV31A alloy exhibits delamination (abrasive wear), oxide development (corrosive wear), and delamination mixed with plastic deformation (adhesive wear). A Three-layered ANN and adapted Fine Gaussian SVM predicted tribological characteristics. In ANN prediction, the maximum R² was 0.99 for CoF and 0.89 for wear rate, respectively. Despite the fact that the study's normal load is constant, machine learning models allow to deduce that temperature and normal load are the main influential parameters in CoF and wear rate, respectively.

Keywords: Dry sliding wear; High-temperature wear; Machine learning; Magnesium alloy; Rare earth elements; Tribological properties.