Microstructure, mechanical properties and ELM based wear loss prediction of plasma sprayed ZrO2-MgO coatings on a magnesium alloy

Abstract

In this study, the surface of AZ91D magnesium alloy was coated with ZrO2-wt.-% 22 MgO by the plasma spray method. The coatings were made at two different current levels (600 and 500 A) and three different spraying distances (120, 130 and 140 mm). The surface roughness was measured by a profilometer and hardness was measured via a microhardness test. Coated cross-sections were examined under an optical microscope (OM) and scanning electron microscope (SEM). The phases formed on the coating surfaces were detected by x-ray diffractometer (XRD). A dry sliding wear test was performed at 5, 7.5 and 10 N normal loads. Mg2Zr5O12, ZrO2, MgO, and Zr formed on the coating layers. Surface roughness and porosity percentages were enhanced by increasing the spray distance and decreasing current. The maximum microhardness value was reached at 1152 (HV0.1), and significant improvements were observed in the wear resistance of the coatings compared with that of the AZ91D. An extreme learning machine (ELM) algorithm, which is one of the machine learning algorithms, was applied to the wear loss data obtained. The success rate for the model designed using the ELM algorithm, was calculated as 0.9287 (R-squared).