Statistical analysis and multiobjective optimization of process parameters in plasma spraying of partially stabilized zirconia

Abstract

Like other manufacturing processes, plasma spraying also has a non-linear behavior due to contribution of many coating parameters. This characteristic makes finding optimal factor combination difficult. Hence, the problem can be solved through effective and strategic statistical methods integrated with systematic experimental data. The purpose of this paper is to study the tribological and mechanical performance of coated surface deduced by plasma spraying process. Taguchi experimental design and analysis of variances (ANOVA) were used to investigate the influence of plasma spraying factors (spraying layers, accelerating voltage, arc current, travel speed, stand-off distance, powder feed rate, carrier gas flow rate, and primary gas flow rate) on oxidant percent and hardness. To find optimal combination of factors to reach maximum oxidation and hardness, gray relational analysis (GRA) coupled with principal component analysis (PCA) are applied on experimental data. Here, the PCA is used to find the most appropriate weighing factors necessary for construction gray relational grade. Results indicated that stand-off-distance is the most significant factor having the greatest influence on gray relational grade, followed by powder feed rate, traverse speed, spraying layers, and primary gas flow rate. Furthermore, GRA coupled with PCA can effectively acquire the optimal combination of plasma spraying parameters and the proposed approach can be a useful tool to increase the coating performance.