A Complete Predictor Methodology for the Plasma Spray Process

Abstract

Abstract To enhance usage of the plasma spray process, a better physical understanding of the process is required, which entails a synergistic mix of analytical and empirical studies. Better understanding can lead to development of optimal thermal spray coatings for future applications. This study presents an analytical method that can be used for these purposes. Experimental and analytical studies were conducted to investigate gas, particle, and coating dynamics and the resulting coating properties in the plasma spray process for the Tribaloy 800 powder system. Historical full-factorial statistically designed experiments were the basis for the analytical-experimental comparisons. The thermal plasma produced by a commercial plasma spray torch was then numerically modeled from the electrodes to the standoff distance in the free plume for sixteen experiments. This information was then used as boundary conditions to solve the plasma/particle interaction problem for the experiments. The predicted temperature and velocity of the droplets at the spray distance were then used as initial conditions to a coating dynamics code. Multiple polynomial regression analysis was then used to establish the sequential relationship between the process parameters (i.e., power, total flow, hydrogen flow), the coating properties (porosity, oxides), and the coating mechanical performance properties (tensile strength, microhardness, superficial hardness). The equations derived from the regression analysis were used to construct a predictor code for the process. The code predicts the process and coating attributes reasonably well. The predicted coating properties exhibit excellent correlation with the actual properties obtained from the experimental studies in the range of the parameter settings.