Artificial intelligence computation to establish relationships between APS process parameters and alumina-titania coating properties

Modeling the behavior of air plasma spray (APS) process, one of the challenges nowadays is to identify the parameter interdependencies, correlations and individual effects on coating properties, characteristics and influences on the in-service properties. APS modeling requires a global approach which considers the relationships between coating characteristics/ in-service properties and process parameters. Such an approach permits to reduce the development costs. This is why a robust methodology is needed to study these interrelated effects. Artificial intelligence based on fuzzy logic and artificial neural network concepts offers the possibility to develop a global approach to predict the coating characteristics so as to reach the required operating parameters. The model considered coating properties (porosity) and established the relationships with power process parameters (arc current intensity, total plasma gas flow rate, hydrogen content) on the basis of artificial intelligence rules. Consequently, the role and the effects of each power process parameter were discriminated. The specific case of the deposition of alumina-titania (Al2O3-TiO2, 13% by weight) by APS was considered.