Image processing and statistical analysis of microstructures of as plasma sprayed Alumina–13 wt.% Titania coatings

Abstract

Plasma spraying of Alumina-13 wt.% Titania coatings in the as sprayed conditions show preferentially stabilized α phase. The morphology of this residual α phase are of two types a partially melted particle type prevailing as rounded or nearly spherical shaped form and the other splats type, a disc shaped caused by the high impact and rapid solidification of the fully molten plume. While the former is totally deleterious to the mechanical properties the latter type of morphology is found to strengthen and toughen the matrix due to the inherent larger stiffness of the residual α phase and also as a result of effective operation of toughening mechanisms such as crack bridging and crack deflection. In this paper an attempt has been made with image analysis to establish the individual constituent features of the complicated as sprayed microstructure. Statistical techniques such as correlation and discriminant analysis were used to group the micro structural features based on the process parameters. Our analysis indicates that for maximum contrast situations in process variables such as extreme values of primary current and widely different substrate material the micro structural features can be easily discriminated and separated into groups. The absence of significant correlation between the micro structural features within a group could also be verified. Our analysis indicates a large negative correlation value for partially melted particles and residual α occurring as splats with the process parameter the primary current of the thermal plasma.