A computational approach for predicting microstructure and mechanical properties of plasma sprayed ceramic coatings from powder to bulk

In this study, a multi-pronged computational approach is developed to predict the effect of spray parameters on aluminum oxide splat formation and mechanical properties of the coatings. Effect of plasma power and substrate preheat temperature on the splat formation and coating's elastic modulus are studied. The splat morphology is investigated using computational fluid dynamics approach. Simulated splat morphologies show a good agreement with the experimentally obtained splats. Three-dimensional coating structure is constructed using the stochastic approach with different simulated splat morphologies, such as the disk, fragmented and fingered shapes, and their volume fractions. Microstructure-based Finite Element Analysis is used to compute the elastic modulus of the coating, which is found to be higher (similar to 12%) than the experimentally obtained value. A correction factor (f(is)) is introduced, which takes into account inter-splat cracks, interface bonding, and other effects like curling of splats and splat sliding. After the correction factor, the computed elastic modulus for simulated coating was comparable to the experimental values. This study shows that the proposed computational approach can predict the mechanical properties of the coating and is promising for developing plasma-sprayed coatings with predictable properties. This methodology can be extended to other materials systems resulting in a desktop manufacturing approach of plasma sprayed coatings with a reduced number of experiments.