| Advanced Statistical Optimization of Parameters at Plasma Spraying of Hydroxyapatite Coatings | ||
| Prof. Lech Pawlowski, S. Dyshlovenko, C. Pierlot, R. Tomaszek and P. Roussel, Thermal Spraying, ENSCL, 8, av. Mendeleiev, Villeneuve d'Ascq, 59650, France Spray process of hydroxyapatite was optimized in two ways: statistical planning of experiments and numerical modelling. Factorial design of 25 experiments was used to find effects of four principal plasma spray parameters, i.e. electric power, plasma forming gas composition, carrier gas flow rate and distance of spraying and laser treatment using two different radiation densities onto microstructure of hydroxyapatite (HA) coatings and powders. The granulometry of powders was tested after spraying into water and compared to the initial one. The crystal phase content of powders and coatings was determined using X–ray diffraction (XRD) quantitative analysis. Numerical simulation of interaction between hydroxyapatite (HA) particles and Ar-H2 plasma was also carried out. A ballistic model was applied to describe the phenomena of exchange of momentum and of heat transfer, including heating, melting and evaporation of particle material. Numerical simulations of different experimental conditions including variations of carrier gas flow rate and distance of spraying were carried out. The data obtained in the simulations at short spraying distance were used consequently to model HA coating growth. The validation of numerical simulation was carried out in two ways. Firstly, a fraction of amorphous phase in sprayed material was predicted and compared with experimental data of semi-quantitative X-ray analysis. The data could be found thanks to a hypothesis that the crystal phase composition in a volume of particle in flight is frozen on impact with substrate or on contact with water and that the liquid material transforms into CaO-P2O3 glass. Secondly, the porosity of coatings generated by the numerical simulations was compared to that obtained for the real deposits. Finally, the size experimental size distribution of powder is compared to the calculated one. | ||