Estimation of CMAS infiltration depth in EB-PVD TBCs: A new constraint model supported with experimental approach

Two standard 7YSZ coatings were deposited by EB-PVD techniques and tested against CMAS infiltration at short time intervals (up to 8 min.) at 1250 degrees C in air. They exhibited different microstructures, i.e. porosities and microstructural features. Two species of CMAS with different compositions were used and their viscosities were determined using the concentric cylinder method and their contact angles were measured using high temperature heating microscopy. The theoretical viscosities, which were calculated using a statistical model based on the chemical composition of the melts, differed from the measured values of the viscosities by one order of magnitude. A large variation in the contact angles within a very short range of temperature (1243-1266 degrees C) was observed as well. The porosity and surface area measurements were performed on both EB-PVD microstructures using the nitrogen physisorption method. Additionally, the produced coatings exhibited porosities of 14.5 and 29.5 percent and the infiltration experiments have shown that the more porous coating provides higher infiltration resistance. The effect of porosity on CMAS infiltration kinetics was investigated and the results elucidate that the porosity network plays a more preeminent role than the amount of porosity. The experimental infiltration results have been compared with calculated infiltration data using a novel mathematical approach proposed in previous studies in which the permeability of the coatings is assessed with two contrasting methods termed concentric pipe and open pipe models. The infiltration was calculated by incorporating the experimentally determined properties such as contact angle, viscosity and porosity. A fitting parameter has been derived from the equations for the geometry factor for both microstructures. The calculated and experimental results are in good agreement with the concentric pipe model supporting the validity of this CMAS infiltration model.