Viscosity of CaO-MgO-Al2O3-SiO2 (CMAS) melts: Experimental measurements and comparison to model calculations

Aircraft operating temperatures are expected to increase with the implementation of ceramic matrix composites as hot-section engine materials to replace some traditional nickel-base superalloy components. With increasing temperature, the threat of component degradation from ingested siliceous debris (collectively referred to as calcium magnesium alumino-silicates - CaO-MgO-Al2O3-SiO2 or CMAS) derived from sources such as sand or volcanic ash becomes a critical challenge to engine longevity. The composition of CMAS can vary widely, leading to variable melting and viscosity behavior that can influence the interaction of these molten silicates with engine components. Several models are available to predict CMAS viscosity, yet little experimental data are available for these types of compositions to verify the accuracy of such models. In this study, experimental viscosities were obtained for CMAS-type melts with variable CaO/SiO2 ratios, ternary CaO-Al2O3-SiO2 and CaO-MgO-SiO2 molten glasses containing 10-30 mol% AlO1.5 or MgO, and CMAS molten glasses containing 5 mol% cationic additions of quinary oxides including Fe2O3, Yb2O3, and TiO2. The experimental data were compared to calculated data obtained by common viscosity models including those developed by FactSage, Giordano et al., and Fluegel. Of the models investigated, the FactSage model most accurately described the melts investigated in this study. Key relationships between CMAS composition and viscosity behavior are discussed, and the implications of viscosity behavior on the interaction of CMAS with engine constituents is considered.