Abstract
SiC/SiC ceramic matrix composites (CMCs) entered commercial jet turbine service in 2016. SiC forms a protective SiO<sub>2</sub> layer at operating temperatures in oxidizing environments, however, H<sub>2</sub>O(g) produced in the combustion environment reacts with the SiO<sub>2</sub> layer to form Si(OH)<sub>4</sub>(g). Environmental barrier coatings (EBCs) are required to protect the underlying CMC from H<sub>2</sub>O(g) in the combustion stream. Rare earth (RE) disilicate (RE<sub>2</sub>Si<sub>2</sub>O<sub>7</sub>) coatings, where RE is either Y or Yb, are state-of-the-art EBCs. These coatings are typically deposited using an air plasma spray (APS) process resulting in a heterogeneous microstructure with multiple phases present such as RE<sub>2</sub>O<sub>3</sub>, RE<sub>2</sub>SiO<sub>5</sub>, and RE<sub>2</sub>Si<sub>2</sub>O<sub>7</sub>. Cracks and pores are also likely to result from this process. This complex microstructure results in thermal expansion and thermochemical response to H<sub>2</sub>O(g) that differ from homogeneous RE<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> materials processed in conventional ways. The aim of this research is to quantify property differences that arise due to the APS process and incorporate them into a lifing model to predict EBC behavior at temperatures, pressures, gas velocities, and times relevant for turbine engine applications.
Thermal expansion of Y<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> material deposited using APS was measured by dilatometry at temperatures between 25 and 1400°C. APS Y<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> had a coefficient of thermal expansion (CTE) higher than phase pure Y<sub>2</sub>Si</sub>2</sub>O<sub>7</sub> tested in the same manner, likely due to the presence of constituent phases such as Y</sub>2</sub>O<sub>3</sub> and Y<sub>2</sub>SiO<sub>5</sub> that have CTE values higher than Y<sub>2</sub>Si<sub>2</sub>O<sub>7</sub>. In addition, anisotropy in CTE was determined for δ-Y<sub>2</sub>Si<sub>2</sub>O<sub>7</sub>, X2-Y<sub>2</sub>SiO<sub>5</sub>, β-Yb<sub>2</sub>Si<sub>2</sub>O<sub>7</sub>, and X2-Yb<sub>2</sub>SiO<sub>5</sub> using
high-temperature XRD performed at the Advanced Photon Source at Argonne National Laboratory. Low expansion planes were found along with large differences in expansion along different axes, especially in the monosilicates.
The thermochemical stability of RE<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> is an important factor in the lifetime prediction of EBCs for SiC/SiC CMCs. Samples of Y<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> processed with SPS have been exposed to H<sub>2</sub>O(g) with a gas velocity of 125-250 m/s at 1200 and 1300°C for 60h and 1400°C for times of 60 – 250 h in a steam-jet furnace. Samples were then observed in plan view and cross section using scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS). Phase pure SPS Y<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> reacted with H<sub>2</sub>O(g) forming Y<sub>2</sub>SiO<sub>5</sub> and porosity, releasing Si(OH)<sub>4</sub>(g). Analysis of the change in Y<sub>2</sub>SiO<sub>5</sub> layer thickness with time has shown parabolic reaction kinetics, suggesting a diffusion limited mechanism. For the first time the reaction of Y<sub>2</sub>SiO<sub>5</sub> with H<sub>2</sub>O(g) to form Y<sub>2</sub>O<sub>3</sub> was also observed.
APS Yb<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> provided by Rolls-Royce was tested in the steam-jet furnace as well, showing reaction between H<sub>2</sub>O(g) and Yb<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> similar to that seen in the Y<sub>2</sub>O<sub>3</sub>-SiO<sub>2</sub> system. Local microstructural features in the APS coatings, such as splats, cracks, and pores, were found to heavily influence SiO<sub>2</sub> depletion.
A hybrid Potts/diffusion model of the phase and microstructural evolution was developed. Currently the model simulates the formation of Y<sub>2</sub>SiO<sub>5</sub> and porosity from phase pure Y<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> with parabolic kinetics. Once calibrated, the model will be used to predict Y<sub>2</sub>Si<sub>2</sub>O<sub>7</sub> reactions with H<sub>2</sub>O(g) and the corresponding microstructural evolution as a function of time, temperature, P<sub>H2O</sub> and gas velocity. The model also allows for the introduction of new phases like the formation of Y<sub>2</sub>O<sub>3</sub> from Y<sub>2</sub>SiO<sub>5</sub> reacting with H<sub>2</sub>O(g) and pores and cracks in the microstructure.
This work together will give a better understanding of how the APS process used in EBC deposition alters material properties compared to phase pure materials and how these alterations must be considered for lifetime prediction of the coatings.