Like the reaction with 7YSZ, the reaction between Na-rich CMAS and 2ZrO2⋅Y2O3(ss) produces reprecipitated ZrO2 grains with greater Y-content than those produced from sand-CMAS attack. Tables 1–3 show the difference between the compositions of the reaction products after attack by the sand-CMAS and Na-rich CMASs. The reaction product’s stabilizer content is considered a significant factor in the CMAS penetration rates because it changes the concentration of Y and Ca available to make apatite that can mitigate CMAS infiltration. Considering this hypothesis, the enriched ZrO2 grains that are stable in the Na-rich CMAS would make TBCs less resistant to penetration from basic CMAS. However, our results show that the Na-rich CMAS (OB =0.72) has larger penetration depths after 1 h and 24 h than the sand-CMAS (OB =0.64). This penetration depth difference is affected by the nature of the crystalline reaction product (Na-rich crystalline phase rather than apatite) and its instability as the reaction proceeds. After 1-min attack by the Na-rich CMAS, a thick layer of a Na-rich crystalline phase is found at the interface between the CMAS and 2ZrO2⋅Y2O3(ss) coating. This phase effectively blocks molten CMAS from infiltrating any grain boundaries and shows negligible coating dissolution. (It should be noted that the Na-rich CMAS is expected to melt below 1200 ◦C and it would penetrate prior to the 1 min dwell at 1340 ◦C is complete.)
However, the Na2O volatility changes how the reaction proceeds (See Table 4 to see how the Na volatility changes the CMAS composition with time during the 7YSZ/CMAS reactions.). After 1 h, the CMAS has circumvented the Na-rich layer and infiltrated the coating. Apatite formation is observed but the shape of the crystals indicates different growth kinetics than those occurring during sand-CMAS attack. The apatite likely forms from a second reaction between the Na-crystals and the coating as the latter phase becomes unstable from the volatility of Na2O. After 24 h, no Na-rich crystals remain and the interaction region looks similar to that caused by the sand-CMAS attack, albeit it has consumed more of the 2ZrO2⋅Y2O3(ss) coating. This reaction evolution caused by the Na-volatility and the resultant phases limits our ability to assess the relationship between the reprecipitated ZrO2 stabilizer content and the infiltration rate.
Table 1
Compositions (cation basis, at%) of reprecipitated fluorite grains found in the reaction regions of the 2ZrO2⋅Y2O3(ss)/TBC cross sections after reactions with sand-CMAS and Na-rich CMAS (1340 ◦C).
Table 2
Compositions (cation basis, at%) of apatite grains found in the reaction regions of the 2ZrO2⋅Y2O3(ss)/TBC cross sections after reactions with sand-CMAS and Na-rich CMAS (1340 ◦C). The theoretical composition of Ca4Y6(SiO4)6O and Ca2Y8(SiO4)6O2 are listed for comparison.
Table 3
EDS compositions (cation basis, at%) of crystalline phases in the 2ZrO2-⋅Y2O3(ss)/Na-rich CMAS interaction region at various exposure times (1340 ◦C).
Table 4
The CMAS compositions (cation basis, at%) after interaction with 7YSZ TBCs at 1340 ◦C for 1 min, 1 h and 24 h measured using EDS.
2ZrO2⋅Y2O3(ss) coatings dramatically arrest both CMAS compositions tested, regardless of the melt’s basicity. This supports previous results that Y3+in sufficient volumes can induce crystallization that successfully mitigates CMAS infiltration. However, the mitigation capability of Y relative to other proposed rare earth cations (e.g. Gd3+and Yb3+) is still uncertain. The criteria for the best cation to mitigate CMAS infiltration are dependent on its ability to make apatite, which is the reaction product known to consume the main CMAS constituents and shown to have favorable growth kinetics. Here, we show that the formation of the Na-rich crystalline phase provides temporary protection from the CMAS attack but relies on subsequent apatite formation to prevent further CMAS penetration. The composition of the formed apatite and the cation concentration in the reprecipitated ZrO2 have both been used to evaluate the cation that is most effective at mitigating CMAS attack. The apatite and reprecipitated ZrO2 compositions have been considered from a thermodynamic perspective but our results suggest that kinetics influence the reaction products. With regards to the apatite phase, the reactions with the sand-CMAS show a shift from Ca4Y6(SiO4)O (Ca/Y =0.67) to Ca2Y8(SiO4)O2 (Ca/Y =0.25) with increased exposure time. The Ca4Y6(SiO4)O phase, which would require more Ca from the melt, is considered unstable because of the high oxygen vacancies required to maintain the structure. The Ca-rich apatite is observed only at short exposure durations with the sand-CMAS, suggesting it is a metastable phase formed due to the kinetics of the reaction. Its metastability is supported by the lack of the Ca-rich apatite phase in the reactions with Na-rich CMAS, where apatite formation is delayed and caused by secondary reactions with the Na-rich crystalline phase. The Y-content in the reprecipitated ZrO2 grains also changes with time in the reactions with sand-CMAS. In the reaction wake, the Y-content increases in the ZrO2 grains at longer exposure times. Given these observations and their assumed effect on the CMAS’s penetration kinetics, the mitigation capability of 2ZrO2⋅Y2O3(ss) appears to degrade as the reaction proceeds. 2ZrO2∙Y2O3(ss) is, thus, more favorable if thermodynamic equilibrium is not achieved as is seen in shorter exposure durations.
The time to reach equilibrium in the reaction region is assumed to be within 24 h at 1340 ◦C, which is a small fraction of an engine’s lifetime. However, the cyclic loading of an operational gas-turbine engine will create thermal stresses that can cause the CMAS and reaction region to break off. Therefore, the exposure time for each CMAS deposit may be short and metastable products may play a larger role when the engine’s heating cycle consists of rapid heating and cooling and short dwell times. Both the kinetics and thermodynamics of the reaction should be assessed to predict the coatings mitigation capability during engine operation.
本文由桑尧热喷涂网收集整理。本站文章未经允许不得转载；如欲转载请注明出处，北京桑尧科技开发有限公司网址：http://www.sunspraying.com/
|
