Materials based on the non-transformable tetragonal phase in the ZrO2-(Y/Yb)O1.5-TaO2.5 systems are of significant interest for next-generation thermal barrier coatings (TBCs) owing to their thermodynamic stability at high temperature, resistance to deleterious phase transformations on thermal cycling, low thermal conductivity, and potentially adequate toughness. The current study investigates fundamental issues pertaining to the implementation of these compositions for TBC applications and associated concerns potentially impacting coating durability. Topics explored include: (i) material processability by electron beam physical vapor deposition (EB-PVD), (ii) TBC microstructure evolution, (iii) relationships between toughness and composition, and (iv) toughening mechanisms, including ferroelastic domain switching.

Key challenges identified in the fabrication of ZrO2-(Y/Yb)O1.5-TaO2.5 TBCs by EB-PVD are compositional variability throughout the coating thickness and the development of suboptimal columnar microstructures. The former prevents deposition of single-phase materials and potentially compromises the phase stability of the structure and the latter degrades the strain tolerance of the coating. Despite these complications, compositions in the two-phase t-ZrO2 + c-ZrO2 and three-phase t-ZrO2 + c-ZrO2 + M’-YTaO4 phase fields can be deposited by EB-PVD such that the columnar microstructure is adequate and the stability of the tetragonal phase is retained for at least 400 hours at 1250°C and 1500°C.

The toughness of the investigated multi-phase compositions, which is expected to influence the durability of thermal barrier coatings based on these materials, is demonstrated to be significantly dependent on composition and microstructure. The influence of composition arises from the relative fractions of the phases present at equilibrium and the differences between their intrinsic toughness. Minimizing the amount of secondary phases (c-ZrO2 and M’-YTaO4), particularly c-ZrO2, is a path to optimize the toughness of the tetragonal-phase material. The role of microstructure is related to the fracture pathway and prevalence of toughening mechanisms such as crack deflection, grain bridging, grain pullout, and microcracking, which had not been previously reported. It is shown for the first time that domains suggestive of ferroelastic switching are contained within non-transformable tetragonal grains. The domains form on growth and/or densification of precursor-derived powder compacts as a result of short-range ordering between Y3+ and Ta5+ cations, as predicted by first-principle calculations. The ordering patterns in the zirconia solid solution are reminiscent of those predicted and observed for the various forms of the yttrium tantalate. However, direct evidence of ferroelastic switching within the process zone of a crack remains elusive. The results suggest that additional understanding of the relationship between toughness and microstructure is required for the implementation of materials based on the non-transformable tetragonal phase in thermal barrier coatings.

Thermal barrier coating (TBC) systems are critical to the performance of gas turbine engines in the aviation and power generation industries. As engine operating temperatures are increased to improve efficiency, ceramic topcoats that depend on a metastable tetragonal phase, namely t’-8YSZ (ZrO2 + 8 ± 0.5 mol% YO1.5), rapidly degrade. Alternative TBC compositions that exhibit combinations of properties superior to 8YSZ are needed to meet the efficiency targets of next generation gas turbines. In this work, two approaches are pursued based on either a single-phase, non-transformable tetragonal structure or multiphase compositions that offer the possibility of adequate toughness and resistance to molten silicates.

Previous laboratory investigations of the CeO2-TiO2-ZrO2 (CeTiSZ) system revealed a relatively large, non-transformable tetragonal field with exceptional tetragonality is stable at 1350°C. In the present work, challenges with transitioning the earlier laboratory results to industrial scale coatings are elucidated. Two important effects not previously encountered in the deposition of current TBCs are identified. Understanding the driving force for segregation during deposition as well as the potential reducibility of the cations upon heating is critical to maintaining stable phase equilibria in thermally sprayed coatings. The reduction of cerium (IV) upon heating places an inherent limit on the temperature capabilities of CeTiSZ coatings.

Contrary to the CeTiSZ system, the ZrO2-YO1.5-TaO2.5 (ZYTO) system does not experience reduction of the cations at high temperatures. A stable, non-transformable tetragonal phase exists at temperatures up to 1500°C, albeit in a much narrower composition range. Although the ZrO2-rich region of the ternary had been studied previously, there was a paucity of information on the phase equilibria in the rest of the diagram. Systematic investigation of the entire YO1.5-TaO2.5 binary phase diagram elucidates the equilibrium phases and serves as a solid foundation for the investigation of the ZYTO ternary. For the first time, isothermal sections of the complete ZYTO ternary are experimentally determined at 1250°C and 1500°C. Two-phase fields in the ZYTO ternary suggest that multiphase compositions with improved combinations of properties may be possible. Specifically, a large two-phase field in which fluorite and YTaO4 (YT) are stable offers the potential to increase the toughness of fluorite, which, with sufficient rare earth content, can mitigate damage caused by molten silicates. Additionally incorporating YT as a second phase has the potential to improve the CMAS resistance of phase stable tetragonal compositions.

Similar to the phase equilibria, there was a dearth of information on the toughness values and associated toughening mechanisms of compositions in the ZrO2-lean portion of the ZYTO phase diagram. Micro-indentation and micro-3-point bend experiments indicate that YT contributes to the toughness of multiphase samples. It is shown for the first time that samples containing YT exhibit microstructural features clearly consistent with domain wall motion, the fundamental mechanism underpinning ferroelastic toughening.

Enhancing the efficiency of gas turbine engines requires higher operation temperatures and the materials capable of surviving the increasingly challenging environment. Ceramic barrier coatings, with carefully engineered microstructures, protect the structural components within the hottest sections of the engine. However, these coatings are susceptible to damage mechanisms arises from the ingestion of siliceous debris, which can melt and deposit on the coating’s surface. Thermomechanical strains develop that are either mitigated or exacerbated by the thermochemical interactions between the coating and the melt. This work investigates these thermochemical interactions and their pertinent kinetics and thermodynamics.

Dense compacts or single crystalline pieces of barrier coating oxides were placed into a semi-infinite 1D diffusion couple geometry with one of two synthetic silicate melts at 1200–1400 °C. Concentration profiles within the melt were obtained and fit to partial differential equations quantitatively describing the coating dissolution rate into the melt and diffusivities therein. Cation diffusivities were most affected by the melt composition, whereas the ratio of rare-earth (RE3+) oxides to ZrO2 or HfO2 most strongly affected the initial detachment rate of barrier oxides into the melt. Ultimately, the dissolution kinetics were sufficiently slow to delay melt saturation and the nucleation of reprecipitated or reaction phases that limit coating degradation. This delay was worse for barrier oxides with low concentration of RE3+ elements. Finite element models—using the gathered kinetic data but applied to small length scales relevant for real coatings—suggest this delay will be controlled primarily by the initial interface detachment rate in practice.

After the initial dissolution transient period, the crystallization of reprecipitated and reaction phases was investigated qualitatively using electron microscopy and chemical analysis techniques. The presence of only a small amount of RE3+ oxide in the dissolving material (e.g., 7%) kinetically hindered the crystallization of reaction products—even those based on Zr4+ or Hf 4+—favoring instead reprecipitated phases, deviating from the expected thermodynamic response predicted by CALPHAD databases. Conversely, those barrier oxides free of RE3+ (e.g., HfO2) more readily crystallized reaction products such as (Zr,Hf)SiO4, or Ca2HfSi4O12; those containing a substantial amount of RE3+ (e.g., Gd2Zr2O7) rapidly crystallized a RE-apatite, nominally Ca2RE8(SiO4)6O2.

Finally, the thermodynamics of Y-Al-Fe-garnet formation, i.e., the solid-solubility limits of substitutional cations Ca2+, Mg2+, Fe2+, and Si4+, their crystallographic site preference, and the competition between garnet and other phases was investigated. Long duration heat treatments afforded equilibrated samples, for which the phase assemblage was analyzed using X-ray diffraction, electron microscopy, and standardized chemical analysis techniques. A key factor in the stability of garnet was the Fe:Al ratio of the system. Indeed, increasing the Fe:Al ratio of the as-synthesized powder significantly increased the Ca2+ and Si4+ solubility and the quantity of garnet present, with a concomitant decrease to the quantity of other important reaction phases such as apatite.

This dissertation advances the understanding of thermochemical interactions between protective barrier coatings and molten silicates, which is critical to design robust coatings. The quantitative kinetic data and thermodynamic information enables computational approaches to coating design.

Increasing the efficiency of gas turbine engines, both for aviation and commercial power generation, means higher operation temperatures and new materials capable of withstanding the increasingly extreme environments in the hot sections of these engines. Ceramic matrix composites have increased temperature capabilities, and particularly porous matrix oxide fiber ceramic composites (OFCCs) are promising materials due to their innate oxidative stability, resistance to thermal shock, cost, and ease of manufacture. However, OFCCs are not without limitations. The fibers and porous matrices have temperature limitations for microstructural stability, and the primary oxide constituents, Al2O3 and especially SiO2, have issues with corrosion and volatilization in high temperature water vapor, a major constituent of combustion atmospheres. Components require thermal and environmental barrier coatings (T/EBCs) with engineered structures and properties to maximize protection. This work studied techniques for building multilayered T/EBC systems for OFCCs and tests their efficacy. Two generations of air plasma sprayed (APS) coated OFCC coupons were supplied by Siemens Corporate Technology for durability assessment. These were compared against in-house coated OFCC coupons using electron beam-physical vapor deposition (EB-PVD) in thermocyclic testing from room temperature to 1200°C. EB-PVD coated samples performed better, with significant delamination seen in the APS coated specimens. The porous matrices of the OFCCs, while necessary for bulk damage tolerance and fiber pullout, complicates adhesion of these barrier coatings. Precursor impregnation and pyrolysis was used to infiltrate uncoated OFCC coupons with both alumina and yttria stabilized zirconia precursor solutions to create a gradient in density to selectively strengthen the matrix region near the surface to be coated, without sacrificing bulk toughness. The gradient was characterized using Vickers microhardness indentation tests of matrix pockets in cross-section. Further EB-PVD coating was performed on hardened, uncoated OFCCs, with additional compositions from 7wt% yttria stabilized zirconia (7YSZ) to more yttria rich compositions of Y4Zr3O12 (YZO) and Y2O3 with better matched thermal expansion to the OFCC. The latter compositions were deposited in bilayer configurations with a thin 7YSZ interlayer to prevent diffusional interaction with the OFCC. Despite morphological irregularities observed in 7YSZ and Y2O3 depositions, especially at deposition temperatures above 1080°C, adherent coatings with desirable columnar microstructures were grown in all three compositions, with deposition temperatures below 1035°C. Finally, select EB-PVD coated OFCCs were thermally cycled in flowing water vapor at 1200°C to assess durability of the coatings and amount of damage to the SiO2 – containing fibers from water vapor ingress through the porous coating. All coatings remained adherent, fiber damage from SiO2 volatilization was observed, especially around coating defects. Specimens with denser interlayers did mitigate some ingress, coating and OFCC surface defects still provided pathways for water vapor. The multilayer thermal and environmental barrier coating systems for OFCCs assembled and tested in this work provides insight into creating needed effective barrier coatings for the implementation of OFCCs into hot gas components of gas turbine engines.

Abstract

Accelerated Design and Characterization of

Oxidation Resistant Co-Based Gamma / Gamma-Prime Superalloys

by

Colin Alistair Stewart

Superalloys are metallic materials comprising predominantly an FCC matrix (γ) with an L12 intermetallic reinforcing precipitate (γ’), capable of load-bearing applications at high temperatures such as the hot section components of gas turbine engines. Recently discovered Co-based γ/γ’ alloys offer a potential avenue for improving engine performance beyond the current limit of commercial Ni-base alloys. To exploit their potential, Co-base alloys must be able to grow a protective, continuous α-Al2O3 scale after brief (minutes) exposure to an oxidizing environment at temperatures on the order of 1100°C. Simultaneously, an alloy must also maintain high thermal stability of the strengthening γ’ phase. The composition space available to achieve such behavior is vast, with commercial Ni-base superalloys typically containing 11+ elements. Instead of a more traditional alloy design process by incremental trial and error, this dissertation employs modern computational modeling in combination with high-throughput experimental techniques, pursuant to the mission statement of the NSF DMREF program.

Combinatorial ion plasma deposition was used to systematically synthesize 234 distinct alloy samples in Co-Ni-W-Al-Cr-Ta space. The alloy phase constitution and oxidation behavior after 1 h in air at 1100°C were investigated. Photostimulated luminescence spectroscopy (PSLS) was employed to rapidly screen for the presence of α-Al2O3 in the oxide scales. While PSLS can adequately detect the formation of α-Al2O3, its ability to do so is moderated by the presence of extraneous, non-desirable oxides with varying thickness and composition. Notably, it was concluded that overlaying CoO was more opaque to the PSLS signal than NiO. Nevertheless, it was shown that a large number of positive PSLS spectra was often related to the overall thickness of the scale, with thinner scales usually characterized by an underlying continuous α-Al2O3 scale. The merging of these synthesis and characterization techniques identified the compositional space of promising oxidation behavior, and the effects of Al, Cr, and to a lesser extent W and Ni, on this behavior.

The experimental results were combined with calculations of the senary phase equilibria using available thermodynamic databases. In this manner, the design space for Co-base γ/γ’ alloys with suitable stability of the reinforcing phase and desirable oxidation resistant were identified. New compositions designated as “DMREF” alloys were synthesized, and were found to achieve a superior combination of properties relative to other Co-base alloys reported to date. Thus, this dissertation has established the oxidation behavior and phase-constitution in composition space relevant to Co γ/γ’ superalloys, designed and generated a novel alloy with markedly improved properties, and demonstrated proof of concept for the strategy of accelerated exploration, design and characterization of new alloy systems.