Materials Science Forum Vols. 595-598

Abstract: Oxidation tests have been undertaken in air on two experimental fourth-generation Nibased superalloys containing, respectively, 3 and 5 wt.% Ru. A Ru-free, third generation alloy (CMSX-10K) was also included in the test programme which used temperatures covering the range 750-1100°C. Where possible, comparisons were also made with literature data on the secondgeneration CMSX-4 alloy and with IN738. After an initial period of transient oxidation, the subsequent parabolic rate constants for the 3%-Ru alloy were similar to those for CMSX-4 at all temperatures above 750°C. At 1000 and 1100°C these were consistent with the formation and growth of an α-alumina layer. At 950°C and lower, protective oxidation was obtained through the formation of an Al-rich spinel rather than alumina but, again, no deleterious influence of Ru additions could be found, at least in the temperature range of 800-950°C. Anomalous behaviour was found, however, at 750°C in that the Ru-bearing alloys oxidised non-protectively through the formation of sub-surface pits in which the γ΄ phase within the alloy was oxidised in situ.

Abstract: Single crystal superalloy TMS82+ and model alloys of Ni-12Al and Ni-6Cr-8Co-12Al were electroplated with 5-6μm of platinum and heat treated at 1000°C for up to 100hrs. In the model alloys the platinum concentration gradient in the interdiffusion region caused uphill diffusion of aluminium. The subsequent surface enrichment led to formation of aluminium-rich γ′ phase. In TMS82+ superalloy inward diffusion of platinum resulted in loss of the γ/γ′ microstructure and formation of new γ and γ′ grains. The initial dissolution of the γ′ cubes was due to the outward diffusion of aluminium. Again, the presence of a platinum gradient in the interdiffusion region resulted in uphill diffusion of aluminium and a net increase in aluminium content at the alloy surface, leading to a near-surface single-phase γ′ layer being formed, however topologically close-packed (TCP) phase formation was not observed.

Abstract: Ceramic Thermal Barrier Coatings (TBCs) have been developed for advanced gas turbine engine components to improve the engine efficiency and reliability. The integrity and reliability of these coatings is of paramount importance. Accurate prediction of service lifetimes for these components relies upon many factors, and is not straightforward as knowledge of the service conditions and accurate input data for modelling are required. The main cause of failure of coatings is through debonding which develops as a consequence of thermally induced strains between the metallic bondcoat and the alumina TGO layers due to the differences in the thermal expansion coefficients of the individual layers. Thermal transients due to the power cycles of turbines will then cause these fractures to grow between the TGO and the bondcoat. When these fractures reach a critical size they can grow rapidly and cause the TBC to spall off. Thermal cycling of TBCs is used therefore to evaluate and rank TBC performance. Within the laboratory this is often conducted under isothermal conditions. Whilst this test method has performed adequately in the past it does not fully simulate service conditions. Work has been underway therefore to develop a more complex test method, which better simulates the service conditions experienced by the TBC. The approach here employs a gas torch to heat the operating face of the TBC whilst cooling the rear of the substrate with compressed air, thereby imparting a heat flux on the specimen. The specimen is then cycled by removing the gas torch and cooling with compressed air on the front and rear faces. Tests have been conducted on a TBC system consisting of an IN738 substrate with a CN334 bondcoat and EBPVD TBC. Thermal cycling tests have been performed under both isothermal and heat flux conditions. During the course of the tests the samples were examined non-destructively using a thermal camera to identify early indications of spallation. This paper reports on the performance of the flame rig equipment and the results from the exposures on the TBC system.

Abstract: The isothermal and cyclic oxidation behaviors of a Ni-based superalloy with singlecrystalline (SC), polycrystalline and nanocrystalline (NC) structures were studied at 1000°C. Results indicated that a uniform oxides scale consisted of external Cr2O3 with little TiO2 and internal continuous Al2O3 formed on SC alloy. A non-uniform external oxide of which some locations were nodule-like scale was formed on surface of cast alloy. The nodule-like parts consisted of TiO2, Cr2O3 and serious internal oxidation of Al, and rest flat surface was a Cr2O3 and Al2O3 layer. A continuous Al2O3 layer formed on the sputtered NC coating. The micro-structure influenced the oxidation mechanism and resulted in different oxide scale formed on three materials, which greatly influenced materials’ oxidation and cyclic-oxidation resistance.

Abstract: Thermal barrier systems, used for turbine blades, are made of a nickel-based superalloy, a nickel aluminide bond coat layer and a ceramic thermal barrier. The aim of the present work is to study the initial stages of oxidation of the AM1/NiAl(Zr) system. It is currently of prime importance to characterize the initial thin oxide layer that covers the bond coat prior to the topcoat deposition. Indeed, the adhesion of the thermal barrier layer and the lifetime of the system are partly influenced by the substrate pre-heating oxidizing treatment. In order to determine the contribution of zirconium during this intermediate temperature range oxidation, the AM1/NiAl(Zr) system was heat treated at 950°C, in two vacuum conditions, that were close to the industrial ones. The compositions of the extreme surface of the nickel aluminide and of the thermally grown oxide were investigated by Xray photoelectron spectroscopy. In particular, these experiments allowed us to detect zirconium at the surface of the system and to determine its oxidation state.

Abstract: Thermally grown oxide (TGO) spallation increases the degradation rate of aluminide protective coatings during thermomechanical cycling. Thermal expansion misfit between TGO, bond coat and substrate, applied mechanical load in the system, and local instabilities are known triggers for spallation. Mechanical tests have been performed on coated and oxidised AM1 superalloy. In situ and post mortem study including digital image analysis and SEM were performed in order to characterise strain fields and associated damage field. Good correlation is found between oxide strain and damage extent.

Abstract: High temperature oxidation / creep deformation behavior of a diffusion barrier coated Hastelloy-X alloy, with large grain size ~500μm, was investigated at 970°C in air with external tensile stress of 22.5, 27.5, 32, and 40MPa. The diffusion barrier coating formed on Hastelloy-X consisted of a duplex structure with an inner diffusion barrier layer of Re-Cr-Ni alloy, and an outer oxidation resistant layer of β-NiAl. Un coated bare Hastelloy-X alloy with same grain size was also examined under the same conditions for comparison. The composition of the as-coated diffusion barrier coating was (15~21)Ni, (33~37)Cr, (30~33)Re, (11~15)Mo, and (9~14)Fe. This composition corresponds to σ-phase in the Ni-Cr-Re ternary system, which is known as a topologically close packed, TCP phase. The composition of this diffusion barrier layer did not change during the experiment. The oxide scales formed after creep testing on the coated and un-coated alloy surfaces were needle-like θ-Al2O3, and Cr2O3 with small amount of FeCr2O4, respectively. Grain boundary oxidation was also found in the subsurface region of the un-coated alloy. The Al2O3 scale exhibited severe spallation, and many cracks were formed perpendicular to the stress direction. However, no spallation or cracks were observed in the Cr2O3. The creep rupture times for the diffusion barrier coated alloy were about 1.5 times longer than those for bare alloy at all creep stress conditions. The fracture surface after rupture indicates that fracture occurred along alloy grain boundaries in both the coated and un-coated alloy substrate. Many cavities and cracks were observed within the diffusion barrier coated alloy substrate. These cavities and cracks tended to propagate from the substrate toward the diffusion barrier layer, and then stopped at the Re-Cr-Ni / β-NiAl interface. Cracks formed in the un-coated alloy initiated at the tip of grain boundary oxides, and propagated into alloy substrate. However no major cavities were observed inside the alloy substrate. The stress index, n, for both specimens was about 6, and this indicates that the deformation mechanism of both samples was dislocation creep. These results suggest that the Re-Cr-Ni diffusion barrier layer acts as a barrier against the movement of dislocations at the interface with the alloy surface.

Abstract: In this study, chromium based alloys containing a hardening NiAl phase are investigated from 1100°C to 1300°C in air. The influence of the NiAl content and of the Al/Ni ratio on microstructure modification and on oxidation behaviour are characterized by metallography and thermogravimetry. Increasing the Al/Ni ratio leads to a higher Al content in the chromium solid solution but does not modify the amount of primary NiAl. At high temperature, and for NiAl content exceeding 16 at%, a duplex oxide layer forms at the surface of the alloys exposed in air, alumina in the inner part, and chromia in the outer part. The oxidation behavior is discussed according to oxidation profile and to the thermodynamic predominance diagram of the involved species.

Abstract: A diffusion barrier between a 4th generation superalloy (MC-NG) and a β-(Ni, Pt)Al has been studied. The used coating process combines Re and NiW electrolytic deposits followed by thermal treatments. The diffusion barrier is composed of a continuous 3 &m thick ReWNi layer under a 10 &m thick β-(Ni, Pt)Al containing W rich precipitates. EDS analysis on as coated samples and on 50h-1100°C-Ar aged samples showed that the Re-NiW layer works as a diffusion barrier. The Al reservoir in the bond coat after aging is higher with the diffusion barrier than without. The concentrations of alloying elements are also lower in the bond coat with the diffusion barrier than without.

Abstract: A duplex layer, outer Pt-modified γ’-Ni3Al + γ-Ni and inner multi-barrier σ- Re(Cr,Ni,W), coating system was formed on a Ni-based single crystal 4th generation superalloy. Oxidation behavior of the coated alloy was investigated under thermo-cycling conditions, and analyzed by EPMA and XRD. During cyclic oxidation 1hr at 1100°C and 20 min at room temperature, a slow growing α-Al2O3 formed for up to 400 cycles and its spallation was rare. The parabolic rate constant of mass change was 6.3x10-16 kg2m-4s-1. The Pt-modified γ’-Ni3Al + γ-Ni contained 19Al, 12Pt, 4Cr, and 3Co in at%, and their concentration profiles were almost flat across the outer layer. The multi-barrier, σ-Re(Cr,Ni,W) contained 40Re, 23Cr, 17Ni, 7Al, 4W, 3.5Mo, and 3Co in at%. Furthermore, the γ’-Ni3Al containing Pt was newly formed between the multibarrier and bulk alloy substrate. It was concluded that the σ-Re(Cr,Ni,W) is compatible with the Ptmodified γ’-Ni3Al in the multi-diffusion barrier coating on Ni-based single crystal, 4th generation superalloy at high temperatures.