Search

Nickel-based alloys continue to play a crucial role for high temperature applications due to their excellent mechanical properties. Despite their high strength, toughness, and creep resistance, these alloys are vulnerable to aggressive oxidation and corrosion upon exposure to air at high temperature, limiting their application space. To promote oxidation resistance, nickel is typically alloyed with aluminum, which enables the formation of a stable, continuous layer of Al2O3 on the alloy surface. A protective Al2O3 layer should develop quickly and then grow slowly at temperature to effectively limit metal loss. Minor alloying elements additions are well-known to reduce Al2O3 scale growth rate. Even small reductions in the growth rate can make a substantial difference in extending the lifetime of a high temperature component. Yet many of the mechanisms by which dopants reduce oxidation rate remain unclear. To develop accurate predictive models of oxidation rate and design highly oxidation resistant alloys for a range of applications, an understanding of the fundamental effects of dopants on oxidation is critical. The current work aims to elucidate the mechanisms by which common dopants, known to improve oxidation resistance of commercial alloys, modify α-Al2O3 growth. Model NiAl and NiCrAl alloys doped with Ti, Y, and Si were oxidized at temperatures from 950-1200 °C, and high-resolution characterization techniques were used to probe the resulting alumina scale microstructure and chemistry. In particular, the effects of minor alloying additions on alumina scale development in the early stages of oxidation and Al/O transport through an established alumina scale after longer exposures were investigated. During the early stages of alumina scale development, metastable Al2O3 phases form first and grow quickly before transforming to α-Al2O3 via a nucleation and growth process. Y additions did not significantly change the transformation rate, while Si and Ti additions were found to accelerate the overall transformation to α-Al2O3 on NiAl by different mechanisms. The addition of Ti decreased the incubation time for nucleation of α-Al2O3, while Si additions accelerated the metastable to α-Al2O3 transformation rate. These findings can inform the development of alloys that quickly form protective α-Al2O3 even at moderate (~800-1000 °C) temperatures, where rapid metastable Al2O3 growth can quickly degrade thin-walled components. Once a fully α-Al2O3 scale is formed, dopants affect the transport of Al and O along the α-Al2O3 grain boundaries, thus altering the α-Al2O3 growth rate. Ti, Y, and Si had different impacts on the individual magnitudes of Al and O fluxes through the scale, resulting in different oxidation rates. Differences in Al and O fluxes can also affect the evolution of α-Al2O3 scale microstructure, that can then impact other aspects of scale growth such as stress development. Accordingly, the new knowledge gained from this work can be used to design alloys that not only have very slow oxidation rates, but develop more adherent scales and longer lifetimes. Through its systematic approach, this work examined specific aspects of the complex oxidation process and clarified the mechanisms by which different dopant elements can affect scale growth. This new information can be used for targeted alloy design and modelling of scale growth processes.

The intermetallic compound, B2 NiAl, is a promising material for high temperature structural applications such as in aviation jet engines or gas turbines, provided that its high temperature mechanical properties can be improved. Although extensive efforts over the last several decades have been devoted toward enhancing ductility through alloying design and reducing impurities, as well as improving high temperature creep strength through precipitation and dispersion strengthening, these efforts have relied on traditional approaches, a combination of large grain size to limit diffusional creep and precipitation/dispersion (50 ~ 100 nm size) strengthening to limit dislocation creep, for high temperature strengthening. While traditional approaches have shown a good improvement from a relatively high temperature strengthening point of view, the size and number density of dispersoids were not able to provide sufficient strength in the high temperature creep regime. Furthermore, details of the interaction mechanism between dislocations and dispersoids are not yet well understood.This study focuses on designing and developing advanced oxide dispersion strengthened (ODS) NiAl intermetallics with improved high temperature creep strength by incorporating a high number density (~1024 m-3) of very thermally stable Y-Ti-O nano-clusters, akin to those recently observed to improve creep strength and radiation resistance in nano-structured ferritic alloys.Advanced ODS NiAl alloys have been produced by mechanical alloying of pre-alloyed Ni-50at%Al with Y2O3 and Ti elemental powders. The milled powders were subsequently consolidated by spark plasma sintering, with the objective of producing very high number densities of nano-sized Y-Ti-O precipitates, along with fine grain size. Advanced experimental characterization techniques, combined with microhardness strength measurement, were used to investigate the material microstructure and strength following processing and to evaluate the thermal stability during an extensive matrix of long-term thermal annealing. In particular, the size, number density and composition of nano-clusters were assessed.While improvements in strength were obtained in the advanced NiAl ODS alloys, and the higher strength persisted through thermal annealing for 100 hrs at 1723K, characterization revealed the presence of Al in the oxide precipitate phases. The Al incorporation is believed detrimental to the formation of a high density of thermally stable Y-Ti-O nanoscale precipitates.

Hot section components in gas turbines can be NiCoCrAlY-coated to provide the component with an Al reservoir that maintains a protective oxide layer on its surface. Over the service life of the component, the coatings degrade by composition and phase changes due to oxidation/hot-corrosion, and multicomponent interdiffusion from and into the superalloy substrate. In this study, the rate of Al interdiffusion into selected Ni-base superalloys using various diffusion couples of two-phase NiCoCrAlY (beta + gamma) and single beta-phase NiAl with the selected alloys is measured. The diffusion couples were examined with an emphasis on the composition-dependence of Al interdiffusion. Microstructural analysis of the NiCoCrAlY vs. superalloys couples is performed to examine the dependence of coatings lifetime on the superalloy composition. The beta-NiAl diffusion couples were analyzed to determine the integrated, apparent and average effective interdiffusion coefficient as a function of superalloy's composition. Concentration profiles were obtained by EPMA of the NiAl vs. superalloy diffusion couples. Findings of this study show that the lifetimes of NiCoCrAlY are heavily dependent on superalloy compositions. The rate of interdiffusion in the diffusion couples is affected by the refractory precipitate phase microstructure structures in the interdiffusion zones as well as by component interactions. The results of the beta-NiAl diffusion couples show that increasing concentrations of Cr, Mo and Ti in the superalloy increase the Al effective interdiffusion coefficient into the superalloy, while increasing concentrations of Al, Ta and W reduce it. Thus NiCoCrAlY-superalloy systems may be designed to produce optimal microstructures in the interdiffusion zone and minimize Al interdiffusion by consideration of these diffusional interactions.

Molecular dynamics simulations of the diffusion process in ordered B2 compounds at high temperature were performed using an embedded atom interatomic potential developed to fit NiAl properties. Diffusion in the bulk occurs through a variety of cyclic mechanisms that accomplish the motion of the vacancy through nearest neighbor jumps restoring order to the alloy at the end of the cycle. The traditionally postulated six-jump cycle is only one of the various cycles observed and some of these are quite complex. Diffusion at the grain boundary mainly takes place through sequences of coordinated nearest neighbor jumps yielding to a rearrangement of the grain boundary structure. Two distinct mechanisms resulting in a structural unit migration of the vacancy are also identified. The results are analyzed in terms of the activation and configuration energies calculated using molecular statics simulations.

Ordered intermetallics based on transition metal aluminides have been proposed as structural materials for advanced aerospace applications. The development of these materials, which have the advantages of low density and high operating temperatures, have been focused on the aluminides of titanium, nickel and iron. Though these materials exhibit attractive properties at elevated temperatures, their utilization is limited due to their propensity for low temperature fracture and susceptibility to decreased ductility due to environmental effects. A major embrittlement mechanism at ambient temperatures in these aluminides has been by the loss of cohesive strength at the interfaces (intergranular failure). This study focuses on this mechanism of failure, by undertaking a systematic study of the energies and structures of specific grain boundaries in some of these compounds. The relaxed atomistic grain boundary structures in B2 aluminides, FeAl, NiAl and CoAl and L1₀ γ-TiAl were investigated using molecular statics and embedded atom potentials in order to explore general trends for a series of B2 compounds as well as TiAl. The potentials used correctly predict the proper mechanism of compositional disorder of these compounds. Using these potentials, point defects, free surface energies and various grain boundary structures of similar energies in three B2 compounds, FeAl, NiAl and CoAl were studied. These B2 alloys exhibited increasing anti-phase boundary energies respectively. The misorientations chosen for detailed study correspond to the Σ5(310) and Σ5(210) boundaries. These boundaries were investigated with consideration given to possible variations in the local chemical composition. The effects of both boundary stoichiometry and bulk stoichiometry on grain boundary energetics were also considered. Defect energies were calculated for boundaries contained in both stoichiometric and off-stoichiometric bulk. The surface energies for these aluminides were also calculated so that trends concerning the cohesive energy of the boundaries could be studied. The implications of stoichiometry, the multiplicity of the boundary structures and possible transformations between them for grain boundary brittleness are also discussed.

The microstructural stability of a directionally solidified (DS) NiAl-9 at.%Mo has was investigated. The DS alloy consisted of faceted rods of Mo in a matrix of NiAl. Two microstructures were produced by varying processing conditions: a cellular microstructure and a columnar-grain microstructure. The cellular microstructure consisted of Mo rods that were branched and bent towards the cell boundary. The columnar grain structure was fault free with the rods completely aligned to the $\langle 001\rangle$ growth direction. Both microstructures were investigated for stability at elevated temperatures during isothermal annealing at temperatures near the melting point $\rm (T\sb{M})$ and thermal fatigue at temperatures near $\rm 0.75T\sb{M}.$ Changes of fiber size and size distribution and number of fibers per unit area were measured as a function of temperature. The branch-free microstructures resisted coarsening for up to 200 h at temperatures up to $\rm 0.97T\sb{M}.$ However, the cells containing branched Mo rods did coarsen, but only at the maximum temperature of the study, $\rm 0.97T\sb{M}.$ The coarsening is explained by fault migration reducing the number of fibers per area. Thermal fatigue tests were conducted on disk-shaped specimens induction heated in an argon gas atmosphere. Heating and cooling cycles were controlled to approximate engine conditions. Although microstructure did not coarsen, the disks deformed plastically. The shape change was related to the microstructure, specifically to deformation at or near the individual cells. Material with the columnar grains deformed in the first few cycles, whereas the cellular material underwent 6000 cycles before appreciable deformation was noted. The conclusion is that the cellular microstructures are more stable than highly oriented ones.