Microstructural properties of superalloys investigated by nanoindentations in an atomic force microscope

The microstructure of nickel-base superalloys with differently shaped γ′ precipitates determines their excellent high-temperature behaviour. A reliable quantitative determination of volume fraction and particle size distribution (PSD) of these precipitates is difficult, since the size of the precipitates is often below 100 nm. With an atomic force microscope (AFM), sections through specimens are analysed with a resolution in the nanometre range, which allows a quantitative determination of the γ′ volume content and PSD for different superalloys. Thus, determined volume fractions for the γ′ phase in the superalloys CMSX-6 and Waspaloy amount to 56% and 27%, respectively. A combination of an AFM with a nanoindentation system also allows the measurement of local mechanical properties such as hardness and elasticity. These quantities are determined for the first time directly on the superalloys CMSX-6 and Waspaloy for the γ′ and matrix phases, separately. The γ′ phase in both alloys shows a significantly higher but different hardness than the matrix phase, whereas the moduli of elasticity are similar. A depth dependence of the hardness was found for very small indentations.     

A high thermal gradient directional solidification method for growing superalloy single crystals

The experiments here were conducted at withdrawal rates of 3 mm/min and 1 mm/min using a CMSX-6 and a CMSX-4 superalloy, respectively. The process was assessed in terms of the thermal gradient (G_L), structural refinement, microsegregation and porosity distribution, and compared to those using a Bridgman process. The G_L of the process was 200–236 K/cm, which was 10–12 times higher than that in the Bridgman process. A more refined microstructure was produced having average primary and secondary dendrite arm spacing values as low as 243 μm and 72 μm, as well as 272 μm and 76 μm in the CMSX-6 and the CMSX-4 castings, respectively. The diameter of γ′ phase in the dendrite core of CMSX-6 and CMSX-4 castings was reduced from 0.8 μm to 0.3 μm and from 1.2 μm to 0.6 μm, respectively. The average areas of (γ′ + γ) eutectic pools became smaller and more homogeneously distributed. The mean pore sizes in the castings were reduced by 57% and 43% for the CMSX-6 and CMSX-4 superalloys, respectively, and the area fractions of the pores in the CMSX-6 and CMSX-4 samples were 16% and 12% of those produced in the Bridgman samples. The segregation coefficients of the major alloying elements were closer to unity than those in the Bridgman process, which indicates that the composition distribution is more uniform.     

Permeation of hydrogen in Ni-based superalloy CMSX-4

Permeation of hydrogen in CMSX-4 superalloy single crystals was studied by the volumetric method in the temperature interval from 573 to 1173 K. The measurements were carried out with two purity grades of hydrogen and for various sample thicknesses to estimate the influence of surface effects upon the permeation rate. The effect of phase transformation in the γ′ particles upon permeation rate was detected.     

Lattice-parameter misfits between the γ and γ′ phases in single crystals of nickel superalloys

X-ray diffractometry has been used to study the transformation of the profiles of X-ray diffraction lines (222), (004), (003) and to determine the values of the lattice misfit between the γ and γ′ phases (γ/γ′ misfit) for single crystals of nickel superalloys ZhS6U, ZhS30, ZhS32, VZhM1, and VZhM4 in the as-cast state and after heat treatment. It has been established that with an increase in the γ/γ′ misfit there increase tetragonal distortions of the fcc lattice of the γ solid solution. For the single crystals of VZhM4 alloyed with rhenium and ruthenium, splitting of the superlattice reflection of the strengthening γ′ phase has been revealed. The structural regularities revealed testify to qualitative changes in the factors that govern the redistribution of refractory alloying elements between the lattices of the γ and γ′ phases, for example, of Mo in the presence of Re and Ru.     

Lattice strain evolution during creep in single-crystal superalloys

In situ neutron diffraction studies are carried out to characterize the micromechanical deformation occurring during tensile creep of a typical single-crystal nickel-based superalloy, CMSX-4. The loading responses of the matrix γ phase and the precipitate γ′ are distinct. Moreover, the behaviour in the tertiary creep regime (in which the γ′ phase remains intact) is qualitatively different from that in the primary creep regime (when γ′ is sheared). In tertiary creep, initial deformation of the matrix leads to a release of misfit between the phases in the (1 0 0), resulting in elastic compression of the γ in the loading direction. The load state then remains fairly constant during creep. During the initial stages of primary creep, elastic compression of the γ phase is observed until at around 2–4% creep strain this compression stabilizes as the (1 0 0) misfit is released. This is the point at which γ′ shearing is thought to begin. Subsequently, the load in the γ increases by around 200 MPa until a maximum is reached at around 8% creep strain. This load is then suddenly released, which may be due to the release of back-stress.     

Increase of misfit during creep of superalloys and its correlation with deformation

In superalloys the loss of coherency during creep results in the increase of misfit of the γ/γ′-interface. The kinetics of this process were measured locally by TEM (Moiré fringes) and X-ray diffraction. Two materials were creep tested (SRR99 and CMSX-4) in two temperature ranges (stable γ′-morphology and rafting), and the morphology changes were quantified. A microstructural model allows calculation of the equilibrium misfit and the increase of plastic strain on the basis of these data. At high temperatures and low stresses the model describes quantitatively creep kinetics up to 30 h. Here the processes controlling primary creep are propagation of dislocation loops along matrix channels and thickening of the matrix channels orientated perpendicular to the load direction.     

The phase-field approach and solute drag modeling of the transition to massive γ → α transformation in binary Fe-C alloys

The transition between diffusion controlled and massive transformation γ →α in Fe–C alloys is investigated by means of phase-field simulations and thermodynamic functions assessed by the Calphad technique as well as diffusional mobilities available in the literature. A gradual variation in properties over an incoherent interface, having a thickness around 1 nm, is assumed. The phase-field simulations are compared with a newly developed technique to model solute drag during phase transformations. Both approaches show qualitatively the same behavior and predict a transition to a massive transformation at a critical temperature below the T₀ line and close to the α/α + γ phase boundary. It is concluded that the quantitative difference between the two predictions stems from different assumptions on how the properties vary across the phase interface yielding a lower dissipation of Gibbs energy by diffusion in the phase-field simulations. The need for more detailed information about the actual variation in interfacial properties is emphasized.     

On the possibility of rhenium clustering in nickel-based superalloys

In order to elucidate the role of this element in superalloy metallurgy, the binding energy of Re–Re pairs and the stability of small Re clusters in the nickel face-centred cubic (fcc) lattice is investigated using ab initio density functional theory. It is shown that the formation of Re–Re nearest neighbour pairs is energetically unfavourable, and that this repulsive energy is dramatically reduced as soon as the solute atoms move further apart from one another. Furthermore, small nearest neighbour and second neighbour Re clusters are found to be unstable. The calculations are repeated for W and Ta, which lie beside Re in the periodic table; the results are essentially the same, except that some Ta–Ta higher order pairs have a positive binding energy, consistent with the Ni–Ta binary phase diagram exhibiting several ordered intermetallics. The predictions show that Re clusters are unstable in fcc Ni and it is unlikely that clustering has a role in improving creep and fatigue properties (the rhenium-effect) in Ni-based superalloys.     

Development of γ phase stacking faults during high temperature creep of Ru-containing single crystal superalloys

An unusual deformation mode involving the formation of intrinsic stacking faults in the γ matrix of experimental Ru-containing γ–γ′ superalloys with high Co and Re contents during high temperature creep at 950 °C/290 MPa has been observed. The morphology, distribution and dependence of these stacking faults on alloy chemistry has been investigated along with their formation mechanism. Additions of Re and Co substantially decrease the stacking fault energy of the γ matrix. The observed stacking faults in the γ matrix form by the dissociation of a/2〈1 1 0〉 matrix dislocations with Burgers vectors perpendicular to the loading direction in the early stages of creep. The dependence of creep properties on elemental additions that influence stacking fault energy is discussed.     

An overview of rhenium effect in single-crystal superalloys

Nickel-based single-crystal superalloys are the key materials for the manufacturing and development of advanced aeroengines. Rhenium is a crucial alloying element in the advanced nickel-based single-crystal superalloys for its special strengthening effects. The addition of Re could effectively enhance the creep properties of the single-crystal superalloys; thus, the content of Re is considered as one of the characteristics in different-generation single-crystal superalloys. Owing to the fundamental importance of rhenium to nickel-based single-crystal superalloys, much progress has been made on understanding of the effect of rhenium in the single-crystal superalloys. While the effect of Re doping on the nickel-based superalloys is well documented, the origins of the so-called rhenium effect are still under debate. In this paper, the effect of Re doping on the single-crystal superalloys and progress in understanding the rhenium effect are reviewed. The characteristics of the d-states occupancy in the electronic structure of Re make it the slowest diffusion elements in the single-crystal superalloys, which is undoubtedly responsible for the rhenium effect, while the postulates of Re cluster and the enrichment of Re at the γ/γ′ interface are still under debate, and the synergistic action of Re with other alloying elements should be further studied. Additionally, the interaction of Re with interfacial dislocations seems to be a promising explanation for the rhenium effect. Finally, the addition of Ru could help suppress topologically close-packed (TCP) phase formation and strengthen the Re doping single-crystal superalloys. Understanding the mechanism of rhenium effect will be beneficial for the effective utilization of Re and the design of low-cost single-crystal superalloys.     

热暴露下预应变对CMSX-4单晶高温合金的显微组织演变的影响

在室温下,对经完全热处理的第二代单晶高温合金CMSX-4实施压缩和拉伸预应变。压缩和拉伸预应变在单晶CMSX-4中产生了剪切带。单晶CMSX-4在950℃下热暴露10h,沿剪切带产生了γ′粒子择优粗化。剪切带上的γ′粒子逐渐侵入γ通道。最后,γ通道沿着剪切带消失。TCP状粒子伴随着γ通道的消失而出现。然而,热暴露10h的普通单晶CMSX-4没有产生TCP沉淀,也没有γ′粒子择优粗化。热暴露100h的预应变CMSX-4沿剪切带产生了γ′粒子和TCP相粒子择优粗化,基体中也有γ′粒子粗化。     

Phase-field modeling of the γ′-coarsening behavior in Ni-based superalloys

Microstructural changes during heat treatment of the Ni-based CMSX-4 and CMSX-6 superalloys have been investigated experimentally and simulated using a phase-field multi-component model incorporating elastic driving forces in the presence of a lattice misfit. Furthermore, a theoretical model of the coarsening of anisotropic particles is proposed for the prediction of the main kinetic parameters of the coarsening process. A comparison of the main characteristics of the microstructural evolution during non-directional γ′-coarsening, provided from both experiments and phase-field simulations, gives a good agreement of the coarsening kinetics of the CMSX-4 superalloy. However, for the CMSX-6 superalloy, phase-field simulation results and theoretical predictions are not entirely consistent with experimental results, and show that additional effects, for example, those caused by plastic deformation, might be a reason for the slow coarsening rate.     

Elemental partitioning of platinum group metal containing Ni-base superalloys using electron microprobe analysis and atom probe tomography

The partitioning of platinum group metal (PGM) additions in polycrystalline Ni-base superalloys has been investigated. Alloys with a baseline composition of 15Al–5Cr–1Re–2Ta–0.1Hf (at.%) with systematic variations in Pt, Ir, Ru and W were cast and heat-treated to produce bimodal two-phase γ–γ′ microstructures. Electron microprobe analysis was used to measure the composition of coarse γ′ particles. Selected alloys were studied using atom probe tomography. Pt and Ru weakly partition to the γ′ and γ phases, respectively, whereas Ir partitions equally between both phases. Pt and Ru do not change partitioning behavior appreciably with additional W and Re. Interestingly, Ir lowers γ′ volume fraction and reduces partitioning of W, Re, Ru, Al and Pt. Tungsten partitioning coefficients exhibit an unusually strong temperature dependence, with increased partitioning of W to the matrix at lower temperatures. The thermodynamics of these PGM-containing systems are analyzed and the implications for partitioning are discussed.     

Evolution of solute clustering in Al–Cu–Mg alloys during secondary ageing

The evolution of an atomistic-level nanostructure during the early stages of secondary ageing of a rapid hardening Al–1.1Cu–1.7Mg (at.%) alloy has been characterized by a combination of atom probe tomography (APT) and transmission electron microscopy. Quantitative APT analysis reveals key changes in the evolution of solute clusters during secondary ageing (T6I4 condition) that correlate with secondary hardness increments. The microstructures that cause peak hardness differ between the T6 and T6I4 tempers – the former is a result of solute clustering as well as the precipitation of GPB zones and S phase, whereas in the latter, secondary ageing promotes only the formation of solute clusters. Cu–Mg clusters with high Mg:Cu ratio have the most strengthening potency during secondary ageing in T6I4 heat treatment.     

Directional solidification of TiAl–Re–Si alloys with aligned γ/α2 lamellar microstructures

The effect of Re additions on the solidification behavior of Ti–Al–Re and Ti–Al–Re–Si alloys was investigated to determine if such additions would be compatible with a seeding technique developed previously to align the γ/α₂ lamellar microstructure. Rhenium additions to TiAl were found to enlarge the primary β composition range. Rhenium also tends to segregate to the β-dendrite cores leading to the formation of the B2 phase. After the L+β→α peritectic reaction, Re tends to segregate to the interdendritic liquid. For the Ti–Al–Re alloys, the maximum Re content found in the lamellar microstructure was near 0.5 at% and the lamellar orientation in master alloys containing this amount of Re could successfully be grown from the seed material. However, such levels of Re had little effect on the measured yield stress either at room temperature or at 1000°C and tensile failure usually occurred by the coalescence of microcracks that nucleated from large silicide particles. Finally, it was found that the γ/α₂ lamellar microstructure of Ti–43.5Al–3Si–0.5Re can successfully be aligned by using an appropriately oriented seed even though β is the primary solidification phase indicating that the α-phase from both the L+β→α peritectic and the L→α+Ti₅Si₃ eutectic reactions must be continuous with that of the seed material.     

Sulfur embrittlement on γ/γ′ interface of Ni-base single crystal superalloys

The impurity sulfur embrittlement on γ/γ′ interface of Ni-base single crystal superalloys has been investigated by first principles quantum mechanics DMol calculations. Using a local sum of vertical (∑BO^v) and horizontal (∑BO^h)Mayer bond orders, we proposed a new method to evaluate the competition between the shear and cohesive strengths of the interface. Coupled with the phenomenological theory of fracture, we define the ratio of R_BO=∑BO^h/∑BO^v to assess the embrittlement trend of the interface related to sulfur doping or sulfur doping combined with Re substitution. It is shown that sulfur increases R_BO by 121% relative to the sulfur-free γ/γ′ interface, which could induce interface embrittlement from the electron bonding point of view. Calculations of both BO and charge density distribution reveal that it is the strong bonds between sulfur and Ni atoms lying within the interface that contribute to the interface embrittlement. The substitution of Re for Al at the γ/γ′ interface results in reduced R_BO, thus relieving the tendency of interface embrittlement. Furthermore, our model on sulfur embrittlement is compared with previous models.     

High-Temperature Corrosion Resistance of Al–Re and Re Coatings

The oxidation behavior in air at 1473 K, and sulfidation behavior in a H2S–H2 gas mixture with a sulfur partial pressure of 10^–2 Pa at 973 K of Al–Re coated CMSX-4 were studied. Investigation on the sulfidation behavior of the Re-coated CMSX-4 was carried out in a H2S–H2 gas mixture with a sulfur partial pressure of 10^–2 Pa at 973 K. The experimental results show that a Re-rich alloy layer was formed between an -Al₂O₃ layer and the inner concentration zone of Ta and Ti for the CMSX-4 single crystal alloy with an Al–Re coating after oxidation. The Re-rich alloy layer containing Cr, W, Ni, Co, and Mo effectively inhibited the outward diffusion of alloying elements and the inward diffusion of Al. The Al/Re-coated CMSX-4 single crystal alloy had excellent sulfidation resistance; the Re-rich alloy layer, containing W, Ti, Ta, and Mo, which formed during the sulfidation process and located between the alumina scale and the CMSX-4 base alloy, plays a role in inhibiting the outward diffusion of alloying elements. The sulfidation resistance of CMSX-4 single-crystal alloy is greatly improved by a Re coating on the CMSX-4 alloy surface; this is attributed to a Re–Cr–W alloy layer that retarded the outward diffusion of cations and the oxide layer containing Ta, Ti, and Al, which inhibited the inward penetration of sulfur.     

Recovery of rhenium from tungsten‑rhenium wire by alkali fusion in KOH-K2CO3 binary molten salt

Tungsten‑rhenium wire is used in thermocouple and lamp filament manufacturing due to its good thermal sensitivity and high temperature plasticity, and many waste wires are generated in processing and after use. This work focuses on the efficient recovery of high value rhenium from tungsten‑rhenium wire waste with a mass composition represented by W95Re5. The main steps for recovery include alkali fusion, recrystallization, hydrogen reducing and washing. First, WRe wire was decomposed by KOH-K₂CO₃ molten salt to produce potassium perrhenate, where the decomposition ratios of W and Re reached 99.36% and 99.80% using a mass ratio of salt to wire of 3:1, m(KOH) of 80% (m representing the mass fraction of KOH in binary salt), a temperature of 800 °C and a reaction time of 60 min. Then, the decomposed product was leached by water, and from the resulting lixivium high purity KReO₄ crystals were obtained by segregation, which had a perfect rhombic dipyramid morphology and average size of 73.26 μm. Last, the material was reduced to Re powder at 350 °C with a H₂ flow rate of 10 L/min. Re powder, with a purity of higher than 99.5% and fine grain size of 19.37 μm, was obtained after washing with acid and water. This method provided a potential economic process for the recovery of waste WRe wire.     

Solute clustering in Al–Cu–Mg alloys during the early stages of elevated temperature ageing

The evolution of atomistic-level nanostructure during the early stages of elevated temperature ageing of rapid hardening (RH) Al–Cu–Mg alloys has been characterised by a combination of atom probe tomography (APT), transmission electron microscopy (TEM) and positron annihilation spectroscopy (PAS). APT analysis confirms that significant dispersions of small solute clusters form during ageing for 60 s at 150 °C. No zone-like precipitate structures were observed by TEM and APT examinations. These small clusters are believed to be responsible for the RH effect. Careful quantitative APT analysis reveals that a high density of Cu–Mg clusters with high Mg:Cu ratio gives the most potent strengthening response. Positron annihilation measurements also show that Cu–Mg clusters provide additional sites for vacancy stabilisation.     

Impact of replacement of Re by W on dislocation slip mediated creeps of γ′-Ni3Al phases

The anomalous flow behavior of γ′-Ni₃Al phases at high temperature is closely related to the cross-slip of 1/2〈110〉{111} super-partial dislocations. Generalized stacking fault energy curves (i.e., Γ-surfaces) along the lowest energy path can provide a great deal of information on the nucleation and movement of dislocations. With the first-principles calculation, the interplay between Re and W, Mo, Ta, Ti doped at preferential sites and their synergetic influence on Γ-surfaces and ideal shear strength (τ_max) in γ′-Ni₃Al phases are investigated. Similar to single Re-addition, the Suzuki segregation of W at stacking faults is demonstrated to enable to impede the movement of 1/6〈112〉{111} Shockley partial dislocations and promote the cross-slip of 1/2〈110〉{111} super-partial dislocations. With the replacement of a part of Re by W, a decreased indicates that the anomalous flow behavior of γ′ phases at high temperature is not as excellent as the double Re-addition, but an increased τ_max means that the creep rupture strength of Ni-based single crystal superalloys can be benefited from this replacement to some extent, especially in the co-segregation of Re and W at Al−Al sites. As the interaction between X1_Al and X2_Al point defects is characterized by an correlation energy function , it is found that both strong attraction and strong repulsion are unfavarable for the improvement of yield strengths of γ′ phase.