Search

The coarsening behavior of Al3Zr precipitates during aging was investigated for two Al–Mg–Zr alloys (Al–3.6Mg–1.2Zr and Al–2.9Mg–2.1Zr, wt%) processed by selective laser melting (SLM). Scanning transmission electron microscopy (STEM) investigations of peak-aged (400 °C, 8 h) samples reveal both continuous (~2 nm in diameter) and discontinuous (~5 nm wide and hundreds of nanometers in length) coherent, secondary L12-Al3Zr precipitates. In-situ STEM experiments showed that aging at 400 °C results in the appearance and growth of both grain-boundary Al3Zr precipitates, and intragranular nanometer-sized spherical Al3Zr precipitates in Zr-rich dendritic arms. Heating to 500 °C resulted in the disappearance of most Al3Zr precipitates and oxide particles. This microstructural evolution sheds light on the evolution of the alloy strength at elevated temperature. For short-term yield tests, as-fabricated samples displayed higher yield strengths than peak-aged samples at temperatures above 150 °C (e.g., 87 vs 24 MPa at 260 °C). This is attributed to coarsening of grain-boundary precipitates during aging, decreasing their ability to inhibit grain-boundary sliding (GBS) of the fine equiaxed grains (~1 µm). For longer term creep tests at 260 °C, both as-fabricated and peak-aged samples displayed near-identical creep behavior during a long-duration (168 h) creep test; by contrast, during a shorter duration creep test (8 h), as-fabricated samples are more creep-resistant than samples previously aged at 260 °C (threshold stresses of ~40 vs. ~14 MPa, respectively). Again, the creep behavior is consistent with coarsening of grain-boundary precipitates, occurring now during long-duration creep tests at 260 °C. An exact creep mechanism could not be isolated due to microstructural changes during testing but is believed to be a combination of GBS and dislocation motion.

Gas-atomized powders of two ternary alloys, Al-3.60Mg-1.18Zr and Al-3.66Mg-1.57Zr (wt.%), were densified via laser powder bed fusion. At energy densities ranging from 123 to 247 J/mm3, as-fabricated components are near-fully densified (relative density 99.2–99.9%) as verified by X-ray tomography. While Mg acts a solid-solution strengthener, Zr creates two types of metastable L12 Al3Zr precipitates, each playing dual roles: (a) sub-micrometer Al3Zr particles form in the melt upon solidification and act as grain refining agents, nucleating fine aluminum grains, which (i) prevent hot-tearing during the rapid solidification inherent to laser melting and (ii) enhance tensile strength (Hall-Petch strengthening) and ductility (influence a heterogenous grain structure) after fabrication; (b) Al3Zr nano-precipitates form in the solid alloy during subsequent aging, which (i) precipitation-strengthen the alloy leading to an increase of >40% in strength over the as-fabricated value, and (ii) promote thermal stability of the fine grain size (and the associated Hall-Petch strengthening) after exposure to high temperature due to the slow kinetics of Al3Zr coarsening (from the sluggish diffusivity of Zr in solid Al-Mg). While the Zr-richer alloy shows higher yield and ultimate tensile strength in the as-fabricated state, both alloys have identical mechanical properties after peak aging. Interconnected bands of fine (∼0.8 μm), equiaxed, isotropic grains and coarser (∼1 × 10 μm), columnar, textured grains – both containing oxide particles and Al3Zr precipitates - provide a combination of high yield strength and high ductility (e.g., ∼354 MPa, and ∼20%, respectively) with isotropic values in both as-fabricated and peak-aged samples, unlike Al-Si alloys processed via laser fusion of commercial Al-Si-based powders. The pre-alloyed, gas-atomized Al-Mg-Zr powders do not contain expensive alloying elements such as Sc, nor do they require blending with a second powder to nucleate fine grains, making them excellent candidates for economical, large-scale additive manufacturing applications.

Al-Er-Sc-Zr-Si alloys, strengthened by L12-ordered, coherent Al3(Er,Sc,Zr) nanoscale precipitates, can be used for automotive and aerospace applications up to 400 °C. Vanadium, due to its small diffusivity in aluminum and its ability to form L12-ordered tri-aluminide precipitates, is a possible micro-alloying addition for further improving the service temperature of these alloys. Moreover, vanadium-containing Al3(Er,Sc,Zr,V) precipitates are anticipated to have a smaller lattice parameter mismatch with the matrix, thereby improving the alloy's coarsening resistance. In this study, the temporal evolution of microstructural and mechanical properties of an Al-0.005Er-0.02Sc-0.07Zr-0.06Si alloy micro-alloyed with V are investigated utilizing isochronal, isothermal and double-aging treatments and compared to the results obtained from an alloy that does not contain V, but otherwise has the same composition. Both isochronal and isothermal aging treatments reveal slower precipitation and coarsening kinetics for the V-containing alloy. A peak microhardness value of ∼600 MPa is obtained after a double-aging treatment at 350 °C/16 h, followed by aging at 400 °C for 12 h. Transmission electron microscopy reveals a duplex-size precipitate microstructure, with the smaller precipitates having a mean radius

A hierarchical microstructure is obtained in an alloy with composition Fe–8.1Al–12.2Cr–1.9Mo–18.2Ni–2.0Ti (wt.%) processed by melt-spinning. The evolution of the precipitation pathways is investigated using transmission electron microscopy (TEM) techniques, atom probe tomography (APT) and first-principles thermodynamic calculations. As-solidified ribbons exhibit a random dispersion of B2-ordered precipitates (NiAl-type) in an Fe-based matrix. Subsequent aging at 700 °C yields nucleation and growth of the L21-phase (Ni2TiAl-type) within the primary B2-precipitates, leading to a microstructure exhibiting three types of hierarchy: (i) a structural hierarchy due to chemical ordering, with a chemically disordered matrix of bcc-Fe (A2), the nearest-neighbor (NN) ordered B2-precipitates (NiAl-type) and the next nearest-neighbor (NNN) ordered L21-precipitates (Ni2TiAl-type) within B2, (ii) a dimensional hierarchy with a continuous bcc-Fe matrix, coherently embedded B2-precipitates, with a size range of 60–200 nm and the coherent precipitate substructure, with L21-phase and dimensions of 15–20 nm. (iii) A spatial hierarchy where B2-precipitates are embedded in the bcc-Fe matrix and L21-precipitates nucleate and grow only within B2-precipitates. In addition, it is verified that the interface between B2 and L21 is coherent and adopts a diffuse structural profile. Monte-Carlo simulations reproduce these observations and it is found that interface energies of B2 and L21 reduce from 50 mJ/m2 at 0 K to 11 mJ/m2 at 973 K. Kinetic-Monte-Carlo simulations support the interpretation of the experimental results that the L21 nucleates within the B2 phase.

The ferritic alloy Fe–10Cr–10Ni–5.5Al–3.4Mo–0.25Zr–0.005B (wt.%), strengthened by coherent B2-structured (Ni, Fe)Al precipitates with a volume fraction of 13 vol.% and average precipitate radius of 62 nm, was subjected to creep in the stress range 30–300 MPa and the temperature range 600–700 °C. The stress dependence of the steady-state strain rate can be represented by a power law with high apparent stress exponents of 6–13 and high apparent activation energies of 510–680 kJ mol−1. Threshold stresses at all studied temperatures were observed, ranging from 69 to 156 MPa, from which a true stress exponent of ∼4 and a true activation energy of 243 ± 37 kJ mol−1 were determined, which are equal to those for dislocation creep and lattice diffusion in the ferritic matrix, respectively. Based on these mechanical results and detailed electron microscopy observations, the creep mechanism was rationalized to be general dislocation climb with repulsive elastic interaction between coherent precipitates and the matrix dislocations.

Shear mixing of the ternary alloy system Ag–Cu–Ni during ball-milling and high-pressure torsion was investigated to elucidate the effects of chemical interactions on phase formation. First, ball-milling of pure Ni with homogeneous Ag67Cu33 alloy powders at room temperature (RT) was studied for average Ni atomic concentrations of 4%, 9%, 15% and 25%. Additional samples with an average composition of Ag50Cu25Ni25 were ball-milled at ∼−15 °C or subjected to high pressure torsion at ∼−125 °C. X-ray diffraction and atom probe tomography measurements showed that Cu largely transferred from the Ag–Cu alloy phase to the Ni-rich phase at all temperatures, but that Ag and Ni did not significantly intermix. The Cu concentration in the steady state, moreover, was surprisingly higher in the Ni-rich phase than in the Ag-rich phase, and it was further enriched at the interphase boundary, even at −125 °C. High-resolution transmission electron microscopy revealed that the sizes of the Ni/Cu precipitates and the grain size of the Ag-rich matrix were reduced to a few nanometers during RT or cryo-ball-milling, which is much finer than those observed after ball-milling of Cu–Ag or Ni–Ag binary powders. These findings illustrate that chemical effects can play an important role in phase formation during severe plastic deformation, but they also show that other kinetic factors can influence the final microstructure as well.

Nanosized precipitates in an Al–0.06Sc at.% alloy containing various microalloying additions were studied, with the goal of developing cost-effective aluminum alloys for high-temperature applications, using micro-hardness, electrical conductivity and atom-probe tomography measurements. Substituting 0.005 at.% Er for the more expensive Sc maintains high ambient-temperature strength, and dramatically improves the high-temperature creep resistance, as anticipated from the increase in lattice parameter mismatch between the α-Al(fcc) matrix and the coherent L1 2 -ordered Al 3 (Sc,Zr,Er) precipitates. A concentration of the slow-diffuser Zr as low as 0.02 at.% is sufficient to provide coarsening resistance at 400 °C (an homologous temperature of 0.72) for up to 66 days by forming a Zr-enriched outer shell encapsulating the precipitates. Finally, adding 0.05 at.% Si enhances ambient-temperature strength by increasing the number density of precipitates, while decreasing the homogenization and peak-aging heat-treatment times, which is caused by the Si atoms accelerating the Er and Sc diffusion kinetics. Si-containing alloys are also cost effective, owing to the existence of Si in commercial purity Al. But addition of Si reduces the precipitate coarsening resistance: the magnitude of this effect is, however, determined by the Si concentration.

The microstructure of a twin-roll-cast Al–4.5Mg–0.28Sc at.% alloy after friction-stir processing, performed at two tool rotational rates, was investigated by atom probe tomography. Outside the stir zone, the peak-aged alloy contains a high number density (∼8.0 × 1023 m−3) of ∼1.5 nm radius Al3Sc (L12) precipitates with a minor Mg content, providing an increase of ∼600 MPa in the Vickers microhardness. In the stir zone of the sample processed at 400 rpm rotational rate, the microhardness increase is mainly due to grain refinement, rather than precipitate strengthening, because the Al3Sc precipitates, with spherical lobed cuboids and platelet-like morphology, grow and coarsen to a 10–20 nm radius. The Sc supersaturation across the stir-processed zone has a concentration gradient, which is higher on the retreating side and lower on the advancing side of the friction-stir tool. Hence, after aging at 290 °C for 22 h, the microhardness increase within the stir zone also displays a gradient due to precipitate strengthening with varying precipitate volume fractions. In the stir zone for the sample processed at 325 rpm rotational rate, the microhardness increase is also predominantly due to grain refinement, as coarse Al3Sc precipitates form heterogeneously at grain boundaries with a platelet-like morphology. The hardness remains unchanged after a 290 °C aging treatment. This is because the Al3Sc precipitates are highly heterogeneously distributed due to a combination of a small Sc supersaturation (0.05 at.%) in the matrix, the existence of dislocations, and a large area per unit volume of grain boundaries (∼4–6 × 106 m−1).

Shear-induced mixing in heterogeneous Cu alloy systems was investigated by molecular dynamics simulation. Each system, which comprised a single spherical particle within a Cu matrix, was subjected to cyclical shearing events to high strains at 100 K. The particles investigated were Cu, Ag, Ni, Fe, Nb, and V. fcc particles were observed to undergo “superdiffusive” mixing with dislocations crossing the particle–matrix interface. The initial rate of mixing in these systems increased quadratically with particle radius. bcc particles showed different behaviors. For Nb and V dislocations did not cross into the particles and the initial rates of mixing increased linearly with particle radius. The Nb particles remained spherical but developed amorphous Cu–Nb shells at the particle–matrix interface. The V particle showed some faceting, but it too formed amorphous layers on non-facetted interfaces. The Fe particle behaved similarly to Nb and V particles initially, but more like fcc particles at high strains. The molecular dynamics results are compared with experimental observations.

The microstructural evolution of dilute Cu-based ternary Cu–Nb alloys during high-temperature annealing was investigated using a combination of transmission electron microscopy, X-ray diffraction and kinetic Monte Carlo (KMC) simulations. The experiments showed that, during annealing, the Cu 90 Nb 10 binary alloy undergoes extensive coarsening at 600 °C, with precipitate sizes increasing to >40 nm. Additions of just 1.5 at.% W to this alloy, however, dramatically suppresses the growth; the average precipitate size in the ternary alloy now increases to only ∼10 nm at 600 °C, and only to 18 nm at 700 °C. On annealing at still higher temperatures, the precipitate size then, surprisingly, decreases. The precipitate size distribution at the higher temperatures, moreover, is bimodal. It is also observed that irradiation has no effect on the microstructure of the ternary alloys above 600 °C. KMC simulations indicate that the saturation of the average precipitate size, followed by its decrease as the annealing temperature is increased and the development of a bimodal size distribution can be explained by competition between thermodynamic and kinetic effects during precipitation in this ternary alloy.

The forced chemical mixing of atoms in model immiscible alloys during severe plastic deformation (SPD) has been investigated as a function of temperature and the heat of mixing using molecular dynamics computer simulations. At low temperatures, A75B25 alloys form solid solutions during SPD for heats of mixing less than ∼20 kJ mol−1, but tend to phase separate at larger values. At high temperatures these alloys show more extensive precipitation, with the precipitate morphology dependent on the heat of mixing. Analysis of the high-temperature mixing kinetics reveals that the precipitation process involves two separate mechanisms. The first derives from long-range diffusion mediated by shear-induced vacancies, while the second is due to local rearrangements of atoms induced by the forced mixing of atoms.

Shock compression of mono- and nanocrystalline (nc) nickel is simulated over a range of pressures (10–80 GPa) and compared with experimental results. Contributions to the strain from the various mechanisms of plastic deformation such as partial dislocations, perfect dislocations and twins are quantified in the nc samples. The effect of stress unloading, a phenomenon often neglected in MD simulations, on dislocation behavior is computed. It is shown that a large fraction of the dislocations generated during compression is annihilated upon unloading. The present analysis resolves a disagreement consistently observed between MD computations and experimental results. Analytical models are applied to predict the critical pressures for the cell-to-stacking-fault transition and the onset of twinning as a function of grain-size and stacking-fault energy (through the addition of tungsten). These predictions are successfully compared with experimental results.