The resolution offered by weak-beam transmission electron microscopy allows for the direct observation of partial dislocations in a large number of intermetallic alloys. In many instances, the comparison of these observations with computer simulated images leads to results that are much more quantitative and descriptive than are possible with ordinary analyses. Examples from several parallel studies are presented: i) comparisons of experimental and simulated images that have been used to characterize extrinsic stacking faults in TiAl are shown, ii) the need to correct the experimentally observed antiphase boundary dissociations (dAPB) in Ni3Al in order to account for the image shifts that occur during microscopy are highlighted, and iii) dissociation distances of less than 2 nm, which were observed with (2g-5g) diffraction conditions, were verified and quantified by comparisons with computer simulated images. Antiphase boundary energies (γAPB) and complex stacking fault energies (γCSF) were calculated from the corrected observations, and it was found that the APB energies did not depend on alloy content, but that the CSF energy of binary Ni3Al is less than it is for a boron containing alloy.

A set of three Embedded Atom Method potentials are presented that produce a stable L1 0structure. All three were nominally fitted to experimental parameters of the γ-TiAl phase; variations in the fit were allowed in order to give potentials with differing values of planar fault energies. These potentials were evaluated and found to produce stable APB(111), APB(100), CSF(111), and SISF(111) faults. Dislocation core structures were computed for 1/2<110>-type dislocations in both edge and screw orientation with all three potentials. The screw orientation of the highest fault energy potential was found to have a non-planar configuration, being spread about equally on two {111} close packed planes. All other cores were found to be planar and spread on a single {111} plane. No significant edge components of strain developed for the screw orientation, suggesting that these dislocations would not appear to be spread in Atomic Resolution TEM. Preliminary evaluations showed that dislocation mobilities were low, requiring stresses on the order of 10-3 μ for motion.

We discuss our experiments and coordinated computational modeling that show the value of magnetic moment measurements for helping to identify the site selection of an additive to an intermetallic compound. Magnetic susceptibility measurements (5K-300K) were used to determine additive (V, Cr, Mn) moments; and the site distribution was determined experimentally by measuring the relative variations of the (001) and (110) x-ray diffraction superlattice lines and comparing to the expected behavior for occupation of either site, or a distribution between sites. This gives us a predicted relationship of measured additive moment to site distribution which was then further validated by comparison to our ab initio calculation of the moments for Ti site occupation (as expected at low additive concentrations from total energy considerations). The moments (calculated using the cluster discrete variational (LCAO) method with isolated impurities) of 1.21, 2.36 and 2.52 μB for V, Cr and Mn, respectively compare favorably to the lowest-concentration experimental values of 1.01 and 2.3 μB for V and Mn, respectively; while the disagreement for Cr with low-concentration experimental moment of 0.55 μB may be due to additive clustering effects not included in the single-additive-atom cluster calculations or to multivalent behavior of Cr. The experimental decrease of V and Cr moment with concentration indicates a shift from Ti site occupancy toward increasing Al site occupancy with increase in V and Cr concentration.

The local chemical order near grain boundaries (GBs) in NisAl was verified by measuring the local degree of order and chemistry with microdiffraction and EDS microanalysis. The composition of alloys are (a) Ni-24at%Al with and without 700ppm wt boron;(b) Ni-25 at%Al with 700ppm wt boron;(c) Ni-26 at%Al with 700ppm wt boron. GBs studied in the present work are high angle boundaries. The results show that there is a local disordered region about 5-10nm wide on the side of the GB in non-stoichiometric Ni3Al with or without boron, and no disordered region in stoichiometric Ni3Al. EDS microanalysis results show that the composition of GB almost keeps the matrix concentration of Ni in stoichiometric Ni3Al with boron, and is Ni-rich in hypostoichiometric and Al-rich in hyperstoichiometric Ni3Al with and without boron. The local disordered region existing in the vicinity of GBs is related to Ni enrichment or depletion near GBs in hypostoichiometric or hyperstoichiometric Ni3Al alloys. Boron seems not to be the main factor to control the local disorder and composition at GBs in Ni3Al alloys. It can be concluded that the local disorder at GBs is not a decisive factor for ductilizing GBs in Ni3Al alloys.

The atom probe field ion microscope (APFIM) has been used to characterize grain boundaries and matrix in NiAl doped with either boron, carbon or beryllium. Boron was observed to segregate to grain boundaries whereas carbon and beryllium did not. Atom probe analyses of the matrix revealed that the matrix was severely depleted of the solute in the boron- and carbon-doped alloys. Field ion imaging and matrix analyses also revealed ultrafine MB2 - and MC-type precipitates ranging in size between 2 and 20 nm in diameter in the boron- and carbon-doped alloys. These precipitates occurred in significant number densities. Atom probe analyses of beryllium-doped NiAl did not reveal ultrafine precipitates and was consistent with the fact that almost all the beryllium was in solid solution. The enormous increase in yield stress in the boron- and carbon-doped alloys is predominantly due to a precipitation hardening effect. The small increase in yield stress in beryllium-doped NiAl is due to a mild substitutional solid solution hardening effect.

The structure of a highly oriented NiAl Σ=5 [001] (310) grain boundary has been imaged by high resolution electron microscopy and analyzed by multislice image calculations. This grain boundary exhibits an asymmetry produced by a rigid body displacement of approximately ½ d130 along the grain boundary, but does not show any measurable expansion of the grain boundary. The match between the simulated and experimental images of this grain boundary is improved by changing the stoichiometry of the boundary, which can be accomplished by adding antisite defects and vacancies. The final model structure incorporates 50% constitutional vacancies on the nickel sites adjacent to the boundary plane and a replacement of the aluminum sites adjacent to the boundary plane with 50% nickel antisite defects and 50% vacancies.

The microstructures of the L12 titanium trialuminides at low temperatures were studied using a number of single crystals with various Al-Ti-Fe compositions, all of which lie in the nominal single phase L12 Field at 1200 °C. Five different second phases were found to be in equilibrium with the L12 matrix, namely , (Al,Ti)3Fe, Al3Ti, Al2FeTi, Ti2NAl and Al2Ti+Fe. Small volume fractions of the first two phases are often seen in compounds containing relatively low Ti contents. The Al2FeTi, a so-called T phase, was observed at relatively high Fe contents. The Ti2NAl does not seems to be sensitive to the Al-Ti-Fe composition, it exists to some extent in all the alloys used in this study. Ti2NAl and the interface with the L12 matrix are found to be brittle and provide the sites for crack initiation. The Al2Ti+Fe phase has been observed in many compounds containing high Ti contents (>25 at.%), and has a large hardening effect. Like binary Al2Ti, the phase possesses a tetragonal structure of the Ga2Hf-type, and forms plates on the cube planes of the LI2 matrix. The crystallographic relation between the Al2Ti and the Ll2 matrix was determined to be (100)p//(100)m.and (010)p//(010)m. Porosity is also commonly seen in these alloys, and has a very destructive impact on ductility.

The morphologies and defect structures of TiCr2 in several Ti-Cr alloys have been examined by optical metallography, x-ray diffraction, and transmission electron microscopy (TEM), in order to explore the room-temperature deformability of the Laves phase TiCr2. The morphology of the Laves phase was found to be dependent upon alloy composition and annealing temperature. Samples deformed by compression have also been studied using TEM. Comparisons of microstructures before and after deformation suggest an increase in twin, stacking fault, and dislocation density within the Laves phase, indicating some but not extensive room-temperature deformability.

The deformation characteristics of the B2 intermetallic alloys NiAl and CoAl have been examined using single crystals deformed at 673 K in the <110> orientation. Slip trace analysis and transmission electron microscopy was used to characterize the deformation process. While many aspects of the deformation are similar, some distinct differences were observed. All of the alloys were found to deform by <001> slip. NiAl was found to deform by slip on a combination of {110} and {100} slip planes with the predominant slip plane being a function of alloy stoichiometry. In contrast, CoAl was found to deform by slip only on {100} planes.

The evolution of dislocation substructures formed in single crystals of NiAl by tension creep testing at temperatures between 1123 K and 1473 K has been studied. Significant differences in the dislocation structure are evident in samples tested in different orientations.

For samples tested in soft orientations, where glide of (001) dislocations is easy, some interacting (001) dislocations can be found. In hard oriented samples, the density of dislocation networks is high and low angle subgrain boundaries are observed. The subgrain size in hard oriented crystals is strongly stress dependent.

In hard oriented crystals, (001) dislocations have no resolved shear stress for glide and must move by the slower process of climb; or by glide of (110) dislocations. The (001) dislocations are captured in networks, leading to a high total dislocation density in hard oriented samples compared to soft oriented crystals, whereas the dislocation density within the volume of subgrains is constant for all orientations.

Using Pettifor Maps as a guide to alloying, scandium additions were made to Ti3Al to try and transform the crystal structure of this intermetallic alloy from ordered hexagonal to ordered face centred cubic. Scandium was found to have a solid solubility limit of about 6.4at.% in Ti3Al and preferentially occupy the titanium sublattice. According to the A3B Pettifor Map, these are criteria which should effect a change in crystal structure, however no such transformation was observed. Scandium additions in excess of the solid solubility limit resulted in the formation of SC2AI at grain boundaries and triple points.

Alloys based on intermetallic compounds in the Ti-Al binary system have been extensively investigated as candidate light weight structural materials for elevated temperature applications. However, one compound in this system, Al2Ti, has been virtually ignored. Al2Ti, like Al3Ti, is expected to possess lower density and better oxidation resistance compared to Ti3Al and TiAl based alloys. Since the structure and stability of Al2Ti are poorly understood, this work was undertaken to determine the compound's equilibrium crystal structure and the microstructures in the as-cast, melt-spun, and annealed conditions. Powder X-ray diffraction shows that annealed melt-spun Al2Ti has the tetragonal Ga2Hf crystal structure containing 24 atoms per unit cell with a = 3.9705 Å, c = 24.322 Å and a calculated density of 3.530 g/cm3. Optical microscopy and scanning electron microscopy with energy dispersive spectroscopy were used to investigate the material's microstructures. Cast Al2Ti has a small volume fraction of TiAl second phase particles. Rapidly solidified and annealed Al2Ti is single phase. Powder processing (melt spinning and hot isostatic pressing) Al2Ti yields material with a finer microstructure, a higher hardness, and increased resistance to room temperature cracking than as-cast material. Additionally, microhardness results indicate that Al2Ti has an increased resistance to cracking at room temperature than identically processed Al3Ti with a similar hardness.

Dislocation configurations produced by room and high temperature deformation of two alloys with compositions Ti-48A1 and Ti-47.5Al-2.5Cr have been compared. At room temperature the major differences between the two alloys is the increased twinning activity and that of 1%2 <112] superdislocations in the chromium containing alloy while the binary alloy contains a larger contribution from <101] superdislocations. At 400°C, the temperature at which a peak in flow stress is observed in both alloys, an increase in twinning activity and emission of 1%2<110] dislocations on (001) planes is observed from twin intersections in both materials. The contribution of superdislocations to the deformation process is limited and their effect on the flow stress increase appears irrelevant. Instead the anomalous strengthening effect has been related to the cross-slip of single dislocations from cube planes, on which they are emitted, to octahedral planes where they can also glide.

There is no generally accepted model of the atypical mechanical properties of L12 alloys such as the flow stress peak. Since its introduction in the early sixties, the so-called Kear-Wilsdorf (KW) lock has played a central role in the analysis of the mechanical properties of this category of alloys. We analyze the mechanical stability of a split configuration, intermediate between the superdislocation fully extended in the primary slip plane and the KW lock. The influence of composition on mechanical properties is stressed, based on experimental determinations of the flow stress dependencies upon load orientation when composition is varied.

We present a model for the anomalous increase of the yield stress exhibited by many L12 compounds. It is based on two thermally activated processes that describe respectively the pinning and unpinning of [101] screw dislocations in the (111) plane. The model explains all the important characteristic features observed in the anomalous regime. We discuss the applications of the model to Ni3Ga and Ni3(Al,Ta).

Alloys based on the B2 compound NiAl have significant potential for applications in hot sections of aircraft engines due to their low density, high melting point, and high thermal conductivity. A major disadvantage of this class of materials is low ductility at ambient temperatures. Recently, it was discovered that small levels of certain elements (eg. Fe, Ga, Mo) result in dramatic improvements in room temperature ductility. In this paper, results are presented from a mechanistic investigation of the “microalloying” effect. Tensile and compression testing as a function of temperature and orientation has been performed on both the binary compound and microalloyed material. Data on ductile to brittle transition temperatures, critical resolved shear stress values as a function of temperature on the different slip systems, and dislocation structures from TEM analysis of the tested specimens are presented. These data are discussed in terms of possible mechanisms for the microalloying effect in NiAl alloys.

A new method of relaxation series is proposed which allows for the measurement of the apparent activation volumes and of the hardening correction terms. This method is successfully applied to the measurement of the effective activation volumes along the stress strain curves. A transition stress between a microplastic and a macroplastic domain can be unambiguously defined on the plots of the effective volume as a function of strain. The corresponding transition stress measured at three temperatures in the range of the flow stress anomaly (293K, 423K, 573K), agrees reasonably well with the values of τ0.2%. The latter stress was usually considered as the critical stress for plastic deformation, in former studies.

We discuss the possibilities of stabilizing (111) slip in NiAl based on recent results of the atomistic structure of dislocation cores in this material. In previous work it was found that the (111) complete dislocation is stable with respect to partials of 1/2 (111) Burgers vector. However, split configurations were found with a very similar energy. In particular, split configurations that are nearly planar were found, indicating that if this type of dislocation could be stabilized they would possibly result in ductile alloys.

We discuss the possible effects of local disorder and deviations from stoichiometry on core structure concentrating on the objective of stabilizing (111)slip. Atomistic simulations performed to study these effects show that local compositional changes and disorder may have significant effects on the core structure. These effects are most important in the regions of the core where the point defects are localized and it is unlikely that such local changes could be used to successfully stabilize (111) slip. Furthermore, it is not clear whether local changes in ordering and compositional states can follow the dislocation as it moves during the deformation process. We therefore propose that the alloying efforts should be directed towards changing the APB energy of the NiAl-base alloy to a lower value.

The orientation and temperature dependence of the operative slip systems and critical resolved shear stress (CRSS) were investigated in Ti3Al single crystals containing 24.4at%Al and 33.0at%Al. Prism {1010}<1210> slip occurs preferentially in comparison with basal (0001)<1210> and pyramidal {1121}<1126> slip. The CRSS for prism and basal slip decreases with increasing temperature, while that for pyramidal slip exhibits positive temperature dependence having an anomalous peak around 500°C. The temperature dependence of the CRSS for these slip systems does not strongly depend on Al content. The ductility of Ti3Al single crystals is influenced by operative slip systems. Coarse slip bands on the basal plane often act as a trigger for crack nucleation. The compositional deviation to the Al-rich side is harmful for ductility of Ti3Al alloys.

The compressive flow behavior of single crystalline LI2 Al67FegTi25 was investigated as a function of temperature and orientation at temperatures from 77K to about 1250K, using specimens with compressive axes orientated near [001], [113], [011], [122] and [111]. The operating slip systems seen in these specimens after 0.4% plastic deformation are predominantly of the octahedral type at all temperatures, even in near-[122] and [111] specimens in which the Schmid factors for the primary cube slip system are larger than that for the primary octahedral slip system. The yield stress increases rapidly with decreasing temperature at low temperatures, while it decreases gradually from room temperature to higher temperatures. The critical resolved shear stress (CRSS) on the [101](111) slip system does not seem to be orientation-dependent over a wide range of temperatures, except at temperatures from 1050K to 1250K where the CRSS exhibits a mild orientation-dependence. Fracture tests at room temperature were also conducted. No special orientation-dependence of the ductility was observed.