Electromagnetic levitation and drop tube processing were used to investigate the solidification behaviour of Co50Cu50 alloys. Deep undercooling was achieved down to 241 K, the alloy melts separated into Co-rich L1-phase and Cu-richL2-phase before solidification. In drop tube processing, the Cu-rich phases disperse morehomogeneously than in EML. This is due to the reduced gravity and high cooling rate during free fall and solidification.

In order to engineer materials suitable for micromagnetic inductor cores we have fabricated and investigated the properties of granular (CoFeB)-SiO2 amorphous thin films. The deposition method, synchronous triple-rf magnetron sputtering onto rotating substrates, induces anisotropy in the film plane. Here we present magnetic hysteresis loops, resistivity and magnetoresistance measurements of two of these films, with different amount of metallic content. Both films have anisotropic resistivity, which is higher for the film with less metallic content. The low-metallic-content film exhibits typical isotropic giant magnetoresistance (GMR), while the film with higher metallic content shows a mixture of GMR and anisotropic magnetoresistance (AMR). The AMR appears at fields below 500 Oe. We believe that this has not been observed before in amorphous samples. The behaviour of the resistivity versus temperature ρ (T) indicates that the sample with lower metallic content consists of isolated grains while the high-metallic-content sample consists of a percolated network. Experimental ρ (T) data for the low-metallic-content sample fit: to the variable-range hopping model at temperatures lower than 221 K and can be described by excitation of the carriers to the mobility edge at higher temperatures.

A new criterion γ, defined as Tx/(Tg+Tl) (where Tx is the onset crystallization temperature, Tg the glass transition temperature and Tl the liquidus temperature), has been proposed for glass formation in bulk metallic glasses (BMGs). The interrelationship between this new parameter and glass-forming ability (GFA) was elaborated and discussed in comparison with other established criteria. It was found that the new criterion γ had a much better correlation with GFA than other established parameters. An approximation of the critical cooling rate for glass formation was also formulated and evaluated in the light of this new parameter.

Bulk metallic glass can be formed only when the critical cooling rate for glass formation is reduced to 100–2 K/sec. However, a cooling rate achievable with molecular dynamics simulation is higher by many orders of magnitude, so the gap has to be abridged by analytical theories. We propose a theory of bulk metallic glass formation based upon our early theories of glass formation composition. The critical concepts include the idea of local glass transition, distributed local glass transition temperatures and coincident local fluctuation for atomic transport. Strong repulsion between small atoms was recognized for the first time as the necessary condition for bulk glass formation.

Asynthesized calculation model for developing metallic glasses has been created by taking into account criteria for the achievement of high glass-forming ability (GFA) and viscosity. The model deals with amorphous-forming composition region (AFCR), crystallization temperature (Tx) and three GFA factors: critical cooling rate (Rc), reduced glass-transition temperature (Tg/Tm) and supercooled liquid region (ΔTx=Tx-Tg) where Tg and Tm are glass transition and melting temperature, respectively. The principle of the model is based on thermodynamic functions for multicomponent systems, i.e., mismatch entropy and mixing enthalpy which express the criteria in terms of the number of elements, atomic size differences and the heat of mixing. By combining these thermodynamic quantities with the Miedema's semi-empirical model, AFCR was calculated, and was compared with the experimental results. The GFA factors were also analyzed from viscosity. The Rc was derived from transformation diagram of metallic glasses for crystallization while ΔTx was calculated by solving a differential equation expressing the change in free volume with temperature. As a result of these analyses, Rc-Tg/Tm and R-ΔTx diagrams were found to fit with the experimental results qualitatively. Furthermore, crystallization temperature (Tx) was also calculated for multicomponent metallic glasses by the modification of the Miedema's binary model for the calculation of Tx. The reduced crystallization temperature (Tx/Tl), where Tl is liquidus temperature, was calculated for evaluating GFA of metallic glasses instead of Tg/Tm. Some of the calculation methods used in the present study have theadvantage giving results as a function of composition; thus, there exists possibility to lead to the prediction of glassy alloys compositions. In this sense, the present calculation methods are completely different to the current method for the development of new metallic glasses relying on the empirical criteria which suggest only appropriate systems and/or elements of the alloysfor the achievement of high GFA. In near future, this kind of calculation technique can be used for the prediction of optimal compositions of the metallic glasses with high GFA.

The influence of different microstructures of La-based fully amorphous samples and its composites on the impact fracture energy were investigated and discussed. Results showed improvement in fracture energy of glassy metals with intermetallic phases, but deteriorated in the presence of dendrite phases and high volume % of crystalline phases.

Based on the available kinetic and thermodynamic data, we compare the kinetic fragility and thermodynamic fragility of different metallic glass forming liquids. The results indicate a correlation between the kinetic and thermodynamic fragility in metallic glass forming liquids, consistent with the energy landscape model which predicts a connection between the kinetic and thermodynamic properties of supercooled liquids. The metal - metalloid glass forming alloys such as PdNiCuP are found to exhibit a distinctively different correlation as compared to early – late transition metal – type metallic glass forming alloys such as ZrTiCuNiBe. For the same thermodynamic fragility the former exhibit a much larger kinetic fragility indicating that the two classes of alloys have a different liquid structure. In addition, the relationship between the kinetic fragility and the Gibbs free energy difference between the undercooled liquid and the crystalline phases has been discussed in both metal metalloid glass forming alloys and early - late transition metal – type metallic glass forming alloys.

High-manganese ferrous-based alloys containing 10–20 at. % Mn have been investigated as prospective iron-based structural amorphous metals with magnetic transition temperatures far below the ambient temperature. Many of these alloys are found to have a high reduced glass transition temperature of 0.6–0.63 and large supercooled liquid region of 40–90 °C. Rod-shaped amorphous samples with diameters reaching 4 mm are obtained by employing simple injection casting. The search for good glass-forming alloys has been guided by an atomistic approach coupled with the realization of low-lying liquidus temperatures via proper alloying. The tensile yield strengths and Vickers hardness of the new amorphous metals far exceed those known in high-strength steel alloys, and the elastic moduli are comparable to those reported for super-austenitic steels. The present high-manganese amorphous Fe-alloys also show promise as very good corrosion-resistant materials.

Effects of atomic-scale open-volume regions in metallic glass structure on the flow and fracture behavior of a Zr-Ti-Ni-Cu-Be bulk metallic glass were examined. Studies of relaxation time scales showed that atomic arrangement processes for viscous flow were significantly retarded with annealing. Plane strain fracture toughness was significantly decreased and fatigue crack crack-growth rates were dramatically increased, indicating degradation of resistance to crack extension as a result of annealing. Fracture morphology completely changed from vein patterns to cleavage-like features with little evidence of plasticity with annealing. The positron lifetime and Doppler broadening experiments revealed decreased open-volume regions as a result of annealing. The loss of stress relief ability by retarded crack tip viscous flow as a result of the anneal-out of open-volume regions is believed to contribute to observed annealing embrittlement.

This work presents investigations on the short-range order of stable and deeply undercooled melts of the pure elements Fe, Zr and Ni. The liquids were undercooled by use of the electromagnetic levitation technique, which was combined with elastic neutron scattering in order to determine the structure factors of the liquids as a function of the temperature. The diffraction experiments provide evidence that a short-range order based on aggregates of icosahedral symmetry prevails in the melt. This short-range order is observed even at temperatures above the melting temperature and gets more pronounced if the liquids are undercooled.

Molecular dynamics simulations of a two-dimensional amorphous solid exhibit strain localization when loaded in uniaxial tension with free boundaries. The degree of localization depends sensitively on the rate of loading and on the existence of surface defects. Regions of both dilation and contraction arise during the shear band slip process. These dilation and contraction events correspond to dynamic free-volume generation and annihilation during shear. Correlations of the dilation and contraction of the material reveal a length scale of 1–4 atomic diameters associated with this physical process. Dilation is observed to result in nanometer scale cavitation.

Metallic glasses lack long-range translational symmetry and have excess volume trapped within their amorphous structure, which has a direct bearing on their physical properties including deformation characteristics. Moreover, the trapped excess free volume is directly correlated to the defect concentration facilitating the possibility to model the temperature and time dependence of the free volume changes during creep as a trade off between defect generation and annihilation. Using differential scanning calorimetry (DSC) analysis the residual free volume of a metallic glass can be characterised based on the glass transition peak height (Δcp). In the present work constant strain rate tests were carried out at the ‘onset’ (Tgon = 685 K) and ‘point of inflection’ (Tgp = 705 K) of the calorimetric glass transition to study the time dependent flow behaviour in Zr55Cu30Al10Ni5 bulk metallic glass. Modelling based on DSC analysis and positron lifetime spectroscopy on samples creep deformed to different plastic strain values corroborate the stress decrease after the peak stress (‘stress overshoot’) occurring in bulk metallic glasses with increasing plastic strain to be associated with a small increase in free volume.

The sub-Tg elevated temperature fatigue propagation behavior of a Zr-based bulk metallic glass was examined. Fatigue crack-growth rates as a function of the applied stress intensity range, ΔK, are reported. Increased testing temperature was found to increase ΔKTH. Examination of crack surfaces at near-threshold growth rates revealed the presence of elongated ridges perpendicular to the crack front at higher temperatures; these ridges were absent in specimens tested at lower temperatures. The size of the features was found to increase with temperature and applied ΔK, finally breaking up at higher growth rates. Mechanisms responsible for the ridge formation are described, and modeled in terms of fluid meniscus instabilities that occur at the fatigue crack tip.

We have used transmission electron microscopy (TEM) to investigate the structure of shear bands produced by bending electron-transparent Zr52.5Cu17.9Ni14.6Al10Ti5 metallic glass specimens. Shear bands were located by comparing the structure of the specimens before and after deformation. The shear band spacing is influenced by the structure of the specimen; portions of the specimen with a significant population of nanocrystals show a smaller separation between shear bands. Quantitative high resolution TEM analysis based on ratio technique has been used to explore the defect structure in shear bands. High density and void-like defects with size of about 1 nm were found in shear bands formed in both amorphous and nanocrystalline areas. A simple model was proposed to explain the formation of these defects.

Processing of foams from bulk metallic glass (BMG) alloys, using melt infiltration techniques, is reported for the first time. Foaming methods based on infiltration of two types of pattern materials are described: investment of a continuous refractory yielding very low relative density structures (5% dense relative to the BMG), and investment of a discontinuous refractory pellet bed yielding higher relative density (50–60% dense relative to the BMG). Both methods are capable of producing foam structures; however high surface area and diminished thermal conductivity, especially in lower density structures, make vitrification of the alloy difficult.

Bulk forms of Refractory Alloy Glasses (RAGs) of the composition Ni60Nb37Sn3 have recently been synthesized as part of a larger project for potential use in structural applications. Modeling efforts of such metallic glasses have traditionally involved the use of hard sphere models without regard to potential electron density fluctuations of individual components. X-ray characterization of these materials (in both reflection and transmission modes) provided scattering data necessary for subsequent radial distribution analysis, which gives structural information such as nearest neighbor distances and packing characteristics. A model radial distribution function (RDF) was constructed based upon a hard sphere space filling assumption and compared to the RDFs obtained from experimental scattering data. The experimental RDFs showed no difference from the model RDF within the limit of experimental error, with regard to nearest neighbor distances and coordination numbers of the first two nearest neighbors.

Additionally, transmission mode scattering experiments from a white beam x-ray source (E = 2 to 40 keV) demonstrated a through thickness amorphous structure of 2mm thick samples. Converted line profiles from the two-dimensional diffraction patterns from this experiment also showed agreement with reflection mode experiments. Samples of Vitreloy-106 (Zr57Nb5Cu15.4Ni12.6Al10) of similar thicknesses were used as standards due to the well-known behavior of this particular class of metallic glasses.

We have investigated failure in bulk metallic glass-forming alloys under dynamic compression. We implemented a recovery technique for the compression Kolsky bar to obtain dynamically deformed, intact specimens at various stages of deformation; this allows us to characterize the development of failure. We have also used high-speed photography to examine the failure process during the recovery experiments. The experimental results indicate that the failure under dynamic loading is somewhat different from that under quasi-static loading. Specimens subjected to quasistatic deformation developed multiple shear bands and substantial plastic deformations, while specimens subjected to dynamic (—strain rate ∼103 s-1) compressive loading fail by fracture along one dominant shear band. The mechanisms of dynamic failure in bulk metallic glasses are discussed on the basis of these experimental results.

We compare the potential energy at zero temperature of a range of crystal structures for a glass-forming binary mixture of Lennard-Jones particles. The lowest energy ordered state consists of coexisting phases of a single component face centered cubic structure and an equimolar cesium chloride structure. An infinite number of layered crystal structures are identified with energies close to this groundstate. We demonstrate that the finite size increase of the energy of the coexisting crystal with incoherent interfaces is sufficient to destabilize this ordered phase in simulations of typical size. Specific local coordination structures are identified as of possible structural significance in the amorphous state. We observe rapid crystal growth in mixtures near the equimolar composition.

The annealing response of amorphous Al based alloy samples were investigated to assess the role of the as-quenched state on primary crystallization of Al nanocrystals(nc). Continuous heating differential scanning calorimetry (DSC) traces of amorphous Al87Ni10Ce3 powders were compared to those from melt spun ribbon (MSR) to examine the effect of sample subdivision on primary crystallization. While the powders exhibited the same onset temperature as MSR, thermal cycling experiments show fine powder sizes reacting at the onset temperature and coarse powder sizes with the lowest melt quench rate transforming at the highest primary reaction temperature. In Al92Sm8 MSR, a kinetics analysis of Al nc distributions indicates a notable effect of the as-quenched state on primary crystallization during isothermal annealing. With Al88Y7Fe5 MSR intense deformation can induce thedevelopment of an Al nc distribution without thermal annealing. In each case examined, the results support the inclusion of quenched in clusters in the analysis of primary crystallization reactions.

X-ray Photoemission spectra of Ni0.40Pd0.40P0.20 and Cu0.40Pd0.40P0.20 in the amorphous states are compared to each other and to first principles calculations of the electronic density of states. The electronic structure is calculated for a previously validated amorphous model of Ni0.40Pd0.40P0.20 and for the same amorphous structure but with Cu replacing Ni. The measured and calculated electronic structures of Ni0.40Pd0.40P0.20 agree and exhibit very little charge transfer to Pd. However, the measured and calculated electronic structures of Cu0.40Pd0.40P0.20 differ considerably. In Cu0.40Pd0.40P0.20 there is a large electron transfer to Pd. The local densities of states on Pd atoms in different CuPdP local environments that arise in the amorphous model shed light on this dramatic charge redistribution.