Superpurity (> 99.995wt% pure) lxxx series Al-0.3wt%Fe-0.lwt% Si alloys with different impurity additions have been melt spun to produce ribbons with microstructures of fine scale ∼0.1–0.2μm diameter secondary phase particles entrained in a higher melting point Al matrix. The melting and resolidification behaviour at 2Kmin−1 of the secondary phase particles while entrained in the solid matrix was monitored using differential scanning calorimetry, to simulate the nucleation of secondary phases in liquid interdendritic regions during the final stages of solidification in casting. The melting and resolidification behaviour of 500ppm V and Zr doped superpurity ribbons is the same as that of a commercial purity ribbon. This indicates that V is responsible for the promotion of the metastable FeAlm phase in the commercial purity ribbon. The presence of FeAlm in direct chill (DC) cast commercial purity alloys indicates that V also influences phase selection in the faster cooled (∼5-10Ks−1) regions of these alloys.

Recently, it was demonstrated that grain refinement in metals can take place through two mechanisms, namely, dynamic nucleation and remelting of initially formed dendrites. In this study, it was found that Ni99.45B0.55 undergoes grain refinement, both by dynamic nucleation or by remelting, depending on the initial bulk undercooling just before crystallization. The nature of the grain refinement process is confirmed by microstructural analysis of the undercooled specimens.

Samples of the Al – 0.5wt.% Cu system were directionally grown under controlled conditions, to study the role played by the instabilities in the process relatives to the microstructure selection for a given values of interface velocity, and thermal gradient. Using an interface quenching technique and metallographic analysis in longitudinal and transversal cuts of the samples, we determine the transition mechanisms between the different stages of the growth, and associate them to the stability of the solidification front. We study the planar to a cellular transition in different conditions, and although the solidification parameters are in good agreement with the perturbations theory, when analyzing the amplitude of the perturbations during the planar to a cellular transition, the same theory is not able to predict certainly the critical wavelength in this case. Also, we found a subcritical behavior during the transition from a planar to a cellular interface for the diluted Al – Cu system, detecting a hysteresis behavior for the amplitude of the perturbations when it is increasing and then decreasing the interface velocity, through the threshold.

Refractory metals and alloys are deeply undercooled in an ultrahigh vacuum drop-tube facility. During rapid solidification, some brightness traces give evidences of two successive phase transformations. First-principles calculations of the structural stability in these systems are developed to appreciate the possibilities of obtaining metastable phases.

The aim of the present paper is to analyse the melting/solidification of pure metals resulting from non-equilibrium conditions. Attention is focused on the direct measurement of the calorimetric signal obtained under isothermal hold of the sample at a temperature close to the equilibrium melting temperature Tm, which results in both melting of an overheated solid and solidification of an undercooled melt. The non-equilibrium transformations are monitored by DSC under isothermal regime, with previous continuous heating/continuous cooling of the sample. The dependence of the calorimetric signal on thermodynamic factors, ΔH and ΔG, is explored. Here ΔH and ΔG are, respectively, the melting enthalpy and Gibbs free energy difference between the crystal and the liquid. In particular, the results of the investigation performed on the melting/solidification behaviour of pure In and Pb are presented.

The mechanisms of layer structure formation in the two phase region of peritectic systems are discussed. Under diffusive growth conditions, a banded structure is predicted within a narrow composition range in the hypoperitectic composition. Experimental studies show the formation of an oscillatory structure in the hyperperitectic region. It is shown that this structure is induced by convection in the melt, and it is a novel structure in which a large single tree-like structure of primary phase is surrounded by the peritectic phase. Basic ideas of convection that lead to this tree-like morphology are described.

Substrate surfaces of α-Fe were laser alloyed with a CW-CO2 laser using Ni2B powder. Single laser scans and laser scans with a 50% overlap between successive passes were carried out to obtain hard surface coatings on the α-Fe substrate. High hardness values in the range of 520 - 600 Hv1 were achieved by laser boriding of α-Fe, compared to a substrate hardness of ∼ 110 Hv1. The improvement in hardness is caused by an off-eutectic microstructure, which consists of Fe dendrites and Fe3B and Fe lamellae. Both phases, Fe3B and Fe, contain Ni in solid solution. For samples treated with a single laser scan the Fe phase solidifies in the bcc structure. In contrast, a 50 % overlap between successive laser passes causes the iron to crystallize in the fcc γ-phase. The difference in crystallographic structure is caused by the amount of Ni dissolved in Fe during laser alloying: only ∼ 9 at.% Ni is contained in the α-Fe phase after a single laser scan, while up to ∼ 40 at.% of Ni is dissolved in the γ-Fe phase after a 50 % overlap between each laser scan due to the nearly complete melting of all injected Ni2B particles. However, the Fe3B phase solidifies during laser alloying independent of the amount of Ni in the orthorhombic Fe3C structure, which is metastable for Fe contents exceeding 45 at.%.

A simple model is proposed to analyze the interface stability of the RE123 superconductor in accordance with the constitutional supercooling criterion. As the single crystal growth of the 123 phase is largely dependent on the growth interface stability, a quantitative analysis has been required. From the numerical analysis for the case of peritectically solidified Sm123, it was clarified that the constitutional supercooling must exist in the liquid when the 123 growth interface comes close to a 211 particle. It could also predict that larger 211 particle radius, smaller volume fraction of the 211 particles, larger growth rate, or smaller imposed temperature gradient cause easy occurrence of the constitutional supercooling. The growth rate and a 211 particle radius are determining parameters. Further consideration of the nucleation at the L/211 interface just ahead of the 123 growth front could describe the 123 growth morphological transition from with the planar interface to the equiaxed blocky.

This novel approach to modeling the steady-state solidification of undercooled pure liquids is based upon first principles. Continuum equations are used to describe a volumetrically averaged, coexisting mixture of solid and liquid in the thin phase transformation zone between regions of pure liquid and pure solid. These equations are coupled with a dynamic equilibrium based rate law that describes temperature dependent phase transformation kinetics. The time scale associated with finite rate phase transformation is found to depend on a solidification activation energy, thermal energy, and the enthalpy of fusion. The model leads naturally to an eigenvalue problem for the solidification front propagation speed. In addition, the variation of the volumetrically averaged liquid fraction across the solidification zone is predicted.

The directional solidification of the transparent binary alloy succinonitrile-poly(ethylene oxide) was studied in an experiment in which solidification speeds of about 2 mm/sec could be reached without loss of the linear temperature gradient. The low diffusivity of the polymer solute allowed the study of the dynamics of rapid solidification using an optical microscope. For both normal and doublonic dendrites we observed a transition to large triangular “superdendrites” above a certain solidification speed and we report measurements of the primary and secondary spacing as a function of the pulling speed. Our measurements suggest that the observed triangular shape is due to a decoupling of primary and secondary growth at large undercooling.

In this report, transition of solidified phases for directionally solidified Fe-Cr-Ni alloys has been investigated in low growth rate range by using Bridgman type furnace. The ferrite-austenite eutectic growth has been confirmed like a plane front growth of ferrite single phase under low growth rate condition. The transition velocity between eutectic and ferrite cell growth has a good agreement with the result of calculation based on the phase selection criterion and the interface temperature calculation for ferrite, austenite and eutectic phases. These results show that the phase prediction by calculating interface temperature can be applied not only to competitive growth between single phases like peritectic systems, but also to eutectic systems. Furthermore, under the condition of eutectic coupled growth to be occurred in steady state, the changes of solidified phases and their morphologies in the initial transient are discussed.

The relationship between solidification microstructure and fluidity in MMC was studied. The composites were fabricated by infiltration of liquid metal into a alumina SAFFIL fibers preform under a gas pressure, using alloys of the AlCuSi system as matrices. The fluidity was measured in terms of classic foundry practice (i.e., the distance of flow liquid metal into the preform, while solidifying). The characterization of solidification microstructure in the cast composite was analysed and correlated with the results of fluidity. The attention was particularly focused on such effects as the presence or absence of selective nucleation, the refinement of certain solidifying phases in the presence of fibers and their influence on microstructure formation and segregation of certain elements present in the liquid at the fiber matrix interface. By comparing reinforced and non reinforced zones, it was shown that the presence of fibers resulted in a refinement of the dendritic arm spacing of the α-Al phase, with nucleation of Si on the fibers and without nucleation of primary Al dendrites. The results were discussed and compared with the microstructures and fluidity test of the unreinforced Al-Cu-Si alloys.

Yttrium iron garnet (YIG) single crystal fibers of nominal composition Y3Fe5O12 were grown by the laser heated pedestal growth (LHPG) technique, a miniaturized floated-zone process. YIG which melts incongruently, was grown at a temperature below the peritectic decomposition temperature under self-adjusting conditions even though it has very narrow solidification region according to the Y2O3-Fe2O3 phase diagram. YIG fibers in diameter ranges from 100 to 750 μm were grown at various growth rates and conditions, and analyzed by x-ray diffraction, electron microprobe, and IR-VIS spectroscopy. Infrared transparent YIG fibers were grown at rates below 12 mm/h in air. At these growth rates, yttrium orthoferrite and iron-oxide inclusions within the YIG fiber, which act as IR scattering centers, were significantly reduced. The transparency of the fibers was more dependent on the growth rate than the stability of the molten zone. Surface ridges containing an Fe-rich composition were observed at all growth rates. These were associated with molten zone instability.

A metastable, disordered bcc phase has been formed from the melt in a Cr-rich alloy of the Nb-Cr-Ti system, where large volume fractions of the Laves phase would develop under equilibrium solidification conditions. X-ray diffraction (XRD) studies and lattice constant determinations confirm that the melt-spun ribbons contain a bcc phase beyond its terminal solid solution limits. Solidification pathways are proposed based upon metastable and equilibrium phase diagrams. Microstructures have been studied using optical microscopy, scanning electron microscopy (SEM), and transmission electron microscopy (TEM). Cross-sectional TEM reveals tiny, dispersed Laves phase precipitates within the metastable bcc grains near the chill (wheel) side. Away from the wheel, the microstructure changes to a dendritic structure of the bcc and C15 Laves phase. Annealing of the melt-spun ribbons produces a fine, two-phase microstructure of the equilibrium phases. Microstructures from the melt-spun and arc-melted processing conditions are compared. The microstructural control afforded through this metastable processing route enables a methodology to tailor phase distributions for optimized toughness in Laves phase alloys.

The kinetics of phase transformation which follows a nucleation-and-growth mechanism was studied by using probability theory. From the calculation of the survival probability for each individual site, the general equation for describing the transformation kinetics, in which the nucleation and growth rates are considered as a function of time and space, is derived. In comparison to the classical derivation by Avrami, the new derivation is logical and transparent. The extension of the treatment by using the definition of the multiple-survival probability leads to the exact solutions of the time dependent grain size distribution functions during transformation. A new understanding of fundamental relationships for the microstructural analysis can be achieved by comparing different kinds of size distribution functions. By applying the principles of the analytical treatment to the simulation, model systems of vast size can be handled for very complicated transformation process.

Nucleation in multi-component systems is poorly understood, particularly when the rates of long-range diffusion are comparable with the rates of attachment at the cluster interface. For illustration, measurements of the time-dependent nucleation rates in silicate and metallic glasses are discussed. A new model for nucleation in partitioning systems, which explains many of devitrification microstructural features in bulk metallic glasses, is presented.

Reaction kinetics in bimetallic multilayers have demonstrated that sharp unidirectional concentration gradient, which develop as interdiffusion proceeds at the interface are able to delay or to suppress nucleation of intermetallics. It has been found that the existence of a critical gradient beyond which nucleation is inhibited is strongly dependent on the mechanism of formation of the embryo [1–7].

A mechanism of nucleation under concentration gradient (∇c) is proposed and treated on the basis of the Fokker–Planck equation for the distribution in the size space.

The influence of the aspect ratio of the embryo on the critical concentration gradient is also studied. Due to the fluctuations of the embryo shape, it is shown that the minimization of the thermodynamic potential leading to the aspect ratio of the embryo is only significant beyond a certain value of the concentration gradient. Application is presented to the nucleation of the compound Ni10Zr7 in an amorphous layer Ni-Zr.

Dissipative regimes may induce athermal processes in addition to the thermally activated single atom addition or subtraction events considered by classical nucleation theory. Displacement cascades or impinging energetic ions could destroy sub-critical precipitates. Similarly, either displacement cascades or energetic knock-on atoms could remove single atoms or groups of atoms from subcritical clusters. Cluster injection is also possible, as in vapor deposition of clusters or in the cores of displacement cascades providing void nuclei.

In this work we have re-examined the classical problem of nucleation and growth. A new model considers the correlations among droplets and naturally incorporates the crossover from the early-stage, nucleation dominated regime to the scaling, late-stage, coarsening regime within a single framework.

We present a new theoretical model of nucleation and growth in term of a novel domainspatial correlation function. This model probes the patterns and spatio-temporal evolution of nucleation and growth process and agrees very well with experimental data. The dynamic domain-spatial correlation function directly and simultaneously explores the transformed volume fraction, the time-dependent domain size distribution function, and the spatial correlation function of domain core centers for the entire process, including the post-nucleation, domain growth and grain formation stages.