Volume 89, Issue 4

A numerical model is presented that describes the microstructure evolution in liquid immiscible alloys under the common action of nucleation, diffusional growth and drop motion during continuous cooling through the miscibility gap. The model is satisfactorily verified by comparison with an analytically solvable case. The numerical results for an Al-5 wt.% Pb alloy cooled with 10,20 and 50 K/s from a temperature above the binodal show that the simultaneous action of Marangoni migration and Stokes settlement governs the macrosegregation of Pb. At low cooling rates a Pb film establishes by Stokes motion at bottom, at high cooling rates the Pb droplets accumulate in the middle part of a cylindrical sample due to the strong Marangoni motion opposing gravity induced sedimentation. The average drop radius does not vary with position in the sample at high cooling rates. At small cooling rates a slight variation with position is predicted.

A method for measuring the liquid–liquid interfacial tension between coexisting phases in metal systems will be presented in detail. First experimental results for the Ga–Hg and Ga–Pb systems were obtained. Measurements in the Ga–Hg system were taken between the monotectic temperature ( T M = 26.7 °C) and 63.1 °C and in the whole range of the miscibility gap in the Ga–Pb system. In addition experimental results for the densities of the coexisting phases in the Ga–Hg system and for the density difference between the coexisting phases in the Ga–Pb system were obtained.

In this part of a series of papers on Fe-, Co- and Ni-rich melt-quenched ribbons containing Zr of Hf, the results of a study of the thermal stability by differential scanning calorimetry, transmission electron microscopy and thermomagnetic measurements will be described for the amorphous Zr–Fe, Zr–Co and Zr–Ni as well as the bcc-Ni(Zr) alloy ribbons. For the amorphous alloys, a general tendency is that both the crystallization temperature and the enthalpy of crystallization decreases in the sequence Fe → Co → Ni. A comparison of two a-Zr 9 Fe 91 ribbons prepared with different quenching rates revealed that for the lower quenching rate the activation energy and the crystallization temperature are slightly smaller due to the larger number of quenched-in nuclei. For the a-Zr 9 Co 91 alloy, the crystallization products were found to be fcc-Co + Zr 6 Co 23 and although a single differential scanning calorimetry (DSC) peak only occurs, transmission electron microscopy (TEM) studies indicated that the precipitation of Zr 6 Co 23 precedes the appearance of fcc-Co. For the Ni-based alloys, the phase transformation occurs at practically the same temperatures for both the amorphous and bcc states and the final phases are the same (fcc-Ni + ZrNi 5 ). The transformation temperatures determined from thermomagnetic measurements agreed well with the corresponding DSC data and the magnetization changes accompanying the phase transformation could be well interpreted on the basis of DSC and TEM results.

The crystallization processes of MgO-B 2 O 3 -SiO 2 slag have been investigated using two methods of heat treatment. Phase separation, which is similar to grape clusters, was observed when the amorphous slag was heat-treated at 800 °C, however, dendrites which are found to be 2MgO · B 2 O 3 , were formed at 850 °C. The micrographs of phase separation and the dendrites show fractal characteristics and their fractal dimensions were calculated using a density-density correlation function to be 1.90 and 1.63, respectively. In order to reveal their formation mechanisms, phase separation and the growth for 2MgO · B 2 O 3 were simulated using the Monte-Carlo method based on spinodal decomposition, nucleation and growth, diffusion-limited aggregation (DLA) and random rain (RR) models in a computer. The simulation results show that phase separation corresponds to nucleation and growth, but the growth for 2MgO · B 2 O 3 depends on the DLA.

The phase equilibria of the ternary Al-AlCo-AlNi system were investigated around the decagonal phase. Isopleths with 70, 71.5 and 72.5 at.% Al and with 10 and 13 at.% Ni are elaborated with the aid of differential thermoanalysis, magnetothermal analysis, optical microscopy and transmission electron microscopy. These are based on results of samples of more than 80 different compositions and on data from other isopleths.

The change of the Gibbs energy of the reaction between group IIIA and VA elements to form a liquid III-V alloy was derived for the Ga-As, In-As, Ga-P and In-P systems by combining the heat content of the alloy with the heat and entropy of formation of the equiatomic compound at a reference temperature. The obtained data were analysed by using a Redlich-Kister polynomial equation and the optimized parameters were derived for a temperature range between 873 and 1523 K and a composition range between x v = 0 and 0.5 (for the Ga-P system between x p = 0 and 0.1). The partial pressures of these liquid alloys were expressed as a function of alloy composition and temperature. Partial pressure and total pressure-composition-temperature diagrams were constructed.

The Dy–Fe–Nd phase diagram at 500 °C was determined with the use of powder X-ray diffraction (XRD), electron probe microanalysis (EPMA), metallography and DTA.The 500 °C isothermal section of this ternary system consists of 10 single-phase regions, 17 two-phase regions and 8 three-phase regions. No ternary compounds have been found. At 500 °C the maximum solid solubilities of Nd in DyFe 3 and DyFe 2 and that of Fe in α-Nd(Dy) are about 1 at.%, 3 at.% and 0.1 at.%, respectively. The existence of the metastable phase NdFe 2 and δ phase of the Dy–Nd binary system are confirmed, but no evidence of the existence of Nd 3 Dy 2 was found. The metastable phase NdFe2 appears in the phase regions which include the stable phase DyFe2 but was not found in the alloys near the composition of NdFe 2 .

The paper presents a review of the main theoretical formulations of the analysis of steep stress gradients by X-ray diffraction and a comparison of calculations with experimentally applied stresses in a four-point bending test on a beryllium specimen. The comparison between theoretical approaches and experiments is therefore limited to uniaxial linear stress gradients. It shows that, in this case, only an evaluation of stress gradient can be obtained from the analysis of X-ray diffraction measurements even when experiments are performed very carefully. In this direction, very complex formulations could often not be necessary. Then, a summary of advantages and drawbacks of each theoretical approach is presented. Finally, for different forms of stress gradients, the paper proposes the most convenient formalism and experimental technique to be used.

A method is presented which applies state of the art nonlinear denoising techniques on measurement data consisting of counts per channel, e. g. EDS line scans. The development of denoising techniques has made great progress during the last decade due to the application of non-linear techniques based on wavelet transformation. Several breakthroughs in this field have made it possible to almost eliminate the noise of measurement data with minimal effect on the signal and without the introduction of artefacts. In particular, sharp features in the original data stay sharp in reconstruction.

Samples of Zr-2.5 wt.% Nb alloy, obtained from pressure tube, were annealed in the temperature range 700– 900 °C up to 100 h to investigate the grain growth behaviour. The grain size increases with annealing time ( t ) and temperature ( T) according to the relationship d ∈ t 0.21 exp((– 387 kJ/mol)/RT). Room temperature hardness increases with the increasing annealing temperature and grain size, which is ascribed to the transformation of β- phase into martensitic α' and the substructure developed during quenching. The grain sizes, produced by further cold working and annealing for short time (< 1 h), and the corresponding hardness values, however, support the Hall-Petch type relationship in the limited cases.