Volume 96, Issue 1

Interdiffusion in Fe –Al alloys was studied by electron-microprobe analysis of the interdiffusion profiles for three series of Fe x Al 1– x /Fe y Al 1– y diffusion couples. Temperatures between 997 and 1447 K and Al contents between 18.0 and 49.5 at.% Al were investigated. Interdiffusion coefficients were obtained from the diffusion profiles by a modified Boltzmann – Matano method taking into account the composition dependence of the molar volume. Al tracer diffusivities for concentrations of 25.5, 33.0, and 48.0 at.% Al were deduced from the interdiffusion coefficients with the help of the Darken – Manning equation using Fe tracer diffusivities available from a previous study in our laboratory and thermodynamic factors calculated on the basis of theoretical models, the latter being verified by an analysis of the Kirkendall effect of the interdiffusion couples. Al self-diffusion is found to be slightly faster than Fe self-diffusion and is comparable to the diffusion of the 65 Zn and 114 m In radiotracers, which substitute Al on the Al-sublattice of B2 Fe –Al and were used in the previous study from our laboratory to approximate Al self-diffusion. Die Interdiffusion von Fe – Al Legierungen wurde an Interdiffusionsprofilen dreier Serien von Fe x Al 1– x /Fe y Al 1– y Diffusionspaaren mittels Elektronenstrahl-Mikrosonde analysiert. Die Untersuchung umfasste Temperaturen im Bereich von 997 bis 1447 K und Al-Gehalte von 18,0 bis 49,5 At.%. Aus den Diffusionsprofilen wurden Interdiffusionskoeffizienten mit Hilfe einer modifizierten Boltzmann – Matano–Methode ermittelt, welche die Konzentrationsabhängigkeit des Molvolumens berücksichtigt. Aus diesen Interdiffusionskoeffizienten wurden Al Tracer-Diffusionskoeffizienten für die Zusammensetzungen 25,5, 33,0 und 48,0 At.% Al durch Anwendung der Darken–Manning Gleichung errechnet. Dazu wurden Fe Tracer-Diffusionskoeffizienten aus einer vorherigen Untersuchung in unserer Arbeitsgruppe sowie thermodynamische Faktoren aus theoretischen Modellen verwendet. Letztere wurden anhand einer Analyse des Kirkendall-Effektes an den Diffusionspaaren verifiziert. Die Selbstdiffusion von Al ist geringfügig schneller als die von Fe und vergleichbar mit der Diffusion von 65 Zn und 114 m In Radiotracern, welche Al auf dessen Untergitter der B2 Fe –Al Legierungen substituieren und in einer früheren Untersuchung in unserer Arbeitsgruppe verwendet wurden, um die Al Selbstdiffusion näherungsweise zu bestimmen.

In the five-component permanent magnets based on Sm 2 Co 17 the formation of the familiar microstructural features (2 : 17 cell, 1 : 5 cell wall, and 1 : 3 lamellae) are elucidated with the help of recent and past information on the phase equilibria and crystal structures. Contrary to Zr-free alloys, in which the 1 : 5 cell-wall phase can form directly, the quinary alloys, to attain peak magnetic coercivity, undergo a transformation within the cell walls from undesirable 2 : 7 + 5 : 19 to high-anisotropy 1 : 5. A dramatic increase in the Cu content of the cell wall accompanies this transformation.

The following unified equation has been elaborated in the present paper, to describe the viscosity of all liquid metals as a function of temperature: ηi=A⋅Mi1/2Vi2/3⋅T1/2⋅exp(B⋅Tm,iT) $${\eta _i} = A \cdot {{M_i^{1/2}} \over {V_i^{2/3}}} \cdot {T^{1/2}} \cdot \exp \left( {B \cdot {{{T_{{\rm{m}},i}}} \over T}} \right)$$ with η i , M i , V i , T m, i being the dynamic viscosity, atomic mass, molar volume and melting point of the given metal i , and T is temperature. The above equation was tested on 101 measured points of 15 selected liquid metals, and the average values of the generally valid parameters were found as: A = (1.80 ± 0.39) · 10 –8 (J/Kmol 1/3 ) 1/2 , B = 2.34 ± 0.20. Based on these parameters, the temperature dependence of viscosity was estimated for 32 liquid metals. The above equation was derived by (i) combining Andrade’s equation with the activation energy concept, and (ii) by combining Andrade’s equation with the free volume concept. It is shown, that the activation energy and the free volume concepts have identical roots and lead to identical results. The above equation is shown to be valid for liquid semi-metals (Si, Ge, Sb, Bi), if their actual melting points are replaced by their corrected melting points, corresponding to (unstable) metallic solid crystals. The ratio of viscosity to surface tension of pure liquid metals is discussed, as well.

The thermodynamic properties of Mg 2 Zn 3 were investigated by calorimetry. The standard entropy of formation at 298 K, ΔfS298∘${{\rm{\Delta }}_{\rm{f}}}S_{298}^ \circ $, was determined from measuring the heat capacity, C p , from near-absolute zero (2 K) to 300 K by the relaxation method. The standard enthalpy of formation at 298 K, ΔfH298∘${{\rm{\Delta }}_{\rm{f}}}H_{298}^ \circ $, was determined by solution calorimetry in hydrochloric acid solution. The standard Gibbs energy of formation at 298 K, ΔfG298∘${{\rm{\Delta }}_{\rm{f}}}G_{298}^ \circ $, was determined from these data. The results obtained were as follows: ΔfH298∘(Mg2Zn3)${{\rm{\Delta }}_{\rm{f}}}H_{298}^ \circ \left( {{\rm{M}}{{\rm{g}}_2}{\rm{Z}}{{\rm{n}}_3}} \right)$ = –69.80 ±20 kJ · mol – 1 ; ΔfS298∘(Mg2Zn3)${{\rm{\Delta }}_{\rm{f}}}S_{298}^ \circ \left( {{\rm{M}}{{\rm{g}}_2}{\rm{Z}}{{\rm{n}}_3}} \right)$ = –10.95 ± 1.80 J · K –1 · mol –1 ; ΔfG298∘(Mg2Zn3)${{\rm{\Delta }}_{\rm{f}}}G_{298}^ \circ \left( {{\rm{M}}{{\rm{g}}_2}{\rm{Z}}{{\rm{n}}_3}} \right)$ = –66.55 ± 20 kJ · mol – 1 . The electronic contribution to the heat capacity of Mg 2 Zn 3 was found to be similar to pure magnesium, indicating that the density of states in the vicinity of the Fermi level follows the free-electron parabolic law.

Using a Calvet type calorimeter, the enthalpy of mixing in the two ternary systems, Al –Sn–Zn and Ag–Sn–Zn was determined at 973 K. In the Al –Sn –Zn system, the investigations were carried out along three cross-sections with constant molar ratios of Sn : Zn = 1 : 2, 1 : 1 and 2 : 1 and an aluminium content between 0 and 100 at.%. For the Ag–Sn –Zn system, the measurements have been done at constant molar ratios of Ag : Sn = 1 : 3, 1 : 1 and 3 : 1 with zinc contents between 0 and 50 at.%. The substitutional solution model of Redlich – Kister according to the CALPHAD method was used for a least-square fit of the experimental data in order to get an analytical expression for the integral enthalpy of mixing. The ternary extrapolation models of Kohler, Muggianu, and Toop were applied on the binary data to compare measured, fitted, and extrapolated values.

In order to clarify the effects of the ageing temperature and the Mg content on precipitation of the β′- and β-phases in Al –Mg alloys, Al-6, 7, 8 and 9 wt.% Mg alloys were solution treated either at 450 °C and isothermally aged at temperatures between 150 and 325 °C. Precipitation of these phases was investigated in detail by optical microscopy and by the measurements of electrical conductivity and micro- Vickers hardness. It was found that, in these alloys, nucleation of precipitates occurred much more rapidly at grain boundaries than within the grains at all ageing temperatures. It was always after grain boundary precipitation had completed that precipitation within the grains began to take place. Time Temperature Transformation curves were therefore constructed both for grain boundary precipitation and for precipitation within the grains. These TTT curves showed that, in all cases, both types of precipitation occurred most rapidly at ageing temperatures between 250 and 275 °C. At ageing temperatures above these nose temperatures, precipitation occurred mainly by rapid growth of a limited number of precipitates, whereas at ageing temperatures below these nose temperatures, precipitation took place by the nucleation of numerous fine precipitates and by their subsequent slow growth. The increase in the Mg content not only accelerated precipitation, but also it increased the amount of precipitation within the grains.

Conventional pressure die casting was used to produce ZA 27 alloy and zinc –aluminum composites with alumina short fibers (Saffil TM ). While the Brinell hardness of the matrix alloy increased with increasing Saffil fiber content, on the other hand the tensile strength decreased. Creep tests were carried out on the 10%– 30% volume Saffil fiber reinforced ZA 27 matrix alloy and fiber-free ZA 27 alloy at 90 °C, 120 °C, 150 °C and stress values of 20 MPa, 40 MPa, 60 MPa. The results of ZA 27 alloys with Saffil fibers showed higher creep resistance than fiber free ZA 27 alloy. Among the MMCs, composite with fiber content 0.30 showed the highest creep resistance. The average stress exponent is 4 and 5.92 and average creep activation energy is 66.7 kJ/mol and 96.4 kJ/mol for ZA 27 and the MMCs, respectively.

The deformation and fracture behavior of polycrystalline Al 2 O 3 and Al 2 O 3 /Nb/Al 2 O 3 composites with perpendicular and inclined interfaces were investigated at room temperature by means of compression tests. It was found that polycrystalline Al 2 O 3 fractured at higher stresses (>500 MPa) than Al 2 O 3 /Nb/Al 2 O 3 composites, and that its failure always showed a split fracture mode. For the Al 2 O 3 /Nb/ Al 2 O 3 specimens with interfaces perpendicular to the loading direction, cracks normally nucleated near the Nb/ Al 2 O 3 interfaces and then propagated along the loading direction, finally leading to a split fracture in the polycrystalline Al 2 O 3 parts. For the Al 2 O 3 /Nb/Al 2 O 3 specimens with interfaces inclined to the loading direction, the split fracture of the polycrystalline Al 2 O 3 parts strongly depends on the orientation of the Al 2 O 3 /Nb interfaces with respect to the loading direction. For all the Al 2 O 3 /Nb/Al 2 O 3 specimens, intergranular cracking was seldom observed independent whether the interfaces are perpendicular or inclined to the loading direction. Based on these results, the fracture mechanisms of Al 2 O 3 and Al 2 O 3 /Nb/Al 2 O 3 specimens are discussed.

The hot-deformation behavior of oxygen-free high-conductivity (OFHC) copper has been studied in the temperature range 300–950 °C and strain rate range 0.001–100 s –1 using hot-compression tests. The dynamics of the behavior is studied with the help of processing maps, and the-rate controlling processes are evaluated using the standard kinetic rate equation. The processing maps exhibited two domains: (1) in the temperature range 600–900 °C and strain rate range 0.01–10 s –1 , (2) in the range 750–950 °C and 30–100 s –1 . From kinetic analysis of the flow stress data, apparent activation energy values of 159 kJ/mole and 75 kJ/mole have been estimated in the above two domains, respectively. These values suggest that dislocation core self-diffusion and grain boundary self-diffusion are the rate-controlling mechanisms. In both these domains, the average grain diameter varies linearly with the Zener–Hollomon parameter.

Light and scanning electron microscopy (SEM), X-ray diffraction analysis, hardness and uniaxial tensile tests have been applied to study the mechanism and kinetics of aging of the Cu-5 wt.% Ni-2.5 wt.% Ti alloy under the influence of static and dynamic conditions. Hardening of alloy during aging is a consequence of precipitation of (Ni, Cu) 3 Ti secondary particles with the hcp crystal structure, i. e., the same as the η phase (Ni 3 Ti). During static (SSA) and dynamic strain aging (DSA), the strength of the alloy is increased compared to quench aging (QA). This effect is most pronounced during the dynamic process. The total concentration of vacancies participating in the process of DSA is increased compared to the vacancy concentration during SSA and QA. This may be explained by the fact that vacancies are not annihilated at dislocations, but their concentration is preserved through the vacancy – precipitate reaction.

An Al-10Ni-8Y (at.%) alloy was atomized by Ar gas atomization in a close-coupled nozzle atomizer. The microstructure, phase composition, and thermal stability of the as-atomized powder were investigated by scanning electron microscopy, transmission electron microscopy, X-ray diffraction analysis and differential scanning calorimetry. Activation energies for different crystallization reactions were calculated using both Kissinger and Ozawa methods. The supercooled liquid region Δ T x (= T x – T g , where T x and T g are crystallization onset temperature and glass transition temperature, respectively) was detected to be in the range of 14 –24 K. Also presented were attempts to determine the critical particle size for amorphous structure.

A model based on the existence of self-associates has been used to compute bulk thermodynamic properties of Bi –Cd, In –Pb, and Ni – Pd binary liquid alloys. Also, a statistical mechanics model has been employed to calculate the surface properties of these three alloys. Our study reveals that the degree of phase separation increases in the order Bi –Cd, In–Pb and Ni –Pd. Essentially, we find that Bi –Cd, like Au–Bi and Bi –Sb alloys, does not belong exclusively to the class of either segregating systems or to the short-range ordered alloys. Furthermore, we were able to predict the temperature dependence of the surface properties for our binary liquid alloys based on the knowlege of the results at their melting points.