Band 89, Heft 7

In this paper we report on the 71 Ge diffusion in Al–(Si,Ge) solid-solution alloys with seven different microalloying concentrations. The temperature, composition, and pressure dependence of diffusion and the pertaining activation enthalpies, enhancement factors, and activation volumes have been determined within the solid-solution phase field. The radiotracer method in combination with serial sectioning by microtome was employed in our experiments. Oxide hold-up effects of the isotope at the surface, which are often encountered in tracer diffusion studies of Al and of Al alloys, have been overcome either by sputter-cleaning of the diffusion samples before evaporating a thin 71 Ge layer or by implantation of 71 Ge. Diffusion anneals were either performed at ambient pressure or in a high-pressure cell. Like in pure aluminium Ge is a ‘normal’ diffuser in both binary and ternary solid solutions. Its activation enthalpies are slightly smaller than those of self-diffusion of the Al solvent. Activation volumes of about one atomic volume are observed which are typical of vacancy-mediated diffusion of ‘normal’ foreign atoms in Al. Solute additions slightly enhance the Ge diffusivity in both binary and ternary solid solutions by similar amounts. The results are discussed together with literature data on the temperature and pressure dependence of the diffusion of various solute atoms in pure Al.

The influence of the microstructure on the fatigue properties of the high-strength beta titanium alloy β-CEZ was investigated by comparing three distinctly different microstructures: a lamellar (β annealed), a bi-modal (α + β processed), and a necklace microstructure (processed through the β transus). Comparing the three microstructures at the same yield strength level of 1200 MPa, the bi-modal microstructure showed the best HCF and LCF properties. Concerning fatigue crack propagation, small surface cracks (microcracks) exhibited the slowest propagation rate in the bi-modal microstructure whereas the propagation rates of large cracks (macrocracks) were similar in the three microstructures, with the tendency that the lamellar microstructure had a slightly higher resistance against macrocrack propagation. The fatigue properties can be explained firstly by differences in the β grain size. In addition, the crack propagation behavior of macrocracks is influenced by the dimension of the α plates as a second parameter.

The Mg-rich phase equilibria in the Mg–Sc binary system have been established by experiments and thermodynamic modelling. The (Mg)-βSc two-phase region is of special interest, because the shape of the diagram proposed in the literature is thermodynamically not possible. A series of experiments have therefore been defined based on preliminary thermodynamic calculations. The c/a ratio of the lattice parameters within the (Mg) solid solution decreases monotonically. The measured liquidus temperatures of selected samples confirmed the literature values. The final thermodynamic modelling reproduces well the equilibria in the Mg-rich part of the Mg–Sc system. Our experimentally supported and calculated phase diagram differs considerably from the proposed diagram in the literature. Open questions still remain concerning the solid state equilibria in the Sc-rich part, where further experimental work is needed.

The thermal expansion coefficients for Al and TiB 2 have been determined by refining the unit cells measured by means of high-temperature powder X-ray diffraction. The lattice expansion coefficients were found to be linear within the measured temperature range 300 to 900 K: 2.79 · 10 –5 K –1 for Al; 3.56 · 10 –6 K –1 for TiB 2 in the a -axis direction and 6.13 · 10 –6 K –1 for TiB 2 in the c -axis direction. These data have been used to calculate the lattice mismatch between the two phases at the melting point of aluminium. This is of interest since TiB 2 particles often are added to Al melts as a nucleating agent in order to induce grain refinement. It was found that the fit is better at the melting temperature of aluminium than at room temperature.

Temperature and strain-amplitude dependent variations of logarithmic decrement ( δ ) in a high damping Mn-20Cu-5Ni-2Fe (at.%) alloy are compared for 4 different processing conditions. The primarily cold-rolled sample shows δ above 0.08 in a wide temperature range of about 100 °C, while solution treating the as-casted sample results in a strain-amplitude sensitive damping behavior, as high as 0.25 at the strain of 2.0 × 10 –4 . The chemical inhomogeneity in primarily cast alloy is reflected by the flattened changes of Young's modulus ( E ), broadened X-ray patterns, and the co-existence of twin plates and tweed microstructure. While the strain-amplitude sensitive damping behavior is related to the large fraction of Mn-rich γ Mn phase, the thin (011) twin plates, intersected with the traces of stacking faults, are considered to be the product of dislocation reaction at the conjunction of intersected (111) plane. Such twins may possess a low activation energy for cyclic movement, and therefore cause a stable damping behavior in the vicinity of room temperature.

The activities of Mg and Al in solid Pd have been measured in the temperature range 1425 to 1725 K using metal-oxide equilibration technique under controlled oxygen partial pressures. Thin foils of Pd were sandwiched between pellets of either MgO or Al 2 O 3 . Oxygen partial pressures were established using H 2 + CO 2 gas mixtures and measured using a solid state cell based on yttria-doped thoria as the electrolyte. The concentration of Mg or Al in solid Pd was determined using inductively coupled plasma – atomic emission spectrometer (ICP-AES) and electron probe microanalysis (EPMA). The activity of Mg or Al was computed from the standard Gibbs energy of formation of MgO or Al 2 O 3 and the oxygen potential established by the gas phase. The very small values for the activity coefficients (2.3 · 10 –7 γMgo $ \gamma _{Mg}^o $ < 7.4 · 10 –5 and 2.2 · 10 –10 γAlo $ \gamma _{Al}^o $ < 3.6 · 10 –8 ) indicate unusually strong interaction between the solute and the solvent, caused primarily by charge transfer from the solute to the solvent, with a smaller contribution from hybridization and covalent bonding involving s, p and d electrons.

The liquid dynamic compaction comprises two consecutive stages, namely, atomization and deposition. The size of the droplets in the atomization stage strongly affects the quality of the product, which on its turn is controlled by the atomization process parameters. This work studied through two- level fractional factorial design the effect of the liquid metal over pressure, the pressure of the atomizing gas, length and diameter of the liquid metal discharge tube, the effective area of the atomizer nozzle gas outlet and the degree of super-heating of the liquid metal on the mean droplet size. The main parameters determining the mean droplet size showed to be the pressure of the atomizing gas and the diameter of the liquid metal discharge tube. The other parameters and their interactions only have a slight effect. To obtain smaller droplets the gas pressure should be maintained high and/or the diameter of the liquid metal discharge tube should be small.

The deformation of bismuth crystals under mechanical stress and simultaneous superposition of a constant magnetic field and an impulse electric current has been investigated experimentally. Two plastifying factors have been identified. The first is a damping elastic ultrasonic vibration caused by the Ampere force. In case of the deformation by twinning, these vibrations result in an increase of the density of twinning wedges around the indentation of a diamond indentor. The second factor is an electric charge on the free surface. The electric charge on the (111) cleavage face increases the mobility of twinning dislocations.

The majority of the particles at the surface of a strip-cast Al- 0.8Fe-0.6Si alloy were identified as β-AlFeSi. There were a relatively small number of α-AlFeSi and Si particles while no evidence for binary Al–Fe phases was available. The center of the strip revealed a similar distribution of intermetallic particles with β-AlFeSi still being the dominant constituent. While Si dissolved rapidly into the aluminium matrix, the interdendritic network, mostly of β stringers, was gradually replaced by a uniform dispersion of α-AlFe- Si particles during heat treatments. The phases identified in deformed and undeformed samples after 8 hour heat treatments in the temperature range 540 to 600 °C were identical. The transformation of β-AlFeSi was accelerated, however, in those samples which have received a cold-rolling pass before the heat treatment. Finally, only the hexagonal α has prevailed in all samples after extended holding at 600 °C.

The influence of the variation of percent carbon equivalent (%CE) of spheroidal graphite (SG) cast iron on the microstructure, hardness, tensile properties, and absorbed energy during impact testing has been investigated. The matrix- phases of the sand and metal mould ingots are ferrite and/or pearlite and/or iron-carbide in the range of %CE used. The ratios of these constituent phases depend on the %CE. The curves of the variation of the tensile properties (strength and ductility) and the impact-energy showed a remarkable increase with increasing %CE from 3.78 up to 4.61 and, thereafter, decreased again with increasing %CE until a value of 5.24. The form of the curve of the variation of hardness with %CE showed inconsistency with the tensile strength curve. This point needs further investigation. The present research presents some mechanisms to explain the form of curves obtained for the variation of properties with %CE.