Günter Frohberg

Diffusion in liquid metals

Abstract By doing experiments in space, it was possible to avoid convection in liquid diffusion experiments for the first time. Due to the improved accuracy it was possible to check experimentally different diffusion theories. Spacelab experiments revealed a square dependence of the self-diffusion coefficient on temperature for different metals (Sn, In, Pb, Sb) and for impurity diffusion and interdiffusion in the In-Sn system. For future microgravity experiments a new shear cell was developed, and shearing was investigated during parabolic flights in model experiments. The method of hydromagnetic damping of convection for ground-based liquid metal diffusion experiments was also tested.

Diffusion in ZrTiCuNiBe bulk glasses at temperatures around the glass transition

Abstract The diffusion coefficients of impurities in two glasses, Zr 41 Ti 14 Cu 12.5 Ni 10 Be 22.5 and Zr 46.75 Ti 8.25 Cu 7.5 Ni 10 Be 27.5 were measured at temperatures around the caloric glass transition temperature. They were determined for annealing times during which there was no decomposition in either of the alloys. For Co, Fe and B the data could not be reasonably fitted by a single Arrhenius relation. A larger effective activation energy is indicated above about 585 K than below for these impurities. No change was observed for the bigger Al atoms. The temperature dependence of the diffusion jump time of the impurities is compared with that of the visco-elastic relaxation of the glasses. From this comparison we suggest that the different activation energies at higher and lower temperatures indicate a change of the diffusion mechanism.

Diffusion experiments in liquid Sn-Bi and Al-Ni systems with a stable density layering using the FOTON shear cell under 1g conditions

In this study we investigated the possibility of reliable diffusion experiments in liquid metals under 1g conditions, instead of expensive μg-experiments. To minimise buoyancy convection we used thick layer diffusion from a binary alloy into apure metal. This can provide a stable density layering, which we have shown to be an important factor for successful 1g-experiments. To avoid the segregation problem and to minimize free surfaces a shear cell was used, which was specially developed for the mission FOTON-M2 and was equipped with reservoirs providing pressure on the liquid samples. Thick layer diffusion experiments from SnBi2.5wt% into Sn and from AlNi3.5wt% into pure Al were performed at 300°Cfor 8h and at 730°Cfor 5h respectively. For each set-up four parallel experiments were performed at the same time. The concentration profiles were obtained by AAS (atom absorption spectroscopy) and the diffusion coefficients were evaluated by fitting with the thick layer solution. For the evaluation a correction method was used for the shear convection and the AAS averaging effect inside a cell. As a result, the obtained concentration curves agreed well with the fitting function. The diffusion coefficients DBi=2.35x10− 9m2/s andDNi=3.81x10− 9m2/s agreed well within the error range with the μg-reference data obtained in the FOTON-12 mission and reference data obtained in the 1g diffusion experiment in a magnetic field. The reproducibility of the diffusion coefficients among four parallel experiments was very good with a standard deviation among four capillaries smaller than 3.1% including the standard temperature deviation. From these results we conclude that buoyancy convection was practically absent and thus the applied method was very effective.

Effect of reversible structural relaxation on diffusion in bulk metallic glasses

Abstract Long time relaxation at temperatures below the calorimetric glass transition causes reversible structural changes in metallic glasses. The resulting enthalpy recovery was measured by means of DSC in Zr 46.8 Ti 8.2 Cu 7.5 Ni 10 Be 27.5 - and Pd 40 Cu 30 Ni 10 P 20 -bulk glass annealed for different times at 553 K and 542 K, respectively. Relaxation times of about 10 6 s and 10 4 s, respectively, were determined. The diffusion coefficients of B, Fe and Co were measured above and below the calorimetric glass transition temperature. Whereas the temperature dependence of these diffusion coefficients in the non-relaxed glasses shows “non-linear” Arrhenius behaviour with a break near the glass transition, the diffusion in the long time relaxed glasses follows a uniform temperature dependence over the entire temperature range with considerably reduced diffusion coefficients below the glass transition. This behaviour can be reversed by annealing the relaxed glasses again at higher temperatures indicating the strong effect of the reversible structural relaxation on the diffusion coefficients.

Effect of reversible structural relaxation on diffusion in a ZrTiCuNiBe bulk glass

Abstract Diffusion coefficients of B and Fe were measured in the Zr 46.8 Ti 8.2 Cu 7.5 Ni 10 Be 27.5 bulk glass which was relaxed at 553 K for up to nine months. While above 600 K the diffusion coefficients in the relaxed and as-cast glass were equal, at and below 553 K the diffusion coefficients in the relaxed glass were significantly lower than in the as-cast glass.

Shear cell development for diffusion experiments in foton-satellite missions and on the ground with consideration of shear-induced convection

A shear cell technique was developed to obtain exact diffusion data. The shear cell in this study was designed for the utilization under μg-conditions, especially in the FOTON-M2 mission, but also under 1g-conditions. To minimize the influence of the shear convection, the cell size, the rotation system and the speed of the discs were optimized. To minimize free surfaces, which can cause Marangoni convection, a reservoir system providing pressure on the liquid was introduced. Using this FOTON shear cell we performed short-time diffusion experiments in the In-Sn system in a parabolic flight and under 1g conditions to investigate the influence of the shear convection quantitatively. As a result, the influence of the shear convection was so small that the mean square diffusion depth caused by the shear convection was in the order of10− 7m2, which is smaller than 1% of the typical value Xdiff2 ≈ 10− 4m2 in a standard diffusion experiment using the FOTON shear cell. By using this result a correction method for the evaluation of the diffusion coefficient was established. In several ground experiments, the FOTON shear cell showed the same diffusion coefficients as from μg reference experiments within the range of errors and no obvious indication of Marangoni convection was detected. From these results we confirmed that the FOTON shear cell can be applied to μg-experiments and ground-based experiments as well.

Ground-Based Diffusion Experiments on Liquid Sn-In Systems Using the Shear Cell Technique of the Satellite Mission Foton-M1

Abstract: This study reported in this paper was aimed at testing the shear cell that was developed for the satellite mission Foton‐M1 to measure diffusion coefficients in liquid metals under microgravity (μg)‐conditions. Thick Layer diffusion experiments were performed in the system Sn90In10 versus Sn under 1 g‐conditions. For this system several μg‐diffusion results are available as reference data. This combination provides a low, but sufficiently stable, density layering throughout the entire experiment, which is important to avoid buoyancy‐driven convection. The experimental results were corrected for the influences of the shear‐induced convection and mixing after the final shearing, both of which are typical for the shear cell technique. As the result, the reproducibility and the reliability of the diffusion coefficients in the ground‐based experiments were within the limits of error of μg‐data. Based on our results we discuss the necessary conditions to avoid buoyancy‐driven convection.

Chemical Diffusion Experiments in AlNiCe-Melts

The long-capillary method was used to measure chemical diffusion in molten AlNiCe alloys. The interdiffusion coefficients were determined for a mean concentration of Al87Ni10Ce3 at 1273 K and for a mean concentration of Al77Ni20Ce3 at 1373 K. The absence of major convection disturbances and of macro-segregation was demonstrated by time-dependent diffusion measurements. An in-situ x-ray monitoring technique for real-time concentration profile determination is presented.

Determination of γ/γ interface types in a γ-TiAl alloy using convergent beam electron diffraction

Abstract The domain orientations and lamellar interfaces play a critical role in the mechanical properties of TiAl based alloys. An umambiguous experimental determination of all six domain orientation in lamellar structure and the interfaces between neighboring γ laths in a (α 2 +γ) two-phase TiAl alloy containing Mn and Nb were analyzed by means of convergent-beam electron diffraction. The frequency of different interface types are reported.

A modified shear cell for mass transport measurements in melts

A graphite shear cell had been developed to measure diffusion coefficients in metal or semiconductor melts. The usable temperature range reaches from the melting point of the material to limits like chemical reaction temperatures with the graphite crucible, respectively, to the vaporization temperature of the liquid. As opposed to existing cells the improvement is in the simple initial shear to start diffusion and the “comb-shearing” process to terminate the diffusion experiment. These mechanisms minimize convection in the melt initialized by the shear movement itself. The reliability of the cell had been verified in diffusion experiments in the system In–Sn, for which microgravity data (without an additional contribution due to gravity-driven natural convection) are present. In addition the mass transport of the pure shearing process had been measured. It was found to be negligible for typical diffusion experiments with t>10 000 s.