For Fe69Ni30Cr1 droplets processed by electromagnetic levitation, the crystallographic texture is strongly correlated with the microstructure seen in optical metallography and with the undercooling at which solidification was initiated. Systematic texture analysis is therefore useful in understanding the development of the microstructure and in particular the mechanism of grain refinement. For initially grain-refined samples, differences in the orientation spread can be attributed to solid-state coarsening.

The techniques of electromagnetic levitation, drop-tube processing and atomization are used to study containerless solidification of the Cu-Ni and Fe-Ni systems. Microstructure-selection maps are developed for the case of droplet solidification where droplet diameter and undercooling are the process parameters. The microstructure varies systematically with undercooling and composition, such that grain-refined microstructures are produced at both low and high undercoolings and dendritic microstructures are produced at intermediate undercoolings. The results are analyzed within a recently developed model for grain refinement which is based on the fragmentation of primary dendrites. The role of cooling rate in determining whether the microstructure transforms in the solid state or is retained as a primary structure is explored. Keywords: Cu-Ni; Dendrites; Droplet solidification; Fe-Ni; Grain refinement; Microstructure-selection maps

The electromagnetic levitation method was applied to conduct in-situ observations of phase selection processes in undercooled metal drops. By means of a high-speed photosensing diode with a sampling rate of 1 MHz focused onto two adjacent areas of a droplet surface, the passing of a solidification wave front can be recorded and thereby the growth velocities calculated. The alloys studied were Fe69Cr31−xNix where x was varied between 12 and 24, and equilibrium are f.c.c. structure (austenitic phase) but when undercooled sufficiently nucleate the b.c.c. (ferritic) phase. The dendrite growth velocities showed a clear break at the transition from the b.c.c. to the f.c.c. profile of the respective growth velocity-undercooling curves. A two-fold explanation is found for the clear break; firstly there is a contrast between the lower undercooling f.c.c. phase which travels relatively faster for the same undercooling. Secondly, the liquidus temperatures for the two phases are offset so that for the metastable b.c.c. phase, the b.c.c.-f.c.c. liquidi temperature difference increases significantly with increasing nickel content. The result is two easily identifiable curves for the separate phase growth modes; f.c.c. at lower undercoolings and metastable b.c.c. at higher undercoolings. It was also found that these alloys show a trend towards increasing undercooling necessary to nucleate the metastable b.c.c. phase with increasing nickel content. A number of nucleation theories have been applied to explain this behaviour, namely Classical Nucleation Theory (CNT) and Diffuse Interface Theory (DIT). Two recent forms for the latter DIT theory have been proposed and both were applied to this system. The results show that by optimising the nucleus composition all predictions were seen to lie much closer to the experimental results. The DIT theory by Gránásy was found to give the closest agreement to experiment after optimisation of the nucleus composition. © 1997 Acta Metallurgica Inc.

The flow field of a free fall gas atomiser has been measured in three dimensions using hot wire anemometry and a pitot tube for confirmation of axial flow speeds. A finite difference based fluid dynamics package has then been used to model the gas flow from the atomiser geometry. Overall a close fit for the velocity magnitudes was obtained for predictions parallel to and along the central axis of the atomiser. An exception was a broader but lower peak along the central axis. Hot aluminium droplets of a distribution reproduced from experimental atomisation were injected into the gas flow at the confluence of the gas jets along the central axis. The modified gas flow pattern is specifically altered, when compared to the original single phase flow, close to the axis where the droplets travel. The resulting mass flowrate at the axial peak flow corresponds to that expected from the law of conservation of momentum. (C) 1996 The Institute of Materials.

Ni–Cu droplets were processed by electromagnetic levitation, in a drop-tube and by atomization. The droplet diameters varied from 30μm to 7mm, and the entire composition range was studied. Most microstructures observed were coarse-dendritic or refined-equiaxed. Microstructure-selection maps have been constructed for the drop-tube and atomization techniques showing the predominant microstructure as a function of composition and droplet size. From the results on levitated droplets, a selection map has been constructed which relates the microstructure to the composition and the melt undercooling reached prior to nucleation. The results are analysed within a recently developed model for grain refinement which is based upon the fragmentation of primary dendrites. The influence of impurities on the microstructure formation in material of commercial purity is investigated.

The gas flow field of a free-fall atomizer was measured in three dimensions using hot wire anemometry and corresponding measurements were taken using a Pitot tube to confirm axial flow velocities. The measurements show good agreement with predictions of fluid flow modelling. 2014 aluminium alloy spray droplets collected in water or cooled methanol at the position of a spray deposition substrate, under typical atomization conditions, show a dendrite cell size typical of gas-quenched droplets for droplet sizes less than 50 μm, while collected droplets less than 200 μm in size show values of cell size significantly smaller than gas-quenched droplets of the same size.