The occurrence of aligned microstructures in directionally solidified aluminum-bismuth alloys

Abstract

In a recent paper, Parr and Johnston ~ reported that an aligned fibrous composite structure did not form in pure binary A1-Bi monotectic alloy, but that it did so with small additions (e .g., 0.2 wt pct) of iron. Examination of monotectic alloys in several other systems 2 has shown that an aligned fibrous structure is generally formed in systems where the liquid miscibility gap is high, i .e. , where there is a large temperature difference between the monotectic horizontal and the upper consolute temperature. In the A1-Bi system the miscibility gap is relatively high ( 6 0 0 K), and we would have expected that the pure binary monotectic alloys should solidify directionally to give an aligned fibrous structure; moreover, we understand from other workers 3 that they cannot confirm the results of Parr and Johnston. Therefore, we have repeated some of the experiments to resolve an apparent anomaly and contradiction. Alloy samples were prepared from pure (>99.95 pct) components by melting under argon in graphite crucibles and were subsequently frozen undirectionally in 3 mm bore alumina tubes. The experimental assembly was very similar to that used by Parr and Johnston, and the traction rate of (1 to 10 /xm s -~) and t e m p e r a t u r e g r ad i en t o f ( -15 K mm -1) at the monotectic temperature of 930 K coincided closely with their growth conditions. We found that the pure binary, A1-3.5 wt pct Bi alloy, solidified at rates below 5 /xm s -~ to give regularly spaced rods of bismuth in an aluminum matrix (Figure 1 (a) and (b). At higher growth rates the rod-like structure broke down to a fine globular dispersion of Bi droplets, and, as in other systems, 2 there was a small range of rate close to 5 / ~ m s -1 where the droplet arrangement was periodic with the plane of the solidification front. We did not observe evidence of rods breaking down to aligned droplets by perturbation behind the monotectic front, and whether this happens in the course of directional growth remains something of an open question. Six monotectic alloys were then prepared with minor additions of iron (concentrations of 0.05 and 0.2 wt pct), and these were solidified at 2 .5 /xm s -~. No perceptible change in the microstructures was detected at low iron content, but with 0.2 wt pct the bismuth occurred as globules and as larger, less regular, and widely spaced rods. A sample quenched during growth revealed (Figure 2) that the growth front was just cellular and the bismuth collected at some of the cell boundaries, rather than forming as fine evenly spaced fibers.