Anomalies in the crystal growth velocities of intermetallic alloys at the hypercooling limit

Abstract

Electrostatic levitation has been used to deeply undercool melts of intermetallic compounds prior to solidification. The solidification front has been monitored by a high-speed camera and the crystal growth velocity has been derived as a function of melt undercooling deltaT. As expected, all alloys under consideration showed an increase of the solidification velocity with increasing undercooling. Remarkable differences appeared in the range of high melt undercooling near the hypercooling limit deltaThyp, which is defined by the relation deltaThyp = deltaHf/cp. (deltaHf: heat of fusion, cp: heat capacity) [1]. CuZr showed a maximum in the solidification velocity and decreasing values for higher undercoolings. (Ni0.3Cu0.7)-Zr showed a plateau at high undercoolings and (Ni0.4Cu0.6)-Zr a large scatter in the solidification velocities in this region. A similar scatter was found in the alloy Ni-(Zr0.5Ti0.5), while NiTi exhibited further increasing velocities at high undercoolings accompanied by an increased scatter. In situ XRD measurements of the latter 4 alloys performed at ESRF (Grenoble) showed that the alloys solidified primarily with the cubic B2 structure [2]. Microstructure investigations of the solidified samples are used to correlate the observed macroscopic solidification fronts directly to the geometry of the grown dendrites and, thus, to the individual dendrite tips. This method reduces the scatter of the measured solidification velocities. At present, it is not clear whether the hypercooling limit is the cause for these anomalies or whether it is a simple coincidence. We discuss our findings in the light of existing theories for non-equilibrium solidification.