ABSTRACT
This equation suggests that for the same values of K 0, u, and D, the transport of impurities from the layer into the liquid zone is controlled by the layer thickness S alone. In the general case, the dimensionless quantity vS/D is the normalized rate of ingot growth in zone remelting, since it includes the main parameters controlling the efficiency of the floating zone refining process. As a rule, D ranges from 10"5 to 10"4 cm2/s, but for most applications, the diffusion coefficient D is unknown.The solidification rate or the rate of travel v of the molten zone along the ingot is usually within 0.1 cm/s; the depth of the diffusion layer 5 is 0.1-0.001 cm and largely depends on the hydrodynamics of the liquid metal in the pool. The velocity of any liquid flow is known to vanish near a solid wall where a turbulent flow of liquid becomes laminar. Therefore, mixing in the bulk of the liquid pool has a weaker effect on the thickness of the diffusion layer than a flow of liquid formed near the solidification surface, directed normal to this surface, and responsible for the direct transport of impurity from the
(9.2)
diffusion layer into the molten zone. Attempts have been made to intensify the diffusion mixing of impurity in the liquid pool by magnetic and electric fields and mechanical stirring in this zone. However, these dynamic impacts affected mainly the stirring of the melt beyond the diffusion zone and changed the effect of liquid on the solidification front.In a joint team with Abramov, Filonenko and Teumin (1967) we investigated the effect of ultrasound on the zone melting of naphthalene (C10Hg) purposefully enriched with impurities of azobenzene (C6H5N) which turned the naphthalene bright orange. In this work, we first demonstrated that the diffusion layer should be sonicated in the precavitation mode producing acoustic streams. If the sound power exceeds the cavitation threshold, the arising cavitation phenomena destroy the interface of the solidifying ingot, thus breaking down the zone refining process. We conducted the experiment with a contact method of sonication of the liquid zone, which unfortunately is inappropriate for most metals refined by zone remelting, including aluminum, because of possible dissolution and destruction of the radiating horn in the melt.For practical purposes we developed a method of sonication through the solid part of the ingot (Eskin, 1988) in which the solidification front serves as a surface which radiates ultrasonic oscillations. In this technique, the impact of ultrasound is absolutely “sterile.”
