Inductive System for Reliable Magnesium Level Detection in a Titanium Reduction Reactor
- PDF / 2,268,387 Bytes
- 8 Pages / 593.972 x 792 pts Page_size
- 5 Downloads / 158 Views
I.
INTRODUCTION
THE Kroll process is presently the main road for the production of Titanium.[1,2] It consists in the metallothermic reduction of titanium tetrachloride (TiCl4), using liquid magnesium as reductant. After the molten magnesium is filled into the reactor, which is situated in a furnace to provide temperature control, the TiCl4 is poured in from the top to react at the surface of the molten magnesium according to the Kroll reduction reaction TiCl4(g) + 2Mg(l) fi Ti(s)+2MgCl2(l). This exothermic reaction takes place over a period of few days. The product is titanium sponge with high porosity which has to be processed further (e.g., by distillation) to remove residual magnesium and MgCl2. At intervals of a couple of hours, the MgCl2 is tapped from the bottom of the reactor. The smooth performance of the process and the final titanium yield and quality are strongly dependent on the control of temperature and reaction rate, and the formation of lower chlorides.[3] Usually, the molten magnesium is supposed to float on the top surface of the molten MgCl2, thereby having contact with the inflowing TiCl4. However, one of the
NICO KRAUTER, SVEN ECKERT, THOMAS GUNDRUM, FRANK STEFANI, and THOMAS WONDRAK are with Helmholtz-Zentrum Dresden-Rossendorf, Bautzner Landstr. 400, 01328, Dresden, Germany. Contact e-mail: [email protected] PETER FRICK, RUSLAN KHALILOV, and ANDREI TEIMURAZOV are with the Institute of Continuous Media Mechanics, Akademika Koroleva 1, Perm, Russia, 614013. Manuscript submitted November, 10 2017.
METALLURGICAL AND MATERIALS TRANSACTIONS B
technological problems of the Kroll process is the occurrence of a so-called non-separation regime, in which the MgCl2, which is only slightly heavier than Mg, ceases to settle at the bottom of the reactor. This regime can gravely hamper the proper continuation of the process.[4,5] At present, up to 5 pct of the production cycles are rejected due to contingencies associated with the disruption of the magnesium chloride settling process and local retort overheating. For the general control of the Kroll process, and in particular for the identification of undesired regimes, various measurement techniques have been utilized. Besides the use of thermocouples at the rim of the retort, various inductive methods have been proposed to identify the position of the upper surface of the liquid Mg, whether it be in direct contact with the inflowing TiCl4 (in the desired regime), or whether it be covered by MgCl2 (in the undesired ‘‘non-separation’’ regime). These inductive methods, as proposed in various Russian patents,[6–8] generally rely on the determination of mutual inductances between coils of various types and positions which are sensitive to the level of the molten Mg[9] (note that mutual inductance methods are widely used in other fields of metallurgy, too[10–12]). While generally successful during the initial stages of the process, the inductive methods turned out to run into problems at later stages. The reason for that failure was identified in the formation of cert
Data Loading...
