화학공학소재연구정보센터
Minerals Engineering, Vol.19, No.15, 1573-1579, 2006
Enhanced separation of mineral sands using the Reflux Classifier
The Reflux Classifier consists of a conventional fluidized bed, with a set of parallel inclined plates immediately above. The fluidized suspension passes into the inclined channels where relatively fast settling particles segregate, slide down the plates and return to the fluidized bed. The slower settling particles pass up through the inclined channels and into the overflow. By increasing the aspect ratio of the inclined channels, the effective sedimentation area of the vessel increases, in turn increasing the hydraulic capacity of the device. Laskovski et al. [Laskovski, D., Duncan, P., Stevenson, P., Zhou, J., Galvin, K.P., 2006. Segregation of hydraulically suspended particles in inclined channels. Chem. Eng. Sci., doi:10.1016/j.ces.2006.08.024], however, have shown that there exists an asymptotic limit to this hydraulic capacity, due to an increased tendency for particle re-suspension within the inclined channels as the aspect ratio increases. Laskovski et al. (2006) showed that the re-suspension facilitates the separation of the particles on the basis of density, reducing the dependence on the particle size. A conventional fluidized bed does not offer the benefits of the mechanism identified by Laskovski et al. (2006). In this study, the Reflux Classifier was used to recover and concentrate the heavy minerals presented in a low grade feed of mineral sands. Enhanced separation was achieved using inclined channels having a large aspect ratio of about 200, thus promoting the particle segregation of the denser particles while also promoting the re-suspension and hence hydraulic conveying of the lower density silica sand. With the aspect ratio of approximately 200, recoveries of heavy minerals in excess of 90% were achieved at solids throughputs of up to 40 t/m(2)h. A heavy mineral recovery of 97% was achieved at a solids throughput of 21 t/m(2)h. The separations resulted in heavy mineral grades approaching 100% in the size range of 90-180 mu m. (C) 2006 Elsevier Ltd. All rights reserved.