Prediction of underflow medium density in dense medium cyclones

The prediction of the underflow medium density, ρu, in a dense medium cyclone is important, both because many empirical and phenomenological models of DMC performance incorporate ρu as an independent variable, and for plant design purposes. This paper reviews the published correlations for ρu, and presents some recent work on the subject. Three approaches have been used: empirical regression equations, correlations based on dimensional reasoning, and methods based on modelling the classification of the medium in the cyclone. The latter method has been successful in some cases, but suffers from a sensitivity to the accurate determination of certain parameters, particularly water recovery. Correlations based on dimensional reasoning give good results within a particular dataset, and show some extrapolative power, but more comprehensive datasets will be required to provide correlations of more general applicability.