Abstract
Data on penetration of particles through single screens in a 0.5 m wide multiple-deck probability screening machine (Mogensen Sizer), using coal and sand—gravel mixtures, were correlated as a function of the operating variables. The correlations were successfully used to predict the output from the machine with multiple decks in place, both in the laboratory and in a field trial at a coal mine. Misplacement of fines due to adherence to larger particles was found to occur and to depend on the nature of the material and on its moisture content. This factor was not taken into account in the correlations and will lead to very significant errors in some circumstances.