Calibration of Textural Models for Liberation Prediction Using Heavy Liquid Separation Results

The models for particle composition prediction that have been put forward are limited to simple textures and lead to an independance of grade with respect to particle size in broken ore. A presentation is made of a modification in texture description; binary textures, in which each subeonstituent has its own breakage characteristics are introduced. The calibration of the textural model can be made using heavy liquid separation results. The application to the case of coal is presented. Coal washability data present spe¬cial difficulty due to the distribution in the specific gravity of coal, which must be taken into account. Coal s.g. distribution is taken as a generalized exponential distribution. Model calibration is then possible and is shown to be in good agreement with experimental data.