Prediction of particle composition distribution after fragmentation of heterogeneous materials

Abstract

A review of problems encountered in developing prediction models for composition distribution for particles obtained by the breakage of multiphase materials is presented. The needs to characterize texture by image analysis, to describe particle production by similar methods, to relate particle production and texture, to use integral geometry methods and stochastic geometrical processes to solve the complex probability equations in the three dimensional space, are stressed. A presentation is made of models developed by other researchers, which illustrates some of their deficiencies. A complete presentation is made of a model which incorporates the required elements. A special case of the model, combining a description of texture based on a Boolean model with primary Poisson polyhedra grains, of particle production calibrated on screen fractions extracted from ground ores, independence of texture and breakage, is presented. Applications of the model to data presented by other researchers give evidence of its value especially in predicting particle composition distribution at sizes similar or coarser than grain sizes.