Predicting the accuracy of mineral phase analysis by X-ray diffraction using Monte Carlo modelling

Abstract

Rapid, on-line measurement of feedstock mineralogy is a highly attractive technology for the mineral processing industry. A Monte Carlo particle transport-based modelling technique has been developed to help design and predict the measurement performance of on-line energy-dispersive X-ray diffraction (EDXRD) analysers. The accuracy of the technique was evaluated by performing quantitative phase analysis on a suite of fifteen synthetic potash ore samples. The diffraction profile of each sample was measured with a laboratory EDXRD analyser and an equivalent profile was simulated in the Monte Carlo package. Linear regression analysis was used to determine the mineral abundances in each sample from both the measured and modelled profiles. Comparison of the results showed that the diffraction profiles and measurement accuracies obtained by simulation agree very well with the measured data.