| Popis: |
Processing behaviour of a mineral resource highly depends on the characteristics of particles such as shape, size and composition. Therefore, comprehensive particle characterization is crucial to understand and optimize processing behaviour to enhance recoveries and reduce waste production. Nevertheless, despite the obvious importance of particle characteristics, current analysis techniques are restricted to two-dimensional (2D) particle characterization. In order to have advance three-dimensional (3D) characterization, this study aims to present a new X-CT methodology for single particle characterization with a special sample preparation method to reduce the X-CT artefacts. A homogenous and dispersed particle sample reduce the X-CT artefacts and ease the segmentation process for individual particle labelling. This labelled data then further used for image processing combined with a new single particle peak analysis method for enhanced mineral classification based on greyscale. Classification of mineral phases for X-CT data was performed with the correlation of Mineral Liberation Analyzer (MLA). All the major mineral phases present in the ore were successfully classified except the gold grains. Characterization using 2D (MLA) and 3D (X-CT) was compared and mineralogy difference of around 2% observed. The effect of particle properties measured by both methods also investigated in processing simulation |
