Experimental validation of a statistical reliability method for the liberation distribution measurement of ore particles. (15th August 2019)

• Record Type: Journal Article
• Authors: Ueda, Takao
• Is Part Of: Minerals engineering. Volume 140(2019)
• Publication Date: 2019-08-15
• Subjects: Statistical reliability -- Mineral liberation -- Liberation distribution -- X-ray computed tomography -- Binary particle; Mines and mineral resources -- Periodicals; Ressources minérales -- Périodiques; Mines and mineral resources; Periodicals; Electronic journals; 622
• DOI: 10.1016/j.mineng.2019.105880
• Languages: English
• ISSNs: 0892-6875

Abstract

Highlights: A statistical reliability method for liberation distribution assessment is examined. The liberation distribution of binary particles is analyzed by X-ray CT. The required statistical reliability is demonstrated through experimentation. Abstract: Accurate assessment of mineral liberation is important in mineral processing. Recently popularized automated liberation analyzers have enabled the rapid measurement of the liberation state of particles. However, the statistical reliability of liberation distribution assessment has not been thoroughly investigated. Recently, a statistical method which correlates particle measurements with the confidence interval of the measured liberation distribution has been proposed. In the present study, this statistical method was validated by a series of experiments using binary particles and X-ray computed tomography (X-ray CT) analysis. Specifically, three types of binary particles composed of a matrix phase (epoxy resin) and grain phase (glass beads or silica sand) were created; the liberation distribution of all the particles was obtained as the mother population, through X-ray CT and post-processing; a certain number of particles determined by the statistical method were selected from the mother population, and the percentage in which the sampled liberation distribution satisfied the designated confidence interval was determined. The percentages ranged from 94% to 97%, which agreed with the required reliability of 95%, demonstrating the effectiveness of the statistical model.