Multivariate Analysis Supporting Factorial Experiments: A Case Study in the Physical Separation of Chromite Fines by a Multigravity Separator

In the present work the concentration of chromite fines by a multigravity separator (MGS) has been studied. The main goal of the experimental work has been the evaluation of the influence of the main MGS process parameters in the enrichment of an Albanian ore. The experimental tests have involved the design of techniques using factor analysis as a support for full or fractional factorial experiments. Therefore, factorial experiments have been performed in order to find the main and interaction effects of the investigated factors on MGS performance: shake frequency, shake amplitude, tilt angle, washwater flow rate, and size fraction. The factor analysis supported factorial experiments in order to reduce the responses considered in the analysis of the variance (ANOVA). This statistical method has been used to identify the truely independent variables of the process. A very strong correlation has been found among Cr2O3 grade and the other elements. For this reason the only responses of the process were the recovery of Cr2O3 in the concentrate and its grade. The results obtained revealed the technical feasibility of the separation of chromite fines by MGS and the advantages that can be obtained using statistical methods in the design of experiments for the study of multifactor processes.