Mineral Processing
Fatemeh Kazemi; Ali akbar Abdollahzadeh
Abstract
This research work aims to explore the intricate mineralogy and texture of the tailing piles of iron ore processing plants to present a particle-based prediction for magnetite recovery. Three samples were taken from different points of tailings piles of an iron ore processing plant. Davis tube tests ...
Read More
This research work aims to explore the intricate mineralogy and texture of the tailing piles of iron ore processing plants to present a particle-based prediction for magnetite recovery. Three samples were taken from different points of tailings piles of an iron ore processing plant. Davis tube tests were performed on each sample under various operating conditions. Process mineralogy studies were conducted to determine the mineralogy modal of the feed and product of each test. An Artificial Neural Network (ANN) model was used to make a model that related the grade and recovery of magnetite in the product to the mineralogy modal of the tailing piles. The magnetite grade and association index of feed, the magnetic intensity, and the water flow rate were the inputs to this network. The grade and magnetite recovery correlation coefficients were 0.954 and 0.86, respectively. The grade of magnetite in the feed emerged as a limiting factor on the grade and recovery of magnetite in concentrate. An increase of one unit in magnetite grade in the feed resulted in a 1.68 decrease in the recovery. The association index changes with the coefficients of -0.173 cause the changes in predicted magnetite recovery in the concentrate.
