Upgrading of a low-grade iron ore using a dry roll magnetic separator

In this thesis, it was aimed to investigate the beneficiation of a low-grade iron ore using a dry-roll magnetic separator. The ore (39.5% Fe content) used herein was collected from Feke/Adana region, Türkiye. The properties of the ore were characterized using Mineral Liberation Analysis (MLA), which determines the following properties: (i) modal mineralogy, (ii) calculated assay, (iii) elemental distribution, and (iv) mineral locking. It was determined that the ore was composed mainly of iron-oxide, quartz, chlorite, muscovite, plagioclase, and other minerals. Goethite and hematite minerals were observed in X-ray diffraction (XRD) measurement. The free surface and liberation degree of iron-oxide mineral detected by MLA were found to be 32.84% and 54.17%, respectively. In the magnetic separation test, the effect of particle size (-1+0.500mm, -0.500+0.300 mm, and - 0.300+0.125 mm), splitter position (43° and 58°), cleaning stage (1 and 2), magnetic field (0.2 T and 0.4 T) on the recovery of the magnetic product from the ore was investigated. It was understood from the experimental results that the magnetic product (-1+0.500 mm) with an iron content of 67.67% can be obtained, but its recovery was not acceptable value (<30%). There was an inverse relationship determined between iron content and recovery value in the magnetic separation test. As such, an increase in the speed of conveyor led to an increase in the iron content in the magnetic product, and a decrease its recovery. The iron content of the magnetic product (-0.500+0.300 and -0.300+0.125 mm) cannot be reached to desired levels (>56% of Fe and >70% of recovery). The experimental results were evaluated by artificial neural network (ANN) method to determine the efficiency of the parameters on the beneficiation test. The Pearson (Pc) correlation showed that the parameters were ordered as follows: conveyor speed, particle size, and splitter position for iron content. Different scenarios were carried out to obtain the magnetic product with desired properties (>56% Fe). It was tried to conduct microwave treatment to increase the magnetic susceptibility (MH) of the ore. It was found that there was no need to perform microwave treatment on the ore with a size of <0.500 mm. Microwave treatment played a key role on the ore with a size of -1+0.500 mm and its MH value increased to 4.30 emu/g from 0.16 emu/g. However, the change in the ore with a size of <0.500 mm was negligible. Finally, a flowsheet was proposed to obtain a magnetic product (-1+0.500 mm) containing 56.65% iron and 90.01% of recovery, whereas the iron content of the mixed magnetic product (-0.500+0.300 mm and -0.300+0.125 mm) was 51.11% and recoveries over 90%. Keywords: Artificial Neural Network (ANN), Dry-Roll Magnetic Separation, Low-grade Iron Ore, Mineral Liberation Analysis (MLA), Microwave Treatment