During the iron ore processing stages of crushing, washing, screening and classification, the lateritic material, rich in alumina and silica, gets accumulated with the ore fines (-10 mm). Therefore, where high grade ore is available, the fines are usually discarded as waste leading to huge iron loss and the problem of dumping. In medium and low grade ores these fines are used in agglomeration that increases the cost of ironmaking due to the alumina content [1]. Therefore iron ore beneficiation in the fines range calls for new processes. In this regard, a novel method was devised at CSIR-NML in collaboration with M N Dastur & Co along the lines of dry ore processing. Although limited, the past work shows that magnetic separation methods are selective and have low throughput [2-4]; whereas the studies on fluidized bed separators show that they are potentially scalable to high throughput [5-9]. In the light of these, a shallow bed of iron ore fines subjected elutriation to separate particles based on density and size, was considered.
The hematite-rich and gangue particles have significant density differences and that can be clearly separated by density based separation methods. However, the particle size variation results in overlapping regimes of factors that cause the particle separation. Therefore the goal is to optimize the separating factors to maximize the yield. As a proof of concept, a benchscale model for dry ore beneficiation was constructed at CSIR-NML. The model consists of an inclined vibrating deck with a mesh plate (100 mesh) on which ore fines are fed at the upper end and as these particles descend down the slope, are subjected to an air blast issuing from beneath the mesh plate; the particles depending on their size and density are thrown up in the air and as they fall, are collected in a tray called collector. The collected particles are analyzed for their iron content. The particle collection behavior is governed by the particle trajectories that is affected by jet air velocity and mesh plate inclination according to the particle size and density. Therefore to visualize and characterize the complex phenomena, the ore separation process is numerically simulated using computational fluid dynamics (CFD) in the present work. The particle inlet condition that is central to simulation of any such fluidization processes is treated rigorously. However, the feeding of ore particles on the mesh plate and their movement due to vibration are not simulated to avoid associated complexities. Therefore, the scope of the present work is only to assess the particle separation behavior without any inter-particle interaction.