Understanding the interaction of multicomponent particles in hydrocyclone classifier using CFD model

In mineral comminution circuits, the feed to hydrocyclone consists of a heterogeneous mixture having varied degrees of liberation and sizes leading to the multi-component particle system. The presence of multicomponent particles exhibits a significant change in the classification performance parameters. In this work, a numerical study is conducted with an artificial mixture of magnetite and silica of different proportions having similar particle size distribution to understand the multicomponent interaction in a laboratory scale hydrocyclone. The heterogeneous feed is injected tangentially would attain different settling rates under the influence of centrifugal, drag and turbulent dispersion forces. In this work, Large Eddy Simulation (LES) is used to resolve the continuous phase turbulence. Modified Algebraic Slip Mixture (ASM) model using the corrected Newtonian rheology of fines to simulate multi-density and multi-size particles of the mixture in a hydrocyclone. The influence of the multi-component particles on classification performance is extracted from the simulation in terms of the performance curve, i.e. recovery to underflow, volume fraction distribution and the key forces acting on each particle in the mixture. An interaction was observed between different components during the classification process, featuring accumulation of heavier particle at wall-side pushing the lighter particles into the inner locus of zero vertical velocity (LZVV) zone. The comparison of separation factors of different size particles presented the higher probability of the magnetite of same size reporting to wall region and hence to underflow. Hence, an increased component cut-size with the increased magnetite proportion, particularly a profound effect on lighter density component's cut-size was observed.