Data-Based Performance Modelling of Hydrocyclone for Processing Iron Ore Fines

In this study, a data driven performance characterization model of hydrocyclone has been developed using multiple experimental data set collected from the published literature pertaining to processing of iron ore fines. The cut size, d50, has been determined for a given cyclone operating conditions using Lagrangian interpolation technique. A reduced efficiency curve has been constructed to map the performance and the functional behaviour has been modeled employing three typical distribution functions, namely, Rosin-Rammler, Exponential and Logistic. All pertinent model parameters have been estimated in accordance with the experimental data sets. It has been observed that all these functions fairly mimic the performance of cyclone for processing iron ore in the particle size range 25-300 m. Rosin-Rammler distribution found to be a better function for fitting the experimental data set in comparison to Exponential and Logistic functions to characterize the performance.