Abstract

The cross-flow air classifier is a dry process device. The use of such separators is being evaluated to replace the wet classification methods in the production of manufactured sand. The influence of the solid feed rate is critical for the correct design of the process. One method to analize the influence of the feed is by using numerical simulation. In this work, the powder classification process is simulated by computational fluid dynamics (CFD). The particle-particle interaction is modeled with a discrete element method (DEM). Simulations are optimized to better represent the experimental data by choosing between two turbulence models and two models for the injection of solids. Dust classification at three different solid rates is simulated. It is established that the amount of solids has a great influence in the classification process. This fact indicates that the solids content is a critical parameter for a proper design of the device.