Stochastic modeling of classifying aerodynamic lenses for separation of airborne particles by material and size

Abstract

A flexible stochastic approach is described to model separation processes, in which air-borne particles are separated via a setup of one or more aerodynamic lens orifices. Varying the size of the orifices, the focusing pressure and the mass flow rate through the lens, leads to focusing of particles with distinct aerodynamic properties on the central axis. In this modeling approach a bivariate transfer function is used to describe the passage probability of particles depending on their size and mass. The distribution of feed particles and the changes of the distribution due to the separation process is described via probability densities. The modeling procedure is applicable to various kinds of separation methods and allows optimization of geometric and operation parameters. To this end, the model utilizes flexibly defined separation performance measures which are illustrated in a case study that considers the separation of Cu particles from SiO2 particles. The spherical particles in the considered virtual mixtures are described by their log-normally distributed diameters and their normally distributed mass densities. Furthermore, the cases are selected in such a manner that the mean aerodynamic diameters of both Cu and SiO2 particles are equal.