A method of comparing the performance of air classifiers

An experimental and theoretical comparison is made between a “zig-zag” air classifier and a straight-duct type. Equations are presented, based on a model already published, for the effective diffusivity of spherical particles in a turbulent air stream. The diffusivity reflects the tendency of particle species to diffuse the “wrong” way, that is to be lost. The lower the value of the effective diffusivity, the more efficient is the classifier. It is shown that the effective diffusivities are lower in the “zig-zag” classifier than in the straight-duct type especially for the “light” species. It is also shown that the effective diffusivities of “light” and “heavy” species are similar in value, indicating the mid-point as the optimum feed location. It is also shown that, at high values of the air/solids ratio, the effective diffusivities can be easily estimated by using mixtures of regular particles, or simply by feeding a single species of spherical particles. The effective diffusivities were found to be independent of the number of stages in the “zig-zag” classifier, and took values similar to freestream diffusion values. The method described represents a very simple test which could be applied to proposed designs of air classifier to enable a comparison of effectiveness to be made.