Enhancement of classification performance for irregular ultrafine particles

The increasing industrial demand for ultrafine powders has heightened the need for precise classification technologies. However, achieving high classification accuracy remains challenging, especially for irregularly shaped particles produced by conventional methods. In this study, we proposed a modified classifier design that incorporates a wire mesh with a grid structure onto the classifier impellers to enhance the classification accuracy of irregular particles. The wire mesh promotes the formation of micro-eddies that induce particle rotation, facilitating irregular particles to approximate the hydrodynamic characteristics of spherical particles. We systematically evaluated the effects of mesh position, layer number, and aperture size on classification performance. The results indicate that a single-layer wire mesh outperforms multi-layer configurations, and optimal performance is achieved when the mesh with medium-sized apertures is positioned externally on the impeller blades. The placement of the wire mesh increases D₁₀, reduces D₉₀ and span at the same D₅₀, steepens the partial classification efficiency curve, and lowers the classification accuracy index (K value). These parameters collectively validate the enhancement in classification accuracy. To further assess classification quality, we introduced two new indicators: the overlapping ratio of the area under the particle size distribution (PSD) curves for fine and coarse fractions, and the fine/coarse powder removal ratio. Both metrics confirm the enhanced classification precision. The improvement is attributed to three synergistic effects: micro-eddy-induced particle rotation, improved airflow uniformity due to added resistance, and the mesh's screening effect. This work offers a practical and theoretical basis for improving the classification accuracy of irregular particles.