Static centrifugal classifiers, such as gas cyclones and hydrocyclones, are crucial in mineral processing for particle separation. While gas cyclones remove particles from gas streams, hydrocyclones function as classifiers but suffer from high bypass in ball-mill circuits. This thesis explores the impact of conic curvature and surface wall roughness in both cyclone and hydrocyclone performance.
Using Computational Fluid Dynamics (CFD) and laboratory experiments with 3D-printed prototypes, the study investigates cyclone behavior under varying surface roughness. A novel expandable chamber method was developed for particle size measurement. CFD simulations of 31 mm gas cyclones reveal that convex conic designs produce coarser cut sizes, enhancing classification efficiency over standard designs. Response Surface Methodology and Monte Carlo simulations further validate convex designs as optimal for coarser classification.
Experimental results confirm that convex gas cyclones reduce fine particle separation while improving the classification of coarser fractions. Further, these insights are transferred to hydrocyclones. In 75 mm hydrocyclones, convex designs mitigate fines bypass, reducing re-grinding costs in ball-mill circuits. Additionally, surface roughness effects differ between gas cyclones and hydrocyclones: rough surfaces lower tangential velocities in gas cyclones, reducing separation efficiency, whereas increased downward axial velocities in rough hydrocyclones improve solids recovery and concentration, benefiting dewatering applications.
This thesis advances cyclone design through CFD and experimental validation, providing insights for more efficient classification and industrial-scale applications.