Response surface optimization study on centrifugal classifier of double cage of vertical roller mills

The traditional vertical roller mills (VRMs) system often has the problem of low efficiency of finished particle collection due to poor matching of various performance parameters of the classifier. To address the issue, this study conducted the gas-solid two-phase flow simulations using Fluent software, and optimized key parameters of the double cage classifier through Box-Behnken response surface methodology. A four-factor, three-level experimental model was established, incorporating rotor blade rotation speed, number of rotor blade, rotor blade installation angles, and static blade installation angles. Fit the regression equations of classifier pressure loss and particle collection efficiency with classifier performance parameters. After optimization, the collection efficiency for 460 mesh particles improved from 80.57% to 95.63%, significantly enhancing the production efficiency of ultrafine particles. The research demonstrates the reliability of the proposed simulation model and response surface methodology. The optimization strategy provides effective guidance for structural design and parameter configuration of classifier systems in the similar VRMs.