Development of models relating vertical stirred mill operating characteristics to screw stirrer wear and their application in devising stirrer wear offset strategies
摘要
Vertical stirred mill performance degrades significantly due to screw stirrer wear. This study proposes a bidirectional prediction and management framework to address this issue. A CFD-DEM coupled model was introduced to analyze the spatial distribution of Archard wear on the screw stirrer. Subsequently, data from a Box-Behnken (BBD) experimental design (45 conditions) revealed the nonlinear coupled effects of five key parameters on mill performance (power draw, normal force, energy efficiency) and Archard wear, including rotational speed, filling rate, pulp density, screw stirrer outer diameter, and coefficient of friction. The core contribution is the establishment of interconnected forward and inverse prediction models. The forward model utilizes operational parameters (rotational speed, filling rate, and pulp density) and stirrer state parameters (coefficient of friction and diameter) to predict current performance metrics and Archard wear. While the inverse model infers changes in stirrer diameter and coefficient of friction from operational parameters, target performance metrics, and measured Archard wear. To maintain the chosen performance attributes constant throughout the screw stirrer’s wear life cycle, the operational management strategies are discussed based on these parametric models, specifically regarding adjustments to control variables such as target filling rate and rotational speed. Specific examples include the development of strategies that preserve selected characteristics, such as the mill’s normal force or power draw.