Model-based control of the dynamics during fine grinding in wet-operated stirred media mills

点击翻译按钮获取中文摘要

Stirred media mills are used in various ultra-fine comminution processes. The selected operating parameters in combination with the suspension properties determine the kinetic energy of the grinding media and, in the case of continuous operation, also the transport of the particles as well as the non-uniform axial distribution of the grinding media within the mill. This determines comminution behavior and power consumption. For ultra-fine comminution of the product particles down to the nanoscale, mills are often circuit-operated with an agitated vessel. Hereby, the particle size distributions in the agitated vessel and mill change dynamically. Due to a constant in-crease in specific surface area with decreasing particle size, particle interactions and their influence on suspension stability and viscosity play a fundamental role in ultra-fine comminution. The increase in suspension viscosity has a direct influence on the relative grinding media velocity and thus on the energy transferred to the product particles. In addition, the suspension viscosity influences the axial distribution and thus the movement behavior of the grinding media. Hence, electrostatically or sterically stabilizing additives must be considered in the process control in addition to the control of, e.g., stirrer tip speed or flow rate. Moreover, with changing particle size, the stress energy required for ideal particle breakage, i.e. the optimal operating point of the mill, shifts towards lower stress energies. When controlling a fine grinding process in a wet-operated mill, the interplay between damping of the grinding media through the increase in viscosity, in-crease in particle strength with decreasing particle size and shift in the optimum operating point due to the increase in stress intensity must be considered. Finally, spontaneous reagglomeration or recrystallization processes place high demands on online measurement technology and process control. Although short-cut models and modelling approaches for stirred media mills exist, there currently is no model that allows for dynamic control of changing fine comminution processes with regard to particle size distribution, optimum energy utilization, or maximum production rate. One possibility of control also pursued in this project is the description via population balance models (PBM) in combination with a Nonlinear Model Predictive Control (NMPC) strategy. Thereby the PBM will be described via a mechanistic comminution modell.