Estudo experimental e numérico da operação de moagem em um moinho de bolas com suspensores

Abstract

In a ball mill, the interaction between the grinding media and the material to be ground will determine the efficiency of the grinding operation. The occurrence of breakage events is the result of numerous collisions between materials and the energy involved in the impacts. The dynamics of particulate material inside the mill can be better understood through experimental studies and numerical simulations. For this purpose, the Lagrangean approach (or DEM) emerges as a consolidated tool, but involves some limitations, such as the need to specify some physical parameters that will determine the correct representation of the particular dynamics. Thus, the present study performed the experimental measurement of these DEM parameters, namely the restitution coefficient and static and rolling friction coefficients, and analyzed the influence of different test conditions on these measurements. Through DEM numerical simulations, it was investigated how the measured values of the DEM parameters influence the number of collisions and collision energy between the grinding media inside the ball mill under different experimental conditions and, consequently, how they can affect the simulation result of the grinding operation. It was observed that different measured parameter values for the same interaction alternating the role of each material on the particle-surface pairs were not the same, with a maximum deviation of 43,8%. This requires a deeper understanding of the parameters application in the DEM model calculations, because the different measured values of these parameters changed the results of number of collisions and collision energy, which may directly affect the predictions of milling performance by numerical simulations. It was also investigated how the liner design of a ball mill, such as the quantity and geometry of the lifters, can affect the number of collisions and collision energy, thus defining their influence on the performance of the grinding operation. Finally, quartzite particles were broken both experimentally and by numerical simulations using the Bonded Particle Model (BPM). In both situations, the conditions of maximum number of collisions and collision energy were employed, and the optimized condition of these two responses simultaneously through the desirability function. It was found that the optimized condition was the one that resulted in the best breaking efficiency of quartzite particles (59,2%), instead of using only the individual maximum conditions of each response “number of collisions” (38,7%) and “collision energy” (47,5%).