Experimental and numerical investigations of particle breakage in grinding processes

Abstract

Grinding plays an important role in many industries and is still poorly understood because of complicated breakage behaviour of particles in the process. The lack of understanding of particle breakage has prevented the development of the fundamental understandings of the grinding process. This work investigated the breakage behaviours of particles under different grinding conditions. Physical experiments and numerical simulations were conducted in two different types of mills: a variable speed impact mill and a horizontal stirred mill. The effects of material characteristics and operation conditions were investigated. For the variable speed impact mill, it is shown that particle shape has a strong effect on the breakage behaviour. The irregular shape particles experience different crack mechanisms comparing with the regular shape particles, leading to the variation in breakage ratio. The breakage ratio increases significantly as the impact energy increases. Multiple impact tests show that the complete breakage can be reached at low impact energy after multiple impactions. Two material properties, resistance to fracture f_mat and threshold fracture energy W_(m,min) which describe the grindability of the materials, can be determined by fitting method. The results show that they are inversely proportional to each other. For the horizontal stirred mill, it is shown that the mill rotation speed plays a crucial role in particle breakage. The relationship between the grinding rate K_p and impact energy E_i in horizontal stirred mill has been analysed and the results show it follows the power law. The study of the relationship between grinding rate K_p and time indicates that the kinetics of the grinding process in horizontal stirred mill is in first order. The effect of increasing of the grinding media size shows a better grinding performance but the cost of power consumption is considerable. The CFD-DEM simulations of particle flow patterns reach an acceptable agreement with experimental results. Therefore, the model can be used to predict the variation of flow patterns of particles in the grinding process in horizontal stirred mill.