Experimental and numerical investigation of particle size and particle strength reduction in high pressure grinding rolls

Abstract

Feed particles experience significant reduction in size and particle strength in high pressure grinding rolls (HPGR) due to strong compression. This paper presents a combined experimental and numerical study of the mechanics of cement particle breakage in a lab-scale HPGR under different conditions. A model based on the discrete element method (DEM) was developed and coupled with a multi-physics model and a particle breakage model to mimic the dynamics of particles in the HPGR. Through careful calibration of the model parameters, the model was able to generate results in good agreements with the experiments in terms of throughput, power consumption, product size distribution and mill productivity. The strength of the product particles, which was characterised by particle fracture energy, was analysed in the simulations and results showed much weaker and broader fracture energy distributions compared with the feed particles. Increasing roll speed increased throughput and power consumption but had little effect on working gap, product size and particle fracture energy. Higher roll speeds also significantly increased the pressure on the floating roll (but not on the fixed roll), which may result in more severe wear on the roll. Increasing pressure in general reduced throughput due to smaller working gaps and finer and weaker product particles. Furthermore, the over-sized (larger than 0.045 mm) product particles were fed into the mill to investigate the effect of multiple grinding. The particles passing through the 2nd grinding were much smaller and weaker, and the simulation results were comparable to the experiments. The study demonstrated that both particle size and particle strength need to be considered properly in DEM models to provide accurate prediction on the performance of HPGR under various condition. • Particle breakage in a lab-scale HPGR was investigated. • Dynamic response of the HPGR and particle breakage were included in the DEM model • The simulations agreed well with the experiments • Roll speed and roll pressure affected the size and strength of the products • Both the size and strength of feeds need to be considered to predict the performance of HPGR