Future Of Grinding Rate Modeling Based On Energy Spectra

In this manuscript, a new approach for modeling the rate of grinding process is explained. A review about grinding rate definitions so far is given and the possible ultimate way for grinding modeling is given as modeling the rate based on energy spectra. The traditional/phenomenological approaches for modeling size distribution of grinding mills give their place to a new generation modeling technique. Developments in the computational methods and necessity to model product size distribution of large mills, larger than ball mill size, paved the way for this novel technique. The traditional grinding rate models are mostly based on the back calculated breakage parameters from the experimental data. However, direct modeling of breakage parameters and hence the prediction of size distribution of the grinding mills may become possible with this approach. The material breakage data and discrete element method (DEM) energy spectra are the main components of this approach. Albeit the focus in this study is on modeling of breakage rate under impact mode of breakage, the abrasion mode of breakage was also tested by carrying out abrasion tests. Impact mode of breakage was analyzed using ultra-fast load cell device (UFLC) and drop weight tester. Dry tumbling test was carried out for abrasion experiments. Obtained material breakage versus energy data are combined with the collision energy spectra to obtain the breakage rate function. In addition, the material breakage versus energy information from the tests is converted into breakage distribution function. The model was verified in a lab scale mill (90 cm diameter-15 cm length). 10 to 15 kg of limestone was ground using 50 mm steel ball at varying conditions, i.e., rpm, volume loading. It is shown that the proposed model predicts all (1, 2, 4 min.) grinding intervals for all experiments highly accurate.