Particle upscaling: Calibration and validation of the discrete element method

The discrete element method (DEM) is a numerical tool often used by engineers and researchers to design, analyse and improve bulk materials handling systems. However, the accuracy of any DEM model depends on the input values assigned to the parameters as determined through a calibration process. Together with this, limitations in computing power to model full scale systems are still limiting the use of DEM by industry. For this reason the particle size is often scaled up although the influence that it might have on the results has not been studied in great detail. In this study the calibration process made use of a large rotating drum to measure the dynamic angle of repose experimentally. The equivalent DEM model made use of particles scaled by a factor of up to 9. A small range of particle-particle coefficients of friction which accurately predicted the angle of repose could be identified for all scaling factors up to 4 (drum/particle diameter ≥25), while larger scaling factors gave inconsistent results. This calibrated parameter set was then used in modelling hopper discharge of the scaled particles. Here the results indicated that the maximum scale factor was independent of the size of the hopper opening and should be less than 1.3 and 2.5 to accurately model the rate of discharge and the velocity field respectively. This demonstrated that the maximum scale factor is application dependent and should be carefully selected. The correlations of Beverloo [1] and Villagran Olivares et al. [2] to predict the rate of discharge were compared to the experimental and DEM results. Beverloo's correlation could be accurately fitted for scaling of the hopper opening and the particles by factors of 2 and 4 respectively, however, the correlation by Villagran Olivares et al. failed to fully take the particle scale into account. It was shown that the parameter in this correlation is particle size dependent. [Display omitted] •Successful calibration based on the dynamic angle of repose•The drum to particle diameter ratio should be more than 25 to be accurate•In hopper discharge, scaling factor should be less than 1.3 for accurate prediction•Limitation in scaling factor independent of hopper opening size