Calibration of the discrete element method and the effect of particle shape

The discrete element method (DEM) is used across a wide range of applications. However, accurate predictions can only be made if the input parameters are carefully selected. In this paper a calibration process is proposed to calibrate the parameter values for crushed rock particles up to 40 mm in size. A large shear box with a diameter of 590 mm was designed and built for this purpose. Confined compression tests were used to determine the particle stiffness and direct shear tests to determine the particle-particle friction coefficient. Two methods were used to create the clump particles: a manual process and an optimised process to-create clumps comprising of 2, 4 or 8 spheres. The clump types were individually calibrated to obtain a unique set of parameter values for each. The angle of repose is often used for the calibration of the particle-particle friction coefficient and was included in the calibration process for comparison with the direct shear test results. The calibration process was validated by modelling anchor pull-out tests and hopper discharge using the different clump types. The results showed that care should be taken when the angle of repose is used to calibrate the particle-particle friction coefficient. It can result in a friction value which is too low for use in other applications, although the angle of repose is accurately predicted. (C) 2016 Elsevier B.V. All rights reserved.