Shape matters: DEM modelling of coarse cohesionless materials for bulk handling and transport

Abstract

Coarse, cohesionless bulk materials with large particle size distributions (PSD), such as limestone, are difficult to handle. Numerical modelling, such as the Discrete Element Method (DEM), can analyse the material’s behaviour and optimise the handling process. Current large-scale DEM models are inaccurate and computationally expensive. In the current work, the effects of irregular particle shapes (shape descriptors and interlocking) on material bulk behaviour are studied with particle shapes modelled in DEM through a novel combination of experiments, including the 3D surface scan, the Blott and Pye shape classification and a newly developed tumbler which can degrade particle angularity. Based on these findings, a new, accurate and efficient multiscale DEM model is developed for coarse limestone. The model calibration and verification are performed using both small- and large-scale experiments, e.g., the inclined surface test, the lifting cylinder test and full-scale grab operations. In addition, the possibility of grab design improvements is investigated using the new DEM model developed here.

Experiments reveal that particles with lower angularity have a smaller angle of repose (AoR), highlighting the influence of particle shape on frictional behaviour. The particle-particle sliding friction coefficient (μs,p-p), determined via inclined surface testing, is high but only suitable for small-scale DEM models. The bulk volume is generated using ten multi-spherical particles representatively depicting the classified shape distribution, and the DEM model is calibrated using a small-scale lifting cylinder and a full-scale grab. Verification of the knife penetration and path confirms precision, and subsequent design optimisation simulations in grab handling achieve up to 13% capacity improvement.