Calibration procedure of Discrete Element Method (DEM) parameters for wet and sticky bulk materials

Abstract

The downtimes associated with blockage events in the materials handling sector have prompted the mining industry to optimise its operations to maximise the profitability of fines and lump products. These blockage events are caused by materials with a high clay or moisture content typically referred to as Wet and Sticky Material (WSM). Numerous methods are commonly used to prevent such blockage events, however, in the majority of cases, they are often too late to prevent the downtime of the system. The Discrete Element Method (DEM) is commonly used to predict and visualise the bulk material flow. DEM is capable of replicating the flow of non-cohesive granular materials with reasonable accuracy, however, when cohesive materials are considered the mechanism becomes much more complex. With the advancement of computational power over the last few decades, it is now more practical to simulate WSMs into DEM.This paper evaluates the capability of three cohesion models to simulate WSMs; the Simplified Johnson-Kendall-Roberts (SJKR) model, the Easo Liquid Bridging model and the Edinburgh Elasto-Plastic Adhesion (EEPA) model. A calibration procedure is proposed where the parameters for each cohesion model are discussed in detail. A series of calibration simulations with systematic parameter variation is undertaken to define a set of calibration matrices. These matrices allow the formation of a parameter database, used for the simulation of on-site applications to optimise plant geometry and other operational parameters. Finally, the numerical model was validated using a lab-scale vertical impact testing facility. Individually, in a transfer chute application, it was determined that none of the 3 cohesion models were suited to WSMs; the SJKR and Easo liquid bridging models were unable to replicate the behaviour, and while the EEPA model was able to replicate the behaviour, extensive solve times deemed it impractical for industrial applications. These limitations were overcome through the coupling of the SJKR and Easo Liquid Bridging models.