Abstract
The discrete element method (DEM) is used extensively for modelling particle systems in a wide range of research and industrial areas. The simulations are typically performed to solve a problem, evaluate novel ideas or elicit knowledge regarding some system. However, there are very few examples in the literature where development and problem-solving strategies for particle-machine systems, in particular, are outlined. In this paper, we propose and evaluate a methodology for efficiently utilising the discrete element method in a problem-solving process. The example case application used to demonstrate this methodology is the feeding of a cone crusher at a mining site in North America. The DEM simulations are performed with the In-house DEM code Demify® developed at the Fraunhofer-Chalmers Centre. Rock particles are modelled using a non-convex polyhedral irregular shape representation with massively parallelised high-performance computation on graphical processing units (GPUs). The results demonstrate how the DEM simulation can aid the engineering team and provide a more robust decision-making process. The final feeding solution resulted in a substantial improvement concerning production and liner wear rate in particular. The liner wear replacement period increased from around 3-4 weeks to 4 months, saving substantial cost, production downtime and energy solely from the liner wear improvement.