Optimisation of calibration parameters for a discrete element method (DEM) model of plant-origin grains with a low elasticity modulus

Abstract

This paper presents the application of the Discrete Element Method (DEM) to describe the physical properties of plant-origin grains with a low elasticity modulus for the needs of numerical simulation of processing processes. The study proposes the modelling of maize grains based on 3D scanning of real grains and further use of the multi-sphere method to fill the numerical model with a conglomerate of elementary spheres. The main aim of the paper is to develop a calibration method based on exploring parameter spaces at points selected using Sobol's grids. As criteria and functional limitations for the calibration, the following were proposed: the slope angle of repose (AoR), the radius of the heaped cone's vertex (Rad), the number of grains, and the slope height. The study results show that for the calibration of the DEM model describing maize grains, test points with a set of the following nine parameters should be used: Poisson's Ratio for grain, Density of grain, Shear Modulus, Coefficient of Restitution for grain-grain, Coefficient of Restitution for grain-material, Coefficient of static friction for grain- grain, Coefficient of static friction for grain-material, Coefficient of rolling friction for grain-grain, and Coefficient of rolling friction for grain-material.