An approach to and validation of maize-seed-assembly modelling based on the discrete element method

Abstract

To build an analysis model of maize seed particles using the discrete element method, the shape, size, density and moisture content of maize seed particles are measured and analysed. It can be seen that within the same variety of maize seed particles, the shape can be classified into horse-tooth, spherical-cone, spheroid, prism, and irregular shape. The percentage of horse-tooth, spherical-cone, and spheroid particles accounts for approximately 90% of the total. The characteristic size of the horse-tooth, spherical-cone, and spheroid particles approximately obey a normal distribution, and there is a certain functional relationship between the sizes. For the same variety of maize seed assemblies, the horse-tooth, spherical-cone, and spheroid particles are selected to build the analysis model. The upper size of the horse-tooth particle and the thickness of the spheroid particle are chosen as the primary sizes and are randomly generated according to their normal distribution. The other characteristic sizes are calculated via their corresponding functional relationships. An analysis model of individual maize seed particles is built using the multi-sphere (MS) method. Taking three varieties of maize seeds as examples in this paper, the results show that the simulated results are close to those obtained experimentally in terms of the piling and “self-flow screening” tests when the horse-tooth, spherical-cone, and spheroid particles are filled with 10 to 14, 18, and 6 sub-spheres, respectively. A preliminary verification is performed of the feasibility and validity of the modelling method of maize seed particle assemblies and individual maize seed particles.