Numerical investigations on random close packings of cylindrical particles with different aspect ratios

摘要

Random close packings (RCP) of different shaped cylindrical particles under optimal vibration conditions were numerically reproduced using discrete element method. Corresponding characteristic macro and micro properties such as packing density, coordination number (CN), radial distribution function (RDF), particle orientation, orientation randomness and forces/stresses as a function of particle aspect ratio (AR) were systematically characterized and analyzed. And the container wall effects on the packing were also comprehensively considered. The results show that with the increase of AR, the packing density of each RCP structure first increases to a maximum (0.70 with AR =1) and then decreases. The RDF analysis shows the particle arrangement of local structure varies with AR. The container wall is only in effect within three layers for equiaxed cylinders (AR = 1) near the boundary and just one layer for other shaped cylinders. The non-equiaxed cylinders especially elongated ones exhibit a strong tendency to horizontal alignment. With the growth of AR, the nematic order parameter of the packings first decreases to a minimum and then increases. Both particle position and orientation distributions demonstrate that the obtained packing structures are random. Force analyses indicate that the strong forces follow an exponential distribution, while those weak forces follow a power-law distribution, which is comparable to the force distributions of spherical particles. Compared with other cylindrical particles, the force orientation distribution in the packing of equiaxed cylinders is more homogenous. The peak position of the hydrostatic stress distribution is the same for different shaped cylinders. Moreover, with the increase of the packing density, the percentage of strong stresses increases. With the AR deviating from 1, the static vertical stresses present a gradually larger saturation stress in the same depth. (C) 2018 Elsevier B.V. All rights reserved.