粒度
离散元法
粒子(生态学)
粒径
工作(物理)
计算机科学
算法
计算科学
生物系统
材料科学
机械
物理
工程类
机械工程
地质学
化学工程
海洋学
生物
操作系统
作者
Tarun De,Jayanta Chakraborty,Jitendra Kumar,Anurag Tripathi,Maitraye Sen,William R. Ketterhagen
标识
DOI:10.1016/j.powtec.2021.117058
摘要
Large scale industrial simulations of bulk solids using the Discrete Element Method involve enormously large number of particles, therefore coarse-graining appears to be an effective technique to reduce the computational demands for such scenarios. However, most of the current coarse-graining techniques use a uniform coarse-grain ratio throughout the equipment. Such uniform coarse-graining severely limits the degree of coarse-graining possible, because at higher coarse-grain ratio, the accuracy of the simulation can be affected. To resolve this issue, a multi-level coarse-graining technique (MCG) has recently been proposed by Queteschiner et al. [Powder Technol. 338 (2018) 614] where the coarse-grain ratio varies spatially. In this work, we propose a simpler particle location based method for multi-level coarse-graining. The new method can be readily implemented using open-source software LIGGGHTS Public (DCS Computing GmbH, Linz, Austria). In this method, the coarse-grained particles disintegrate into particles of higher resolution in the refinement step at the desired spatial location and the opposite occurs during the coarsening step. The efficiency of this new coarse-graining method has been demonstrated successfully using discharge from a conical hopper as an example. Compared to uniform coarse-graining, the computational time is reduced by more than an order of magnitude without compromising on accuracy.
科研通智能强力驱动
Strongly Powered by AbleSci AI