A hybrid scaling coarse-graining method based on a computational fluid dynamics-discrete element method
Li Longwei,Li Jian,Li Shichang,Dai Zhangjun,Chen Shanxiong,Wei Xiaoyang
DOI: https://doi.org/10.1007/s40571-024-00859-z
2024-11-03
Computational Particle Mechanics
Abstract:A computational fluid dynamics-discrete element method (CFD-DEM) is an important method for simulating the interaction and movement of fluid and particulate materials. Its ability to simulate the mechanical behavior of particulate materials has led to its widespread research and application. However, due to limitations in computer computing power, CFD-DEM is limited in the number of particles it can simulate, making it difficult to achieve simulations at an engineering scale. To solve this issue, this study proposes a hybrid scaling coarse-graining method (HSCGM). This method significantly reduces the number of particles by replacing a collection of small particles with a single large particle. Additionally, the principles of particle motion balance, energy conservation, and the exact scaling model are used to determine the accurate relationship for the interaction force between coarse-grained particles. Finally, the accuracy and efficiency of the calculations are analyzed through Ergun and settling tests. The results show that the HSCGM more accurately simulates the interaction forces between particles and their motion behavior, while significantly improving computational efficiency. The advantages and disadvantages of other fluid–solid coupling methods are also discussed. The HSCGM further advances the application prospects of CFD-DEM at an engineering scale.
mechanics,mathematics, interdisciplinary applications