刘轩廷
北京大学
A cross-scale CFD-DEM for particulate flow
Hybrid CFD-DEM, which integrates the characteristics of un-resolved, semi-resolved, and fully resolved methods, is employed for high-fidelity multi-resolution simulations of cross-scale polydisperse particle-fluid flow systems. Its core principle involves selecting appropriate methods based on the grid size to particle diameter ratio ($\Delta/d$). For instance, the semi-resolved method is applied when $0.1<\Delta/d<3$. This range covers a wide spectrum of particle sizes, with the maximum-to-minimum particle diameter ratio ($d_{max}/d_{min}$) reaching up to 30. In most studies, a uniform grid is employed to maintain the size ratio $\Delta/d$, ensuring that medium-sized particles fall within the $1<\Delta/d<3$ range to achieve optimal results. However, in practical industrial applications, the grid size is dictated by the Reynolds number and specific geometry, whereas particle sizes are predetermined by the operational scenario. Consequently, some particles inevitably fall within the $0.1<\Delta/d<1$ range, where the applicability and reliability of the semi-resolved method remain uncertain. In this study, we first develop and validate a hybrid CFD-DEM model. Subsequently, we examine and discuss the applicability and reliability of the semi-resolved method based on the hybrid CFD-DEM results. The findings indicate that the semi-resolved method underestimates the particle wake field to some extent; when the particle diameter ratio $d_{max}/d_{min}$ is relatively large, the background field reconstruction method may introduce stability concerns; furthermore, for particles with smaller size ratios $\Delta/d$, the volume fraction calculation approach must be carefully chosen to avoid inaccurate drag force predictions
刘轩廷,北京大学博士生,导师为刘谋斌教授。主要研究方向为增材制造、颗粒流动的多相多物理场数值建模与仿真。下·