Mengtao Han*, Royzo Ooka, Hideki Kikumoto. Lattice Boltzmann method-based large-eddy simulation of indoor isothermal airflow. (SCI检索)
发布时间：2021-03-10
点击次数：
- 论文类型：
- 期刊论文
- 发表刊物：
- International Journal of Heat and Mass Transfer
- 收录刊物：
- SCI
- 学科门类：
- 工学
- 一级学科：
- 动力工程及工程热物理
- 文献类型：
- J
- 卷号：
- 130
- 页面范围：
- 700–709
- 关键字：
- lattice Boltzmann method; Finite volume method; Large-eddy simulation; Indoor turbulent flow; Parallel efficiency
- DOI码：
- 10.1016/j.ijheatmasstransfer.2018.10.137
- 发表时间：
- 2018-09-05
- 影响因子：
- 5.584
- 摘要：
- As a relatively new computational fluid simulation method, the Lattice Boltzmann method (LBM) has recently garnered widespread attention in various engineering fields. The LBM-based large-eddy simulation (LBM-LES) model is commonly used to predict flows with high Reynolds numbers, and is considered to yield a prediction accuracy comparable to that of the traditional finite volume method (FVM-LES). Nonetheless, in the indoor environment, a detailed benchmark for the LBM-LES accuracy is currently underdeveloped, and its computational time has not been sufficiently compared with that of FVM-LES. In this study, simulations of an indoor isothermal-forced convection benchmark case were carried out in the forms of LBM-LES and FVM-LES, using different grid resolutions, relaxation time schemes, and discrete velocity schemes of the LBM, with the aim of verifying the prediction accuracy of LBM-LES and investigating its consistency with FVM-LES in indoor turbulent flow. Their computational speeds and parallel computational efficiencies were also compared. The results demonstrate that LBM-LES can achieve the same level of indoor-turbulent-flow prediction accuracy as FVM-LES; however, it requires a more refined grid system (in this study, its grid width was half of that of FVM-LES). Furthermore, the relaxation time and discrete velocity schemes of LBM barely influenced the accuracy of the indoor flow. For the same level of accuracy, their computational speeds approached the same level, even though the LBM-LES required an eight times larger mesh quantity than the FVM-LES; however, the LBM-LES computational speed can surpass that of the FVM-LES when more cores are utilized because of its outstanding parallel efficiency.
- 备注：
- 中科院1区TOP，JCR1区
- 上一条：Mengtao Han*, Royzo Ooka, Hideki Kikumoto. Validation of lattice Boltzmann method-based large-eddy simulation applied to wind flow around single 1:1:2 building model. (SCI检索)
- 下一条：Mengtao Han, Hong Chen*. Effect of external air-conditioner units’ heat release modes and positions on energy consumption in large public buildings. (SCI检索)