A scalable implicit finite element solver for the large eddy simulation of 3D incompressible turbulent flows on unstructured meshes
文献类型：会议论文
| 作者 | Zi-Ju Liao; Rongliang Chen; Zhengzheng Yan; Xiao-Chuan Cai |
| 出版日期 | 2016 |
| 会议名称 | The 20th IMACS WORLD CONGRESS |
| 会议地点 | 厦门 |
| 英文摘要 | Large eddy simulation (LES) has exhibited more and more advantages in simulating engineering turbulent flows. As the increasing computational speed and power of supercomputers, there is a growing interest in the scalable parallel algorithms for LES. In particular, such methods designed on unstructured meshes are extremely demanding. In this paper, we present a parallel fully implicit unstructured solver for the LES of 3D incompressible turbulent flows. For the spatial approximation of the problem, we use a stabilized Galerkin finite element method, while we employ the implicit BDF scheme for the time discretization. A parallel scalable domain decomposition method named Newton-Krylov-Schwarz algorithm is introduced for solving the nonlinear algebraic system arising from the discretization. We validate the proposed numerical scheme for a benchmark lid-driven cavity flow problem and the flow passing a rectangular cylinder at high Reynolds numbers. We then apply the proposed method to the large eddy simulation of turbulent flows around a full size high-speed train with realistic geometry and operating conditions. The numerical results show that the algorithm is both accurate and efficient, and exhibit a good scalability with tens of millions of degrees of freedom on a supercomputer with up to 4096 processors. |
| 收录类别 | 其他 |
| 语种 | 英语 |
| 源URL | [http://ir.siat.ac.cn:8080/handle/172644/10313]  |
| 专题 | 深圳先进技术研究院_数字所 |
| 作者单位 | 2016 |
| 推荐引用方式 GB/T 7714 | Zi-Ju Liao,Rongliang Chen,Zhengzheng Yan,et al. A scalable implicit finite element solver for the large eddy simulation of 3D incompressible turbulent flows on unstructured meshes[C]. 见:The 20th IMACS WORLD CONGRESS. 厦门. |
入库方式： OAI收割
来源：深圳先进技术研究院
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。