| 项目编号: | 1604224
|
| 项目名称: | Quantifying Numerical Dissipation in CFD Codes |
| 作者: | Julian Domaradzki
|
| 承担单位: | University of Southern California
|
| 批准年: | 2016
|
| 开始日期: | 2016-07-01
|
| 结束日期: | 2019-06-30
|
| 资助金额: | 300264
|
| 资助来源: | US-NSF
|
| 项目类别: | Standard Grant
|
| 国家: | US
|
| 语种: | 英语
|
| 特色学科分类: | Engineering - Chemical, Bioengineering, Environmental, and Transport Systems
|
| 英文关键词: | cfd
; truncation error
; arbitrary cfd code
; cfd user
; cfd software
; simulation
|
| 英文摘要: | PI: Domaradzki, Julian
Proposal Number: 1604224
Results of numerical simulations for fluid flows obtained using Computational Fluid Dynamics (CFD) are always contaminated by errors caused by the representation of the governing equations through a set of equations that can be solved numerically by computers. This project is about predicting these errors and developing simulation procedures to account for them in CFD software.
Truncation errors in CFD can only be neglected if all scales are well resolved by a given mesh and time-step size. However, for most turbulent flows at high Reynolds numbers resolving all physically relevant scales is not feasible computationally. Simulating such flows often requires to model the contributions of unresolved scales by turbulence modeling procedures, leading to Reynolds-averaged Navier-Stokes (RANS) simulations or large-eddy simulations (LES), known as explicit turbulence modeling. In recent years, it has been observed that the truncation errors at spatial resolutions used in such methods may be of the same order as explicit turbulence modeling terms, i.e., the truncation errors can be of a similar magnitude as physical effects one attempts to model. Yet despite the observed importance of the truncation errors in CFD quantifying their effects remains a challenging, and largely unresolved task. The truncation errors can be analyzed analytically, but the analysis is often limited to one dimensional model problems because of the complications of three dimensionality, iterative procedures, and nonlocal pressure effects make it impossible otherwise. Because of that CFD normally employs a number of other, indirect verification/validation procedures, e.g., comparison with experimental data or running simulations on progressively refined meshes to establish mesh independent results. This significantly increases required computational effort. It is thus clear that if information about truncation errors were more easily accessible it would provide an important means of improving all CFD analyzes. The primary goal of the proposed work is to develop, validate, and implement in a software package a general computational procedure that allows a CFD user to estimate dissipative properties of the truncation error for an arbitrary CFD code in an arbitrary simulation case, either as a concurrent part of a given simulation or as a post-processing step. This research could impact the use of CFD for modeling flows in many applications, e.g., around airplanes and ground vehicles; flows in internal combustion engines, turbomachinery, and in rockets; blood flow in a human body; heating, air conditioning, and ventilation flows in buildings; atmospheric and oceanic flows responsible for weather, climate, and spreading of pollutants. In addition to college student training, K-12 educational activities are planned that take advantage of existing initiatives within the School of Engineering at the University of South California. |
| 资源类型: | 项目
|
| 标识符: | http://119.78.100.158/handle/2HF3EXSE/91990
|
| Appears in Collections: | 全球变化的国际研究计划
科学计划与规划
|
|
There are no files associated with this item.
|
| Recommended Citation: | |
Julian Domaradzki. Quantifying Numerical Dissipation in CFD Codes. 2016-01-01.
|
|
|