Novel subgrid-scale modelling for large-eddy simulation of turbulent flow and dispersion in urban environments

近年来，由于意外或故意将有害物质释放到城市环境中而导致的释放到大气中的污染物的顺风扩散的环境和毒理学影响已成为日益重要的问题。与这一重要问题相关的主要挑战涉及获得更深入的了解动态演变的流动结构与复杂的建筑物几何形状在城市环境中的相互作用，以及耦合的动量，热能和被动标量传输过程。与基于雷诺平均Navier-Stokes（RANS）方法的传统方法相比，该研究旨在开发新的动态线性和非线性亚网格尺度（SGS）应力和标量通量模型，用于城市流动和扩散的高性能大涡模拟（LES）。为了达到本研究的目的，本研究分为两个部分。第1部分侧重于预测的主动和被动标量的湍流输送的LES方法的基本方面;和第2部分侧重于应用先进的SGS建模方法，在第1部分中开发的改进LES的湍流输送的危险物质在复杂的城市环境。为了验证新的LES方法在这项研究中开发的，得到的数值结果将与实验和直接数值模拟（DNS）数据进行比较。预计拟议中的研究将提供一个系统的比较研究现有的和拟议的SGS应力和标量通量模型;推进LES的方法，高性能计算的城市流量和分散;并导致更深入的物理理解的亚网格尺度运动及其相互作用的相干流结构内和上方的城市冠层。预计拟议的研究还将提供机会，在机械和环境工程、先进科学计算以及国防科学和技术等领域培训高素质人才。

The environmental and toxicological impact of the downwind dispersion of contaminants released into the atmosphere, resulting from an accidental or deliberate release of a hazardous material into an urban environment, has become an increasingly important problem in recent years. The major challenge associated with this important problem involves obtaining a deeper understanding of the interaction of the dynamically evolving flow structures with the complex building geometry in an urban environment, as well as the coupling of the momentum, thermal energy and passive scalar transport processes. In contrast to conventional methods based on the Reynolds averaged Navier-Stokes (RANS) approach, the proposed research aims at developing novel dynamic linear and nonlinear subgrid-scale (SGS) stress and scalar-flux models for high-performance large-eddy simulation (LES) of urban flow and dispersion. In order to achieve the goal of this research, this proposal is divided into two parts. Part 1 focuses on the fundamental aspects of the LES methodology for prediction of turbulent transport of active and passive scalars; and Part 2 focuses on application of the advanced SGS modelling approaches developed in Part 1 for improved LES of turbulent transport of hazardous substances in complex urban environments. In order to validate the novel LES approaches developed in this research, the obtained numerical results will be compared with experimental and direct numerical simulation (DNS) data. It is anticipated that the proposed research will provide a systematic comparative study of the existing and proposed SGS stress and scalar-flux models; advance the methodology of LES for high-performance computing of urban flow and dispersion; and lead to a deeper physical understanding of the subgrid-scale motions and their interactions with the coherent flow structures within and above the urban canopy. It is also anticipated that the proposed research will provide an opportunity to train highly qualified personnel in areas related to mechanical and environmental engineering, advanced scientific computing, and defence science and technology.