Development of Adaptive CFD Tools for Study of Tornadic Wind Field

除美国外，加拿大比其他任何国家都更频繁地报道龙卷风。这种灾难性的天气可能造成巨大的财产损失，甚至夺去无数人的生命。由于依靠现场观测和实验室实验进行龙卷风动力学研究通常成本高、限制多且耗时长，因此以大涡模拟（LES）或雷诺平均法（RANS）为主要手段的计算机模拟已成为揭示龙卷风风场复杂特征的一个更有吸引力的研究方向。然而，这两种湍流模型都有各自的缺点，如果在整个龙卷风模拟过程中只采用一种湍流模型，这可能会限制计算机建模方法的适用性。此外，大多数龙卷风数值模拟是使用商业软件包进行的，这些软件包通常无法生成适合于各种随时间变化的龙卷风情况的适当自适应网格。为了克服龙卷风建模界面临的这些主要瓶颈，本研究将包括两个目标：(1)基于部分解析数值模拟（PRNS）概念开发一种新的龙卷风模型，使LES、RANS和直接Navier-Stokes （DNS）方法统一为一个无缝的混合体，更适合各种龙卷风风场的模拟需要；(2)推导出一种全新的基于三维ranss的后验误差估计框架，该框架可以独立指导有限元网格适应不同的时变龙卷风情景，而无需逐案选择自适应标准。这两种流的结果将在商业计算流体动力学（CFD）代码中实现，借助用户设计的一套广泛的子程序，针对不同部署的表面结构配置的各种龙卷风风场进行自适应模拟。使用自适应模拟结果，龙卷风对地面建筑物的影响将更准确地进行检查，从而为改进建筑设计提供有价值的见解和指导，以提高建筑的抗风能力。

Tornadoes are more frequently reported in Canada than any other country except the United States. This type of disastrous weather may inflict colossal property damage and even take numerous human lives. Since the study of tornado dynamics relying on field observations and laboratory experiments is usually expensive, restrictive, and time-consuming, computer simulation mainly via the large eddy simulation (LES) or the Reynolds average Navier-Stokes (RANS) method has become a more attractive research direction in shedding light on the intricate characteristics of a tornadic wind field. However, these two turbulence models have their respective drawbacks, which might confine the applicability of the computer modeling approach if only one turbulence model is employed throughout a tornado simulation. Also, most tornado numerical simulations are performed using commercial software packages that are often incapable of generating properly adaptive meshes suitable for various time-dependent tornado cases. In order to overcome such major bottlenecks faced by the tornado modeling community, this research will be composed of two objectives: (1) developing a novel tornado model based on the partially resolved numerical simulation (PRNS) concept so that the LES, RANS and the Direct Navier-Stokes (DNS) methods can be unified into a seamless hybrid, more suitably accommodating the need for simulation of various tornadic wind fields; and (2) deriving a brand new 3-D RANS-based a posteriori error estimation framework that can independently guide a finite element mesh to be adapted for different time-dependent tornado scenarios without needing to select adaptive criteria on a case-by-case basis. The outcome of the two streams will be implemented in commercial computational fluid dynamics (CFD) code with the aid of an extensive suite of user-designed subroutines targeted at adaptive simulation of a variety of tornadic wind fields over configurations of differently deployed surface constructions. Using the adaptive simulation results, the impact of tornadoes on ground buildings will be more accurately examined, leading to valuable insight and guidance for improving design of constructions towards better wind-resistant capabilities.