Viscous Extension of the Classical Theory of Unsteady Aerodynamics

目前最先进的翼面绕流流体动力学模型要么计算成本太高，不能用于初步设计阶段，要么过于简化，忽略了一些基本的物理特性。这项建议的目标是通过开发有效的空气动力学模型来填补这一空白，该模型捕捉到了流动动力学的主要物理原理。依靠基本非定常空气动力学的基本知识和数据科学的最新进展(例如，机器学习技术)的独特组合，该项目将产生更真实的空气动力载荷计算。发展起来的理论对于下一代飞行器(飞行出租车、微型飞行器和无人机)和风力涡轮机的初步设计将是不可或缺的，这将促进国家安全和经济繁荣。该项目将包括在这项研究的基础上开发一门关于非定常空气动力学的研究生课程，正在计划开发一本关于这一主题的教科书，本科生将通过与空气动力学相关的高级项目的方式参与进来。认识到升力产生和涡量产生本质上是粘性过程，提出了非定常空气动力学经典理论的粘性延伸。这一扩展将通过将位势理论与一种特殊的边界层理论相结合来实现，该理论解决了后缘附近的流动细节。这种方法将在不需要任何辅助条件(即类库塔条件)的情况下提供环流动力学。此外，为了推广这种方法，将使用机器学习工具从高保真模拟中提取有关循环动力学的信息。这种建模方法可以由蝴蝶结架构在概念上描述：采用高维输入空间(例如，来自高保真模拟的流场)到更小、更紧凑的低维核心模型(例如，一维循环动力学)的压缩映射。然后，势流理论，与准确的循环动力学，将被用来恢复流场的高维输出空间。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

The current state-of-the-art of fluid dynamic models of the flow over wing sections are either computationally too expensive to be used in preliminary design phases or overly simplified, neglecting some essential physics. The objective of this proposal is to fill this gap by developing efficient aerodynamic models that capture the main physics of the flow dynamics. Relying on a unique combination of the fundamental knowledge of basic unsteady aerodynamics and the recent advances in data science (for example, machine learning techniques), this project will yield more realistic computations of the aerodynamic loads. The developed theory will be indispensable for preliminary design of next generation flying vehicles (flying taxis, micro-air-vehicles, and drones) and wind turbines, which will improve national security and economic prosperity. The project will include the development of a graduate course on unsteady aerodynamics based on this research, plans are underway to develop a textbook on the subject, and undergraduate students will be involved byway of aerodynamic-related senior projects.Realizing that the lift generation and vorticity production are essentially viscous processes, a viscous extension of the classical theory of unsteady aerodynamics is proposed. This extension will be achieved by coupling the potential theory with a special boundary layer theory that resolves the flow details in the vicinity of the trailing edge. This approach will provide the circulation dynamics without any need for an auxiliary condition (i.e., Kutta-like condition). Moreover, to generalize this approach, machine learning tools will be employed to extract information about the circulation dynamics from high-fidelity simulations. This modeling approach can be conceptually described by a bow tie architecture: a compressive mapping taking a high-dimensional input space (e.g., flow field from high-fidelity simulation) to a smaller, more compact lower-dimensional core model (e.g., one-dimensional circulation dynamics). Then, the potential flow theory, with the accurate circulation dynamics, will be used to recover a high-dimensional output space of the flow field.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.