Prediction and Control of Turbulence Based on Turbulent Elementary Vortices

为了认识湍流中的初等涡并将其应用于工程问题，本文采用直接数值模拟的方法研究了各种湍流中初等涡的特性。结果表明，初等涡旋的特性与湍流类型和雷诺数无关，在高雷诺数湍流中，初等涡旋会形成大规模的团簇。平面通道流动的大尺度DNS表明，近壁面区域涡是普遍存在的，远壁面区域涡呈条纹状结构。此外，在数值实验中，这种结构并不持续，而是出现了卷胞涡，其计算域的展向宽度受到严格限制。并阐明了湍流中初等涡的标量输运机理和标量表面的分形特征。利用DNS进行了壁面湍流主动反馈控制的研究。对控制效果的雷诺数依赖性的研究表明，所提出的控制方法对于实际应用中出现的高雷诺数流也是有效的。结果表明，现有的控制算法对湍流管道流动也是有效的，并证明了仅基于壁面可测量信息的减阻可能性。开发并评价了一种由阵列传感器和放置在矩阵中的可变形膜型磁致动器组成的反馈控制系统原型。用DNS研究了添加剂、壁面操纵和压力梯度对湍流流场的调节。分析了颗粒流中颗粒分布和颗粒旋转不均匀性的影响、织物种植的影响以及加速或减速的空间发展。为了建立新的SGS模型，利用高雷诺数湍流的DNS研究了湍流的层次结构和不同尺度间的能量传递机理。少

To understand elementary vortex in turbulence and to apply it for engineering problems, characteristics of the elementary vortex in various turbulent flows are investigated by direct numerical simulations. It is shown that the characteristics of the elementary vortex are independent of a type of turbulent flows and Reynolds number, and that they create large-scale clusters in high Reynolds number turbulence. Large scale DNS of plane-channel flow reveals that vortices in near-wall region are universal and that the far-wall field is characterized by streaky structures. In addition, such a structure was not sustained but roll-cell vortices appeared in the numerical experiment, in which spanwise width of computational domain was strictly restricted. Scalar transport mechanism by the elementary vortex and fractal characteristics of scalar surface in turbulence are also clarified. R&D studies on active feedback control of wall turbulence are performed by using DNS. Investigation on the Reyno … More lds number dependency of the control effect reveals the effectiveness of the proposed control for high Reynolds number flows appearing in real applications, too. It is shown that the existing control algorithms are also valid for a turbulent pipe flow, and we demonstrate the possibility of drag reduction based on the information measurable on the walls only. A prototype feedback control system with arrayed sensor and deformable-membrane-type magnetic actuators placed in a matrix is also developed and evaluated. Turbulent flow fields modulated by additives, manipulation at the wall or pressure gradient were investigated by means of DNS. Namely, influences of the non-uniformity of particle distribution and particle rotation in particle-laden flows, effect of fabric planting, and spatially developing field with acceleration or deceleration were analyzed. To develop new SGS model, hierarchical structure of turbulence and energy transfer mechanism between different scales are investigated using DNS of high Reynolds number turbulence. Less