从根子上实现气动计算湍流模型的突破

湍流是制约我国航空航天飞行器设计的卡脖子问题。在计算流体动力学（ CFD ）日益成为飞行器设计的主要手段时，如何构建湍流的数学物理模型是最突出的基础科学难题。我们以复杂系统的新认识为基础，从理论、实验、计算等多方面形成一个突破湍流世纪难题的新思路。理论上，我们构建了结构系综理论新框架，在物理思想上将经典的朗道平均场理论推广到非平衡系统，形成基于广义李群变换的多层结构模型；在方法上创新发展了序函数分析方法，从精确数值模拟和实验测量的经验数据中提炼湍流场的物理结构参数。这一工作在探索飞行器相关的壁湍流的物理对称性原理方面取得了圆满的成功，首次获得规范壁湍流的解析解，确认了新的普适卡门常数0.45。同时，我们以钱学森系统工程的思想为指导，组织了湍流实验和计算平台的建设，针对飞行器典型流场实现了实验、计算和理论的联合攻关，并组织实施了工程计算平台的建设。这些努力为新一代工程湍流模型的研制奠定了坚实的基础。

Breakthrough in Modeling Turbulence for Aerody-namics Computations

It is known that turbulence model is the most diffi-cult bottleneck for achieving high quality aerodynamics computations , or computational fluid dynamics （ CFD） . Here, we report the results of a 973 project, targeting in improving the understanding of mean flow properties of turbulence from a synthesized cooperation of theoreti-cal, experimental and computational efforts .Theoreti-cally , we have developed a new approach , called struc-tural ensemble dynamics （ SED） , which has achieved a breakthrough in finding solutions to modeling wall-bounded flows relevant to aerodynamics computations . The SED generalizes the Landau＆#39;s mean-field theory to non-equilibrium situations by developing a concept of order function which obeys proper Lie-group symmetry. The theory predicts a multi-layer structure for general wall-bounded flows and develops a series of methods quantifying the multi-layer structure , yielding analytic prediction for mean velocity , mean kinetic energy and mean temperature .The predictions are in very good a-greement with empirical data from high precision experi-mental measurements and numerical simulations; the accuracy is unprecedented （99%）.The most outstand-ing achievement is the identification of universal Kar-man constant, 0.45, for canonical wall-bounded flows （ channel, pipe and turbulent boundary layer ）.The 973 project has also established a series of research platforms for performing compressible turbulence experi-ments and computations , and for aerodynamics CFD computations .These efforts have paved the way to de-veloping more sophisticated （ faster and more accurate ） CFD tools.