Investigation of the dynamics of large-scale circulation patterns in turbulent convection cells at very large aspect ratios with direct numerical simulations and experiments in compressed sulfur hexafluoride

自然界和技术中的许多湍流对流过程都存在于扩展的层或盒中，并显示出规则有序流动模式的层次，尽管相应的瑞利数和雷诺数表明了一个充分发展的湍流。在这个项目中，我们想要详细研究这种大尺度结构形成的时空动态。因此，压缩六氟化硫实验将与基于谱元法的大规模并行直接数值模拟相结合。这两种方法都将允许在瑞利数的非常大的纵横比系统中进行对流研究，这在以前是无法实现的。分析将澄清这种大尺度环流模式是否是从对流和弱非线性状态开始一直持续到完全湍流状态的遗迹。我们想确定模式演变的时间尺度，并研究它们的瑞利数依赖关系。该实验使我们能够研究相对于非布辛尼斯克效应的循环模式的稳健性。在模拟的基础上，我们希望开发振幅模型，该模型可以通过几个主导自由度来描述这些模式，从而在技术应用中产生对这些模式的控制的可能性。

Many turbulent convection processes in nature and technology are present in extended layers or boxes and show hierarchies of regular ordered flow patterns although the corresponding Rayleigh and Reynolds numbers suggest a fully developed turbulence. In this project we want to study this spatio-temporal dynamics of large-scale structure formation in detail. Therefore, experiments with compressed sulfur hexafluoride will be combined with massively parallel direct numerical simulations based on a spectral element method. Both approaches will allow for convection studies in very large aspect ratio systems at Rayleigh numbers that were not accessible before. The analyses will clarify if such large-scale circulation patterns are relics persisting all the way from the onset of convection and the weakly nonlinear regime into the full turbulent regime. We want to identify the time scales on which the patterns evolve and investigate their Rayleigh number dependence. The experiment allows us to study the robustness of the circulation patterns with respect to non-Boussinesq effects. Based on the simulations, we want to develop amplitude models which can describe the patterns by means of a few dominant degrees of freedom giving rise to the possibility of a control of these patterns in technological applications.