Thermal convection at high aspect ratio: The local distribution and the dynamics of dissipative processes

大多数包含对流换热的地球物理和工程过程的特点是流体层的横向和垂直范围之间的比例很大。然而，关于热对流的实验和数值工作仍然集中在构型上，在这些构型中，热对流比相当小(等于或小于1)。特别是，这是因为获得类似的高瑞利数的实验必须非常高，并且通常不能用相应的宽度来补偿这种大的高度。因此，实际上不可能完全评估通常基于一个没有横向约束的一维模型的现象学理论。申请人希望通过在同时满足高深宽比(8或更大)和高瑞利数(高达10{12})的实验中测量热通量和热耗散率来克服这一点。与前一个项目不同，他们在新的工作中将重点放在这两个量的动力学上。它是从Scheel和Schumacher[J.Schumacher和J.D.Scheel的直接数值模拟(DNS)中得知的。湍流对流槽中羽流碰撞引起的极端耗散事件。太棒了。E 94,043104(2016年)]特别是，热耗散率在当地和时间上都有非常强烈的波动。现在，这将首次在实验中得到验证。由于测量覆盖的“观察期”可能比可比的数字要长得多，测量数据的统计不确定度在第二部分将小得多，申请者将提出以下问题：单个点(在湍流Rayleigh-Bénard对流中)的本地和时间平均测量数据对整个相应水平面的代表性如何。对于小于或等于1的长宽比，过去一直在研究这个问题，并以不具代表性的方式回答。对于大长宽比，侧壁只影响流体层的一小部分外部部分，申请者将通过测量多个水平面上不同位置的热耗散率来回答这个问题。因此，这些测量还将提供局部湍流热通量的分布。

The majority of geophysical and engineering processes incorporating convective heat transfer is characterized by a large ratio between the lateral and the vertical extent of the fluid layer. However, experimental and numerical work on thermal convection still focuses on configurations, where this ratio is rather small (equal or smaller than one). In particular, this is due to the fact that experiments achieving comparable high Rayleigh numbers must be very high and quite often this large height cannot be compensated by a respective width. Thus, it is actually not possible to fully evaluate phenomenological theories that are usually based one a one-dimensional model without lateral confinement. The applicants wish to contribute to overcome this by measurements of the heat flux and the thermal dissipation rate in an experiment meeting both, high aspect ratio (eight or larger) as well as high Rayleigh number (up to 10{12}). Unlike in the preceding project, they will focus in the new work on the dynamics of these two quantities. It is known from direct numerical simulations (DNS) by Scheel and Schumacher [J. Schumacher and J. D. Scheel. Extreme dissipation event due to plume collision in a turbulent convection cell. Phys. Rev. E 94, 043104 (2016)] that, in particular, the thermal dissipation rate fluctuates very strongly as well locally as in time. Now, this will be validated in an experiment for the first time. Because, measurements potentially cover a much longer “observation period” than comparable DNS, the statistical uncertainty of the measurement data will be significantly smaller In a second part, the applicants will pursue the following question: How representative are local and time-averaged measurement data at a single point (in turbulent Rayleigh-Bénard convection) for the entire respective horizontal plane. For aspect ratios smaller or equal to one, this problem is always studied in the past, and it is answered with “not representative”. For large aspect ratios, where the sidewall affects only a little outer fraction of the fluid layer, the applicants will answer this question by measurements of the thermal dissipation rate at various positions in multiple horizontal planes. These measurements will, thus, provide also the distribution of the local turbulent heat flux.