Direct numerical simulation of buoyant-convectively driven gas transfer across gas-liquid interfaces

拟议的项目涉及的气体物质通过气-液界面的传输驱动的对流不稳定性。在浮力驱动的对流气体传输中发挥作用的物理机制还没有得到很好的理解，尽管它们对全球热收支和环境重要的气体循环（包括温室气体循环）有重大贡献。 虽然在文献中已经报道了许多经验关系来预测气体传输速度，近表面湍流场和浮力驱动流中的界面气体通量之间的相互作用的动力学尚未得到充分的描述。由于这种低扩散（高施密特数）物质的界面传质的特征在于界面附近的扩散层非常薄，并且在其他区域瞬时出现陡峭的浓度梯度，因此进行详细的实验室测量是极其困难的。同时，现有的直接数值模拟（DNS）-受限于对解析所有运动尺度所需的计算资源的高需求-主要限于低施密特数（通常小于10）和/或低雷诺数。我们建议进行直接的数值模拟浮力驱动的气体传输使用一个专门设计的数值代码的标量对流和扩散的离散化。据我们所知，本研究将是第一个进行直接数值模拟的传质驱动的对流不稳定性在现实的高施密特数高达500。详细的数据将使我们能够首先确定波动的浓度场和速度场之间的相关性-这个量今天在实验室实验中极难高保真地测量。在那些能够测量该量的少数实验中，可信的测量通常限于不太靠近界面且不太深入流体主体的区域。 此外，上述领域的三维时间和空间分辨测量不存在这种流动。拟议项目的主要影响将是在全球CO2预算或水体中氧气再曝气的框架内更可靠地预测气体转移率。更具体地说，我们将确定作为施密特和瑞利数的函数的界面传质的标度律。在一个应用中（如遥感），这一结果将允许重建给定气体的质量传递速率，从重新调整的热场的信息单独。

The proposed project deals with the transfer of gaseous substances across a gas-liquid interface driven by buoyant-convective instability. The physical mechanisms that play a role in buoyancy-driven convective gas transfer are not well understood, despite of their significant contributions to the global heat budget and environmentally important gas cycles (including green-house gas cycles). Though numerous empirical relations to predict the gas transfer velocity have been reported in the literature, the dynamics of the interaction between the near-surface turbulent field and the interfacial gas flux in buoyancy driven flow are yet to be fully described. As the interfacial mass transfer of such low-diffusive (high Schmidt number) substances is characterized by very thin diffusive layers near the interface and the instantaneous occurrence of steep concentration gradients in other regions, performing detailed laboratory measurements is extremely difficult. At the same time existing direct numerical simulations (DNS) - constrained by the high demand on computational resources needed to resolve all scales of motion - are mostly limited to low Schmidt numbers (typically less than 10) and/or low Reynolds numbers. We propose to perform direct numerical simulations of buoyancy-driven gas transfer using a specifically-designed numerical code for the discretization of scalar convection and diffusion. To our knowledge, the present study will be the first to perform direct numerical simulation of mass transfer driven by a buoyant-convective instability at realistically high Schmidt numbers up to 500. The detailed data will allow us to determine first and foremost the correlation between the fluctuating concentration and velocity fields -- a quantity which today is extremely difficult to measure with high fidelity in laboratory experiments. In those few experiments which are able to measure this quantity, the trustworthy measurement is typically restricted to a region not too close to the interface and not too deep into the bulk of the fluid. Furthermore, three-dimensional time- and space-resolved measurements of the aforementioned fields do not exist for this flow. The main impact of the proposed project will be a more reliable prediction of the gas transfer rates in the framework of e.g. the global CO2 budget or the oxygen re-aeration in water bodies. More specifically, we will determine the scaling law of the interfacial mass transfer as a function of the Schmidt and Rayleigh numbers. In an application (such as remote sensing) this result will allow to reconstruct the mass transfer rate of a given gas from rescaling of the information on the thermal field alone.