盐指型双扩散湍流二维直接数值模拟研究

Two-dimensional direct numerical simulations of fingering double diffusive convection turbulence

针对海洋相关盐指型双扩散湍流开展了二维直接数值模拟研究,特别关注了四种不同物性参数组合情况下,传输特性和流动结构随瑞利数和密度比的变化趋势。结果表明:温度努塞尔特数、盐度努塞尔特数和流动雷诺数随瑞利数的整体变化规律受密度比的影响较小,但其具体数值随密度比的增加而减小;密度比的影响幅度随着组分扩散率比值,即刘易斯数的增大而减小;相关变化规律可由幂律模型描述,且基于数值模拟结果本文确定了四种物性参数组合所对应的幂律指数。对于大多数算例,流动主导结构为盐指。对具有小刘易斯数和大瑞利数的数个算例,主导结构变为类对流涡结构。对所有物性参数组合,盐指水平宽度随瑞利数变化趋势与线性稳定性分析给出的结果类似。这些结果对理解海洋盐指双扩散湍流、以及实验和模拟结果应用于真实海洋流动等方面有重要的意义。

Double diffusive convection in the fingering regime is a commonly observed phenomenon in the (sub-) tropic ocean. It often happens when both temperature and salinity decrease with depth in the upper water,and plays a significant role in the vertical mixing and transport of heat and salt. In this study, we conduct twodimensional direct numerical simulations of oceanic double diffusive convection in the fingering regime for four different combinations of component properties. We focus on the transport behaviors and flow structures, and their dependences on the control parameters, especially the influences of the Prandtl and Schmidt numbers.Results suggest that the overall trends of the dependences of heat and salt Nusselt numbers and Reynolds number on the Rayleigh number do not change for different density ratio, but their absolute values decrease as the density ratio becomes larger. This influence of density ratio becomes weaker as the Lewis number, i.e., the ratio between the two diffusivities increases. The dependences can be described by power-law scalings. For most of cases, the dominant structures are slender salt fingers. For several cases with small Lewis number and large Rayleigh number, the dominant structures change to large rolls which is similar to the convection turbulence. The dependence of finger width on Rayleigh number matches the prediction of linear stability analysis. All these findings provide useful information for understanding oceanic double diffusive convection, and extending the experimental and numerical findings to real actual problems.