A novel predictive dual scale model to accurately and efficiently simulate phase interfaces in turbulent flows

具有相界面的湍流几乎发生在我们日常生活的各个方面，从使用液体燃料的能源系统到药物喷雾。其中许多流动涉及将液体雾化成形成喷雾的小液滴。整个系统的性能在很大程度上取决于这种喷雾的质量。例如，在飞机或内燃机中，改善喷雾质量可以提高效率并减少污染物的产生。在医用喷雾剂中，精确控制气溶胶的喷雾滴大小可以显著改善药物进入肺部的输送。遗憾的是，目前还没有从第一性原理推导出的有效模型来预测和模拟湍流中复杂的喷雾形成过程。因此，增量设计改进理念在许多应用中很普遍，因为全新设计的实验非常昂贵，而且数值模拟缺乏湍流雾化过程的预测模型，因此需要与现有的实验数据进行调整。本项目的目标是建立一个有效的预测模型来模拟湍流中的雾化过程。这样的模型有可能开创一种大胆的新设计理念，以全新的设计为目标，因为使用新模型进行预选可行性研究的成本将大大低于实验研究。该项目还包括重要的教育活动，包括本科生研究人员的参与，以及通过校内活动和校外课堂参观，向当地一流学校进行推广。新模型基于双尺度方法的思想，规避了传统湍流建模方法的局限性，这些方法依赖于从大尺度到小尺度的级联过程的存在。这种级联可能不是雾化相界面的主导，因为在小尺度上占主导地位的表面张力既可以产生也可以湮灭小尺度的界面结构。提出的双尺度方法考虑了这一点，采用多尺度分解，结合了直接数值模拟的特点，有效地解决了小尺度和大尺度的大涡模拟。这种方法固有地结合了作用于多个长度和时间尺度上的不同力和雾化机制之间的相互作用和竞争，并通过对已解决的尺度相界面几何形状应用显式过滤器，实现了需要在更大尺度上建模的项的显式关闭。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Turbulent flows with phase interfaces occur in almost every aspect of our daily lives, ranging from energy systems using liquid fuels to pharmaceutical sprays. Many of these flows involve atomizing a liquid into small-scale drops forming a spray. The performance of the overall system depends strongly on the quality of this spray. For example, in aircraft or internal combustion engines improvements in the quality of the spray can result in efficiency gains and reduction in pollutant production. In medical sprays, precisely controlling the spray drop sizes of aerosols can significantly improve the delivery of pharmaceuticals into the lung. Unfortunately, no efficient model derived from first principle currently exists that can predict and simulate the complex process of spray formation in turbulent flows. Thus an incremental design improvement philosophy is prevalent in many applications because experiments of radically new designs are very costly and numerical simulations lack a predictive model for turbulent atomization processes and thus require tuning with existing experimental data. The goal of this project is to develop an efficient predictive model to simulate atomization processes in turbulent flows. Such a model has the potential to initiate a bold new design philosophy that targets radically new designs because the cost of pre-selection feasibility studies using the new model would be significantly lower than experimental studies. The project also incorporates significant educational activities, including involvement of undergraduate researchers and outreach targeting local Title-One schools through on-campus activities and off-campus classroom visits.The new model is based on the idea of a dual scale approach, circumventing the limitations of classical modeling approaches for turbulent flows that rely on the existence of a cascade process from large to small scales. Such a cascade is likely not dominant for atomizing phase interfaces since surface tension forces that are dominant on the small scale can both generate and annihilate small scale interface structures. The proposed dual scale approach takes this into account by employing a multi-scale decomposition that combines features of direct numerical simulations resolving the small scales with large eddy simulation for the larger scales in an efficient way. This approach inherently incorporates the interaction and competition between different forces and atomization mechanisms acting on multiple length and time scales and enables the explicit closure of terms requiring modeling on the larger scales by applying explicit filters to the resolved scale phase interface geometry.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.