CDS&E: Extracting Physics from High-Fidelity Simulations of Atomization using Geometric and Topological Data Analysis

许多应用程序模拟在发动机、消防喷头、喷漆系统和制造过程中发现的雾化喷雾。 这些模拟产生了大量的数据集，完全描述了喷雾。 然而，这些系统的设计受到知识和基于物理的雾化过程模型的缺乏的阻碍。 这项工作将结合联合收割机几何和拓扑数据分析与这些丰富的模拟结果，创造新的知识如何在喷雾液体打破，并最终导致改进的雾化模型。 该项目包括开发一个教学模块，将几何技术引入本科和研究生数值方法课程，这将使未来的工程师接触到来自不同学科的想法。该项目的目标是利用几何和拓扑数据分析的进步，结合高保真模拟结果来量化雾化是如何发生的。 第一个目标是将提取算法添加到高级计算流体动力学代码中，以产生包括液体结构形状和周围流场的雾化事件数据库。第二个目标将扩展几何和拓扑描述符，以量化的液体结构和局部流场的形状。描述符将提供一种方法来索引和搜索从高保真模拟的预先计算的数据库内的定量相似的解体事件。第三个目标是利用关于原子化过程的新知识，发展基于物理学的、减少的三阶模型。 在高层次上，工作耦合几何和拓扑分析与高保真度的模拟结果，提供一个新的视角的过程。从这个角度来看，这项工作将允许在涉及物体形状变化的演变的广泛领域取得进展。该奖项反映了NSF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。

Many applications simulate atomizing sprays that are found in engines, fire sprinklers, spray painting systems, and manufacturing processes. These simulations produce massive datasets that fully describe the spray. However, the design of these systems is hindered by the lack of knowledge and physics-based models of the atomization process. This work will combine geometric and topological data analysis with these rich simulations results to create new knowledge on how the liquid in a spray breaks up and ultimately lead to improved atomization models. The project includes the development of a teaching module that will bring geometric technics into undergraduate and graduate numerical methods courses which will expose future engineers to ideas from different disciplines. The goal of the project is to leverage advances in geometric and topological data analysis combined with high-fidelity simulation results to quantify how atomization occurs. The first objective will add extraction algorithms into an advance computational fluid dynamics code to produce a database of atomization events including the shape of and flow field around liquid structures. The second objective will extend geometric and topological descriptors to quantify the shapes of liquid structures and local flow fields. The descriptors will provide a way to index and search for quantitatively similar breakup events within the precomputed database from the high-fidelity simulations. The third objective will develop physics­based, reduced­order models using new knowledge on the process of atomization. At a high level, the work couples geometric and topological analysis with high­fidelity simulation results to provide a new perspective on a process. With this view, the work will allow for advances in a wide range of fields that involve the evolution of objects with changing shape.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.