Collaborative Research: Physics-Informed Background-Oriented Schlieren Tomography of Wildfire-Relevant Combustion

随着气候变化使世界大部分地区变得更热、更干燥，野火正变得越来越普遍。除了破坏自然资源和动物栖息地外，发生在人口附近的野火还会对建筑物造成毁灭性破坏，造成严重的发病率、死亡率和经济损失。野火难以预测和控制，因为它们是混乱的、三维的（3D），对当地大气条件极其敏感。尽管它们对野火的特征和控制很重要，但大火中大规模漩涡、条纹和漩涡的形成、演变和相互作用仍然知之甚少。需要新的测量策略来描述和理解野火动态并改进预测模型。该项目将开发一种光学传感器，该传感器使用相机网络，可以对野火相关燃烧过程进行时间分辨率、体积（4D）测量。这项工作将开发、验证并应用一种基于背景定向纹影断层扫描（BOST）的新型定量4D流可视化诊断方法。通过使用流场的简化表示和开发用于层析重建的物理信息关闭，该研究将生成实验室规模野火事件替代物的精确四维重建。实验重建将与相同火灾的模拟进行比较，以了解夹带/拓扑相互作用。现有的BOST算法采用非物理闭包的多步骤求解一组耦合逆问题，并使用未知域的粗表示。所开发的方法将在一个有效的小波基一步完成这些计算。此外，这项工作将产生一个基于流体动力学和燃烧控制方程的“物理信息”BOST闭合，提高精度并促进温度和速度等附加场的量化。模拟和测量将使用简单的固体燃料火灾进行交叉验证。此外，改进后的BOST传感器将有助于首次在复杂的野火相关配置中全面表征火灾动力学，从而更全面地了解火灾。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。

Wildfires are becoming more and more common as climate change renders much of the world hotter and drier. Besides the destruction of natural resources and animal habitats, wildfires that occur near human populations cause devastating damage to structures, resulting in significant morbidity, mortality, and financial losses. Wildfires are difficult to anticipate and control because they are chaotic, three-dimensional (3D), and extremely sensitive to local atmospheric conditions. Despite their importance to the characterization and control of wildfires, the formation, evolution, and interaction of large-scale vortices, streaks, and whirls in fires remain poorly understood. New measurement strategies are needed to characterize and understand wildfire dynamics and improve predictive models. This project will develop an optical sensor using a network of cameras that can perform time-resolved, volumetric (4D) measurements of wildfire-relevant combustion processes.This work will develop, validate, and apply a novel, quantitative 4D flow visualization diagnostic based on background-oriented schlieren tomography (BOST). By employing a parsimonious representation of the flow field and developing a physics-informed closure for tomographic reconstruction, the research will produce accurate 4D reconstructions of laboratory-scale wildfire event surrogates. Experimental reconstructions will be compared to simulations of the same fires to understand entrainment/topology interactions. Existing BOST algorithms solve a set of coupled inverse problems in multiple steps with non-physical closures and use a coarse representation of the unknown field. The developed method will perform these calculations in one step with an efficient wavelet basis. Further, this work will produce a “physics-informed” BOST closure based on the governing equations of fluid dynamics and combustion, enhancing accuracy and facilitating the quantification of additional fields like temperature and velocity. Simulations and measurements will be cross validated using simple solid fuel fires. Moreover, the refined BOST sensor will facilitate comprehensive characterization of fire dynamics in complex, wildfire-relevant configurations for the first time, leading to a more complete understanding.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.