The physical mechanism leading to flame local extinction remains a key issue to be further understood. An analysis of large eddy simulation (LES) data with presumed probability density function (PDF) based closure (Chen et al., 2020, Combust. Flame, vol. 212, pp. 415) indicated the presence of localised breaks of the flame front along the stoichiometric line. These observations and their relation to local quenching of burning fluid particles, together with the possible physical mechanisms and conditions allowing their appearance in LES with a simple flamelet model, are investigated in this work using a combined Lagrangian-Eulerian analysis. The Sidney/Sandia piloted jet flames with compositionally inhomogeneous inlet and increasing bulk speeds, amounting to respectively 70 and 90% of the experimental blow-off velocity, are used for this analysis. Passive flow tracers are first seeded in the inlet streams and tracked for their lifetime. The critical scenario observed in the Lagrangian analysis, i.e., burning particles crossing extinction holes on the stoichiometric iso-surface, is then investigated using the Eulerian control-volume approach. For the 70% blow-off case the observed flame front breaks/extinction holes are due to cold and inhomogeneous reactants that are cast onto the stoichiometric iso-surface by large vortices initiated in the jet/pilot shear layer. In this case an extinction hole forms only when the strain effect is accompanied by strong subgrid mixing. This mechanism is captured by the unstrained flamelets model due to the ability of the LES to resolve large-scale strain and considers the SGS mixture fraction variance weakening effect on the reaction rate through the flamelet manifold. Only at 90% blow-off speed the expected limitation of the underlying combustion model assumption become apparent, where the amount of local extinctions predicted by the LES is underestimated compared to the experiment. In this case flame front breaks are still observed in the LES and are caused by a stronger vortex/strain interaction yet without the aid of mixture fraction variance. The reasons for these different behaviours and their implications from a physical and modelling point of view are discussed in this study.

导致火焰局部灭绝的物理机制仍然是一个有待进一步了解的关键问题。基于假设概率密度函数 (PDF) 的闭包分析大涡模拟 (LES) 数据 (Chen et al. , 2020, Combust. Flame，卷。212, pp. 415) 表明火焰前沿沿化学计量线存在局部断裂。这些观察结果及其与燃烧流体颗粒局部淬火的关系，以及允许它们出现在具有简单火焰模型的 LES 中的可能物理机制和条件，在这项工作中使用组合拉格朗日-欧拉分析进行了研究。Sidney/Sandia 引导的喷射火焰具有成分不均匀的入口和增加的体积速度，分别达到实验吹气速度的 70% 和 90%，用于该分析。被动流示踪剂首先在入口流中播种，并在其生命周期内进行跟踪。在拉格朗日分析中观察到的临界情景，即燃烧粒子穿过化学计量等值面上的消光孔，然后使用欧拉控制体积方法进行研究。对于 70% 的吹散情况，观察到的火焰前沿破裂/消光孔是由于冷的和不均匀的反应物，这些反应物被喷射/引导剪切层中引发的大涡流投射到化学计量等值面上。在这种情况下，仅当应变效应伴随着强烈的亚网格混合时才会形成消光孔。由于 LES 解决大规模应变的能力，并且考虑了 SGS 混合分数方差对通过小火焰歧管的反应速率的弱化影响，该机制被无应变小火焰模型捕获。只有在 90% 的排放速度下，潜在燃烧模型假设的预期限制才变得明显，与实验相比，LES 预测的局部消光量被低估了。在这种情况下，在 LES 中仍然观察到火焰前沿断裂，并且是由更强的涡旋/应变相互作用引起的，但没有混合物分数变化的帮助。本研究从物理和建模的角度讨论了这些不同行为的原因及其影响。