Abstract This paper presents a mathematical modelling analysis of two propagating flames inside small and large enclosures of different geometrical configurations via computational fluid dynamics. Fuels considered are hydrogen and methane respectively. As the reactivity of the two fuels vary, time histories of over-pressure and flame position vary accordingly. The level of turbulence produced due to impinging the propagating flame with obstructions drives the location of flow amongst the combustion regime diagram and thus controls the validity of combustion models used. In the present study, the large eddy simulation (LES) technique as fitted in ANSYS Fluent is employed. The turbulent flame speed closure (TFC) is selected with revisiting the validity of its standard formulation. As the interaction of the propagating flame with obstructing walls is considered to be a key feature on determining the turbulence levels attained, it was sought to compare two models, namely the Dynamic Smagorinsky-Lilly and WALE models. Despite the fact that small scale experimental facilities are more adequate for producing detailed diagnostics of the deflagration scenario experimentally as well as allowing LES analysis with a reasonable mesh size within computational economy, it was recognized that the terminal objective is to evaluate the level of performance when dealing with large scale cases. The agreement between the predictions and set of experimental measurements used suggests a reasonable level of qualitative agreement for all different cases, providing a better understanding of models used to simulate the phenomena involved with discrepancies spotting light on areas of potential future improvements.

中文翻译：

---

小规模和大规模湍流环境中的易燃燃料建模

摘要 本文通过计算流体动力学对不同几何结构的小型和大型外壳内的两种传播火焰进行了数学建模分析。所考虑的燃料分别是氢气和甲烷。由于两种燃料的反应性不同，超压和火焰位置的时间历程也相应变化。由于障碍物撞击正在传播的火焰而产生的湍流水平驱动了燃烧状态图中流动的位置，从而控制了所用燃烧模型的有效性。在本研究中，采用了 ANSYS Fluent 中的大涡模拟 (LES) 技术。选择湍流火焰速度闭合 (TFC) 并重新审视其标准公式的有效性。由于传播火焰与障碍壁的相互作用被认为是确定所达到的湍流水平的关键特征，因此试图比较两个模型，即 Dynamic Smagorinsky-Lilly 模型和 WALE 模型。尽管小规模实验设施更适合通过实验产生爆燃情景的详细诊断以及允许在计算经济范围内以合理的网格大小进行 LES 分析，但人们认识到最终目标是评估性能水平处理大规模案件。所使用的预测和一组实验测量之间的一致性表明所有不同情况下定性一致性的合理水平，