LES of premixed jet flames subjected to extreme turbulence using flamelet-generated manifolds: a comparison of unstrained and strained flamelets

Abstract

Two piloted premixed CH${}_4$/air jet flames with extreme turbulence are modelled using the Flamelet-Generated Manifold (FGM) method in the context of Large Eddy Simulation (LES). Free-propagating and reactant-to-product laminar flames are adopted to build unstrained and strained chemical databases, respectively. Detailed comparison of the LES-FGM results using the unstrained/strained databases is conducted which is validated against the direct numerical simulation results. The results show that whether including straining effects in the database will not influence the modelling of temperature and major species (e.g. CH${}_4$, O${}_2$ and CO${}_2$), but using a strained database is essential for predictions of intermediate minor species (e.g. H${}_2$, CH${}_2$O, OH and CO). The main reason is that the major species and temperature are almost insensitive to the straining due to the supportive effects from the pilot flame, which also tentatively explains why the unstrained database can be employed to model these strong turbulent flames. In contrast, the intermediate minor species are more prone to the straining, so that using a strained database is crucial. The preferential diffusion effects on the species and temperature predictions are examined to be negligible. It is proposed that adding extra transport equations for these minor species also leads to satisfactory predictions if using an unstrained database, which helps to include the straining effects on the minor species. This study is expected to be significant in high-fidelity simulations of strong turbulent flames using flamelet-tabulated combustion models.

Keywords:

• large-eddy simulation
• flamelet-generated manifold
• strong turbulent flame
• intermediate minor species
• straining effects

Acknowledgments

The authors sincerely acknowledge the support from combustion group of Prof. Jeroen van Oijen in Eindhoven University of Technology, and thanks to the support from Prof. Wang Han in Beihang University. The DNS data provided by Prof. Haiou Wang in Zhejiang University is gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study is supported by National Natural Science Foundation of China (52206169), and China Postdoctoral Science Foundation (2022M712523).