Adaptive mesh large eddy simulations of transitional jet diffusion flames in crossflow

We leverage simultaneous in situ measurements of absorption-weighted temperature and water mole fraction from near-infrared dual comb spectroscopy (DCS) to validate adaptive mesh large eddy simulations (LES) of a transitional methane-air jet diffusion flame in crossflow. The configuration is that of a slot burner positioned in a wind tunnel test section. To enable direct comparisons with DCS data, we create synthetic absorption spectra from the LES and nonlinearly fit the spectra to known absorption models, thereby providing absorption- weighted temperature and water mole fractions for both the LES and DCS. Due to the grid dependence of the widely used eddy dissipation model (EDM) in the near-field diffusion controlled regime, we formulate and test an extended EDM that describes fuel and oxidizer mixing rates based on Fickian diffusion and the local mixture fraction. Using the modified model, we obtain good agreement between LES and DCS measurements at a wide range of locations downstream of the burner. We also investigate the transition behavior of both the native EDM and the modified model, with the latter demonstrating an earlier - and more accurate - transition to turbulence. We further find that mixing and diffusion rates of fuel and oxidizer have amore significant effect than the diffusivity of product species on downstream distributions of water mole fraction for the configuration studied here. Ultimately, we show that fuel and oxidizer mixing rates in the near-field laminar flame region play a dominant role in product species and temperature distributions, primarily through heat release due to combustion, providing insights into computational modeling requirements for adaptive mesh simulations of transitional diffusion flames.