Laminar flame speed modeling of pre-vaporized jet fuel/hydrogen mixtures under engine conditions

Xiaoxu Zhang,Yuxuan Chen,Jian Liu,Hua Zhou,Lingyun Hou,Zhuyin Ren
DOI: https://doi.org/10.1016/j.fuel.2024.133149
IF: 7.4
2024-09-23
Fuel
Abstract:Hydrogen combustion stands out as a key transition strategy to a carbon–neutral society. As hydrogen features unique properties, it poses great challenges for flame stabilization in hydrogen-fueled combustors. This study firstly uses the laminar flame speed (LFS) data of jet fuel/H 2 mixture measured by experiment to validate the one-dimensional laminar flame simulations with a 56-species mechanism for n-dodecane/H 2 mixtures. Then the coupled thermal-chemical effects on n-dodecane/H 2 flames are analyzed. It is found that hydrogen may strongly affect NTC-related flame structure under engine-relevant pressure and temperature. In addition, the unburnt temperature and pressure are found to exhibit strong coupling effects on LFS for n-dodecane/H 2 flames which further complicates the modelling of LFS for n-dodecane/H 2 mixtures. Both analytical and ANN models have been constructed for pre-vaporized n-dodecane/H 2 flames under a wide range of thermodynamic conditions including pressure, equivalence ratio, and hydrogen content. The former one utilizes a double-Gaussian formulation, in which the hydrogen content is directly incorporated to account for the effect of hydrogen doping on LFS. Isentropic compression with variable compression efficiency is applied to correlate unburnt temperature and pressure. The latter one applies a three-layer ANN with the pressure, equivalence ratio and hydrogen content as the input and the LFS as the output. Both models demonstrate reasonable predictive capabilities for LFS compared to one-dimensional simulations of freely propagating flames. The analytical and ANN models provide complementary modeling strategies to capture the impacts of hydrogen doping on LFS under engine conditions. The analytic model is more efficient than the ANN model, while the ANN model provides higher accuracy.
energy & fuels,engineering, chemical
What problem does this paper attempt to address?