| dc.description.abstract | With air traffic expected to grow 5% annually until the year 2030, alternative fuels such as hydrogen are being investigated in order to tackle the current environmental crisis. Due to safety concerns, future hydrogen combustion chambers will require new designs of injection systems and are expected to operate under multimode combustion regimes. From a large-eddy-simulation (LES) perspective, a prerequisite for the shift toward new hydrogen combustion chamber technologies is a robust turbulent combustion model capable of functioning in non-premixed conditions. Turbulent combustion modeling using flame front filtering is a well-developed strategy in premixed combustion (filtered-tabulated chemistry for large-Eddy-simulation (F-TACLES)). This approach has been extended to non-premixed flames however, it suffers from high flame filter size sensitivity. Moreover, thin hydrogen flame fronts will result in lower resolution on the LES grid, potentially amplifying this issue. In order to address the feasibility of the non-premixed F-TACLES model applied to hydrogen fuel, simple one-dimensional and two-dimensional laminar counterflow diffusion flames are computed. The model is then tested on the three-dimensional Sandia hydrogen jet flame with a Reynolds number of 10,000. Simulations and a priori tests show that tabulated subgrid-scale correction terms are stiff and can result in nonphysical results, however the model is capable of correctly reproducing non-premixed flame structures for small filter sizes. | |
