There is a need for predictive combustion simulations to reach climate neutrality with future aero-engine designs, but addressing hydrogen combustion makes it complex because of its special properties, which include the particularly high burning velocity, wide flammability limits, and high diffusivity. The latter can lead to thermo-diffusive flame instabilities in premixed systems, which strongly impact fuel consumption speeds in both laminar and turbulent flames. Also, non-premixed combustion is impacted by hydrogen’s high diffusivity and special combustion chemistry.
After the stringent emission goals declared in “Flightpath 2050” and due to safety risks when working with fully premixed flames (risk of flashback), industry partners such as Rolls-Royce aim to investigate further the possibility of injecting both reactants separately inside the combustion chamber, by using Rich-Quench-Lean (RQL) concepts adapted to H2 combustion. However, hydrogen flames are eager to anchor in the injector lips, giving rise to higher nitrogen oxide (NOx) formation typical of diffusion flames. In order to diminish emissions, it is possible to move into a lifted flame concept by setting the appropriate parameters with a dual-swirl injector.
Previously, the combination of flamelet-based tabulated manifolds coupled with Large-Eddy Simulation (LES) showed success with hydrocarbon fuels. Consequently, the objective of this work is to verify current modeling strategies and their effectivity in predicting the thermochemical states linked to a lifted partially-premixed turbulent flame, by taking as reference an academic experimental configuration. Besides, improvements to the current modeling strategies are suggested based on the numerical results.
For further information on the project tasks and partner institutions, find HESTIA here (opens in new tab).
Methodology
Further developed CFD models: The simulations are performed using the CFD solver PRECISE-UNS developed and maintained by Rolls-Royce. For this project RANS (k-ω SST), SAS and LES simulations will be consecutively performed. In the context of LES, the computational grid acts as a spatial filter. The non-resolved subgrid viscosity is modeled using the σ-model by Nicoud et al. The Artificially Thickened Flame (ATF) model adapted to hydrogen particularities will be employed to simulate the turbulent flame macrostructure.
Tabulation strategies: Combustion models using a flamelet approach are available in the CFD solver and are always further developed at Rolls-Royce and STFS. The challenge consists in analyzing the necessary dimensions to add into the flamelet manifold (enthalpy, curvature) in order to model the thermo-chemical states present in the H2-air project partners' combustor concepts. Heat losses implementation and boundary conditions treatment will be equally analyzed, so as to approach the best possible the experimental trends.
Key Scientific Takeaways
- The overall model must be capable of reproducing the global physics of the configuration
- Importance of subgrid-scale (SGS) modeling and heat losses
- Improve SGS modeling applied to H2-air flames
- Assess flamelet tabulation effectiveness applied to H2-air flames
- Consider reduced complexity configurations mimicking the involved physics
Funding and cooperation
The project receives financial support through HORIZON Research and Innovation (EU funding), project number 101056865. The project HESTIA gathers 18 universities and research centers, as well as the 6 European aero-engine manufacturers, to significantly prepare, coherently and robustly, for the future development of environmentally friendly combustion chambers. The calculations for this research were conducted on the Lichtenberg high-performance computer at TU Darmstadt.
