Accelerating Hydrogen Combustion Simulations in Aero-Engines With Advanced CFD and GPGPU Techniques

This study employs advanced Computational Fluid Dynamics techniques to simulate hydrogen combustion in an industrial hydrogen injector for gas turbines. We combine Large Eddy Simulation (LES) and the Dynamic Thickened Flame model (DTFLES) to capture detailed turbulent combustion phenomena. The high computational cost associated with finite-rate chemical kinetics is mitigated by the use of General-Purpose Graphics Processing Unit (GPGPU) computing, accelerating the integration of the chemical Ordinary Differential Equations (ODEs). Numerical results are validated against experimental data, demonstrating the feasibility and efficiency of high-fidelity hydrogen combustion simulations. This approach significantly reduces simulation time while maintaining accuracy. The numerical results presented in this work offer critical insights into the flow mixing dynamics and combustion process in an industrial hydrogen injector, thus paving the way for cleaner and more efficient energy solutions.