NUMERICAL INVESTIGATION OF 100% PREMIXED HYDROGEN COMBUSTOR AT GAS TURBINES CONDITIONS USING DETAILED CHEMISTRY

The combustion properties of hydrogen make premixed hydrogen-air flames prone to flashback. Several combustor concepts have been proposed and studied in the past few years to tackle the problem of flame flashback in premixed high hydrogen fuel combustors. This study looks at one of the concepts which uses the Aerodynamically Trapped Vortex to stabilize the flame. Burner concepts based on trapped vortex flame stabilization have a higher resistance towards flame blowout than conventional swirl stabilized burners. This work looks at the flow and flame behavior in the proposed Aerodynamically Trapped Vortex Combustor for 100% premixed hydrogen operation. Numerical simulations for the analysis were performed with the commercial CFD simulation package AVL FIRETM. The flow field characterization was focused on the investigation of the influence of both the inlet velocity and inlet turbulence intensity on the mean velocity, wall velocity gradient and turbulence intensity in the combustor. To study the flame stabilization mechanism, reactive simulations were performed at two fuel equivalence ratios. The combustion regime of the flame, turbulent flame speed and temperature distribution in the combustor were quantified from the simulation results. Combustion is modelled using a detailed chemistry solver with the k - epsilon turbulence model to resolve turbulence. No additional turbulence-chemistry interaction model is used in the current research. To reduce chemistry computational time, the multi-zone method is employed. To capture the effect of preferential diffusion, two approaches were used to quantify the diffusion coefficient of each species. The diffusion coefficients were calculated using both mixture averaged approach and the multi component diffusion approach. The proposed design for the Aerodynamically Trapped Vortex combustor was able to stabilize a 100% premixed hydrogen flame without flashback for the simulated conditions.