Numerical investigation of potential cause of instabilities in a hydrogen micromix injector array

Hydrogen micromix combustion is a promising concept to reduce the environmental impact of both aero and land-based gas turbines by delivering carbon-free and ultra-low-NOx combustion. The high-reactivity and wide flammability limits of hydrogen in a micromix combustor can produce short and small diffusion flames at lean overall equivalence ratios.

There is limited published information on the instabilities of such hydrogen micromix combustors. Diffusion flames are less prone to flashback and autoignition problems than premixed flames as well as combustion dynamics issues. However, with the high laminar flame speed of hydrogen, lean fuel air ratio (FAR) and very compact flames, the risk of combustion dynamics for micromix flames should not be neglected. In addition, the multi-segment array arrangement of the injectors could result in both potential causes and possible solutions to the instabilities within the combustor.

This paper employs numerical simulations to investigate potential sources of instabilities in micromix flames by modelling an extended array of injectors, represented by either single or multiple injectors with appropriate boundary conditions at elevated pressure and temperature. Both RANS and LES simulations were performed and used to derive the Flame Transfer Function (FTF) of the micromix flames to inform lower order thermoacoustic modelling of micromix combustion. LES simulations indicate that the gain of the FTF is lower than predicted from the RANS simulations indicating a lower risk of high frequency thermoacoustic issues than suggested by RANS.

When LES simulations are conducted for certain representative configurations it is observed that there are persistent high-frequency instabilities due to the interaction of the flames. This phenomenon is not observed when only a single injector is modelled. LES simulations for two injectors are conducted with various geometries and radial boundary conditions to identify the cause of the instabilities. It is concluded that the observed high-frequency instabilities are related to aerodynamic jet instabilities enhanced by both aerodynamic and acoustic feedback and key geometric features affecting the occurrence of the instabilities are identified. Only transient simulations such as LES are able to capture such effects and RANS simulations typically used in early stage design will not identify this issue.