Structure and dynamics of a reacting jet injected into a vitiated crossflow in a staged combustion system

Secondary injection of the fuel, also referred to as staged combustion, is being studied by gas turbine manufacturers as a means of increasing the power output of the gas turbine systems with minimal contribution to NO x emission. A reacting jet issuing into a high pressure vitiated cross flow operating at gas turbine relevant conditions was investigated as a means of secondary injection. In this application rapid mixing and chemical reaction in the near field of jet injection is desirable. In this work, advanced diagnostic measurements were performed on an experimental representation of such a system, with a transverse jet injection into a swirling vitiated crossflow. High repetition rate simultaneous particle image velocimetry (PIV) and OH planar laser-induced fluorescence (PLIF) were performed at five measurement planes perpendicular to the jet axis. Transverse jets composed of premixed natural gas and H2 diluted with N2 were injected through a tube protruding into the crossflow. The influence of the nature of vitiated crossflow, swirling or uniform, on to the reacting jets and its corresponding influence on flame stabilization mechanism was investigated. The vitiated crossflow is produced by a low swirl burner (LSB) that imparted a swirling component to the crossflow and a bluff-body burner produced a uniform vitiated crossflow. The crossflow exhibits considerable swirl at the location of the transverse jet injection. The PIV measurements clearly demonstrate the influence of a swirling/ uniform crossflow on the jet. The jet-to-crossflow momentum flux ratio was varied to study the corresponding effect on the flow field. Two momentum flux ratios, J=3 and J=8 were employed to study the effect of momentum flux ratio on the stabilization of reaction fronts. The time averaged flow field shows a steady wake vortex very similar to that seen in the wake of a cylindrical bluff body which helps to stabilize the reaction zone within the wake of the jet. Jet with J = 8 had a deeper penetration into the crossflow as compared to J = 3 jet. Velocity field for a reacting/non-reacting jet in swirling crossflow exhibits higher in-plane velocity gradients as compared to jets in uniform crossflow. The vorticity field is also found to be weaker in case of jets in uniform crossflow as a result there is delay in the formation of the wake vortex structure. The HRR data acquisition also provided temporally resolved information on the transient structure of the wake flow associated with the reacting jet in crossflow. The wake Strouhal number calculation provides a better physical insight into the influence of jet velocity profile and nature of crossflow. A decrease in wake Strouhal number is noticed with an increase in nozzle separation distance. The effect of near-field heat release is also apparent from the wake Strouhal number. It is higher for a reacting jet as compared to that of a non-reacting jet owing to increase in rate of dilatation due to heat release. Based on the experimental data, it can be stated that wake vortices play a significant role in stabilizing a steady reaction zone within the near-wake region of the jet. The time averaged OH-PLIF images show a broad region of OH distribution in the wake of the jet. The measurements provided qualitative as well as quantitative information on the evolution of complex flow structures and transient events such as re-ignition, local extinction and vortex-flame interactions in the turbulent reacting flow. There is a noticeable difference in the flame structure of a H2/N2 flame as compared to a premixed natural gas flame. A thin flame front in the windward side of the jet is apparent for a H2/N2 flame. Due to the higher propensity of strain rate induced extinctions and lower flame speeds of a natural gas flame, a stable reaction zone is seen only in the jet wake. Thus, such high-data-rate measurements provide significantly improved understanding of the complex flow-field and flame stabilization mechanisms in a turbulent reacting flow. Such data-sets are critical for the development of high fidelity turbulence-chemistry interaction models