No abstract

Gas turbine combustors with lean combustion injectors are prone to thermo-acoustic/combustion instabilities. Several passive techniques have been developed to control combustion instabilities, such as using Helmholtz resonators or viscous dampers using perforated liners that have potential for broadband acoustic damping. In this paper the role of single-walled cooling liners is considered in the damping of acoustic waves and on the flame transfer function in a sample bluff-body burner. Three liner geometries are considered: no bias flow (solid liner), normal effusion holes, and grazing effusion holes at 25° inclination. Cold flow experiments with speaker forcing are carried out to characterise the absorption properties of the liner and compared with an acoustic network model. The results show that whereas the bulk of the acoustic losses is due to the vortex recirculation zones, the liners contribute significantly to the absorption over a wide area of the frequency range. The flame transfer function gain is measured as a function of bias flow for a given operating condition of the burner. The experiments show that for the geometry considered, the global flame transfer function is little affected by cooling except in the case of the normal flow holes. Further analysis shows that whereas the total flame transfer function is not affected, the flame heat release becomes more spatially distributed along the axial length, and a 1D flame response shows distinct modes corresponding to the modal heat release locations.

In the present study, two measurement techniques are adopted to evaluate the fuel-air mixing under atmospheric conditions using an industrial fuel-air pre-mixer. These techniques are CO2 mixing and Planar Laser Induced Fluorescence (PLIF) in water. In these techniques, CO2 and fluorescent dye are injected as fuel simulants. CO2 measurements are used to validate PLIF in water. In the CO2 technique, CO2 concentrations are converted to fuel mass fractions whereas, in the PLIF technique, a modified post processing method is used to convert the LIF signal into fuel mass fraction. The experiments are conducted at the same Reynolds number and momentum flux ratio for two injection strategies. To study the effect of the flow aerodynamics on the mixing results, high speed PIV measurements are conducted in water at the same Reynolds number. A comparison of fuel concentrations measured with the CO2 and PLIF techniques shows good quantitative agreement at all momentum flux ratios. However, deviations between the two techniques are observed at high fuel concentration gradients. The unsteady mixing is evaluated using PLIF technique with high temporal resolution. Analysis of PIV and PLIF data shows that unsteady mixing is lower at regions of high fluctuations in velocity. Moreover, it is found that there is high unsteady mixing at locations where there is high concentration gradient.

A novel airblast injector is designed for gas turbine combustors. Unlike standard pressure swirl and prefilming/non-prefilming air blast atomizers, the novel injector is designed to improve the fuel injection delivery to the injector and improve atomization of the fuel by using a porous stainless steel tube. There are three swirling air streams in the injector. The liquid fuel is injected through the porous tube, with 7 micron porosity, between the swirling air streams, viz. an inner swirling air through the tube and the other two swirling air streams merging downstream of the tube. The swirl vane angles and the air split ratio are selected to increase the amount of air through the injector and facilitate the atomization process. The liquid fuel is injected through the outer surface of the porous tube, due to the permeability of the tube, produces a thin liquid sheet on the inner surface of the tube. The atomization occurs by surface stripping of the liquid sheet. The advantage of such an injector is that it produces a liquid sheet with uniform thickness around the circumference of the tube under all liquid loading. The porous tube also increases the surface area of contact between the fuel and air and produces a fine spray at engine idle conditions. An experimental approach is adopted in the present study to characterize the spray and aerodynamics of the injector for Jet-A and Gas-To-Liquid (GTL) fuels at atmospheric conditions. The effect of flare height on the Sauter Mean Diameter (SMD) is also studied. Spray characterization, droplet size and volume flux are investigated with PDI measurements. The effect of pressure drop and fuel properties on SMD distribution is analyzed. Velocity profiles at downstream of the injector are obtained from LDV measurements, and the velocity profile at the exit of the injector is also analyzed. A central toroidal recirculation zone (CTRZ) is observed at the exit of the injector. The effect of different configurations of the injector on spray characteristics is studied. A correlation for SMD is obtained.

A generic novel injector was designed for multi-Lean Direct Injection (M-LDI) combustors. One of the drawbacks of the conventional pressure swirl and prefilming type airblast atomizers is the difficulty of obtaining a uniform symmetric spray under all operating conditions. Micro-channels are needed inside the injector for uniformly distributing the fuel. The problem of non-uniformity is magnified in smaller sized injectors. The non-uniform liquid sheet causes local fuel rich/lean zones leading to higher NOx emissions. To overcome these problems, a novel fuel injector was designed to improve the fuel delivery to the injector by using a porous stainless steel material with 30 μm porosity. The porous tube also acts as a prefilming surface. Liquid and gaseous fuels can be injected through the injector. In the present study, gaseous fuel was injected to investigate injector fuel-air mixing performance. The gaseous fuel was injected through a porous tube between two radial-radial swirling air streams to facilitate fuel-air mixing. The advantage of this injector is that it increases the contact surface area between the fuel-air at the fuel injection point. The increased contact area enhances fuel-air mixing. Fuel-air mixing and combustion studies were carried out for both gaseous and liquid fuel. Flame visualization, and emissions measurements were carried out inside the exit of the combustor. The measurements were carried out at atmospheric conditions under fuel lean conditions. Natural gas was used as a fuel in these experiments. Fuel-air mixing studies were carried out at different equivalence ratios with and without confinement. The mass fraction distributions were measured at different downstream locations from the injector exit. Flame characterization was carried out by chemiluminescence at different equivalence ratios and inlet air temperatures. Symmetry of the flame, flame length and heat release distribution were analyzed from the flame images. The effects of inlet air temperature and combustion flame temperature on emissions was studied. Emissions were corrected to 15% O2 concentration. NOx emissions increase with inlet air temperature and flame temperature. Effect of flame temperature on NOx concentration is more significant than effect of inlet air temperature. Fuel-air mixing profile was used to obtain mass fraction Probability Density Function (pdf). The pdfs were used for simulations in Chemkin Pro. The measured emissions concentrations at the exit of the injector was compared with simulations. In Chemkin model, a network model with several PSRs (perfectly stirred reactor) were utilized, followed by a mixer and a PFR (plug flow reactor). The comparison between the simulations and the experimental results was investigated.