No abstract

Nitric oxide formation in gas turbine combustion depends on four key factors: flame stabilization, heat transfer, fuel–air mixing and combustion instability. The design of modern gas turbine burners requires delicate compromises between fuel efficiency, emissions of oxides of nitrogen (NOx) and combustion stability. Burner designs allowing substantial NOx reduction are often prone to combustion oscillations. These oscillations also change the NOx fields. Being able to predict not only the main species field in a burner but also the pollutant and the oscillation levels is now a major challenge for combustion modelling. This must include a realistic treatment of unsteady acoustic phenomena (which create instabilities) and also of heat transfer mechanisms (convection and radiation) which control NOx generation.In this work, large-eddy simulation (LES) is applied to a realistic gas turbine combustion chamber configuration where pure methane is injected through multiple holes in a cone-shaped burner. In addition to a non-reactive simulation, this article presents three reactive simulations and compares them to experimental results. The first reactive simulation neglects effects of cooling air on flame stabilization and heat losses by radiation and convection. The second reactive simulation shows how cooling air and heat transfer affect nitric oxide emissions. Finally, the third reactive simulation shows the effects of combustion instability on nitric oxide emissions. Additionally, the combustion instability is analysed in detail, including the evaluation of the terms in the acoustic energy equation and the identification of the mechanism driving the oscillation.Results confirm that LES of gas turbine combustion requires not only an accurate chemical scheme and realistic heat transfer models but also a proper description of the acoustics in order to predict nitric oxide emissions and pressure oscillation levels simultaneously.

Computational fluid dynamics (CFD) modelers require high-quality experimental data sets for validation of their numerical tools. Preferred features for numerical simulations of a sooting, turbulent test case flame are simplicity (no pilot flame), well-defined boundary conditions and sufficient soot production. This paper proposes a non-premixed C 2 H 4 /air turbulent jet flame to fill this role and presents an extensive database for soot model validation. The sooting turbulent jet flame has a total visible flame length of approximately 400 mm and a fuel-jet Reynolds number of 10,000. The flame has a measured lift-off height of 26 mm which acts as a sensitive marker for CFD model validation, while this novel compiled experimental database of soot properties, temperature and velocity maps are useful for the validation of kinetic soot models and numerical flame simulations. Due to the relatively simple burner design which produces a flame with sufficient soot concentration while meeting modelers' needs with respect to boundary conditions and flame specifications as well as the present lack of a sooting "standard flame", this flame is suggested as a new reference turbulent sooting flame. The flame characterization presented here involved a variety of optical diagnostics including quantitative 2D laser-induced incandescence (2D-LII), shifted-vibrational coherent anti-Stokes Raman spectroscopy (SV-CARS) and particle image velocimetry (PIV). Producing an accurate and comprehensive characterization of a transient sooting flame was challenging and required optimization of these diagnostics. In this respect, we present the first simultaneous, instantaneous PIV and LII measurements in a heavily sooting flame environment. Simultaneous soot and flow field measurements can provide new insights into the interaction between a turbulent vortex and flame chemistry, especially since soot structures in turbulent flames are known to be small and often treated in a statistical manner.

A sooting C 2 H 4 /air jet diffusion flame was investigated experimentally by laser measuring techniques and the results are compared to CFD calculations. The target flame (C 2 H 4 10.4 g/min, bulk exit velocity 44 m/s, RE = 10000) exhibits well-defined boundary conditions and presents a good test case for model validation. Flow velocity, temperature and soot volume fraction in this flame has been measured previously. In this paper, further experimental results from Raman scattering and laser-induced fluorescence (LIF) measurements are presented to expand the validation data base. Raman scattering is used to measure the fuel/air mixing prior to combustion, while LIF of PAHs monitors the soot precursor region and successive planar OH-LIF serves to map the flame front position and its statistics. Furthermore, a numerical simulation of this flame was performed based on the DLR in-house code THETA. Within the scope of the test case presented here, the code combines a relatively detailed description of the gas phase kinetics coupled with a detailed yet computation-efficient soot model, suitable for CFD applications. This model has been designed to predict soot for a variety of fuels and flames with good accuracy at relatively low computational costs. Universal model parameters are applied, which requires no tuning for the dependence of test case or fuel. The experimental and numerical results are compared and discussed with special emphasis on the pre-flame region of the jet and up to the downstream position where significant soot concentrations are present. Validation shows the general applicability of the CFD code with implemented soot model to rather complex systems like the target sooting turbulent jet flame. Identified discrepancies are analyzed and can be explained, while opening up the field for future optimization of parts of the CFD code.

Soot formation and oxidation were investigated in swirl flames operated with ethylene/air at elevated pressure in a gas turbine model combustor with optical access. Coherent anti-Stokes Raman scattering was used for temperature measurements, laser-induced incandescence for soot concentration and laserinduced fluorescence for the determination of OH radical distributions. A major focus of the experiments was the investigation of the influence of the injection of secondary oxidation air into the fuel-rich product gas of the primary combustion zone. Soot is mainly present in tiny filament-like regions left without OH signal. In the 3 bar flame with oxidation air injection these are found in a region separating the primary combustion zone, fed by combustion air and ethylene, and the secondary combustion induced by oxidation air and unburned hydrocarbons (UHC) that are transported into the inner recirculation zone. The different behavior of flames with and without oxidation air is most pronounced in the inner recirculation zone that is strongly influenced by the oxidation air admixture. This is reflected by changed OH distributions, mean temperatures and the shape of the temperature pdfs and results in significantly different soot distributions. The combined temperature statistics and correlated OH and soot distributions acquired at 3 and 5 bar are well suited to support the understanding of soot formation and oxidation and are expected to be a valuable input to soot model validation.

The accuracy of laser-induced incandescence (LII) measurements is significantly influenced by the calibration process and the laser profile degradation due to beam steering. Additionally, the wavelength used for extinction measurements, needed for LII calibration, is critical and should be kept as high as possible in order to avoid light absorption by molecular species in the flame. The influence of beam steering on the LII measurement was studied in turbulent sooting C 2 H 4 /air flames at different pressures. While inhomogeneities in the laser profile become smoothed out in time averaged measurements, especially at higher pressure, the corresponding single shot beam profiles reveal an increasing effect of beam steering. In the current configuration it was observed that the resulting local laser fluence remains within certain limits (30% to 200%) of the original value. A sufficiently high incident laser fluence can thus prevent the local fluence from dropping below the LII threshold value of approximately 0.3 J/cm 2 at the cost of increased soot surface vaporization. A spatial resolution in the dimension of the sheet thickness of below 1 mm cannot be guaranteed at increased pressure of 9 bars due to beam steering. A feasibility study in a combustor at technical conditions demonstrates the influence of both effects beam steering and choice of calibration wavelength and led to the conclusion that, however, a shot to shot calibration of LII with simultaneously measured extinction can be realized.

No abstract

Mean and instantaneous flow fields were derived for sooting pressurized swirl flames, operated with ethylene/air in an aero-engine model combustor. Stereo particle image velocimetry served to deduce three velocity components and to identify locations of soot based on soot scattering. The measurements complement those of other quantities in the same flames published recently. Flow fields determined for cold and reactive conditions confirm conclusions drawn from application of other laser-based diagnostics: soot is mainly formed in the inner recirculation zone which recirculates reactive, hot unburnt reaction products, and partly transported into the high-velocity in-flow regions. Oxidation air injected after two thirds of the combustor forms a stagnation zone close to the combustor axis and splits into a portion flowing downstream towards the combustor exit and one transported upstream thereby affecting the local gas composition and temperatures in the inner recirculation zone. Analysis of the instantaneous images by proper orthogonal decomposition reveals the existence of a precessing vortex core which impacts the soot distribution. Presence of soot in high-velocity/high strain rate regions where soot formation is unlikely to occur can be explained as a result of transport. Flow field characterization and the correlation with soot presence, in complement of existing data, are expected to provide a valuable contribution to soot model validation.