ASME Conference Presenter Attendance Policy and Archival Proceedings

2013;():V01AT00A001. doi:10.1115/GT2013-NS1A.

This online compilation of papers from the ASME Turbo Expo 2013: Turbine Technical Conference and Exposition (GT2013) represents the archival version of the Conference Proceedings. According to ASME’s conference presenter attendance policy, if a paper is not presented at the Conference, the paper will not be published in the official archival Proceedings, which are registered with the Library of Congress and are submitted for abstracting and indexing. The paper also will not be published in The ASME Digital Collection and may not be cited as a published paper.

Commentary by Dr. Valentin Fuster

Combustion, Fuels and Emissions

2013;():V01AT04A001. doi:10.1115/GT2013-94027.

For predicting primary atomization a numerical code has been developed based on the Lagrangian Smoothed Particle Hydrodynamics (SPH) method. The advantage of this approach is the inherent interface advection. In contrast to commonly used grid based methods such as the Volume of Fluid (VoF) or Level Set method there is no need for costly and approximative interface tracking or reconstruction techniques which are required to avoid interface diffusion. It has been demonstrated by various test cases that the SPH method is capable to correctly predict single — as well as multiphase flows including the effect of surface tension. The goal of this work is to further develop the methodology with the intention to simulate primary atomization within airblast atomizers of jet engines. The authors present two test cases relevant for the simulation of primary atomization. The shear-driven deformation of a fuel droplet in a gaseous flow has been investigated and compared to data from literature. Moreover, the liquid film disintegration at the trailing edge of a planar prefilming airblast atomizer has been studied. The geometry has been derived from an existing test rig, where extensive experimental data have been acquired. Resulting droplet sizes and shear-off frequencies for different geometrical setups have been analyzed and compared to the experiment. The results reveal the promising performance of this new method for predicting primary atomization.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A002. doi:10.1115/GT2013-94033.

As the dominating parameters influencing the Sauter mean diameter of the spray produced by a prefilming airblast atomizer, the air velocity, liquid surface tension and atomizing edge thickness could be identified.

Correlations for the prediction of the droplet sizes produced by prefilming airblast atomizers are compared to droplet sizes measured close to the atomizing edge. The measurements were performed using three geometrical variants of a planar atomizer over a wide range of operating conditions. The diagnostics are based on a particle and ligament tracking technique, that enables simultaneous measurement of the liquid blobs and ligaments formed at the atomizing edge and the droplets in the primary breakup region of the atomizer.

The comparison between measured and calculated droplet diameter indicates, that most of the correlations are capable of reproducing the correct tendency. However, since the measurement data of most correlations were collected in a region where secondary atomization effects can obscure the initial droplet sizes in the primary breakup region, the droplet sizes are generally predicted too small.

Topics: Drops
Commentary by Dr. Valentin Fuster
2013;():V01AT04A003. doi:10.1115/GT2013-94064.

Thermoacoustic instabilities are a major concern in gas turbine combustion chambers today. In the last decades research interest in thermoacoustic instabilities has focused on low frequencies. The feedback mechanisms related to longitudinal modes are for the most part understood. Transverse modes, though, have not been studied to a large extent in the past. However, interest has been rising in the last few years. But little is known about the thermoacoustic feedback of high-frequency instabilities. Our previous publications characterized the flow and the flame at the eigenfrequency of high-frequency instabilities. There, a feedback mechanism was derived from the experimental results and discussed: the acoustic velocity leads to a periodic displacement of the flame resulting in a positive contribution to the Rayleigh criterion. Thus, the thermoacoustic feedback couples to the acoustic velocity, but not to the pressure or a periodic vortex formation. Different means can be derived from the model to influence high-frequency instabilities: Helmholtz dampers are used to shift the onset of instabilities to increased thermal power. With loudspeakers naturally stable operating points are excited. Stopping the excitation and evaluating the signal, decay rates are analyzed. Decay rates — i.e. stability margins — are compared for different operating conditions. Switching from perfect premixing to technical premixing, the radial profile of the fuel-to-air ratio can be changed. The influence of a lean core flow compared to a homogeneous mixture on the feedback is investigated and its impact on the instabilities is compared to the model. The observations reflect, what is predicted by the model. Velocity coupling, at least a significant part of the feedback mechanism for transverse high-frequency instabilities, is supported by the experimental results.

Topics: Damping , Feedback , Flames
Commentary by Dr. Valentin Fuster
2013;():V01AT04A004. doi:10.1115/GT2013-94121.

The present work explores the capability of the transported PDF (probability density function) method to predict nitric oxide (NO) formation in turbulent combustion. To this end a hybrid finite-volume/Lagrangian Monte-Carlo method is implemented into the THETA code of the German Aerospace Center (DLR). In this hybrid approach the transported PDF method governs the evolution of the thermochemical variables, whereas the flow field evolution is computed with a RANS (Reynolds-Averaged Navier Stokes) method. The method is used to compute a turbulent hydrogen-air flame and a methane-air flame and computational results are compared to experimental data. In order to assess the advantages of the transported PDF method, the flame computations are repeated with the “laminar chemistry” approach as well as with an “assumed PDF” method, which are both computationally cheaper. The present study reveals that the transported PDF method provides the highest accuracy in predicting the overall flame structure and nitric oxide formation.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A005. doi:10.1115/GT2013-94129.

It is widely recognized that the fuel/air mixing process is a critical factor in improving combustion efficiency and in minimizing pollutants such as NOx. Enhancement of fuel/air mixing can lead to lower pollutant emissions and greater efficiency. However, swirling flows in lean combustors play the role of fuel/air mixing and flame stability. The complex fluid dynamic phenomena encountered in swirling two-phase flow contribute to the difficulty in complete understanding the different processes occurring in combustors. Fortunately, Optical and laser-based visualization techniques available in our lab are important non-intrusive tools for visualizing flow process, especially for fuel injection and fuel/air mixing. To provide for a better understanding of effects of counter-rotating flow on droplets in atomization process, this study is a detailed characterization of the spray generated by an airblast atomizer by planar laser sheet imaging method.

Optical facility for spray diagnostics with fuel Planar Laser Induced Fluorescence (fuel-PLIF) method for fuel distribution, and Particle Image Velocity (PIV) method for velocity of droplets, is used to evaluate the performance of an air-blast atomizer. The results show that the performance of secondary atomization is influenced by swirling flow and primary atomization simultaneously, the swirling flow exhibits significant influence on the droplet size and space distribution relative to that of primary atomization. The primary swirling air reopens the spray cone generated by pressure-swirl atomizer, and the secondary swirling air affects the fuel distribution by forming the recirculation zone. The results provide critical information for design and development of combustion chamber.

Topics: Lasers , Imaging
Commentary by Dr. Valentin Fuster
2013;():V01AT04A006. doi:10.1115/GT2013-94131.

A Design for Thermo-Acoustic Stability (DeTAS) procedure is presented, that aims at selecting a most stable burner geometry for a given combustor. It is based on the premise that a thermo-acoustic stability model of the combustor can be formulated and that a burner design exists, which has geometric design parameters that sufficiently influence the dynamics of the flame. Describing the burner and flame dynamics in dependence of the geometrical parameters an optimization procedure involving a linear stability model of the target combustor maximizes the damping and thereby yields the optimal geometrical parameters. To demonstrate the procedure on an existing annular combustor a generic burner design was developed that features two geometrical parameters that can easily be adjusted. To provide the data base for the DeTAS procedure static and dynamical properties of burner and flame were measured for three by three configurations at a fixed operation point. The data is presented and discussed. It is found that the chosen design exhibits a significant variability of the flame dynamics in dependence of the geometrical parameters indicating that a DeTAS should be possible for the targeted annular combustor.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A007. doi:10.1115/GT2013-94160.

The prediction of the limit-cycle amplitude of thermoacoustically unstable practical gas turbine combustion systems remains a challenge for the gas turbine industry. The present study uses an experimentally obtained Flame Describing Function (FDF) for the determination of the thermoacoustic oscillation frequency and amplitude. In contrast to other studies, which investigated perfectly premixed laminar or marginally turbulent flames, this study deals with a highly turbulent swirl flame with spatial and temporal fuel–air unmixedness in an order of practical interest. A partially premixed swirl-stabilized flame is investigated at a Reynolds number of approximately 35 000. The Multi-Microphone-Method is used to determine the amplitude dependent transfer function of the flame as well as the transfer function of the burner and the acoustic response of the boundary conditions. The results are compared to OH* chemiluminescence measurements, which show a significant deviation in terms of the flame transfer function gain due to equivalence ratio fluctuations. The measured transfer functions are incorporated into a thermoacoustic modeling framework to determine frequency and amplitude of the self-excited limit cycle oscillation. Measurements were made for various lengths of the exhaust gas tube to verify the results for different frequencies and amplitudes. Good agreement is found for the entire range of combustor lengths investigated. The error between model and experimental results is thoroughly assessed.

Topics: Pressure , Flames
Commentary by Dr. Valentin Fuster
2013;():V01AT04A008. doi:10.1115/GT2013-94183.

The influence of changes in fuel composition and heating value on the performance of an industrial gas turbine combustor was investigated. The combustor tested was a single cannular combustor for Siemens SGT-400 13.4 MW dry low emission (DLE) engine. Ignition, engine starting, emissions, combustion dynamics and flash back through burner metal temperature monitoring were among the parameters investigated to evaluate the impact of fuel flexibility on combustor performance.

Lean ignition and extinction limits were measured for three fuels with different heat values in term of Wobbe Index (WI): 25, 28.9 and 45 MJ/Sm3 (natural gas). The test results show that the air fuel ratio (AFR) at lean ignition/extinction limits decreases and the margin between the two limits tends to be smaller as fuel heat value decreases. Engine start tests were also performed with a lower heating value fuel and results were found to be comparable to those for engine starting with natural gas.

The combustor was further tested in a high pressure air facility at real engine operating conditions with different fuels covering WIs from 17.5 to 70 MJ/Sm3. The variation in fuel composition and heating value was achieved in a gas mixing plant by blending natural gas with CO2, CO, N2 and H2 (for the fuel with WI lower than natural gas) and C3H8 (for the fuel with WI higher than natural gas).

Test results show that a benefit in NOx reduction can be seen for the lower WI fuels without H2 presence in the fuel and there are no adverse impacts on combustor performance except for the requirement of higher fuel supply pressure, however, this can be easily resolved by minor modification through the fuel injection design.

Test results for the H2 enriched and higher WI fuels show that NOx, combustion dynamics and flash back have been adversely affected and major change in burner design is required. For the H2 enriched fuel, the effect of CO and H2 on combustor performance was also investigated for the fuels having a fixed WI of 29 MJ/Sm3. It is found that H2 dominates the adverse impact on combustor performance. The chemical kinetic study shows that H2 has significant effect on flame speed change and CO has significant effect on flame temperature change.

Although the tests were performed on the Siemens SGT-400 combustion system, the results provide general guidance for the challenge of industrial gas turbine burner design for fuel flexibility.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A009. doi:10.1115/GT2013-94224.

A combustion instability in a combustor typical of aero-engines is analyzed and modeled thanks to a low order Helmholtz solver. A Dynamic Mode Decomposition (DMD) is first applied to the Large Eddy Simulation (LES) database. The mode with the highest amplitude shares the same frequency of oscillation as the experiment (approx. 350 Hz) and it shows the presence of large entropy spots generated within the combustion chamber and convected down to the exit nozzle. The lowest purely acoustic mode being in the range 650–700 Hz, it is postulated that the instability observed around 350 Hz stems from a mixed entropy/acoustic mode where the acoustic generation associated with the entropy spots being convected throughout the choked nozzle plays a key role. A Delayed Entropy Coupled Boundary Condition is then derived in order to account for this interaction in the framework of a Helmholtz solver where the baseline flow is assumed at rest. When fed with appropriate transfer functions to model the entropy generation and convection from the flame to the exit, the Helmholtz solver proves able to predict the presence of an unstable mode around 350 Hz, in agreement with both the LES and the experiments. This finding supports the idea that the instability observed in the combustor is indeed driven by the entropy/acoustic coupling.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A010. doi:10.1115/GT2013-94258.

In modern industrial gas turbines swirling flow is widely used for stabilizing flames at the transition from the burner to the combustor. In premixed combustion systems using highly reactive fuels, flashback due to combustion induced vortex breakdown (CIVB) has been observed frequently when swirl was present. This paper focuses on the effect of low swirl intensities on the flashback propensity and the predominant flashback mechanisms in a hydrogen-air tube burner. An existing test rig with a vertical quartz tube and a generic swirl generator has been used. At the tube exit the flame was stabilized in the free atmosphere. The turbulent flashback limits were measured for hydrogen-air mixtures at atmospheric conditions over a broad range of equivalence ratios for both non-swirling and swirling flow. The upstream flame propagation during flashback was observed through the OH*-chemiluminescence captured by two synchronized intensified high-speed cameras in a 90° arrangement, both looking at the flame from the side. In addition to that, a high-speed particle image velocimetry (PIV) system was used to insert a horizontal laser sheet into the vertical tube in order to investigate the propagation path of the leading flame tip through a time series of Mie-scattering images from the bottom. As expected, it turned out that the flame always flashes back along the wall boundary layer for non-swirling flow. For swirling flow it could be shown that again only boundary layer flashback takes place for equivalence ratios lower than ϕ≈0.75. In this rather lean region, the resistance against flashback is improved compared to non-swirling flow due to higher wall velocity gradients. For higher equivalence ratios, flashback is initiated through CIVB. That is, the flame enters the tube on the burner centerline until its tail gets in touch with the burner walls. Subsequently, there is a shift in flashback mechanism and the flame propagates further upstream along the wall boundary layer. For the given setup and these near-stoichiometric mixture compositions, this resulted in a significantly increased flashback propensity when compared with non-swirling flames. The present studies showed that imposing low swirl upon the burner flow can improve the resistance against boundary layer flashback for low and moderate equivalence ratios, whereas the change to the CIVB mechanism deteriorates the performance for high equivalence ratios.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A011. doi:10.1115/GT2013-94266.

Premixed combustion systems show potential to meet future regulations on nitrogen oxide emissions. However, premixed systems always involve the risk of flame flashback into the premixing section. From a gas turbine manufacturer’s point of view it is desirable to broaden the safe operating range, in particular with respect to flame flashback. It has been reported in the literature that flashback along the wall boundary layer represents the most critical failure mechanism for many burner configurations using hydrogen-rich fuels. This paper focuses on the effect of fluid injection into the wall boundary layer on the flashback propensity of a hydrogen-air jet burner. For this purpose, an experiment with a tube burner was designed, where the flame is anchored in the free atmosphere at the burner exit. Pure air was injected through an annular gap at three streamwise locations upstream of the stable flame position and at two different angles — perpendicular and at 45° to the main flow direction, respectively. The turbulent flashback limits for fully premixed hydrogen-air mixtures at atmospheric conditions were measured for a variety of equivalence ratios and different amounts of air injection. It turned out that there is a significant increase in flashback stability with injection devices close to the burner exit. The main reason for this behavior is the dilution of the mixture in the near-wall region and the resulting reduction of the flame speed. The positive effect diminishes quickly with increasing distance between flame and injection location due to considerable mixing of injected flow and main flow. This has been verified by RANS simulations. The simulations also showed that the momentum generated by the injection into the boundary layer has negligible influence on the flashback limits. It was also found that the fluid injection is not capable of stopping the upstream flame propagation once the flame has entered the tube. A probable explanation for this effect is the change from open flame holding at the burner exit to the confined flame situation inside the tube during flashback. The latter is known to substantially increase the flashback propensity, which cannot be counteracted by air injection.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A012. doi:10.1115/GT2013-94280.

A series of experiments have been conducted to study the aerodynamic characteristics of a confined swirling flow generated by multiple rad-rad swirlers arranged linearly. The rad-rad swirlers used in this study are identical, and consist of an inner, primary swirler generating counter-clockwise rotation and an outer, secondary swirler generating clockwise rotation. A two-component Laser Doppler Velocimetry (LDV) system was employed to measure the velocity in the flow field. Initial measurements were conducted on unconfined and confined flow generated by a single swirler to serve as the baseline reference for the multi-swirler arrangements. Tests were conducted for 3 and 5 swirlers arranged in a line, with a spacing of 2D between the swirler centers, where D is the swirler exit diameter. An additional 5 swirler configuration was tested, where the exit plane of the center swirler was shifted 3.2 mm (1/8 inch) in the streamwise direction.

The flow field generated by the multi-swirler arrangement is very complex, due to the interaction between the swirling jets of adjacent swirlers. The number of swirlers is seen to have a clear impact on the entire flow structure, as well as each recirculation zone. For the 3 swirler arrangement, a weak CTRZ is observed for the center swirler, whereas strong CTRZs are observed for the two outer swirlers. For the 5 swirler arrangement, the CTRZ pattern for the 3 inner swirlers is the same strong-weak-strong as seen for the 3 swirler arrangements, with weak CTRZs observed for the two outer swirlers. Higher interaction between swirlers is observed for the 5 swirler arrangement, as compared to the case with 3 swirlers. Since the swirlers are identical, the region between swirlers features merging of two opposing swirling jets, producing high turbulence intensity in the near field region.

For the case with the offset center swirler, the swirling jet from this swirler did not merge with its neighbors in the near field region. This resulted in strong CTRZ for the center swirler, accompanied by weaker CTRZs at its immediate neighbors, which is reverse of the CTRZ strength pattern observed for the initial 5 swirler arrangement.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A013. doi:10.1115/GT2013-94282.

The predictive capabilities of two candidate approaches for CO predictions were assessed for a series of aircraft gas turbine combustors. The first approach involved constructing a large reduced order reactor network coupled with a detailed, 500 species n-dodecane mechanism to simulate the combustion process. The second approach was the traditional RANS based CFD using two finite rate based combustion models in FLUENT. A four step Jet-A global mechanism was developed in-house and was used in the CFD simulations. The global mechanism was validated against the detailed Jet-A mechanism published by Dagaut in 2006 and was able to reproduce the flame speed and species profiles satisfactorily over the range of relevant operating temperatures and pressures. The calibration combustors comprised seven configurations with identical fuel nozzles but different swirlers, dome effusion, liner and quench jet air flow splits. It was found that the CFD approach was better at capturing the trend of rig data than the reactor network approach and was able to capture most of the variations seen in the measurement. The improvement in prediction was attributed mainly to the more accurate global mechanism which results in more accurate kinetic calculation in CFD.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A014. doi:10.1115/GT2013-94284.

This paper presents an experimental study of acoustically forced bluff body stabilized flames, motivated by the problem of combustion instabilities. The goal of the work is to better understand the flame and flow behavior as functions of the proximity of the acoustic frequency to natural hydrodynamic instability frequencies. It is well known that iso-density, high Re bluff body wakes are globally unstable, exhibiting the Von Karman vortex street. In reacting flows, however, this global mode is suppressed if the density ratio across the flame is sufficiently high. Thus, the density ratio is an important parameter that influences the global mode growthrate. In this study, the flame was longitudinally forced over a range of hydrodynamic global mode to forcing frequency ratios, density ratios, and forcing amplitudes. Longitudinal forcing leads to the symmetric rollup of the two separating shear layers. When the forcing frequency is in the vicinity of the wake’s global mode frequency, the global mode locks into the forcing frequency, and the symmetric shear layers quickly stagger as they convect downstream, leading to a large scale, sinuous flapping of the wake and flame. The axial position at which staggering occurs is a function of the forcing amplitude and the proximity of the forcing frequency to the global mode frequency.

The lock-in phenomenon amplifies the flame’s motion at the forcing frequency. However, if the vortices stagger to a fully sinuous structure, this causes a significant reduction in the flame’s oscillatory heat release through phase cancellation of the upper and lower flame branches. Therefore, if a low density ratio flame is subjected to longitudinal acoustic forcing near its global mode frequency, it will respond with weaker heat release fluctuations than it would away from lock-in. This is true even though the local degree of flame flapping is quite significant. Thus, the results of this study show some phenomena that contradict conventional notions, namely that forcing a globally unstable flow near its global mode frequencies can lead to diminished local heat release oscillations.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A015. doi:10.1115/GT2013-94299.

The low NOX emission technology has become an important feature of advanced aviation engine. A wide range of applications attempt to take advantage of the fact that staged combustion operating under lean-premixed-prevaporized (LPP) conditions can significantly decrease pollution emissions and improve combustion efficiency. In this paper a scheme with fuel centrally staged and multi-point injection is proposed. The mixing of fuel and air is improved, and the flame temperature is typically low in combustion zone, minimizing the formation of nitrogen oxides (NOX), especially thermal NOX. In terms of the field distribution of equivalence ratio and temperature obtained from Computational Fluid Dynamics (CFD), a chemical reactor network (CRN), including several different ideal reactor, namely perfectly stirred reactor (PSR) and plug flow reactor (PFR), is constructed to simulate the combustion process. The influences of the pilot equivalence ratio and percentage of pilot/main fuel on NOX and carbon monoxide (CO) emissions were studied by Chemical CRN model. Then the NOX emission in the staged combustor was researched experimentally. The effects of the amount of pilot fuel and primary fuel on pollution emissions were obtained by using gas analyzer. Finally, the effects of pilot fuel proportion on NOX emission were discussed in detail by comparing predicts of CRN and experimental results.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A016. doi:10.1115/GT2013-94307.

In gas turbine operation a turbulent flame is employed. Thus, better understanding of the turbulent flame propagation is the key for further optimisation of turbine combustors and reduction of the environmental footprint. As turbulent flames are exposed to stretch, the effect of flame-stretch interaction must be better understood especially at higher pressures.

In present study, turbulent burning velocity of two mixtures, hydrogen/air and propane/air, with negative and positive Ma, respectively are experimentally investigated in fan-stirred explosion vessel. For the investigation an optical laser method is employed based on the Mie-scattering of the laser light by smoke particles. Within this study the influence of initial parameters as initial pressure and turbulence intensity on the flame front propagation is investigated by giving special attention on influence of Ma variation. The experiments were performed at three different pressures 1, 2, 4 bar. The RMS fluctuation velocity was varied in the range of 0–2.77 m/s. The observed results are compared and discussed in detail.

Topics: Pressure , Turbulence , Flames
Commentary by Dr. Valentin Fuster
2013;():V01AT04A017. doi:10.1115/GT2013-94309.

The first part of this work presents a comparison of predictions obtained with several two-equation type RANS turbulence models commonly used in industry against experimental data obtained by Whitelaw et al [1]. All examined models yield a relatively poor match in the flow region very close to the wall; agreement with the measurements improves significantly when moving further away from the wall. This concerns both the internal normal stress profiles and the average velocity profiles, the latter show improved prediction of the recirculation zone area when moving further into the main stream. Downstream behaviour for both models shows an excellent match more than 6 diameters away from the jet inlet, defined as the region after which the flow essentially resumes its normal duct behaviour[1].

Expanding upon these RANS results, another series of simulations using LES modelling with the standard Smagorinsky SGS model was conducted using the same grid and compared to the RANS-based results. Although performance in the most complex flow areas was slightly improved over RANS, this was at the cost of an increase of computation time by almost a factor of 6.

The next stage involved developing a code based on the model for two-phase flow described in [2] to predict the atomisation pattern for a non-vaporising (or “cold”) flow based on the parameters of the previous simulations. This model implements transport equations for the liquid mass fraction and the average surface area per unit mass along with an equation for average density; resulting in an entirely Eulerian model which can be used to predict atomisation from first principles. Current work consists in development of additional source terms describing vaporisation in a strongly turbulent environment and further coupling with a combustion model applicable to the combustion chamber of an industrial gas turbine.

Topics: Turbulence , Modeling
Commentary by Dr. Valentin Fuster
2013;():V01AT04A018. doi:10.1115/GT2013-94312.

Screech is a high frequency oscillation that is usually characterized by instabilities caused by large-scale coherent flow structures in the wake of bluff-body flameholders and shear layers. Such oscillations can lead to changes in flame surface area which can cause the flame to burn unsteadily, but also couple with the acoustic modes and inherent fluid-mechanical instabilities that are present in the system.

In this study, the flame response to hydrodynamic oscillations is analyzed in a controlled manner using high-fidelity Computational Fluid Dynamics (CFD) with an unsteady Reynolds-averaged Navier-Stokes approach. The response of a premixed flame with and without transverse velocity forcing is analyzed. When unforced, the flame is shown to exhibit a self-excitation that is attributed to the anti-symmetric shedding of vortices in the wake of the flameholder. The flame is also forced using two different kinds of low-amplitude out-of-phase inlet velocity forcing signals. The first forcing method is harmonic forcing with a single characteristic frequency, while the second forcing method involves a broadband forcing signal with frequencies in the range of 500–1000 Hz.

For the harmonic forcing method, the flame is perturbed only lightly about its mean position and exhibits a limit cycle oscillation that is characteristic of the forcing frequency. For the broadband forcing method, larger changes in the flame surface area and detachment of the flame sheet can be seen. Transition to a complicated trajectory in the phase space is observed. When analyzed systematically with system identification methods, the CFD results, expressed in the form of the Flame Transfer Function (FTF) are capable of elucidating the flame response to the imposed perturbation. The FTF also serves to identify, both spatially and temporally, regions where the flame responds linearly and nonlinearly. Locking-in between the flame’s natural self-excited frequency and the subharmonic frequencies of the broadband forcing signal is found to alter the dynamical behaviour of the flame.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A019. doi:10.1115/GT2013-94317.

Gas turbine combustor designers now routinely use high-fidelity reactive computational fluid dynamics (CFD) analyses to gain valuable insight into the complex reactive flow-field and pollutant formation process. But, a large number of such computationally expensive CFD analyses are generally required to arrive at an acceptable combustor configuration. Therefore, given the practical limits on available computational resources and time, traditional combustor design methodologies using only high-fidelity CFD analyses need further improvement. To address this, a combustor design strategy using multifidelity co-Kriging response surface model (RSM) is developed and applied for the design of a two-dimensional test combustor problem in the spatial domain using steady-state Reynolds-averaged Navier Stokes (RANS) formulation. The design and optimization problem is set-up for two geometric variables and a single-objective, NOx concentration, as it is of current interest to the combustor design community. The developed multi-fidelity strategy is also assessed for performance against high-fidelity Kriging RSM strategy. This study demonstrates that the multi-fidelity design strategy can obtain good designs with up to ten times less effort than a full grid sampling search plan. However, the multi-fidelity co-Kriging strategy does not outperform the high-fidelity Kriging strategy for the given spatial domain problem.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A020. doi:10.1115/GT2013-94328.

The objective of this study is to investigate the sensitivity and accuracy of the combustible flow field prediction for the LIMOUSINE combustor with regards to choices in computational mesh and turbulent combustion model. The LIMOUSINE combustor is a partially premixed bluff body stabilized natural gas combustor designed to operate at 40–80 kW and atmospheric pressure and used to study combustion instabilities. The transient simulation of a turbulent combusting flow with the purpose to study thermo-acoustic instabilities is a very time consuming process. For that reason the meshing approach leading to accurate numerical prediction, known sensitivity, and reduced amount of mesh elements is important. Since the numerical dissipation (and dispersion) is highly dependent on, and affected by, the geometrical mesh quality, it is of high importance to control the mesh distribution and element size across the numerical model. Typically, the structural mesh topology allows using much less grid elements compared to the unstructured grid, however an unstructured mesh is favorable for flows in complex geometries. To explore computational stability and accuracy, the numerical dissipation of the cold flow with mixing of fuel and air is studied first in the absence of the combustion process. Thereafter the studies are extended to combustible flows using standard available ANSYS-CFX combustion models. To validate the predicted variable fields of the combustor’s transient reactive flows, the numerical results for dynamic pressure and temperature variations, resolved under structured and unstructured mesh conditions, are compared with experimental data. The obtained results show minor dependence on the used mesh in the velocity and pressure profiles of the investigated grids under non-reacting conditions. More significant differences are observed in the mixing behavior of air and fuel flows. Here the numerical dissipation of the (unstructured) tetrahedral mesh topology is higher than in the case of the (structured) hexahedral mesh. For that reason, the combusting flow resolved with the use of the hexahedral mesh presents better agreement with experimental data and demands less computational effort. Finally in the paper the performance of the combustion model for reacting flow as a function of mesh configuration is presented, and the main issues of the applied combustion modeling are reviewed.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A021. doi:10.1115/GT2013-94344.

In this paper, we develop a linear technique that predicts how the stability of a thermo-acoustic system changes due to the action of a generic passive feedback device or a generic change in the base state. From this, one can calculate the passive device or base state change that most stabilizes the system. This theoretical framework, based on adjoint equations, is applied to two types of Rijke tube. The first contains an electrically-heated hot wire and the second contains a diffusion flame. Both heat sources are assumed to be compact so that the acoustic and heat release models can be decoupled. We find that the most effective passive control device is an adiabatic mesh placed at the downstream end of the Rijke tube. We also investigate the effects of a second hot wire and a local variation of the cross-sectional area but find that both affect the frequency more than the growth rate. This application of adjoint sensitivity analysis opens up new possibilities for the passive control of thermo-acoustic oscillations. For example, the influence of base state changes can be combined with other constraints, such as that the total heat release rate remains constant, in order to show how an unstable thermo-acoustic system should be changed in order to make it stable.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A022. doi:10.1115/GT2013-94350.

Bluff body stabilised non-premixed flames are usually used as pilot flames in lean-premixed combustors. Experiments are conducted to investigate the characteristics of the flame. Typical flame modes are investigated in both stable and unstable conditions. The flow structures, the reaction zone, and the dynamics of unstable flames are measured with PIV, ICCD and a high speed camera, respectively, based on which the inherent mechanisms that influence the configuration and stabilisation of the flame are analysed. Stable flames are apparently influenced by the mixing characteristics in the recirculation zone. Flame detachment, a typical phenomenon of stable flames in a turbulent air flow, can be explained by the distribution of fuel concentration in the recirculation zone. The Reynolds number of air has different effects on different parts of the flame, which results in three unstable flame modes at different Reynolds numbers of air. These results could be helpful for the design of stable burners in practice.

Topics: Fuels , Flames
Commentary by Dr. Valentin Fuster
2013;():V01AT04A023. doi:10.1115/GT2013-94355.

Gas turbines emissions, NOX in particular, have negative impact on the environment. To limit the emissions gas turbine burners are constantly improved. In this work, a fourth generation SIT (Siemens Industrial Turbomachinery) burner is studied to gain information about the formation of NOX emissions. The gas mixtures for the full burner are limited to natural gas with different nitrogen dilutions. The dilutions vary from undiluted to Wobbe index 40 and 30 MJ/m3. In addition to the full burner, the central body (the RPL – Rich/Pilot/Lean) is investigated. Methane is used to characterize standard gas turbine operation, and a non-standard fuel is explored using a generic syngas (67.5 % Hydrogen, 22.5 % Carbon monoxide and 10 % Methane). Both these gases are also investigated after dilution with nitrogen to a Wobbe index of 15 MJ/m3. The experiments are performed in a high-pressure facility. The pressures for the central body burner are 3, 6 and 9 bar. For the full burner the pressures are 3, 4.5 and 6 bar. The combustion air is preheated to 650 K. The emission measurements are sampled with an emission probe at the end of the combustor liner, and analyzed in an emission rack. The results are compared with previous investigations made at atmospheric conditions.

The burner is modeled using a PSR and plug flow network to show which reaction paths are important in the formation of emissions for the burner under the experimental conditions.

The measurement results show that the NOX concentration increases with pressure and flame temperature. With increasing dilution the NOX concentration is decreased. For rich mixtures PSR calculations show that the NOX concentration decreases with pressure.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A024. doi:10.1115/GT2013-94363.

In the present paper a computational analysis of a confined premixed turbulent methane/air jet flame is presented. In this scope, chemistry is reduced by the use of the Flamelet Generated Manifold (FGM) method [1, 2], and the fluid flow is modeled in a RANS context. In the FGM technique the reaction progress of the flame is generally described by a few control variables, for which a transport equation is solved during runtime. The flamelet system is computed in a pre-processing stage, and a manifold with all the information about combustion is stored in a tabulated form. In the present implementation the reaction evolution is described by the reaction progress variable, the heat loss is described by the enthalpy and the turbulence effect on the reaction is represented by the progress variable variance. The turbulence-chemistry interaction is considered through the use of a presumed pdf approach.

A generic lab scale burner for high-velocity preheated jets is used for validation [3, 4]. It consists of a rectangular confinement, and an off-center positioning of the jet nozzle enables flame stabilization by recirculation of hot combustion products. The inlet speed is appropriately high, in order to be close to the blow out limit. Flame structures were visualized by OH* chemiluminescence imaging and planar laser-induced fluorescence of the OH radical. Laser Raman scattering was used to determine concentrations of the major species and the temperature. Velocity fields were measured with particle image velocimetry.

The important effect of conductive heat loss to the walls is included in the FGM chemistry reduction method in a RANS context, in order to predict the evolution and description of a turbulent jet flame in high Reynolds number flow conditions. Comparisons of various mean fields (velocities, temperatures) with RANS results are shown. The use of FGM as a combustion model shows that combustion features at gas turbine conditions can be satisfactorily reproduced with a reasonable computational effort.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A025. doi:10.1115/GT2013-94371.

A new compact injection system design for piloted lean combustion has been developed to reduce the pollutant emissions in aero engines. The system includes an integrated premixing zone to achieve a homogenous fuel distribution, so that peak temperatures in the combustor are avoided. This leads to low NOx emissions at lean conditions. The risks of flame flashback and auto ignition have been considered in the design and neither of them has been detected by the performed tests. To avoid the formation of a recirculation zone within the mixing zone an axial air jet has been introduced. This axial jet also works as an air assisted pilot fuel atomizer, which is a major innovation as compared to other lean injection systems using pressure-swirl atomizers for the pilot fuel like e.g. the PERM (Partial Evaporation and Rapid Mixing) concept developed in a previous research program [1], [2]. The main fuel injection of the current configuration is performed by four circumferentially arranged pressure swirl atomizers, which is also an alternative approach compared to previous concepts. The emission performance of the injection system using kerosene Jet A-1 has been investigated in a tubular combustor with air inlet temperatures up to 733 K and combustor pressures up to 10 bar. The dependencies of pilot fuel split, air to fuel ratio, combustor pressure and air inlet temperature on emissions have been determined. Over a wide range of operating conditions a low amount of pollutant emissions are achieved and the stability range is broadened by the pilot fuel injection. The flame structure has been analyzed by OH* chemiluminescence measurements. The Abel transformation technique has been applied to the images to generate the radial distribution. The main flame is lifted and its shape remains similar for different combustor pressures. The lift off height with only pilot fuel injection decreases with increasing combustor pressure and the flame shape is changing. This behavior is explained based on the effects of combustor pressure on fuel atomization, droplet traces and the distribution of evaporated fuel. The development and testing have been conducted in cooperation of AVIO and Karlsruhe Institute of Technology in the frame of the European Commission co-financed research project TECC-AE (Technology Enhancement for Clean Combustion in Aero Engines).

Topics: Pressure , Flames , Emissions
Commentary by Dr. Valentin Fuster
2013;():V01AT04A026. doi:10.1115/GT2013-94372.

Reducing the weight and decreasing pressure losses of aviation gas turbine engines improves the thrust-to-weight ratio and improves efficiency. In ultra-compact combustors (UCCs), engine length is reduced and pressure losses are decreased by merging a combustor with adjacent components using a systems engineering approach. High-pressure turbine inlet vanes can be placed in a combustor to form a UCC. Eliminating the compressor outlet guide vanes (OGVs) and maintaining swirl through the diffuser can result in further reduction in engine length and weight. Cycle analysis indicates that a 2.4% improvement in engine weight and a 0.8% increase in thrust-specific fuel consumption are possible when a UCC is used. Experiments and analysis were performed in an effort to understand key physical and chemical processes within a trapped-vortex UCC. Experiments were performed using a combustor operating at pressures in the range of 520–1030 kPa (75–150 psi) and inlet temperature of 480–620 K (865–1120 °R). The primary reaction zone is in a single trapped-vortex cavity where the equivalence ratio was varied from 0.7 to 1.8. Combustion efficiencies and NOx emissions were measured and exit temperature profiles obtained, for various air loadings, cavity equivalence ratios, and configurations with and without turbine inlet vanes. A combined diffuser-flameholder (CDF) was used in configurations without vanes to study the interaction of cavity and core flows. Higher combustion efficiency was achieved when the forward-to-aft momentum ratios of the air jets in the cavity were near unity or higher. Discrete jets of air immediately above the cavity result in the highest combustion efficiency. The air jets reinforce the vortex structure within the cavity, as confirmed through coherent structure velocimetry of high-speed images. A more uniform temperature profile was observed at the combustor exit when a CDF is used instead of vanes. This is the result of increased mass transport along the face of the flame holder. Emission indices of NOx were between 3.5 and 6.5 g/kgfuel for all test conditions. Ultra-compact combustors (with a single cavity) can be run with higher air loadings than those employed in previous testing with a trapped-vortex combustor (two cavities) with similar combustion efficiencies being maintained. The results of this study suggest that the length of combustors and adjacent components can be reduced by employing a systems level approach.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A027. doi:10.1115/GT2013-94378.

This paper presents the extension and optimization of the Siemens Industrial Turbomachinery Ltd Dry Low Emission combustion system from the existing SGT-300 Single-Shaft turbine to the new SGT-300 Twin-Shaft engine. The SGT-300 Twin-Shaft combustion development follows the Siemens Product Development Process and the new engine is now validated for introduction to the market.

Different designs are tested and optimized at full engine pressure and temperature conditions in Lincoln, UK Siemens combustion high pressure rig facility. Optimized combustion design is installed to the new design SGT-300 Twin-Shaft engine to validate in the Siemens gas turbine test bed in Lincoln, UK.

Blocker bars have been designed and applied to the combustion rig to capture the acoustic signature of the nozzle guide vane. Traverse is performed in the high pressure rig to identity the combustor exit hot gas temperature map.

Emission turndown to 50% load with NOx < 9 ppm and CO < 10 ppm capability is one of the major Key Performance Indicators. The SGT-300 Twin-Shaft combustion system is optimized for excellent emission behavior in this load range. Additionally, some of the compressor delivery air is bypassed around the combustor to exhaust through the air bleed system at part load.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A028. doi:10.1115/GT2013-94384.

A swirl burner was designed to experimentally study the impact of spark location on ignition efficiency and detailed ignition scenarios until flame stabilization or blow-off were established, following experimental observations. Premixed and non-premixed configurations were investigated for the same turbulent flow, in order to evaluate the fuel heterogeneities on ignition efficiency. Attention was paid to providing accurate data on cold flow velocity field statistics (obtained by stereoscopic PIV) and fuel mole fraction field statistics (obtained by PLIF on acetone). Ignition probability maps were established for all conditions by using laser-induced spark for a constant level of deposited energy. No systematic correlations were observed between local flow properties and ignition probability, which leads to the conclusion that history of the flame kernel inside the combustion chamber, must be taken into account to fully explain the ignition mechanism. From this conclusion, ignition scenarios were built using fast flame visualization and dynamic pressure record. Different steps of the ignition process were identified according to the location of the spark.

In order to evaluate ignition probability according to spark location and flow conditions (velocity, turbulence and mixing), we extended the predictive model of Neophytou et al. [1], with some modifications, to examine whether it can be applied to ignition of swirling premixed flames. Flame particles are emitted by the spark and tracked in the flow with a Langevin equation by using non-reactive velocity fields obtained by PIV. Physical criteria are proposed to represent flame particles generation, expansion and extinction. Results indicate a relatively good agreement with the experimental database and the ignition scenarios are also well reproduced.

Topics: Flames , Ignition
Commentary by Dr. Valentin Fuster
2013;():V01AT04A029. doi:10.1115/GT2013-94404.

The paper presents a one-dimensional approach to assess the reduction potential of NOx emissions for lean premixed gas turbine combustion systems. NOx emissions from these systems are known to be mainly caused by high temperatures; not only from an averaged perspective but especially related to poor mixing quality of fuel and air.

The method separates the NOx chemistry in the flame front zone and the post flame zone (slow reaction). A one-dimensional treatment enables the use of detailed chemistry. A look up table parameterized by reaction progress and equivalence ratio is used to improve the computational efficiency.

The influence of mixing quality is taken into account by a probability density function of the fuel element based equivalence ratio, which itself translates into a temperature distribution. Hence, the NOx source terms are a function of reaction progress and equivalence ratio. The reaction progress is considered by means of the two-zone approach. Based on unsteady CFD data, the evolution of the probability density function with residence time has been analyzed.

Two types of definitions of an unmixedness quantity are considered. One definition accounts for spatial as well as temporal fluctuations and the other is based on the mean spatial distribution. They are determined at the location of the flame front.

The paper presents a comparison of the modeled results with experimental data. A validation and application have shown very good quantitative and qualitative agreement with the measurements. The comparison of the unmixedness definitions has proven the necessity of unsteady simulations. A general emissions - unmixedness correlation can be derived for a given combustion system.

Topics: Gas turbines
Commentary by Dr. Valentin Fuster
2013;():V01AT04A030. doi:10.1115/GT2013-94411.

The design of a modern aero-engine combustor is a highly complex and multi-disciplinary task. The combustor design is strongly driven by severe emission regulations and ACARE 2020/2050 goals. Furthermore, new designs have to be developed within short turn-around times. This paper describes a novel approach of an automated preliminary aero-thermal design process of a rich-burn combustor combining 1D, 2D and 3D design tools in order to speed up the design loop and provide improved combustor designs in an early design stage.

The automated design process includes a knowledge-based preliminary design tool, an 1D network solver, a parametric 3D geometry model, a meshing tool, and 3D-CFD analysis. At first, a preliminary combustor design is created based on industrial in-house design rules. The preliminary design tool provides a 2D geometry model and cooling layout. It is coupled with an 1D network solver to calculate the air distribution inside the combustor. The design process includes two state-of-the-art combustor cooling schemes, effusion cooling and impingement effusion cooling. An air flow model for both cooling schemes is created within the network, respectively. The computed air distribution is subsequently used to generate boundary conditions for a 3D-CFD analysis. To perform the CFD calculations, a parametric 3D geometry model of a combustor sector has been developed based on a 2D preliminary design which takes into account mixing port properties, fuel injector, and combustor wall cooling. After an automated meshing 3D-CFD computations are performed. As a result, quick automatic estimation of combustor emissions, size and efficiency can be obtained within the design process. A CFD parameter study of a mixing port variation and their effect on the emissions of NOx and soot is performed using the described layout process.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A031. doi:10.1115/GT2013-94420.

The combustion of hydrogen-rich fuels (> 80 % vol. H2), relevant for gas turbine cycles with “pre-combustion” carbon capture, creates great challenges in the application of standard lean premix combustion technology. The significant higher flame speed and drastically reduced auto-ignition delay time of hydrogen compared to those of natural gas, which is normally burned in gas turbines, increase the risk of higher NOX emissions and material damage due to flashback. Combustion concepts for gas turbines operating on hydrogen fuel need to be adapted to assure safe and low-emission combustion.

A rich/lean (R/L) combustion concept with integrated heat transfer that addresses the challenges of hydrogen combustion has been investigated. A sub-scale, staged burner with full optical access has been designed and tested at gas turbine relevant conditions (flame temperature of 1750 K, preheat temperature of 400 °C and a pressure of 8 bar).

Results of the burner tests have confirmed the capability of the rich/lean staged concept to reduce the NOx emissions for undiluted hydrogen fuel. The NOx emissions were reduced from 165 ppm measured without staging (fuel pre-conversion) to 23 ppm for an R/L design having a fuel-rich hydrogen pre-conversion of 50 % at a constant power of 8.7 kW.

In the realized R/L concept the products of the first rich stage, which is ignited by a Pt/Pd catalyst (under a laminar flow, Re ≈ 1900) are combusted in a diffusion-flame-like lean stage (turbulent flow Re ≈ 18500) without any flashback risk. The optical accessibility of the reactor has allowed insight into the combustion processes of both stages. Applying OH-LIF and OH*-chemiluminescence optical techniques, it was shown that mainly homogeneous reactions at rich conditions take place in the first stage, questioning the importance of a catalyst in the system, and opening a wide range of optimization possibilities.

The promising results obtained in this study suggest that such a rich/lean staged burner with integrated heat transfer could help to develop a new generation of gas turbine burners for safe and clean combustion of H2-rich fuels.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A032. doi:10.1115/GT2013-94443.

Experiments were performed on the central body rich-pilot-lean (RPL) burner of a Siemens Industrial Turbomachinary 4th generation DLE combustor to observe the combustion changes that may occur when using fuels other than natural gas. Measurements were taken of temperatures at multiple points along the RPL body while hydroxyl (OH) radical distribution extending from the dump plane of the burner was imaged by planar laser induced fluorescence (PLIF). The RPL burner was run using four fuels; methane, a generic syngas (67.5% H2, 22.5% CO and 10% CH4) and dilutions of these with nitrogen to a Wobbe index of 15 MJ/m3. Each of the fuels was operated at several equivalence ratios ranging from ϕ = 0.80 to ϕ = 1.80, for combustion pressures of 3, 6 and 9 bar. It was found that the flame position in the RPL, determined from temperature measurement at the thermocouple positions, was dependent on the fuel, equivalence ratio and to a lesser extent pressure. A link was established between the OH distribution in the post burner region and RPL temperature profiles based on the expected flame behavior inside the RPL. For all measurement points some combustion occurred within the burner volume, indicated by thermocouples at the burner exit.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A033. doi:10.1115/GT2013-94446.

The primary breakup under high-altitude relight conditions is investigated in this study where ambient pressure is as low as 0.4 bar and air, fuel and engine parts are as cold as 265 K. The primary breakup is crucial for the fuel atomization. As of today, the phenomena dictating the primary breakup are not fully understood. Direct Numerical Simulations (DNS) of liquid breakup under realistic conditions and geometries are hardly possible. The embedded DNS (eDNS) approach represents a reliable numerical tool to fill this gap.

The concept consists of three steps: a geometry simplification, the generation of realistic boundary conditions for the DNS and the DNS of the breakup region. The realistic annular airblast atomizer geometry is simplified to a Y-shaped channel representing a planar geometry. Inside this domain the eDNS is located. The eDNS domain requires the generation of boundary conditions. A Large Eddy Simulation (LES) of the entire Y-shaped channel and a Reynolds-Averaged Navier-Stokes Simulation (RANS) of the liquid wall film are performed prior to the DNS. All parameters are stored transiently on all virtual DNS planes. These variables are then mapped to the DNS. Thus, high-quality boundary conditions are generated. The Volume-of-Fluid (VOF) method is used to solve for the two-phase flow.

The results provide a qualitative insight into the primary breakup under realistic high-altitude relight conditions. Instantaneous snapshots in time illustrate the behavior of the liquid wall film along the prefilmer lip and illustrate the breakup process. It is seen that a slight variation of the surface tension force has a strong impact on the appearance of the primary breakup. Case 1 with the surface tension corresponding to kerosene at 293 K indicates large flow structures that are separated from the liquid sheet. By lowering the surface tension related to kerosene at 363 K, the breakup is dominated by numerous small structures and droplets.

This study proves the applicability of the eDNS concept for investigating breakup processes as the transient nature of the phase interface behavior can be captured. At this time, the authors only present a qualitative insight which can be explained by the lack of quantitative data. The approach offers the potential of simulating realistic annular highly-swirled airblast atomizer geometries under realistic conditions.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A034. doi:10.1115/GT2013-94451.

Accurate characterization of swirled flames is a key point in the development of more efficient and safer aeronautical engines. The task is even more challenging for spray injection systems. On the one side, spray interacts with both turbulence and flame, eventually affecting the flame dynamics. On the other side, spray flame structure is highly complex due to equivalence ratio inhomogeneities caused by the evaporation process. Introducing detailed chemistry in numerical simulations, necessary for the prediction of flame stabilization, ignition and pollutant concentration, is then essential but extremely expensive in terms of CPU time. In this context, tabulated chemistry methods, expressly developed to account for detailed chemistry at a reduced computational cost in Large Eddy Simulation of turbulent gaseous flames, are attractive. The objective of this work is to propose a first computation of a swirled spray flame stabilized in an actual turbojet injection system using tabulated chemistry. A Large Eddy Simulation of an experimental benchmark, representative of an industrial swirl two-phase air/kerosene injection system, is performed using a standard tabulated chemistry method. The numerical results are compared to the experimental database in terms of mean and fluctuating axial velocity. The reactive two-phase flow is deeper investigated focusing on the flame structure and dynamics.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A035. doi:10.1115/GT2013-94484.

This work describes measurements and analysis of the turbulent consumption speeds, ST,GC, of H2/CO fuel blends. We report measurements of ST,GC at pressures and normalized turbulence intensities, u′rms/SL,0 up to 20 atm and 1800, respectively for a variety of H2/CO mixtures and equivalence ratios. In addition, we present correlations of these data using laminar burning velocities of highly stretched flames, SL,max, derived from quasi-steady leading points models. These analyses show that SL,max can be used to correlate data over a broad range of fuel compositions, but do not capture the pressure sensitivity of ST,GC. We suggest that these pressure effects are more fundamentally a manifestation of non-quasi-steady behavior in the mass burning rate at the flame leading points.

Topics: Fuels , Turbulence , Hydrogen
Commentary by Dr. Valentin Fuster
2013;():V01AT04A036. doi:10.1115/GT2013-94495.

In the recent years a great interest has been devoted to the understanding of the nonlinear dynamics characterizing the thermoacoustic combustion instabilities. Although linear techniques are able to predict whether the non-oscillating steady state of a thermoacoustic system is “asymptotically” stable (without oscillations) or unstable (increasing oscillations), a thermoacoustic system can reach a permanent oscillating state (the so called “limit cycle”), even when it is linearly stable, if a sufficiently large impulse occurs. A nonlinear analysis is able to predict the existence of this oscillating state and the nature of the bifurcation process.

The aim of this work is to investigate the behavior of gas turbine combustion chambers in presence of nonlinear flame models. The bifurcation diagrams, obtained by using a continuation technique in the frequency domain, give the amplitude of the oscillations as a function of a chosen flame parameter. The Helmholtz equation is used to model the combustion chamber and nonlinear terms are introduced in the flame model, starting from the classical k–τ formulation. A three-dimensional finite element method (FEM) is used for discretization of the computational domain and a solver of quadratic eigenvalue problems is combined with Newton technique in order to identify the points of the bifurcation diagram. First, a simple Rijke tube configuration, as can be found in the literature, is examined in order to obtain bifurcation diagrams. Then, the nonlinear analysis is extended to simplified annular configurations. The obtained results show how the nonlinear behavior is influenced by varying some control parameters, such as the time delay, yielding useful indications to designers and experimentalists.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A037. doi:10.1115/GT2013-94570.

A comprehensive assessment of emissions characteristics of the 1st, N and N+1 generation rich-dome combustion products has been done to identify the lowest emissions products. Focus of this paper is on the large rich-dome engines with its potential application for the (N+3) and (N+4) mixers with inspirational target takeoff NOxEI of 5 at 55 OPR.

A total of ten engine models of the 1st generation were selected in addition to eight recently certified large engines. After evaluating several choices for conducting comparative assessment, the following three expressions were proposed for average takeoff NOxEI, idle COEI and HCEI entitlements, respectively:

Display Formula


Display Formula

Idle COEIL=815.36Takeoff NOxEIL1.159

Display Formula

Idle HCEIL=0.15×Idle COEIL-2.0

In regard to application of the rich-dome technology to the (N+2) cycle based (N+3) mixers, the author tentatively gives it low probability of success barring success story stemming from Lee et al. [2012].

Commentary by Dr. Valentin Fuster
2013;():V01AT04A038. doi:10.1115/GT2013-94571.

Comprehensive assessment of the medium size rich-dome engines was conducted leading to the following emissions correlations:

Display Formula

LTO NOx=1.129×OPR1.0899withR2=0.9248

Takeoff NOxEI given by

Display Formula


Display Formula


These correlations may be compared with the following for the CFM56 Tech Insertion:

Display Formula

Takeoff NOxEICFM_TI=0.0744×OPR1.7151

Display Formula

Idle COEICFM_TI=396.42Takeoff NOxEI0.814

Display Formula

Idle HCEICFM_TI=0.1609×Idle COEI-3.1959

TALON II takeoff NOxEI data are reproduced well by:

Display Formula

NOxEITALON II=0.0167×OPR2.1403

TALON II gives 10% lower NOx at 26 OPR and its NOx is comparable with the CFM_TI at 34 OPR.

The CFM DAC technology is competitive with LEC’s for the low rated thrust engines. However, interaction between the two domes leads to early quenching with resultant higher idle COEI plateau. On the other hand, the 40 OPR lean DAC gave 25% higher NOx than LEC. Moreover, lean DAC (Gen-1) impacted fuel burn adversely making its likelihood to continue as product discouraging.

The second generation lean dome technology initially kicked off under NASA sponsorship with significantly larger funding support from the CFMI and GE Aviation (GEA) led to successful introduction of TAPS into products (GEnx-1B and Gen-2B) with potential applications in other future GEA engines.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A039. doi:10.1115/GT2013-94572.

Comprehensive assessment of the small rich-dome engines was conducted leading to the following emissions correlations:

NOxEILEC = 0.02991 × OPR1.9791

RMSE = 3.0%

NOxEITALON II = 0.01666 × OPR2.1403

RMSE = 2.0%

NOxEICFMTI = 0.06763 × OPR1.7458

RMSE = 2.1%

NOxEICF34 = 0.0541 × OPR1.7917R2 = 0.9794

RMSE = 2.4%

NOxEISM = 0.04782 × OPR1.8388

RMSE = 4.2%

NOxEIAll = 0.03856 × OPR1.9058

RMSE = 3.9%

The best of the small engines’ gaseous emissions, albeit at lower takeoff pressure ratios, were shown to be very competitive with the best of medium and large size engines.

Axially-staged combustion with partially premixed jets in crossflow was identified as a promising concept to pursue for the (N+3) technology mixers.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A040. doi:10.1115/GT2013-94585.

Limited fuel resources, increasing energy demand, and stringent emission regulations are drivers to evaluate process off-gases or process waste streams as fuels for power generation. Often these process waste streams have low energy content and their operability in gas turbines leads to issues such as unstable or incomplete combustion and changes in acoustic response. Due to above reasons, these fuels cannot be used directly without modifications or efficiency penalties in gas turbine engines.

To enable the use of the wide variety of ultra-low and low Btu fuels in gas turbine engines, a rich catalytic lean burn (RCL®) combustion system was developed and tested in a subscale high pressure (10 atm.) rig. Previous work has shown promise with fuels such as blast furnace gas (BFG) with Lower Heating Value (LHV) of 3.1 MJ/Nm3 (85 Btu/scf). The current testing extends the limits of RCL® operability to other weak fuels by further modifying and improving the injector to achieve enhanced flame stability. Fuels containing low methane content such as weak natural gas with an LHV of 6.5 MJ/Nm3 (180 Btu/scf) to fuels containing higher methane content such as landfill gas with an LHV of 21.1 MJ/Nm3 (580 Btu/scf) were tested. These fuels demonstrated improved combustion stability with an extended turndown (defined as the difference between catalytic and non-catalytic lean blow out) of 140°C–170°C (280°F–340°F) with CO and NOx emissions lower than 5 ppm corrected to 15% O2.

Topics: Fuels
Commentary by Dr. Valentin Fuster
2013;():V01AT04A041. doi:10.1115/GT2013-94599.

A multi-objective design optimisation study has been carried out with the objectives to improve the overall efficiency of the device and to reduce the fuel consumption for the proposed micro-scale combustor design configuration. In a previous study we identified the topology of the combustion chamber that produced improved behaviour of the device in terms of the above design criteria. We now extend our design approach, and we propose a new configuration by the addition of a micro-cooling channel that will improve the thermal behaviour of the design as previously suggested in literature. Our initial numerical results revealed an improvement of 2.6% in the combustion efficiency when we applied the micro-cooling channel to an optimum design configuration we identified from our earlier multi-objective optimisation study, and under the same operating conditions.

The computational modelling of the combustion process is implemented in the commercial computational fluid dynamics package ANSYS-CFX using Finite Rate Chemistry and a single step hydrogen-air reaction. With this model we try to balance good accuracy of the combustion solution and at the same time practicality within the context of an optimisation process. The whole design system comprises also the ANSYS-ICEM CFD package for the automatic geometry and mesh generation and the Multi-Objective Tabu Search algorithm for the design space exploration. We model the design problem with 5 geometrical parameters and 3 operational parameters subject to 5 design constraints that secure practicality and feasibility of the new optimum design configurations. The final results demonstrate the reliability and efficiency of the developed computational design system and most importantly we assess the practicality and manufacturability of the revealed optimum design configurations of micro-combustor devices.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A042. doi:10.1115/GT2013-94617.

More efficient and powerful gas turbine engines can be designed using constant-volume combustors that may involve ignition of a combustible mixture using a hot gas jet, subsequent flame and pressure-wave propagation, and their interactions. Accurate prediction of three-dimensional transient turbulent combustion is computationally challenging. To resolve propagating turbulent combustion, predict ignition, and track pressure waves accurately requires techniques to minimize the numerical cell count and kinetics calculation times. This study of shock-flame interaction (SFI) used detailed chemistry that includes low-temperature ignition reactions. Computational cells with similar temperatures and composition were grouped as ‘zones’ where kinetics are solved only once per zone per time step, using average values of species concentrations and thermodynamic properties for that zone. This avoids expensive kinetic calculations in every computational cell, with considerable speedup. A relatively coarser initial mesh was refined selectively and automatically, based on predicted velocity and temperature gradients, tracking propagating pressure waves and flames. The time step is variable, limited by the local speed of sound, to ensure accurate wave propagation. These techniques, previously validated for non-premixed, premixed and multiple-fuel turbulent combustion in industrial IC engines, are applied to study SFI during premixed combustion in a long constant-volume combustor.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A043. doi:10.1115/GT2013-94635.

The development of integrated gasification combined cycle (IGCC) systems provides cost-effective and environmentally sound options for meeting future coal-utilizing power generation needs in the world. The combustion of gasified coal fuel significantly influences overall performance of IGCC power generation. Experimental measurements are carried out on a non-premixed model combustor, equipped with a double-swirled syngas burner. Planar laser-induced fluorescence (PLIF) of OH radical measurement is adopted to identify main reaction zones and burnt gas regions as well. Together with the temperature and emission measurements during the exhaust section, some important characteristics of the syngas flame are investigated overall. In this paper, the effects of the CO/H2 molar ratio consisting of syngas fuel are investigated under different humidity. With the increase of CO/H2 ratios, the concentration field of OH radicals is gradually away from the nozzle exit, and the nozzle exit almost no existence of OH radicals, forming a typical lifted flame. In addition, fluorescent signal strength of OH radicals pronounced weakening, the flame gradually appeared W type distribution and more and more obvious with the increased of humidification amount. At the same time the average exhaust temperature of combustor CO and NOx missions almost no change. The study can provide a reliable database for high moisture gas turbine combustor design and combustion numerical simulation.

Topics: Syngas
Commentary by Dr. Valentin Fuster
2013;():V01AT04A044. doi:10.1115/GT2013-94650.

An understanding of the fundamental combustion properties of alternative fuels is essential for their adoption as replacements for non-renewable sources. In this study, three different biojet fuel mixtures are directly compared to conventional Jet A-1 on the basis of laminar flame speed and vapor pressure. The biofuel is derived from camelina oil and hydrotreated to ensure consistent elemental composition with conventional aviation fuel, yielding a bioderived synthetic paraffinic kerosene (Bio-SPK). Two considered blends are biofuel and Jet A-1 mixtures, while the third is a 90% Bio-SPK mixture with 10% aromatic additives. Premixed flame speed measurements are conducted at an unburned temperature of 400K and atmospheric pressure using a jet-wall stagnation flame apparatus. Since the laminar flame speed cannot be studied experimentally, a strained (or reference) flame speed is used as the basis for the initial comparison. Only by using an appropriate surrogate fuel were the reference flame speed measurements extrapolated to zero flame strain, accomplished using a direct comparison of simulations to experiments. This method has been previously shown to yield results consistent with non-linear extrapolations. Vapor pressure measurements of the biojet blends are also made from 25 to 200°C using an isoteniscope. Finally, the biojet blends are compared to the Jet A-1 benchmark on the basis of laminar flame speed at different equivalence ratios, as well as on the basis of vapor pressure over a wide temperature range, and the suitability of a binary laminar flame speed surrogate for these biojet fuels is discussed.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A045. doi:10.1115/GT2013-94674.

Thermo-acoustic instabilities in gas turbine combustors can prevent the implementation of modern combustion concepts, which are essential for higher efficiency and lower emissions. Perforated combustor liners, especially in combination with a bias flow through the liner, are able to suppress the instabilities by increasing the acoustic losses of the system. Some insight into the parameter dependencies of the acoustic absorption has been gained by means of atmospheric testing at ambient temperature. The next step towards realistic testing conditions is taking into account high temperature and high pressure, which increases the effort of the experimental tests and the complexity of their analysis significantly. Tests in a real combustor can serve as a quality check of a given liner design, but are not appropriate for parameter studies. So far, numerical models accurate enough to enable the design of hot stream liners are simply not available, so that the experimental investigation of the liner’s dependency on temperature and pressure is essential for the transfer of laboratory scale results to a real engine application.

A new test rig has been designed to overcome these problems. The Hot Acoustic Test rig (HAT) enables the study of the influence of pressure and temperature on the damping performance in an acoustically well defined environment, although the high temperature and high pressure conditions are challenging in terms of accurate acoustic measurements.

This paper introduces the Hot Acoustic Test rig with its features and limitations and shows first examples of test results. The focus lies on the hardware, instrumentation, and analysis techniques that are necessary to obtain high quality acoustic data in hot and pressurized flow environments.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A046. doi:10.1115/GT2013-94677.

With increasing focus on environmental effects and the need for fuel diversity in gas turbines, good liquid atomization is increasingly important. It is known that impinging atomization is able to produce fine drops by impingement of fast liquid jets. However, the atomization characteristics deteriorate at lower injection velocities. In this study, for improving atomization characteristics under a wide range of injection velocity, an effective technique is verified utilizing a small amount of gas (microjet) injection. The microjet is supplied from a pressurized reservoir independent of the liquid supply system, and it is injected from the center of the liquid nozzles toward the impingement point. To clarify the flow field and the mechanism of the effectiveness, experimental visualizations and drop size measurements are carried out. It is found that atomization is remarkably promoted when the dynamic pressure of microjet overcomes that of the liquid at the impingement point. By the microjet injection with only 1% of liquid mass flow rate, Sauter mean diameter (SMD) becomes one-tenth of the original SMD. In addition, optimized atomization efficiency is successfully achieved when the dynamic pressure of the microjet is two times that of the liquid at the impingement point.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A047. doi:10.1115/GT2013-94678.

On the topic of CO2 capture from gas turbines, exhaust gas recirculation (EGR) is a commonly discussed method to increase CO2 concentration at a gas turbine outlet to make the CO2 capture process more efficient. This paper presents the influence of the recirculation on heat release rate and emissions. The investigation is made using the commercial RANS solver ANSYS CFX coupled with an in-house code for a hybrid transported PDF/RANS simulation using detailed chemistry of GRI 3.0.

Initially an investigation on reactivity was made using numerical calculation of laminar flame speed. It is found that exhaust gas recirculation has only a minor effect on reactivity in lean premixed combustion. Therefore, the operation point of the combustor can be kept constant with and without EGR.

Simulations of the combustor with exhaust gas recirculation using the hybrid PDF/RANS with GRI 3.0 show a minor influence of NO and NO2 doping of the vitiated air on the flame speed and the doping delays heat release slightly. CO doping has no effect on heat release rate. CO emissions at combustor exit remain unaffected by NO, CO or NO2 doping.

Seeding the vitiated air with 50ppm nitric oxides reveal that any NO2 present in the vitiated air is reduced to NO in the flame. NO2 emissions increase with NO2 doping but are still 2 magnitudes lower than NO emissions. It is found that NO is reduced by 3% due to of NO reburn. Based on literature data it is concluded that there is a deficit of the GRI 3.0 reaction mechanism. Experimental data taken from literature reveal of NO reburn by approximately 20%. Therefore emission data of nitric oxides of flames that should show a considerable reburn effect should be used with caution, while heat release and CO emissions are predicted more accurately.

It is shown, that with the model created for the generic gas turbine combustor it is possible to study the effects of exhaust gas recirculation on the combustion process in detail and resolve detailed kinetic effects.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A048. doi:10.1115/GT2013-94681.

The design of a gas turbine combustion chamber integrates multiple contradicting objectives. Among all the parameters available to the engineers, the number of fuel injection systems and their spacing are crucial information which need to be fixed early on in the design phase. Indeed, such choices not only impact the cost and size of the combustor but they also affect the operability of the future engine. One key objective behind these parameters is the ignition time delay needed for the whole combustion chamber to successfully light. To gather knowledge in the ignition process that takes place in real gas turbine engines, current research orient towards the development of experimental facilities that complement high fidelity unsteady numerical simulations. In this context, a multi-injectors experimental set-up located at CORIA (France) is used to validate Large Eddy Simulation (LES) tools developed by CERFACS, IFP-EN and CORIA (France). Preliminary validations against experimental data show that for a given inter-injector distance, LES stationary and ignition transient predictions are very promising and recover the main features found in the experiment. Exit mean and root mean square velocity profiles of the steady flow are in good agreement with measurements obtained for all injectors at multiple axial locations. The simulation of the ignition transient phase well captures global events such as the propagation of the flame front from one injector to its neighbors and the related mechanisms. Improvement is however still needed to recover the proper ignition time of the whole burner.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A049. doi:10.1115/GT2013-94690.

In this paper the mechanical interaction between pressure oscillations induced by combustion dynamics in a laboratory scale combustor and the structural vibrations of the combustor wall (the liner) is investigated. The combustor has operation conditions where the feedback loop between unsteady heat release and the acoustic field in the combustion chamber is unstable. This drives the amplitudes of the acoustic pressure oscillations to very high magnitudes. Under these circumstances, there is increased risk of failure due to fatigue. To investigate the two way feedback loop between the acoustic field and the structural vibrations of the liner in a gas turbine combustor, a laboratory-size combustor has been developed. The combustor has a rectangular cross section and operates under lean, partially-premixed conditions. The flame is stabilized using a triangular bluff body. To achieve a strong liner vibration feedback, the stiffness of the liner walls is made very low. In the framework of the COPAGT project, transient numerical simulations of the coupled fluid-structure interaction (FSI) system in the combustor are carried out and the results are validated by measurements. A partitioned approach is utilized, where the fluid and structural domain are calculated by different solvers. Coupling is achieved by a two-way data exchange between the fluid and structural domain. Investigated is an operating point in limit cycle oscillation with high amplitude. The numerical results show some agreement with the experimental data, but also show some aspects for improvement. The applied procedure is suitable to reproduce the coupling of the pressure with the structure, but the structural model performance needs to be improved. Neither for the measurements nor for the simulation any influence of the structural vibration is visible in the pressure spectrum, which can be seen as the major characteristic for a closed feedback loop between structural vibration and acoustic field.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A050. doi:10.1115/GT2013-94696.

Liquid jets in cross air flows are widely used and play an important role in propulsion systems, such as ramjet combustors. In this paper, experiments were carried out to investigate the properties of the primary breakup of liquid jets in subsonic transverse airflows at low Weber number. The test ranges included crossflow Weber numbers of 0.5–6.7, liquid-to-gas momentum ratios of 3–120, and Ohnesorge number of 0.0086. Four different injectors with diameter 0.4mm, 0.5mm, 0.6mm and 1mm have been used. A high speed camera was used to observe the jet column breakup process. Results show that the surface wavelength decreases not only with the increase of the gas Weber number but also with the increase of the momentum ratio. The breakup length decreases with the increase of the gas Weber number, in addition to its increase with the increase of the momentum ratio. The droplet diameter decreases with the increase of both the gas Weber number and momentum ratio, although the gas Weber number will dominate the breakup process. The surface wavelength, breakup length, and droplet diameter were also analyzed with to obtain semi-theoretical correlations.

Topics: Jets
Commentary by Dr. Valentin Fuster
2013;():V01AT04A051. doi:10.1115/GT2013-94704.

This paper focuses on optimizing an innovative annular Lean Premixed staged burner, following the Trapped Vortex Combustor concept. The latter consists of a lean main flame stabilized by passing past a rich cavity pilot flame. Unfortunately, this configuration is highly sensitive to combustion instabilities and the flame is not well stabilized. This work consists of adjusting aerodynamic variables, chemical parameters and burner geometry to reach a “low-NOx” operation while reducing other pollutants and getting a stable flame. Results show that stability is reached when mass transfers between main and cavity zones are reduced. Then, the main bulk velocity is increased to reduce the cavity thermal expansion, due to the hot gas expansion. In addition, the cavity flow rate is reduced to prevent from penetrating and disturbing the main flow. Re-arranging injections in the cavity also avoid local unsteady equivalence ratios, which creates an unsteady heat release and combustion with pulses. Regarding NOx, a leaner main flame combined with a sufficiently rich cavity mixture creates local stoichiometric zones at the interface between the cavity and the main zone. The latter point is found to be a good anchoring mechanism. Compared with the original configuration, a stable point of operation is found: acoustic energy is reduced by an order of 100, NOx level is less than 0.4 g/kgfuel, CO is cut by 93% with no more Unburned Hydro-Carbons.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A052. doi:10.1115/GT2013-94738.

Technology to support design for the next generation of aircraft fuel injectors is being developed by NASA’s Fundamental Aeronautics Program to reduce emissions, increase efficiency, and enable fuel flexible aircraft. The design of these fuel injectors can be aided by measuring the vaporization rates of jet fuel droplets in realistic, burning sprays. Currently, in this environment, no instantaneous vaporization rate measurement techniques have been applied. After surveying techniques for potential development for this application, two techniques were identified: stimulated Raman scattering (SRS) and droplet lasing spectroscopy (DLS). Plans were developed for the modification of these techniques for this specific application. Developments of the SRS technique were tested including measurement of a water droplet diameter change rate. SRS spectra were also collected from jet fuel droplets.

Topics: Combustion , Fuels , Drops , Sprays
Commentary by Dr. Valentin Fuster
2013;():V01AT04A053. doi:10.1115/GT2013-94763.

At the Institute of Thermodynamics, Technical University of Munich a large scale atmospheric combustion test rig has been designed and set up. The experimental setup is comprised of two burning zones: A first zone consists of 16 burners providing vitiated air at 1776K, into which a secondary fuel-air mixture jet is injected and ignited by the hot cross flow. The phenomenon is known in the literature as a reacting jet in hot cross flow. The hot data is compared to the cold case in order to show differences in the flow field due to flame propagation.

For evaluating the flow field several experimental analyses have been applied so far (OH*, High-Speed PIV, Mixture Analysis). The focus of this paper is on the momentum ratios J = 4–10 with Jet Reynolds Numbers between 20,000 and 80,000. For the cold case the flow field is measured and compared with the reacting jet. In the injector the air and the natural gas are perfectly premixed. The equivalence ratio of the jet is varied over a wide range of mixtures (ϕ = 0.05–0.77) resulting in an adiabatic flame temperature of the jet between 800 and 2200K. As the pictures of the chemiluminescence analysis show the jet gas ignites immediately upon entering the hot cross flow. The distinct influence of the equivalence ratio on the flame length and shape can be seen in the data. The trajectory of the flame penetrates further into the channel compared to the trajectory of the cold case caused by the reaction in the flame and its resulting gas expansion. Due to the large diameter of the jet in the experiment the origins of the dominant flow patterns are obtained with high spatial resolution. Following this, flame anchoring mechanisms at different operation points are derived.

Topics: Flames , Ignition , Cross-flow
Commentary by Dr. Valentin Fuster
2013;():V01AT04A054. doi:10.1115/GT2013-94769.

The effects of Nanosecond Repetitively Pulsed (NRP) plasma discharges on the dynamics of a swirl-stabilized lean premixed flame are investigated experimentally. Voltage pulses of 8-kV amplitude and 10-ns duration are applied at a repetition rate of 30 kHz. The average electric power deposited by the plasma is limited to 40 W, corresponding to less than 1 % of the thermal power of 4 kW released by the flame. The investigation is carried out with a dedicated experimental setup that allows for studies of the flame dynamics with applied plasma discharges. A loudspeaker is used to perturb the flame acoustically, and the discharges are generated between a central pin electrode and the rim of the injection tube. Velocity and CH* chemiluminescence signals are used to determine the flame transfer function assuming that plasma discharges do not affect the correlation between CH* emission and heat release rate fluctuations. Phase-locked images of the CH* emission were recorded to assess the effect of the plasma on the oscillation of the flame. The results show a strong influence of the NRP discharges on the flame response to acoustic perturbations, thus opening interesting perspectives for combustion control. An interpretation of the modifications observed in the transfer function of the flame is proposed by taking into account the thermal and chemical effects of the discharges. It is then demonstrated that by applying NRP discharges at unstable conditions, the oscillation amplitudes can be reduced by an order of magnitude, thus effectively stabilizing the system.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A055. doi:10.1115/GT2013-94771.

In the future low pollution power generation can be achieved by application of hydrogen as a possible alternative gas turbine fuel if the hydrogen is produced by renewable energy sources such as wind energy or biomass. The utilization of existing IGCC power plant technology with the combination of low cost coal as a bridge to renewable energy sources such as biomass can support the international effort to reduce the environmental impact of electricity generation. Against this background the dry low NOx Micromix combustion principle for hydrogen is developed for years to significantly reduce NOx emissions. This combustion principle is based on cross-flow mixing of air and gaseous hydrogen and burns in multiple miniaturized diffusion-type flames. The two advantages of this principle are the inherent safety against flash-back and the low NOx concentrations due to a very short residence time of reactants in the flame region of the micro-flames.

The paper presents experimental results showing the significant reduction of NOx emissions at high equivalence ratios and at simultaneously increased energy density under preheated atmospheric conditions. Furthermore the paper presents the feasibility of enlarged Micromix hydrogen injectors reducing the number of required injectors of a full-scale Micromix combustion chamber while maintaining the thermal energy output with significantly low NOx formation.

The experimental investigations are accompanied by 3D numerical reacting flow simulations based on a simplified hydrogen combustion model. Comparison with experimental results shows good agreement with respect to flame structure, shape and anchoring position. Thus, the experimental and numerical results highlight further potential of the Micromix combustion principle for low NOx combustion of hydrogen in industrial gas turbine applications.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A056. doi:10.1115/GT2013-94782.

In the current study, the influence of pressure and steam on the emission formation in a premixed natural gas flame is investigated at pressures between 1.5 bar and 9 bar.

A premixed, swirl-stabilized combustor is developed that provides a stable flame up to very high steam contents. Combustion tests are conducted at different pressure levels for equivalence ratios from lean blowout to near-stoichiometric conditions and steam-to-air mass ratios from 0% to 25%.

A reactor network is developed to model the combustion process. The simulation results match the measured NOx and CO concentrations very well for all operating conditions. The reactor network is used for a detailed investigation of the influence of steam and pressure on the NOx formation pathways.

In the experiments, adding 20% steam reduces NOx and CO emissions to below 10 ppm at all tested pressures up to near-stoichiometric conditions. Pressure scaling laws are derived: CO changes with a pressure exponent of approximately −0.5 that is not noticeably affected by the steam. For the NOx emissions, the exponent increases with equivalence ratio from 0.1 to 0.65 at dry conditions. At a steam-to-air mass ratio of 20%, the NOx pressure exponent is reduced to −0.1 to +0.25.

The numerical analysis reveals that steam has a strong effect on the combustion chemistry. The reduction in NOx emissions is mainly caused by lower concentrations of atomic oxygen at steam-diluted conditions, constraining the thermal pathway.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A057. doi:10.1115/GT2013-94795.

There is a current need to reduce turbine engine system size and weight. An Ultra Compact Combustor (UCC) is a system that attempts to accomplish this by integrating the compressor, combustor, and turbine stages. The main feature of this system is a high gravity-loaded combustor that burns the fuel in the circumferential direction around a hybrid vane row. One consequence of this design is that the UCC needs the combustion process to complete within the vane passage. A second issue is that combining the compressor and turbine results in large flow swirl and high Mach number in the core flow. Combined, these two issues lead to the potential for high Rayleigh losses for this design. This investigation used MATLAB to develop an area ruling inside the passage that allowed the flow to slow to a reasonable Mach number during the heat release process and then accelerate to the desired turbine rotor inlet Mach once combustion has finished. Results showed that it was possible to reduce Rayleigh losses in this manner, but equilibrium between Rayleigh losses and aerodynamic losses must be met. This model was incorporated into a design of a test rig. Reductions of more than 5% in Rayleigh losses in the experimental rig are expected.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A058. doi:10.1115/GT2013-94796.

For lean burn combustor development in low emission aero-engines, the pilot stage of the fuel injector plays a key role with respect to stability, operability, NOx emissions, and smoke production. Therefore it is of considerable interest to characterize the pilot module in terms of pilot zone mixing, fuel placement, flow field and interaction with the main stage.

This contribution focusses on the investigation of soot formation during pilot-only operation. Optical test methods were applied in an optically accessible single sector rig at engine idle conditions. Using planar laser-induced incandescence (LII), the distribution of soot and its dependence on air/fuel ratio, as well as geometric injector parameters, was studied. The data shows that below a certain air/fuel ratio, an increase of soot production occurs. This is in agreement with smoke number measurements in a standard single sector flame tube rig without optical access. Reaction zones were identified using chemiluminescence of OH radicals. In addition, the injector flow field was investigated with PIV. A hypothesis regarding the mechanism of pilot smoke formation was made based on these findings. This along with further investigations will form the basis for developing strategies for smoke improvement at elevated pilot only conditions.

Topics: Lasers , Ejectors
Commentary by Dr. Valentin Fuster
2013;():V01AT04A059. doi:10.1115/GT2013-94797.

An increasing demand is being put on the fuel as a heat sink in modern aircraft. In the end, the fuel flows through the atomizer which on the one hand is the hottest part of its thermal history, but on the other hand the most critical for resisting deposition. Most studies have concentrated on the chemistry of deposition, and in recent years there have been modeling efforts. Deposition is really the end product of a coupling between heat transfer to the fuel, chemical reactions to form insoluble gums, followed by the transport of these gums to the surface to form deposits. There is conflicting evidence and theory in the literature concerning the effect of turbulence on deposition, i.e., whether deposition increases or decreases with increasing Reynolds number. This paper demonstrates through a heat transfer analysis that the effect of Reynolds number depends upon the boundary/initial conditions. If the flow is heated from the surface, deposition decreases with increasing Reynolds number; however, for isothermal flows, i.e., preheated, deposition will increase with Reynolds number.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A060. doi:10.1115/GT2013-94813.

A skeletal chemical kinetic model for jet fuel combustion, comprising four representative fuel components, is presented. The sub model for the three components, toluene, methyl cyclohexane (MCH) and n-dodecane, is deduced from a detailed model for jet fuel surrogate proposed by Wang et al. [Wang et al., 2010]. The reduction is based on a species sensitivity approach, herein referred to as Alternate Species Elimination (ASE). The sub model for the fourth component, iso-octane, is established through semi-detailed kinetic modeling, considering existing reactions and species of the smaller hydrocarbon systems as well as species and reactions pertinent to the n-dodecane system.

The performance of the resulting model is assessed by comparing predictions of ignition delay times and laminar burning velocities with those of the detailed model. It is shown that the skeletal model retains the predictive ability of the detailed model with respect to the three components, n-dodecane, MCH and toluene. The complementary iso-octane sub model is also found to reasonably predict high-temperature ignition delay times and laminar burning velocities.

The four component skeletal model is tested against shock tube ignition data and laminar burning velocities of jet fuel surrogates. It is observed that high-temperature ignition is fairly well predicted while low-temperature ignition delay times are longer than experimentally observed. While the predictions of laminar burning velocities of atmospheric flames of jet fuels at 400 K are reasonable, slower flames are predicted at higher temperatures. The proposed skeletal model has 192 species and 1291 reactions, compared to the detailed multi-component model, with 348 species and 2163 elementary reactions, albeit without iso-octane. This results in improvement in the associated computational costs for combustion analysis. Further development of the skeletal model is needed to improve its prediction ability over a wider range of combustion properties and thermodynamic conditions.

Topics: Combustion , Jet fuels
Commentary by Dr. Valentin Fuster
2013;():V01AT04A061. doi:10.1115/GT2013-94822.

Flashback is a key challenge for low NOx premixed combustion of high hydrogen content fuels. Previous work has systematically investigated the impact of fuel composition on flashback propensity, and noted that burner tip temperature played an important role on flashback, yet did not quantify any specific effect. The present work further investigates the coupling of flashback with burner tip temperature and leads to models for flashback propensity as a function of parameters studied. To achieve this, a jet burner configuration with interchangeable burner materials was developed along with automated flashback detection and rim temperature monitoring. An inline heater provides preheated air up to 810 K. Key observations include that for a given condition, tip temperature of a quartz burner at flashback is higher than that of a stainless burner. As a reasult, the flashback propensity of a quartz tube is about double of that of a stainless tube. A polynomial model based on analysis of variance is presented and shows that, if the tip temperature is introduced as a parameter, better correlations result. A physical model is developed and illustates that the critical velocity gradient is proportional to the laminar flame speed computed using the measured tip temperature. Addition of multiple parameters further refined the prediction of the flashback propensity, and the effects of materials are discussed qualitatively using a simple heat transfer analysis.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A062. doi:10.1115/GT2013-94823.

Flashback is a key operability issue for low emission premixed combustion systems operated on high hydrogen content fuels. Previous work investigated fuel composition impacts on flashback propensity and found that burner tip temperature was important in correlating flashback data in premixed jet flames. An enclosure around the jet flame was found to enhance the flame-burner rim interaction. The present study further addresses these issues using a jet burner with various geometric configurations and interchangeable materials. Systematic studies addressing the quantitative influence of various parameters such as tip temperature, burner material, enclosure size, and burner diameter on flashback propensity were carried out. A comprehensive overview of the flashback limits for all conditions tested in the current study as well as those published previously is given. The collective results indicate that the burner materials, tip temperature and flame confinement play significant roles for flashback propensity and thus help explain previous scatter in flashback data. Furthermore, the present work indicates that the upstream flame propagation during flashback is affected by the burner material. The material with lower thermal conductivity yields larger flashback propensity but slower flame regression inside the tube. These observations can be potentially exploited to minimize the negative impacts of flashback in practical applications.

Topics: Temperature , Flames
Commentary by Dr. Valentin Fuster
2013;():V01AT04A063. doi:10.1115/GT2013-94842.

Lean premixed swirling flames stabilized at the sudden expansion of a combustion chamber generally take a V or a M shape, depending on the burner geometry and flow conditions. It is not rare to observe transitions between these shapes as operating conditions of the combustor are modified, but the governing mechanisms of these transitions are not well understood. An experimental study is conducted to analyze the transition from an initially V-shaped flame to a M-shaped flame for swirling flames fed by CH4/H2/air mixtures anchored on a central bluff body when flow conditions and geometrical elements of the combustor are modified. Two situations are identified depending whether strong flame wall interactions take place at the quartz tube confining the flame. When the V-flame front is impinging and quenched at the combustion chamber wall, the transition to a M-shape is triggered by a flashback mechanism along the wall followed by the propagation of the flame tip along the outer shear-layer of the swirling jet in the direction of the external rim of the burner. This mechanism is controlled by the mixture Lewis number, a Karlovitz number based on the velocity gradient at the combustor wall, the swirl imparted to the flow and the cross section area ratio between the injection unit and combustion chamber defining a confinement ratio. Experiments conducted at a given mixture Lewis number indicate that the onset of flashback is determined by decreasing the Karlovitz number under a critical value. It is shown that this critical value decreases when the Lewis number increases, but it also depends on the confinement ratio and swirl number. In the absence of direct interaction between the V-flame tip and the chamber wall, the situation differs and this flashback mechanism along the combustion chamber wall ceases. Attempts are made in this case to identify the governing parameters triggering the V- to M-shape transition for small confinement ratios.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A064. doi:10.1115/GT2013-94844.

Combustors of modern gas turbines for power generation and mechanical drive are predominantly operated in premixed mode, which is sensitive to coupling between flame dynamics and combustor acoustics. In practice, combustor flames tend to drive instabilities at certain eigenfrequencies of the systems, according to the classical Rayleigh criterion. In order to guarantee combustor stability in the entire operation range of the engine it has to be avoided under all circumstances that the flame excites the system beyond its damping potential. One option to accomplish this is to provide sufficient damping capabilities of the combustor system so that the decay of acoustic energy inside the system always exceeds the excitation provided by the flame. Experimental methods for the determination of combustor damping rates therefore may become a valuable tool for combustor design in the future. In the past, methods with different accuracy, complexity and capabilities have been developed to gain experimental access to decay rates of the acoustic energy inside combustor systems.

In this study we compare accuracy and capabilities of three different time-domain methods that allow the determination of pressure decay rates from experimental dynamic pressure traces: A simple exponential fit to the measured dynamic pressure, a method based on the decay of acoustic energy and a newly developed statistical method are examined. In the first step, the methods are tested using artificially generated test signals. The simple signals decaying exponentially with the known rate α are of pure sinusoidal shape and have discrete frequencies. As practical dynamic pressure traces are in general corrupted with noise, in the second step of the analysis a certain amount of random noise is added to the test signal. The last step of the analysis involves realistic pressure signals obtained from a simple duct with throughflow. The results obtained from the different methods are compared with each other and differences regarding performance, accuracy, robustness as well as computational costs are presented.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A065. doi:10.1115/GT2013-94854.

The impact of fuel composition, engine power (idle and full power) and operation mode (cold and hot idle) on the gaseous emissions, particle number and mass concentrations and size distributions from an aircraft auxiliary power unit (APU) was investigated. A re-commissioned Artouste MK113 APU engine was used. The engine was run at three operational modes: i.e. approximately 6 minutes at idle (cold idle) after stabilized from start, 6 minutes at full power and then returning to idle again (hot idle) for 6 minutes. All operating parameters of the engine were monitored and recorded. The engine exhaust particle measurements and gaseous emissions were taken at three operating modes.

Five alternative fuels/blending components were tested and compared to neat conventional JetA1 fuel either in pure or blended forms. These fuels varied in their compositions in terms of H/C ratio, density and other properties. A Scanning Mobility Particle Sizer (SMPS) with a Nano-Differential Mobility Analyzer (NDMA) was used to determine the number and mass concentration and size distribution of engine exhaust in the size range from 5 nm to 160 nm. The influence of fuel elemental ratio (H/C), engine power and cold/hot operation on particle number and mass size distribution was investigated. The results show that there was a good correlation between fuels H/C ratio and particle concentrations, particle size and distributions characteristics. The engine at hot idle produced ∼20% less particles compare to the results at cold idle. The alternative fuel blends produced less particles than JetA1 fuel. The testing fuels produced similar levels of NOx, slight reductions in CO and remarkable reductions in UHC compared to JetA1.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A066. doi:10.1115/GT2013-94859.

In most dry low NOx combustor designs of stationary gas turbines the front panel impingement cooling air is directly injected into the combustor primary zone. This air partially mixes with the swirling flow of premixed reactants from the burner and reduces the effective equivalence ratio in the flame. However, local unmixedness and the lean equivalence ratio are supposed to have a major impact on combustion performance. Overall goal of this investigation is to answer the question whether the cooling air injection into the primary combustor zone has a beneficial effect on combustion stability and NOx emissions or not. The flame stabilization of a typical swirl burner with and without front panel cooling air injection is studied in detail under atmospheric conditions close to the lean blowout limit (LBO) in a full scale single burner combustion test rig. Based on previous isothermal investigations a typical injection configuration is implemented for the combustion tests. Isothermal results of experimental studies in a water test rig adopting high speed planar laser-induced fluorescence (HSPLIF) reveal the spatial and temporal mixing characteristics for the experimental setup studied under atmospheric combustion. This paper focuses on the effects of cooling air injection on both flame dynamics and emissions in the reacting case. To reveal dependencies of cooling air injection on combustion stability and NOx emissions, the amount of injected cooling air is varied. OH*-chemiluminescence measurements are applied to characterize the impact of cooling air injection on the flame front. Emissions are collected for different cooling air concentrations, both global measurements at the chamber exit and local measurements in the region of the flame front close to the burner exit. The effect of cooling air injection on pulsation level is investigated by evaluating the dynamic pressure in the combustor. The flame stabilization at the burner exit changes with an increasing degree of dilution with cooling air. Depending on the amount of cooling only a specific share of the additional air participates in the combustion process.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A067. doi:10.1115/GT2013-94865.

Lean blow out (LBO) has a big impact on emission formation at part load of gas turbines, where flame temperature is low and flame stabilization is an issue. With improved combustion behavior at LBO conditions the operation flexibility of a silo gas turbine can be increased within the scope of retrofitting. In multi burner arrangements a part of the preheated air designated for combustion is used for impingement cooling of the burner front panel and subsequently injected into the primary combustion zone. In this region of flame stabilization air and unburned fuel as well as burned products are mixed to sustain stable combustion. The object of this study is to determine the level of dilution of the flow field by the cooling air with the focus on the conditions below LBO that can impair flame stability. The question addressed in this paper is how mixing of the front panel cooling air with the incoming reactants and the combustion products in multi burner arrangements can be computed in a numerically efficient way. As test case for the methodology the local distribution of cooling air in a silo combustor is presented. In this numerical study mixing processes of air-fuel mixture and cooling air as well as aerodynamic interaction of adjacent burners in a multi burner systems are investigated using isothermal Reynolds Averaged Navier Stokes (RANS) simulations. Former published single burner water channel experiments and Large Eddy Simulations (LES) [1] serve as a baseline. Single burner RANS simulations are done and compared to measurement and LES to validate the velocity and scalar fields. A Schmidt number variation is used to modify the mixing process in the RANS single burner calculations. Based on the LES the single burner is modified to address the multi burner conditions and calculated with LES and RANS. Finally the multi burner system is computed with the settings applied in the single burner configuration. Using the symmetry of the investigated burner matrix an efficient methodology is implemented that allows computation of one sixth of a silo combustor. The results expose a strong burner-burner interaction of the recirculation zones and in contrast to the single burner configuration regions of concentrated cooling air.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A068. doi:10.1115/GT2013-94866.

This work presents a study of the effect of the inflow condition on the flame flashback performance of a gas turbine burner. A generic swirl burner for basic combustion research on engine scale is investigated both under atmospheric conditions in a combustion test rig and numerically to reveal the impact of inflow conditions on the burner stability. Flashback resistance is examined with highly reactive hydrogen fuel and numerical studies with isothermal large eddy simulations (LES) are performed to investigate transient flow field data. Earlier publications showed excellent flashback resistance of a down scaled burner version of similar design, which was tested in a rig with strongly restricted cross sectional inflow area. An influence of the test rig setup on the flashback limits was not expected. However, the results presented in the paper reveal that the inflow conditions at the swirler and the distribution of axial velocity inside the swirler are crucial for flame stability. The inflow conditions upstream of the swirler were modified to redistribute the axial velocity field inside the swirler. Velocity fluctuations both inside the swirler and downstream of the burner outlet were reduced and consequently the susceptibility to perturbations in the flow field. This measure prevents the formation and propagation of local zones of negative axial velocity upstream of the flame position and increases the robustness of the flow field. After modification of the inflow condition the excellent flashback limit data of the down scaled burner was fully reproduced.

Topics: Inflow
Commentary by Dr. Valentin Fuster
2013;():V01AT04A069. doi:10.1115/GT2013-94877.

In order to reduce NOx emissions, modern gas turbines are often equipped with lean burn combustion systems, where the engine operates near the lean blow-out limits. One of the most critical issues of lean combustion technology is the onset of combustion instabilities related to a coupling between pressure oscillations and thermal fluctuations excited by the unsteady heat release. In this work a thermoacoustic analysis of a full annular combustor developed by AVIO is discussed. The system is equipped with an advanced PERM (Partially Evaporating and Rapid Mixing) injection system based on a piloted lean burn spray flame generated by a pre-filming atomizer. Combustor walls are based on multi-perforated liners to control metal temperature: these devices are also recognized as very effective sound absorbers, thus in innovative lean combustors they could represent a good means both for wall cooling and damping combustion instabilities. The performed analysis is based on the resolution of the eigenvalue problem related to an inhomogeneous wave equation which includes a source term representing heat release fluctuations (the so called Flame Transfer Function, FTF) in the flame region using a three-dimensional FEM code. A model representing the entire combustor was assembled including all the acoustically relevant geometrical features. In particular, the acoustic effect of multi-perforated liners was introduced by modeling the corresponding surfaces with an equivalent internal impedance. Different simulations with and without the presence of the flame were performed analyzing the influence of the multi-perforated liners. Furthermore, different modeling approaches for the FTF were examined and compared with each other. Comparisons with available experimental data showed a good agreement in terms of resonant frequencies in the case of passive simulations. On the other hand, when the presence of the flame is considered, comparisons with experiments showed the inadequacy of FTFs commonly used for premixed combustion and thus the necessity of an improved FTF, more suitable for liquid fueled gas turbines where the evaporation process could play an important role in the flame heat release fluctuations.

Commentary by Dr. Valentin Fuster
2013;():V01AT04A070. doi:10.1115/GT2013-94888.

A 9-Point Lean Direct low emissions combustor concept was utilized to evaluate gaseous emissions performance of two bio-derived alternative jet fuels and a JP-8 fuel for comparison. Gaseous emissions were measured in a flame tube operating at inlet temperatures from 650 up to 1030 F, pressures of 150, 250, and 350 psia, and a range of fuel/air ratios. The alternative fuels consisted of a Hydroprocessed Esters and Fatty Acids Fuel made from tallow and a second bio derived fuel produced from direct fermentation of sugar.

Topics: Fuels , Emissions
Commentary by Dr. Valentin Fuster
2013;():V01AT04A071. doi:10.1115/GT2013-94918.

This article presents two dimensional (2D) and three-dimensional (3D) computational analysis of rotating detonation combustion (RDC) in annular chambers using the commercial computational fluid dynamics (CFD) solver ANSYS-Fluent V13. The applicability of ANSYS-Fluent to predict the predominant phenomena taking place in the combustion chamber of a rotating detonation combustor is assessed. Simulations are performed for stoichiometric Hydrogen-Air combustion using two different chemical mechanisms. First, a widely used one-step reaction mechanism that uses mass fraction of the reactant as a progress variable, then a reduced chemical mechanism for H2-Air combustion including NOx chemistry was employed. Time dependent 2D and 3D simulations are carried out by solving Euler equations for compressible flows coupled with chemical reactions. Fluent user defined functions (UDF) were constructed and integrated into the commercial CFD solver in order to model the micro nozzle and slot injection system for fuel and oxidizer, respectively. Predicted pressure and temperature fields and detonation wave velocities are compared for the two reaction mechanisms. Curvature effects on the properties of transverse detonation waves are studied by comparing the 2D and 3D simulations. The effects of diffusion terms on RDC phenomena are assessed by solving full Navier-Stokes equations and comparing the results with those from Euler equations. Computational results are compared with experimentally measured pressure data obtained from the literature. Results show that the detonation wave velocity is over predicted in all the simulations. However, good agreement between computational and experimental data for the pressure field and transverse detonation wave structure proves adequate capabilities of ANSYS-Fluent to predict the main physical characteristics of RDC operation. Finally, various improvements for RDC modeling are postulated, particularly for better prediction of wave velocity.

Commentary by Dr. Valentin Fuster

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In