Non-premixed Mode Research Articles

Numerical investigation of localised forced ignition of pulverised coal particle-laden mixtures: A Direct Numerical Simulation (DNS) analysis

Localised forced ignition of mono-disperse pulverised coal particle-laden mixtures has been analysed based on three-dimensional Direct Numerical Simulations for the carrier phase with simplified chemistry for the combustion of volatile gases. The coal particles are treated as point sources and tracked in a Lagrangian manner. The coupling between Eulerian gaseous and Lagrangian particulate phases has been achieved by appropriate source terms in the mass, momentum, energy and species conservation equations. A detailed parametric analysis has been carried out to analyse the effects of particle equivalence ratio Φp (which is defined based on the total available primary volatile fuel in the particulate phase), root-mean-square of turbulent velocity u′ and particle diameter dp on the early stages of combustion. Both non-premixed and premixed modes of combustion have been observed in the reaction zone for the flames resulting from localised ignition. An increase in Φp is found to be detrimental for sustaining combustion, whereas a reduction in particle size may adversely affect the extent of burning. It has been found that an increase in u′ increases the rate of mixing of devolatilised fuel with the surrounding air, which, though beneficial for sustaining combustion, increases the heat transfer rate from the hot gas kernel, thus leading to flame extinction for high values of u′. Detailed physical explanations have been provided to explain the observed effects of Φp, root-mean-square turbulent velocity fluctuation u′ and particle diameter dp on combustion of coal particle-laden mixtures following successful localised forced ignition.

On the Axisymmetric Counterflow Flame Simulations: Is There an Optimal Nozzle Diameter and Separation Distance to Apply Quasi One-Dimensional Theory?

Two-dimensional axisymmetric numerical analysis of counterflow flames was employed to better understand the applicability of the quasi one-dimensional theory of Seshadri and Williams to flames produced by small diameter convergent nozzles. The computational domain considered included the convergent sections of two opposed nozzles as well as the surrounding inert annular co-flows. For computational efficiency, the fuel-oxidizer system of diluted hydrogen versus air in non-premixed flame mode, with a detailed chemical kinetic model and mixture-averaged transport property description, was considered. With an increase of nozzle diameter from 6.5 mm to larger values (with plug flow velocity profiles at the nozzle exit), the influence of the radial terms on eigenvalue and scalar variables has been compared. Error metric on nozzle diameter effects is presented with comparison to typical experimental measurement uncertainties. The analysis also showed that, for nozzle separation distances below the free-floating limit, the self-similar function in quasi one-dimensional formulation can be preserved by specifying radial velocity boundary conditions, as long as the radial gradients are negligible.

Impact of pressure on high intensity colorless distributed combustion

In recent years, colorless distributed combustion (CDC) has been shown to provide ultra-low pollutants emission, enhanced stability, fuel flexibility and thermal field uniformity. To achieve CDC conditions, fuel–air mixture must be properly prepared prior to the mixture ignition. In this paper, the impact of moderate pressure increase on a test combustor is examined with emphasis on pollutants emission under near CDC conditions. Increase in pressure at constant mass flow rate resulted in a significant increase in NO emission and departure from CDC conditions. This is a result of reduction of the volume flow rate and injection velocity with pressure increase, as well as an increase in the apparent residence time of gases within the combustor. These conditions led to less than adequate mixing and more residence time of gases in the post flame zone to form NOx. CO emissions were reduced as longer residence time allowed for complete conversion of CO to CO2. Also, the increase in pressure enhanced chemical kinetics and suppressed dissociation. Increasing the combustor pressure with constant heat release intensity, where mass flow rates are increased to keep the apparent residence time constant, showed a slight increase (1PPM) of NO under premixed conditions and significant reduction of CO. OH∗ chemiluminescence showed that the reaction zone did not change with increase in pressure under premixed combustion. For non-premixed combustion, the reaction zone intensity increased and was more concentrated with increase in pressure to result in higher emissions as compared to the premixed case. In both cases, CO emission was significantly reduced due to the increased heat load and temperatures within the combustor. Relations describing change of NO and CO emissions with pressure were used to predict the combustor emissions at higher pressures. Difference between premixed and non-premixed modes revealed the need for more adequate mixing, necessary for CDC condition. Increase in pressure without considering flow velocities and flowfield can lead to departure from CDC conditions with simultaneous increase in emissions.

Investigation of flame propagation in a partially premixed jet by high-speed-Stereo-PIV and acetone-PLIF

This paper presents an experimental study of flame propagation through a partially-premixed flow following ignition. A combination of simultaneous high-speed acetone planar laser induced fluorescence (PLIF) and stereoscopic particle image velocimetry (SPIV) was utilized for time-resolved measurements of mixture fraction, flow field and flame position. This provides access to the major quantities needed to characterize non-premixed flames. High quality mixture fraction measurements with signal-to-noise ratios up to 120 for unity mixture fraction were made feasible using a combination of a conventional high-speed laser at 10kHz for LIF excitation and a wavelet based de-noising algorithm to reject camera noise. It was observed that flame propagation in the far-field of a partially-premixed jet takes place in a premixed mode, with the flame propagating through highly stratified mixtures until it approaches locations containing mixtures outside the flammability limits. In these areas the flame recedes and further propagation is controlled by mixing processes of air and fuel. Even though flame propagation is then mixing-controlled, the flame is not observed to switch into a non-premixed mode. Instead, mixing ahead of the flame takes place until locally premixed flammable mixtures are recovered for subsequent flame propagation.

Stability analysis of a swirl spray combustor based on flame describing function

Self-sustained combustion oscillations observed at limit cycles in a swirl combustor equipped with a series of steam assisted liquid fuel injectors are analyzed with the flame describing function (FDF) framework. It is first shown that for globally fuel lean injection conditions, the spray flames investigated burn in a non-premixed mode. A non-sooty and a sooty regime are explored. Their frequency response to acoustic forcing is characterized by exploiting the OH∗ chemiluminescence signal and a velocity signal measured by Laser Doppler velocimetry at the air injection unit outlet. It is found that the FDF of these non-premixed swirl spray flames features a distinct response compared to the FDF of lean premixed swirl flames. The gain and phase lag of these FDF are both a strong function of the perturbation amplitude. At a fixed forcing frequency, the FDF phase lag increases with the perturbation level for the non-sooty flame investigated and it decreases for the sooty flame case. The sooty flame has also a much lower cut-off frequency for the FDF gain than the non-sooty flame. It is then shown that these features are essential to reproduce the correct instability bands and oscillation frequencies observed at limit cycles in the experiments. For the sooty flame, combustion is always found stable in agreement with the low cut-off frequency found for the FDF. A linear analysis for the non-sooty flame fails to capture the low frequency instabilities observed in the experiments that have the largest oscillation levels. Only the use of the FDF yields the correct dynamical states observed in the combustor. It is further concluded that the OH∗ chemiluminescence signal may safely be used to infer the frequency response of these non-premixed swirl spray flames at globally fuel lean injection conditions.

Modelling of turbulent lifted jet flames using flamelets: a priori assessment and a posteriori validation

This study focuses on the modelling of turbulent lifted jet flames using flamelets and a presumed Probability Density Function (PDF) approach with interest in both flame lift-off height and flame brush structure. First, flamelet models used to capture contributions from premixed and non-premixed modes of the partially premixed combustion in the lifted jet flame are assessed using a Direct Numerical Simulation (DNS) data for a turbulent lifted hydrogen jet flame. The joint PDFs of mixture fraction Z and progress variable c, including their statistical correlation, are obtained using a copula method, which is also validated using the DNS data. The statistically independent PDFs are found to be generally inadequate to represent the joint PDFs from the DNS data. The effects of Z–c correlation and the contribution from the non-premixed combustion mode on the flame lift-off height are studied systematically by including one effect at a time in the simulations used for a posteriori validation. A simple model including the effects of chemical kinetics and scalar dissipation rate is suggested and used for non-premixed combustion contributions. The results clearly show that both Z–c correlation and non-premixed combustion effects are required in the premixed flamelets approach to get good agreement with the measured flame lift-off heights as a function of jet velocity. The flame brush structure reported in earlier experimental studies is also captured reasonably well for various axial positions. It seems that flame stabilisation is influenced by both premixed and non-premixed combustion modes, and their mutual influences.

Hydroxyl radical distribution in distributed reaction combustion condition

Distributed combustion has been shown to provide significant performance enhancement in terms of ultra low pollutants emission and enhanced combustion stability. In our quest to achieve distributed conditions, several different distributed combustion geometries have been investigated with focus on the resultant pollutants emission and OH* chemiluminescence. In this paper, Planar Laser Induced Fluorescence technique was used to excite OH radicals. OH-PLIF diagnostic was employed to examine the extent of distributed reaction in the combustor. Q1(5) and Q1(6) activation lines were used to examine lean premixed and non-premixed flames. OH* is used as a flame marker, indicating the location of the reaction zone. A comparison between OH-PLIF and OH* chemiluminescence showed OH* chemiluminescence to identify areas of high OH concentration while OH-PLIF technique detected OH in low temperature regions. OH-PLIF experiments showed that OH radicals exist in the mixture prior to ignition, with the highest signal to reside in the same general areas of the combustor where OH* chemiluminescence signal was recorded. OH-PLIF data for premixed combustion showed the location of reaction zone downstream of the air entry jet location while the non-premixed flame showed a more distributed OH concentration across the combustor. The OH-PLIF indicated that OH radical occupy a large portion of the combustor for non-premixed mode while the OH* chemiluminescence signal showed that the thermally excited OH* is less distributed which is conjectured to provide higher NO emission.

Role of thermal intensity on operational characteristics of ultra-low emission colorless distributed combustion

This paper examines the development of ultra-low emission colorless distributed combustion (CDC) for gas turbines, operating at thermal intensity in the range of 5–453MW/m3atm. Higher thermal intensity combustors are desirable for increased performance with minimal increase in hardware costs in terms of both weight and volume of gas turbine combustors. Most land based gas turbine combustors operate at thermal intensity of about 15MW/m3atm but operation at higher intensity can provide increased performance. Design of high thermal intensity CDC combustor requires control of critical parameters, such as gas recirculation, fuel/oxidizer mixing, air preheats, and residence time distribution characteristics via proper selection of different air and fuel injection configurations to achieve desirable combustion characteristics. Initially, various flow field configurations were investigated at a low thermal intensity of 5MW/m3atm. This effort was followed by investigations at higher thermal intensity ranges of 20–40MW/m3atm by reducing the combustor volume for a forward flow configuration under more favorable novel non-premixed conditions. Further investigations were performed for a simpler combustor having single injection ports for air and fuel with detailed investigation of various flow field configurations performed at a thermal intensity of 28MW/m3atm. Further reduction in combustion volume resulted in thermal intensity of 57MW/m3atm. For the flow configurations investigated, reverse cross-flow configuration was found to give more favorable results with air injected from exit end and fuel injected in a cross-flow direction to the air flow. This reverse cross-flow geometry was investigated in detail by further reducing the combustor volume. Air injection diameter was increased to reduce the pressure drop across the combustor. The combustor was investigated at thermal intensity in the range of 53–85MW/m3atm. This geometry resulted in 4ppm NO emission and CO emissions of about 27ppm under novel non-premixed flow conditions at thermal intensity of 53MW/m3atm. The pressure drop at this operational point was less than 5% to meet the conventional combustor requirements. Further reduction in combustor volume resulted in very high thermal intensity in the range of 156–198MW/m3atm and at the most desirable operational point NO and CO emissions were 8ppm and 82ppm, respectively under novel non-premixed mode at a thermal intensity of 170MW/m3atm. Even further reduction in combustor volume resulted in combustor operation at ultra-high thermal intensity of 283–453MW/m3atm and at the desirable operating point the NO and CO emissions were 4ppm and 115ppm respectively at a thermal intensity of 340MW/m3atm. The sequential efforts reveal the possibility of significantly increasing the thermal intensity while still achieving desirable emission levels.

Novel mixing for ultra-high thermal intensity distributed combustion

Ultra-high thermal intensity colorless distributed combustion has been examined for our quest to achieve near zero pollutants emission and enhanced performance. Reverse-cross flow configuration has been investigated at thermal intensities in the range of 270–420MW/m3-atm with specific focus on the exhaust emissions, distribution of intermediate radical species and flow field within the combustor. Novel simplified geometry is used for easy transition to practical ultra-high thermal intensity gas turbine engine applications. The combustion intensities demonstrated here are significantly higher than that used in current gas turbine engines. Numerical simulations under non-reacting conditions are used to understand the effects of fuel injection in reverse-cross flow configuration. Ultra-low NO emissions were achieved for both the premixed (2ppm) and non-premixed (5ppm) mode at thermal intensity of 317MW/m3-atm. The CO emission of about 100ppm and UHC emission of less than 10ppm, revealed very high combustion efficiency at high heat release intensities. Novel mixing technique is also investigated to further decrease the emissions and enhance reaction distribution. In this case the fuel jet is diluted with a portion of the required air while a portion of the fuel is introduced with the air jet. The equivalence ratio of each of the air jet and fuel jet were kept well outside of the methane air flammability limits, for mitigating the possibility of flame flashback. This approach has demonstrated near zero NO emission, equivalent to that encountered in premixed combustion mode. Numerical simulation and OH* chemiluminescence showed that the mixing and combustion behavior is dictated by the high momentum of the diluted fuel jet suggesting favorably mixed oxidizer prior to ignition that results in lower NO emissions. Results obtained with different air jet and fuel jet equivalence ratios on the emissions of NO and CO, and OH* chemiluminescence are presented with view to further develop ultra-high thermal intensity zero emission combustors.

Effects of Electric Field on the SHS Flame Propagation of the Si-C System, Examined by the Use of the Heterogeneous Theory

Relevant to the self-propagating high-temperature synthesis (SHS) process, an analytical study has been conducted to investigate the effects of electric field on the combustion behavior because the electric field is indispensable for systems with weak exothermic reactions to sustain flame propagation. In the present study, use has been made of the heterogeneous theory which can satisfactorily account for the premixed mode of the bulk flame propagation supported by the nonpremixed mode of particle consumption. It has been confirmed that, even for the SHS flame propagation under electric field, being well recognized to be facilitated, there exists a limit of flammability, due to heat loss, as is the case for the usual SHS flame propagation. Since the heat loss is closely related to the representative sizes of particles and compacted specimen, this identification provides useful insight into manipulating the SHS flame propagation under electric field, by presenting appropriate combinations of those sizes. A fair degree of agreement has been demonstrated through conducting an experimental comparison, as far as the trend and the approximate magnitude are concerned, suggesting that an essential feature has been captured by the present study.

Mixture Preparation Effects on Distributed Combustion for Gas Turbine Applications

Distributed combustion is now known to provide significantly improved performance of gas turbine combustors. Key features of distributed combustion include uniform thermal field in the entire combustion chamber for significantly improved pattern factor and avoidance of hot-spot regions that promote thermal NOx emissions, negligible emissions of hydrocarbons and soot, low noise, and reduced air cooling requirements for turbine blades. Distributed combustion requires controlled mixing between the injected air, fuel, and hot reactive gasses from within the combustor prior to mixture ignition. The mixing process impacts spontaneous ignition of the mixture to result in improved distributed combustion reactions. Distributed reactions can be achieved in premixed, partially premixed, or nonpremixed modes of combustor operation with sufficient entrainment of hot and active species present in the combustion zone and their rapid turbulent mixing with the reactants. Distributed combustion with swirl is investigated here to further explore the beneficial aspects of such combustion under relevant gas turbine combustion conditions. The near term goal is to develop a high intensity combustor with ultralow emissions of NOx and CO, and a much improved pattern factor and eventual goal of near zero emission combustor. Experimental results are reported for a cylindrical geometry combustor for different modes of fuel injection with emphasis on the resulting pollutants emission. In all the cases, air was injected tangentially to impart swirl to the flow inside the combustor. Ultra low NOx emissions were found for both the premixed and nonpremixed combustion modes for the geometries investigated here. Results showed very low levels of NO (∼10 ppm) and CO (∼21 ppm) emissions under nonpremixed mode of combustion with air preheats at an equivalence ratio of 0.6 and a moderate heat release intensity of 27 MW/m3-atm. Results are also reported on lean stability limits and OH* chemiluminescence under different fuel injection scenarios for determining the extent of distribution combustion conditions. Numerical simulations have also been performed to help develop an understanding of the mixing process for better understanding of ignition and combustion.

Fuel dilution and liquid fuel operational effects on ultra-high thermal intensity distributed combustor

Ultra-high thermal intensity colorless distributed combustion has been examined for methane and methane diluted with air, nitrogen or carbon dioxide gas and liquid fuel (ethanol) burning in air at a fixed thermal load. Fuel was diluted with inert gases and air to simulate the combustion of low heating value fuels with special focus on the emission of pollutants. Reverse cross-flow configuration has been investigated at very high thermal intensities in the range of 156–198MW/m3atm with specific focus on the exhaust emissions, and distribution of intermediate radical species emission and flowfield from within the combustor using novel but simplified geometry for easy transition to practical applications in ultra-high thermal intensity gas turbine engine applications. The ultra-high combustion intensity demonstrated here is significantly higher than that used in current stationary gas turbine engines. Numerical simulations are also performed under non-reacting conditions to understand the effects of fuel injection in cross flow configuration for both undiluted and diluted fuel cases. Ultra low NOx emissions are achieved for both the novel premixed (3ppm) and non-premixed (8ppm) mode using methane as fuel at equivalence ratio of 0.6. Carbon monoxide level of about 100ppm was obtained in both novel premixed and non-premixed modes of combustion at equivalence ratio of 0.6. Dilution of fuel significantly reduced NO emissions to 2ppm from 8ppm and slightly increased CO emissions in non-premixed mode at equivalence ratio of 0.6. Dilution of fuel suggested much lower OH* suggesting favorably mixed oxidizer prior to ignition that results in lower NO emissions. Liquid fuel (ethanol) was also examined and very low NO emission of about 6ppm was obtained in direct-injection mode and 2ppm in premixed–prevaporized mode at equivalence ratio of 0.6. CO emission of about 200ppm was observed in both the modes at equivalence ratio of 0.6.

High intensity colorless distributed combustion for ultra low emissions and enhanced performance

Combustion characteristics of colorless distributed combustion have been investigated for application to gas turbine combustors. Very high intensity distributed combustion has been shown for application to stationary gas turbine engines. Various configurations examined have revealed reverse cross-flow mode to be more favorable for desirable combustion characteristics. The reverse-cross flow geometry is further investigated experimentally at range thermal intensities from 53 to 85 MW/m 3 atm with specific focus on exhaust emissions, radical emission, global flame photographs and flowfield using novel but simplified geometry for easy transition to applications in gas turbine engine applications. The high combustion intensity demonstrated here is higher than that used in present stationary gas turbine engines. Numerical simulations are also performed and compared with the experiments for the new design configuration under non-reacting conditions. Ultra low NO x emissions are achieved for both the novel premixed (1 ppm) and non-premixed (4 ppm) combustion modes reported here. Carbon monoxide levels of about 30 ppm are achieved in both novel premixed and non-premixed modes of combustion with a pressure drop of less than 5% across the combustor at the favorable condition. Almost no visible flame color in the reaction zones are observed for both novel premixed and non-premixed modes with volume distributed combustion so that this mode is termed as colorless distributed combustion. This mode of volume distributed colorless combustion is dramatically different than that used in contemporary gas turbine combustion operating under lean premixed, lean direct injection or rich burn, quick quench lean burn gas turbine combustion.

Capabilities and limitations of multi-regime flamelet combustion models

Flamelet combustion models typically assume that burning occurs in either a fully premixed or a fully non-premixed mode. These assumptions tend to limit the applicability of the models to single-regime combustors. Efforts aimed at reducing this limitation have introduced multi-regime approaches that account for different types of mixing and chemistry interactions. In this study a multi-regime model is applied to two laminar n-heptane flames in an effort to characterize the capabilities and limitations of the approach. Both a 2-D laminar triple flame and a 2-D laminar counter-flow diffusion flame are numerically simulated using the multi-regime model. Data for comparison is generated by additionally simulating the flames using finite rate chemistry, a purely premixed flamelet model, and a purely non-premixed flamelet model. Simulations demonstrate that the multi-regime approach functions as desired, and tends to access flamelets from the appropriate regime under both non-premixed and premixed conditions. Some important differences between the flamelet solutions and finite rate solution are observed, however. These differences are caused by the finite rate solution deviating away from the assumed flamelet manifolds, rather than by inadequate regime predictions. In the analyses of these simulations, an emphasis is placed on understanding the formation of the pollutant species NO. It is shown that even when the local combustion regime is correctly predicted, small deviations from an assumed flamelet manifold can lead to changes in the NO production rate. The simulation results confirm that multi-regime flamelet models are applicable to a wide variety of reacting flows, but the results also help to characterize the limitations of these models.

Swirling distributed combustion for clean energy conversion in gas turbine applications

Distributed combustion provides significant performance improvement of gas turbine combustors. Key features of distributed combustion includes uniform thermal field in the entire combustion chamber, thus avoiding hot-spot regions that promote NO x emissions (from thermal NO x ) and significantly improved pattern factor. Rapid mixing between the injected fuel and hot oxidizer has been carefully explored for spontaneous ignition of the mixture to achieve distributed combustion reactions. Distributed reactions can be achieved in premixed, partially premixed or non-premixed modes of combustor operation with sufficient entrainment of hot and active species present in the flame and their rapid turbulent mixing with the reactants. Distributed combustion with swirl is investigated here for our quest to explore the beneficial aspects of such flows on clean combustion in simulated gas turbine combustion conditions. The goal is to develop high intensity combustor with ultra low emissions of NO and CO, and much improved pattern factor. Experimental results are reported from a cylindrical geometry combustor with different modes of fuel injection and gas exit stream location in the combustor. In all the configurations, air was injected tangentially to impart swirl to the flow inside the combustor. Ultra-low NO x emissions were found for both the premixed and non-premixed combustion modes for the geometries investigated here. Swirling flow configuration, wherein the product gas exits axially resulted in characteristics closest to premixed combustion mode. Change in fuel injection location resulted in changing the combustion characteristics from traditional diffusion mode to distributed combustion regime. Results showed very low levels of NO (∼3 PPM) and CO (∼70 PPM) emissions even at rather high equivalence ratio of 0.7 at a high heat release intensity of 36 MW/m 3-atm with non-premixed mode of combustion. Results are also reported on lean stability limit and OH * chemiluminescence under both premixed and non-premixed conditions for determining the extent of distribution combustion conditions.

Hydrogen addition effects on methane–air colorless distributed combustion flames

In this investigation the role of hydrogen addition in a reverse flow configuration, consisting of both non-premixed and premixed combustion modes, have been examined for the CDC flames. In the non-premixed configuration the air injection port is positioned at combustor exit end while the fuel injection port is positioned on the side so that the fuel is injected in cross-flow with respect to air injection. The thermal intensity of the flames investigated is 85 MW/m 3 atm to simulate high thermal intensity gas turbine combustion conditions. The results are presented on the global flame signatures, exhaust emissions, and radical emissions using experiments and flowfield using numerical simulations. Ultra low NO x emissions are found for both the premixed and non-premixed combustion modes. Addition of hydrogen to methane fuel resulted in only a slight increase of NO emission, significant decrease of CO emission and extended the lean operational limit of the combustor.

Combustion Dynamics of Plasma-Enhanced Premixed and Nonpremixed Flames

Combustion dynamics are investigated for plasma-enhanced methane-air flames in premixed and nonpremixed configurations using a transient arc dc plasmatron. Planar laser-induced fluorescence images of hydroxyl (OH) and carbon monoxide (CO) radicals are obtained over a range of equivalence ratios (φ = 0.7 - 1.3), flow rates (6-18 LPM), and plasma powers (100-900 mA) to monitor radical propagation and in situ fuel reforming. The flow rates presented here are outside the range of normal flame stability. In the nonpremixed mode, the fuel is injected separately as a coflow around the plasma discharge, resulting in a unique two-cone flame front geometry, and the flame stability is mainly dependent on the flow dynamics. For premixed flames, partial oxidation occurs inside the chamber, resulting in higher energy conversion efficiencies, and stability is shown to be sensitive to the combustion chemistry. Both configurations are significantly influenced by in situ fuel reforming at higher plasma powers.

Investigation of reverse flow distributed combustion for gas turbine application

In this paper reverse flow modes of colorless distributed combustion (CDC) have been investigated for application to gas turbine combustors. Rapid mixing between the injected fuel and hot oxidizer has been carefully explored for spontaneous ignition of the mixture to achieve distributed combustion reactions. Distributed reactions can be achieved in premixed, partially premixed or non-premixed modes of combustor operation with sufficient entrainment of burned gases and faster turbulent mixing between the reactants. In the present investigation reverse flow modes consisting of three configurations at thermal intensity of 28 MW/m 3-atm and five configurations at thermal intensity of 57 MW/m 3-atm have been investigated and these high thermal loadings represent characteristic gas turbine combustion conditions. In all the configurations the air injection port is positioned at the combustor exit end, whereas the location of fuel injection ports is changed to give different configurations. The results are presented on the exhaust emissions and radical emissions using experiments, and evaluation of flowfield using numerical simulations. Ultra-low NO x emissions were found for both the premixed and non-premixed combustion modes investigated here. Cross-flow configuration, wherein the fuel is injected at high velocity cross stream to the air jet resulted in characteristics closest to premixed combustion mode. Change in fuel injection location resulted in changing the combustion characteristics from closer to diffusion mode to distributed regime. This feature is beneficial for part load operation where higher stability limit is desirable.

Development of high intensity CDC combustor for gas turbine engines

Colorless distributed combustion (CDC) has been demonstrated to provide ultra-low emission of NO x and CO, improved pattern factor and reduced combustion noise in high intensity gas turbine combustors. The key feature to achieve CDC is the controlled flow distribution, reduce ignition delay, and high speed injection of air and fuel jets and their controlled mixing to promote distributed reaction zone in the entire combustion volume without any flame stabilizer. Large gas recirculation and high turbulent mixing rates are desirable to achieve distributed reactions thus avoiding hot spot zones in the flame. The high temperature air combustion (HiTAC) technology has been successfully demonstrated in industrial furnaces which inherently possess low heat release intensity. However, gas turbine combustors operate at high heat release intensity and this result in many challenges for combustor design, which include lower residence time, high flow velocity and difficulty to contain the flame within a given volume. The focus here is on colorless distributed combustion for stationary gas turbine applications. In the first part of investigation effect of fuel injection diameter and air injection diameter is investigated in detail to elucidate the effect fuel/air mixing and gas recirculation on characteristics of CDC at relatively lower heat release intensity of 5 MW/m 3 atm. Based on favorable conditions at lower heat release intensity the effect of confinement size (reduction in combustor volume at same heat load) is investigated to examine heat release intensity up to 40 MW/m 3 atm. Three confinement sizes with same length and different diameters resulting in heat release intensity of 20 MW/m 3 atm, 30 MW/m 3 atm and 40 MW/m 3 atm have been investigated. Both non-premixed and premixed modes were examined for the range of heat release intensities. The heat load for the combustor was 25 kW with methane fuel. The air and fuel injection temperature was at normal 300 K. The combustor was operated at 1 atm pressure. The results were evaluated for flow field, fuel/air mixing and gas recirculation from numerical simulations and global flame images, and emissions of NO, CO from experiments. It was observed that the larger air injection diameter resulted in significantly higher levels of NO and CO whereas increase in fuel injection diameter had minimal effect on the NO and resulted in small increase of CO emissions. Increase in heat release intensity had minimal effect on NO emissions, however it resulted in significantly higher CO emissions. The premixed combustion mode resulted in ultra-low NO levels (<1 ppm) and NO emission as low as 5 ppm was obtained with the non-premixed flame mode.

Large-Eddy Simulation of oxygen/methane flames under transcritical conditions

Large-Eddy Simulation of transcritical combustion is considered in this article. This analysis is motivated by rocket propulsion applications where cryogenic oxygen is injected at low temperature, below the critical value, in a high pressure environment where the pressure is above critical. Problems encountered in this range are related to the description of real-gas effects and in particular to the representation of the very large density changes which occur when a transcritical fluid evolves from low to high temperatures. The simulation also requires a suitable spatially filtered combustion model which can be used to determine the volumetric reaction rates in a situation where reaction essentially takes place in a non-premixed mode. A model is proposed and validated in the “H3” atmospheric flame configuration. This model is then used in a transcritical version of the AVBP LES flow solver. Results of calculations carried out for transcritical oxygen/supercritical methane injection are in good agreement with recent experimental data and provide additional information on the flame structure and flow configuration.