Thomas Mosbach

An experimental investigation of kerosene droplet breakup by laser-induced blast waves

The work presented in this paper intends to deepen our understanding of the mechanisms involved in the spark ignition of liquid fuel sprays. An experimental study is presented regarding the ignition of monodisperse droplet chains of Jet A-1 aviation kerosene in a generic model combustor under well-defined boundary Conditions. Breakdowns created by focused laser radiation were used as ignition sparks. They featured rapid spatial expansion, resulting in the formation of spherical blast waves in the surrounding air. The focus of this study lay on the effect of the blast waves on the fuel droplets. Blast wave trajectories were investigated by Schlieren imaging. Their interaction with kerosene droplets was observed with a high speed camera via a long distance microscope; the droplets were visualized by laser-induced Mie scattering, Droplets within a distance of 10 mm from the breakdown position were deformed and disintegrated by the aerodynamic forces of the postshock flow field. Different breakup modes were observed, depending on the distance from the breakdown position: Catastrophic breakup was observed at a 5 mm distance, resonant breakup was observed at a 10 mm distance. Breakup by blast waves from ignition sparks is expected to be a crucial mechanism for spray ignition because it supports evaporation. Weber number calculations revealed that the breakup modes observed under lab conditions will also appear in aviation gas turbines at high altitude relight conditions.

Investigation of GTL-Like Jet Fuel Composition on GT Engine Altitude Ignition and Combustion Performance: Part II—Detailed Diagnostics

The ignition and combustion performance of different synthetic paraffinic kerosenes (SPKs) under simulated altitude relight conditions were investigated at the altitude relight test rig at the Rolls-Royce Strategic Research Centre in Derby. The conditions corresponded to a low stratospheric flight altitude between 25,000 and 30,000 feet. The combustor under test was a twin-sector representation of an advanced gas turbine combustor and fuel injector. Five different SPKs and Jet A-1 were tested at different mass flow rates of air and fuel, and at two different sub-atmospheric air pressures and temperatures. The fuel temperature was kept approximately constant. Simultaneous high-speed imaging of the OH* and CH* chemiluminescence, and of the broadband luminosity was used to visualize both the transient flame initiation phenomena and the combustion behavior of the steady burning flames. In addition, flame luminosity spectra were recorded with a spectrometer to obtain spectrally resolved information concerning the different chemiluminescence bands and the soot luminosity. These investigations were performed in conjunction with the comparative evaluation of the ignition and stability regimes of the five SPKs, which is the subject of a separate complementary paper [1]. We found that the observed flame initiation phenomena, the overall combustion behavior and the different ratios of the chemiluminescence from the OH*, CH* and C2 * radicals were not strongly dependent on the fuels investigated. But, the SPK flames showed for all combustor operating conditions significantly lower soot luminosities than the corresponding Jet A-1 flames, indicating a potential benefit of the SPK fuels.Copyright

Experimental Analysis of Altitude Relight Under Realistic Conditions Using Laser and High-Speed Video Techniques

The altitude relight performance of a lean fuel injector and combustor was investigated at the altitude relight test rig at the Rolls-Royce Strategic Research Centre (SRC) in Derby. The studies were performed for different mass flow rates of air and kerosene, a combustor temperature and pressure of 278 K and 0.5 bar, respectively. Good optical access to the combustion chamber enabled the application of optical and laser measuring techniques. High-speed video imaging in the UV and visible wavelength range at a frame rate of 3.5 kHz was used to visualize the temporal development of the flame kernel. The observed differences between the UV and visible flame emissions demonstrate the different origins of the luminosity, i.e. OH* chemiluminescence and soot radiation. Further, laser-induced fluorescence of kerosene and OH radicals was applied at a frame rate of 5 Hz to visualize the fuel distribution and regions of hot and reacting mixtures. For two exemplary flames with different mass flow rates and fuel-to-air ratios, the steady burning flames after successful ignition are characterized in this paper by the distributions of kerosene, OH*, OH and soot luminosity. An example of the flame kernel development for a successful ignition is given by an image sequence from a high-speed video recording of the chemiluminescence. The importance of the upstream movement of the flame kernel as a condition preceding successful flame stabilization is identified.Copyright

Validation of an Ignition and Flame Propagation Model for Multiphase Flows

This paper presents a numerical investigation of a generic lab scale combustor with focus on the ignition characteristics. The test case has been examined thoroughly in a comprehensive measurement campaign to provide a detailed set of boundary conditions and a profound data base of results. The experimental setup comprises five parallel-aligned mono-disperse droplet chains which are ignited, using a focused laser beam. One aspect of the experimental study is the ignitability with respect to the imposed boundary conditions. The second covers the growth and the propagation of the flame after the establishment of an initial kernel. The outcome of the numerical simulations is compared to the experimental results which allows an in-depth assessment of the employed numerical models. The chemistry and, thus, the flame propagation behavior is captured by a turbulent flame speed closure approach with an adaptation to render the model suitable to multiphase flows. For the dispersed phase a Lagrangian particle tracking scheme is employed in combination with a continuous thermodynamics fuel model for the evaporation. The overall good agreement demonstrates the capability of a multiphase flow CFD solver in the field of ignition modeling.Copyright

Fuel Influence on Targeted Gas Turbine Combustion Properties: Part II — Detailed Results

The paper presents the results from a study that formed part of a bilateral project between DLR-VT and Sasol Technology Fuels Research aimed at investigating the potential influence of physical and chemical fuel properties on ignition and extinction limits within heterogeneous gas turbine combustion. The threshold of flame extinction and re-ignition behaviour of a range of alternative fuels was investigated in a representative aero-combustor sector to determine the relative influence of physical properties and chemical reaction timescales.A matrix of eight test fuels was selected for use during the study and included conventional crude-derived Jet A-1, synthetic paraffinic kerosene, linear paraffinic solvents, aromatic solvents and pure compounds. All test fuels were characterised through full specification analyses, distillation profiles and two-dimensional gas chromatography.The ignition and extinction behaviour of the test fuel matrix was evaluated under simulated altitude conditions at the Rolls-Royce Strategic Research Centre’s sub-atmospheric altitude ignition facility in Derby, UK. A twin sector segment of a Rich Quench Lean (RQL) combustor was employed with fuel supplied to a single burner. Combustor air inlet conditions were controlled to 41.4 kPa and 265 K. Fuel temperature was controlled to 288 K.In addition to the standard extinction and ignition detection systems, optical diagnostics were applied during the test programme. Simultaneous high-speed imaging of the OH* chemiluminescence, and broadband flame luminosity was employed to capture the main reaction zones, the global heat release and distribution of radiative soot particles respectively.Lean extinction points were determined using both a photodiode as well as from the OH* chemiluminescence data. The position of extinction and overall combustor ignition and extinction timescales were determined. The diagnostic methodology that was used to obtain the results reported in this paper is discussed in greater detail in a separate complementary paper.All eight fuels, including the fully synthetic Jet A-1 fuels that formed part of the test matrix, yielded performance that was comparable to that obtained with conventional crude-derived Jet A-1.© 2014 ASME

Ignition and Flame Propagation along Planar Monodisperse Droplet Streams

An experimental and numerical study is presented concerning the ignition of a fuel spray under well defined conditions. The development of a flame kernel that follows the generation of a plasma by a focused laser pulse and the subsequent flame propagation along five co-planar monodisperse streams of fuel droplets is investigated. High-speed video of the broadband luminosity and simultaneous fuel- and OH-PLIF provide qualitative and quantitative experimental data. Numerical simulations have shown that the focused laser pulse not only provides an energy/radical source for the ignition but is also responsible for transforming the droplets present in a small region into fuel vapor. Actually, without the quasi-instantaneous phase change there would not be enough vapor fuel to sustain the flame kernel under these time scales and surrounding conditions. The simulations are performed using an Eulerian-Lagrangian turbulent spray combustion code. Detailed chemistry is used to compute the laminar flame speed under relevant conditions, and a turbulent flame closure model (TFC) is then adopted for turbulence-chemistry interaction.

Conditional Moment Closure LES Modelling of an Aero-Engine Combustor at Relight Conditions

A Conditional Moment Closure (CMC) approach embedded in an LES CFD framework is presented for simulation of the reactive flow field of an aero-engine combustor operating at altitude relight conditions. Before application to the combustor geometry, the CMC model was validated on the standard lab-scale Sandia flame D. For the combustor simulation, a global mechanism for n-heptane was used along with a Lagrangian approach for the spray, to which a secondary break-up model was applied. The simulation modelled a multi-sector sub-atmospheric rig that was used to verify the altitude relight capability of the combustor. A comprehensive suite of diagnostics was applied to the rig test, including high-speed OH and kerosene PLIF as well as high speed OH* chemiluminescence. The CMC-based CFD simulation was able to predict well the position of the flame front and fuel distribution at the low pressure, low temperature conditions typical of altitude relight. Furthermore, the simulation of the ignition showed strong similarities with OH* chemiluminescence measurements of the event. An EBU-based LES was run too and showed to be unable to capture the flame front as well as the CMC model could. This work demonstrates that CMC LES can be an effective tool to support assessment of the relight capability of aero-engine combustors.Copyright

A numerical investigation of the ignition characteristics of a spray flame under atmospheric conditions

This study presents a numerical investigation of the characteristics of a transient spray ame ignited by laser-induced breakdown. The simulations are carried out under atmospheric conditions within well de ned boundary conditions. The multiphase set-up comprises a kerosene blend as the liquid fuel provided by an air assisted nozzle with droplet velocities up to 40 m/s and an air coow with velocities ranging from 1.0 m/s to 4.5 m/s. An initial ame kernel is created by a laser-induced breakdown at the spray cone edge. Onsetting vaporization and further propagation of the ame, largely on the outskirt of the ame cone surface lead to a growing and convected self-sustaining ame. Results concerning the ame growth and the ame center position, representing the major general characteristics of the instationary combustion process are provided. As experimental data is available for the same set-up and boundary conditions, the simulation outcome is further compared with regard to its qualitative and quantitative agreement. The study shows that the employed numerical tools for predicting transient combustion are of very satisfying quality and can subsequently be used for more complex scenarios to evaluate questions concerning ame kernel development and igniter location.

Fuel Influence on Targeted Gas Turbine Combustion Properties: Part I — Detailed Diagnostics

The threshold combustion performance of different fuel formulations under simulated altitude relight conditions were investigated in the altitude relight test facility located at the Rolls-Royce plc. Strategic Research Centre in Derby, UK. The combustor employed was a twin-sector representation of an RQL gas turbine combustor. Eight fuels including conventional crude-derived Jet A-1 kerosene, synthetic paraffinic kerosenes (SPKs), linear paraffinic solvents, aromatic solvents and pure compounds were tested. The combustor was operated at sub-atmospheric air pressure of 41 kPa and air temperature of 265 K. The temperature of all fuels was regulated to 288 K. The combustor operating conditions corresponded to a low stratospheric flight altitude near 9 kilometres.The experimental work at the Rolls-Royce (RR) test-rig consisted of classical relight envelope ignition and extinction tests, and ancillary optical measurements: Simultaneous high-speed imaging of the OH* chemiluminescence and of the soot luminosity was used to visualize both the transient combustion phenomena and the combustion behaviour of the steady burning flames. Flame luminosity spectra were also simultaneously recorded with a spectrometer to obtain information about the different combustion intermediates and about the thermal soot radiation curve. This paper presents first results from the analysis of the weak extinction measurements. Further detailed test fuel results are the subject of a separate complementary paper [1].It was found in general that the determined weak extinction parameters were not strongly dependent on the fuels investigated, however at the leading edge of the OH* chemiluminescence intensity development in the pre-flame region fuel-related differences were observed.Copyright

Fuel Composition Influence on Gas Turbine Ignition and Combustion Performance

The influence of different jet fuel compositions on aviation gas turbine combustion performance was investigated. Eight fuels including conventional crude-derived Jet A-1 kerosene, fully synthetic Jet fuel, synthetic paraffinic kerosenes, linear paraffinic solvents, aromatic solvents and pure compounds were tested. The tests were performed in the altitude relight test facility located at the Rolls-Royce Strategic Research Centre in Derby (UK). The combustor employed was a twin-sector representation of an RQL gas turbine combustor. The combustor was operated at sub-atmospheric air pressure of 41 kPa and air temperature of 265 K. The temperature of the fuels was regulated to 288 K. The combustor operating conditions corresponded to a simulated low stratospheric flight altitude near 9,000 metres.The experimental work at the Rolls-Royce (RR) test-rig consisted of classical relight envelope ignition and extinction tests, and ancillary optical measurements: Simultaneous high-speed imaging of the OH* chemiluminescence and of the soot luminescence was applied to obtain spatial and temporal resolved insight into the ongoing processes. Optical emission spectroscopy was also applied simultaneously to obtain spectral and temporal resolved insight into the flame luminescence.First results from the analysis of the OH* chemiluminescence and detailed fuel analysis results were presented in previous papers [1, 2]. This article presents further results from the analysis of the soot luminescence imaging and flame spectra.It was found in general that the combustion performance of all test fuel formulations was comparable to regular Jet A-1 kerosene. Fuel related deviations, if existent, are found to be small.Copyright