Two-point time-series measurementsof hydroxyl concentration in a turbulent nonpremixed flame

Quantitative two-point hydroxyl time-series measurements have been performed in a turbulent nonpremixed flame by using two-point picosecond time-resolved laser-induced fluorescence. The current diagnostic system has been improved from its preliminary version to address optical aberrations and fluorescence lifetime fluctuations. In particular, with a newly designed collection system, the aberration-limited blur spot is reduced from 6 mm to 180 microm. Additional photon-counting channels enable the recovery of absolute OH concentrations through a triple-bin integration method. The present research represents what we believe to be the first application of this two-point technique to turbulent flames. Two independent schemes have been applied to remove uncorrelated noise in the derived two-point statistics. We show that optical aberrations can have a significant effect on space-time correlations. However, the sampling rate and fluctuations in the fluorescence lifetime barely affect the spatial autocorrelation function and thus the integral length scale. Such length scales for hydroxyl are found to rise linearly with increasing axial distance at peak [OH] locations. Along the jet centerline, the integral length scale varies slightly below the flame tip but increases rapidly above the flame tip. In addition, the OH length scale demonstrates the same trend as the OH time scale along the jet centerline, but the opposite trend at peak [OH] locations.     

Analysis of laser-induced-fluorescence carbon monoxide measurements in turbulent nonpremixed flames

The influence of fluctuating concentrations and temperature on the laser-induced-fluorescence (LIF) measurement of CO in turbulent flames is described, under conditions in which the fluorescence and the temperature are measured independently. The analysis shows that correlations between CO concentration and temperature can bias the averaged mole fraction extracted from LIF measurements. The magnitude of the bias can exceed the order of the average CO mole fraction. Further, LIF measurements of CO concentrations in a turbulent, nonpremixed, natural gas flame are described. The averaged CO mole fractions are derived from the fluorescence measurements by the use of flame temperatures independently measured by coherent anti-Stokes Raman spectroscopy. Analysis of the fluctuations in measured temperature and fluorescence indicates that temperature and CO concentrations in flame regions with intensive mixing are indeed correlated. In the flame regions where burnout of CO has ceased, the LIF measurements of the CO mole fraction correspond to the probe measurements in exhaust.     

Quantitative technique for imaging mixture fraction, temperature, and the hydroxyl radical in turbulent diffusion flames

A technique for obtaining simultaneous quantitative images of the hydroxyl radical, OH, temperature, mixture fraction, and scalar dissipation rates in turbulent diffusion flames is described. Mixture fraction is obtained from images of Rayleigh and fuel Raman scattering. We quantified the OH laser-induced fluorescence (LIF) images using detailed calibration and a correction for quenching and population distribution effects based on the simultaneous mixture fraction and temperature images. This correction was derived from calculations of laminar counterflow diffusion flames for identical fuel mixtures. These laminar flame computations are further used to estimate the errors in the measured OH concentrations. The technique is applied to piloted, nonpremixed flames over a range of jet velocities. The measured mixture fraction, temperature, and OH concentrations are in good agreement with those obtained earlier in similar flames using the single-point Raman/Rayleigh/LIF technique.     

Quantitative temperature measurements in high-pressure flames with multiline NO-LIF thermometry

An accurate temperature measurement technique for steady, high-pressure flames is investigated using excitation wavelength-scanned laser-induced fluorescence (LIF) within the nitric oxide (NO) A-X(0, 0) band, and demonstration experiments are performed in premixed methane/air flames at pressures between 1 and 60 bars with a fuel/air ratio of 0.9. Excitation spectra are simulated with a computational spectral simulation program (LIFSim) and fit to the experimental data to extract gas temperature. The LIF scan range was chosen to provide sensitivity over a wide temperature range and to minimize LIF interference from oxygen. The fitting method is robust against elastic scattering and broadband LIF interference from other species, and yields absolute, calibration-free temperature measurements. Because of loss of structure in the excitation spectra at high pressures, background signal intensities were determined using a NO addition method that simultaneously yields nascent NO concentrations in the postflame gases. In addition, fluorescence emission spectra were also analyzed to quantify the contribution of background signal and to investigate interference in the detection band-width. The NO-LIF temperatures are in good agreement with intrusive single-color pyrometry. The proposed thermometry method could provide a useful tool for studing high-pressure flame chemistry as well as provide a standard to evaluate and validate fast-imaging thermometry techniques for practical diagnostics of high-pressure combustion systems.     

Quantitative species measurements in microgravity flames with near-IR diode lasers

Absolute concentrations of water vapor are measured in microgravity (μ-g), nonpremixed methane, and propane jet flames with diode-laser wavelength modulation spectroscopy. These experiments are performed in the 2.2-s μ-g drop facility at the NASA Lewis Research Center. Abel inversion methods are used to determine time-dependent radial profiles from eight line-of-sight projections across the flames. At all measured heights above the nozzle, water vapor spatial distributions in μ-g flames are much wider than their 1-g counterparts. Radial growth of the water signal continues throughout the drop, verifying earlier suggestions that a steady state is not reached during the duration of the test, despite a quasi-steady flame shape. Large amounts of water vapor are observed at larger radii, at odds with visual (video) observations and numerical predictions.     

Temperature field measurements of small,nonpremixed flames with use of an Abel inversion of holographic interferograms

Interferometry has been used for many years as a semi-quantitative image-based diagnostic for combustion research. In this paper, we use image-plane, double-pulse holographic interferograms of axisymmetric flames to infer their radial temperature distribution. An Abel inversion is performed on the fringe data to account for line-of-sight integration through the flame. The sensitivity of nonresonant refractive diagnostics decreases inversely with temperature, and the accuracy of the technique is discussed in this context. A small, nonpremixed capillary flame is investigated, and the temperatures inferred from interferometry are compared with those obtained with N2 coherent anti-Stokes Raman spectroscopy thermometry. Additionally, the thermal field of a burning monodisperse methanol droplet stream is investigated interferometrically. Because of their small size, both of these flames challenge the performance limit of temperature interferometery.     

A three-component LDV system for measurements of higher statistical moments in turbulent diffusion flames

A three-component LDV system was built with the following features: 1. Three colors, 2. Forward scattering, 3. Direction determination, 4. Avalanche photodiodes, 5. Transient recorder, 6. Fast Fourier transform. With this equipment measurements in a free air jet and a turbulent methane nitrogen diffusion flame were made. The following quantities were determined on the axis and on some levels: velocities u_i, spreading rate r_(0.5·uc)/d, centerline velocity-decay u₀/u_c, Reynolds-stress tensor u′_iu′_j and all third order moments u′_iu′_ju′_k. These quantities were normalized and compared with literature data and with Reynolds-stress model calculations. One of the main objectives of these measurements was the supply of exit velocity profiles because the predictions below the similarity region depend crucially on the boundary conditions. In this paper details of the equipment and of the data processing are explained and new results are shown.     

Temperature imaging in nonpremixed flames by joint filtered Rayleigh and Raman scattering

Joint fuel Raman and filtered Rayleigh-scattering (FRS) imaging is demonstrated in a laminar methane-air diffusion flame. These experiments are, to our knowledge, the first reported extension of the FRS technique to nonpremixed combustion. This joint imaging approach allows for correction of the FRS images for the large variations in Rayleigh cross section that occur in diffusion flames and for a secondary measurement of fuel mole fraction. The temperature-dependent filtered Rayleigh cross sections are computed with a six-moment kinetic model for calculation of major-species Rayleigh-Brillouin line shapes and a flamelet-based model for physically judicious estimates of gas-phase chemical composition. Shot-averaged temperatures, fuel mole fractions, and fuel number densities from steady and vortex-strained diffusion flames stabilized on a Wolfhard-Parker slot burner are presented, and a detailed uncertainty analysis reveals that the FRS-measured temperatures are accurate to within +/- 4.5 to 6% of the local absolute temperature.     

Rate-equation model for quantitative concentration measurements in flames with picosecond pump-probe absorption spectroscopy

Measurement of radical concentrations is important in understanding the chemical kinetics involved in combustion. Application of optical techniques allows for the nonintrusive determination of specific radical concentrations. One of the most challenging problems for investigators is to obtain flame data that are independent of the collisional environment. We seek to obviate this difficulty by the use of picosecond pump-probe absorption spectroscopy. A picosecond pump-probe absorption model is developed by rate-equation analysis. Implications are discussed for a laser-pulse width that is much smaller than the excited-state lifetime of the absorbing atom or molecule. The possibility of quantitative, quenching-independent concentration measurements is discussed, and detection limits for atomic sodium and the hydroxyl radical are estimated. For a three-level absorber-emitter, the model leads to a novel pump-probe strategy, called dual-beam asynchronous optical sampling, that can be used to obtain both the electronic quenching-rate coefficient and the doublet mixing-rate coefficient during a single measurement. We discuss the successful demonstration of the technique in a companion paper [Appl. Opt. 34, XXX (1995)].     

Quantitative imaging of temperature and OH in turbulent diffusion flames by using a single laser source

We describe a diagnostic technique for obtaining quantitative, simultaneous, and instantaneous images of temperature and the concentration of the hydroxyl radical OH in turbulent flames. The technique uses a single laser source and a single intensified CCD camera. A stoichiometric premixed flame is used for calibration. We use detailed calculations of laminar flames of similar fuels to estimate the effects of quenching and ground-state population on the OH signal. A factor combining both effects is generated as a function of temperature. We validate the technique by comparing measured temperature and OH number density with calculated values in laminar diffusion flames. Absolute errors of 10-20% and 20-30% are estimated on the measured temperature and OH number density, respectively. The technique is applicable to regions of the flames where the Rayleigh cross section is close to that of air.     

Numerical simulation of turbulent gas flames in tubes

Computational fluid dynamics (CFD) is an emerging technique to predict possible consequences of gas explosion and it is often considered a powerful and accurate tool to obtain detailed results. However, systematic analyses of the reliability of this approach to real-scale industrial configurations are still needed. Furthermore, few experimental data are available for comparison and validation.In this work, a set of well documented experimental data related to the flame acceleration obtained within obstacle-filled tubes filled with flammable gas-air mixtures, has been simulated. In these experiments, terminal steady flame speeds corresponding to different propagation regimes were observed, thus, allowing a clear and prompt characterisation of the numerical results with respect to numerical parameters, as grid definition, geometrical parameters, as blockage ratio and to mixture parameters, as mixture reactivity.The CFD code AutoReagas was used for the simulations. Numerical predictions were compared with available experimental data and some insights into the code accuracy were determined. Computational results are satisfactory for the relatively slower turbulent deflagration regimes and became fair when choking regime is observed, whereas transition to quasi-detonation or Chapman-Jogouet (CJ) were never predicted.     

Quantitative two-dimensional instantaneous Raman concentration measurements in a laminar methane jet

We have achieved quantitative two-dimensional Raman measurements of the concentration of methane in a laminar methane jet into nitrogen without multipassing the incident laser sheet with a coaxial flash-lamp-pumped dye laser. The measurements are compared with the results of direct numerical simulation for the particular flow field. We conclude that the accuracy of the technique is determined by limitations in the dynamic range and in the spatial resolution of the data acquired with an intensified camera.     

Transported-PDF (IEM, EMST) micromixing models in a hydrogen-air nonpremixed turbulent flame

The topic of this study is the numerical simulation of a turbulent non-premixed hydrogen flame with different micromixing models in order to investigate their predictive capability. The two micromixing models are compared. The first model is the interaction by exchange with the mean (IEM) model (Dopazo and O’Brien Acta Astronaut 1(9–10):1239, 1974). The second one is the Euclidean minimum spanning tree (EMST) model (Subramaniam and Pope Combust Flame 115(4):487, 1998). The dynamic model for the mixing time-scale is handled, by computing the individual time-scales for the reactive scalars in each cell during the simulation course using the Fluent/MM-INTAS CFD code. The predictions are validated against experimental data provided by Raman and Laser Doppler anemometry measurements for a turbulent jet Hydrogen-air diffusion flame. The interaction of turbulence-chemistry is handled with TPDF method. The chemical model used here consists of 11 chemical species and 23 reactions. Comparisons with experimental data demonstrate that predictions based on the EMST model are slightly better. The EMST improves largely the precision of the results to the detriment of the RAM and the CPU performances. Overall, profile predictions of mixture fraction, flame temperature and major species are in reasonable agreement with experimental data.     

Systematic errors in optical-flow velocimetry for turbulent flows and flames

Optical-flow (OF) velocimetry is based on extracting velocity information from two-dimensional scalar images and represents an unseeded alternative to particle-image velocimetry in turbulent flows. The performance of the technique is examined by direct comparison with simultaneous particle-image velocimetry in both an isothermal turbulent flow and a turbulent flame by use of acetone-OH laser-induced fluorescence. Two representative region-based correlation OF algorithms are applied to assess the general accuracy of the technique. Systematic discrepancies between particle-imaging velocimetry and OF velocimetry are identified with increasing distance from the center line, indicating potential limitations of the current OF techniques. Directional errors are present at all radial positions, with differences in excess of 10 degrees being typical. An experimental measurement setup is described that allows the simultaneous measurement of Mie scattering from seed particles and laser-induced fluorescence on the same CCD camera at two distinct times for validation studies.     

Simultaneous vibrational and pure rotational coherent anti-stokes Raman spectroscopy for temperature and multispecies concentration measurements demonstrated in sooting flames

The potential of measuring temperature and multiple species concentrations (N2, O2, CO) by use of combined vibrational coherent anti-Stokes Raman spectroscopy (CARS) and pure rotational CARS has been investigated. This was achieved with only one Nd:YAG laser and one dye laser together with a single spectrograph and CCD camera. From measurements in premixed sooting C2H4-air flames it was possible to evaluate temperatures from both vibrational CARS and rotational CARS spectra, O2 concentration from the rotational CARS spectra, and CO concentration from the vibrational CARS spectra. Quantitative results from premixed sooting C2H4-air flames are presented, and the uncertainties in the results as well as the possibility of extending the combined CARS technique for probing of additional species are discussed.     

Experimental data base for numerical simulations of turbulent diffusion flames

The objective of the presented measurements is to provide an experimental data base for comparison with numerical simulation results of turbulent H₂-air diffusion flames. Additionally, the date base may also be used for a proof of new measurement techniques, when the same flame conditions are applied. The data base contains time and spatial resolved data on all three velocity components, all Reynolds-stress tensor components, temperature, mixture fraction, species concentrations, higher statistical moments of these quantities and probability density functions for three different flames. The data are given as original measurement data in dependence on flame conditions and location in the flame, as absolute and normalized data and as evaluated data, like anisotropy tensor. The measurements are made to improve the understanding of turbulent transport processes under the influence of combustion and to help the effort to couple the turbulence and combustion model. A Laser-Doppler-Velocimeter was used to obtain three velocity components simultaneously. Temperature was measured with spontaneous Raman-Rayleigh spectroscopy and Coherent Anti-Stokes Raman spectroscopy, separately, while species concentrations and mixture fraction are measured with spontaneous Raman-Rayleigh spectroscopy. Measurements are done from nozzle exit into the self-preserving region up to x/d=100 so that the whole flow field including all boundary conditions are quantified for numerical prediction. A mixture of hydrogen and nitrogen with a mole ratio of 1:1 is used as fuel. Reynolds number and Froude number are set at different values. This complete data set is available upon request.     

Measurements of hydroxyl concentrations and lifetimes in laminar flames using picosecond time-resolved laser-induced fluorescence

Picosecond time-resolved laser-induced fluorescence (PITLIF) can potentially be used to obtain measurements of minor species concentrations in rapidly fluctuating flames. Previous studies demonstrated this potential for atomic sodium by monitoring the temporal fluorescence signal with both an equivalent-time and a real-time sampling method. In this developmental study, PITLIF is used to determine hydroxyl concentrations in laminar CH(4)-O(2)-N(2) flames by the measurement of both the integrated fluorescence signal and the fluorescence lifetime. The quenching environment can be monitored with real-time sampling, and thus the necessary quenching rate coefficient is obtained in 348 us, which is fast enough for use in many turbulent flows. Fluorescence lifetimes of OH are also measured at different equivalence ratios in laminar flames by the use of the equivalent-time sampling technique. These results compare favorably with predicted lifetimes based on relevant quenching cross sections and calculated species concentrations.     

Quantitative concentration measurements of atomic sodium in an atmospheric hydrocarbon flame with asynchronous optical sampling

We report the development of a pump-probe instrument that uses a high-repetition-rate (82-MHz) picosecond laser. To maximize laser power and to minimize jitter between the pump- and the probe-pulse trains, we choose the asynchronous optical sampling (ASOPS) configuration. Verification of the method is obtained through concentration measurements of atomic sodium in an atmospheric methane-air flame. For the first time to our knowledge, ASOPS measurements are made on a quantitative basis. This is accomplished by calibration of the sodium concentration with atomic absorption spectroscopy. ASOPS measurements are taken at a rate of 155.7 kHz with only 128 averages, resulting in a corresponding detection limit of 5 × 10(9) cm(-3). The quenching-rate coefficient is obtained in a single measurement with a variation of ASOPS, which we call dual-beam ASOPS. The value of this coefficient is in excellent agreement with literature values for the present flame conditions. Based on our quantitative results for detection of atomic sodium, a detection limit of 2 × 10(17) cm(-3) is predicted for the Q(1) (9) line of A (2)Σ(+) (v = 0)-X(2)II (v = 0) hydroxyl at 2000 K. Although this value is too large for practical flame studies, a number of improvements that should lower the ASOPS detection limit are suggested.     

Beam shaping for CARS measurements in turbulent environments

This paper describes a new technique to mitigate the effect of beam steering on CARS measurements in turbulent, variable density environments. The new approach combines planar BOXCARS phase-matching with elliptical shaping of one of the beams to generate a signal robust to beam steering, while keeping the same spatial resolution. Numerical and experimental results are provided to demonstrate the effectiveness of this approach. One experiment investigates the effect of beam shaping in the presence of a controlled and well quantified displacement of the beams at the focal plane. Another experiment, more qualitative, proves the effectiveness of the technique in the presence of severe beam steering due to turbulence.