Opposed-flow Flame Code Research Articles

Dual-pump CARS temperature and major species concentration measurements in counter-flow methane flames using narrowband pump and broadband Stokes lasers

Dual-pump coherent anti-Stokes Raman scattering (CARS) is used to measure temperature and species profiles in representative non-premixed and partially-premixed CH 4/O 2/N 2 flames. A new laser system has been developed to generate a tunable single-frequency beam for the second pump beam in the dual-pump N 2–CO 2 CARS process. The second harmonic output (∼532 nm) from an injection-seeded Nd:YAG laser is used as one of the narrowband pump beams. The second single-longitudinal-mode pump beam centered near 561 nm is generated using an injection-seeded optical parametric oscillator, consisting of two non-linear β-BBO crystals, pumped using the third harmonic output (∼355 nm) of the same Nd:YAG laser. A broadband dye laser (BBDL), pumped using the second harmonic output of an unseeded Nd:YAG laser, is employed to produce the Stokes beam centered near 607 nm with full-width-at-half-maximum of ∼250 cm −1. The three beams are focused between two opposing nozzles of a counter-flow burner facility to measure temperature and major species concentrations in a variety of CH 4/O 2/N 2 non-premixed and partially-premixed flames stabilized at a global strain rate of 20 s −1 at atmospheric-pressure. For the non-premixed flames, excellent agreement is observed between the measured profiles of temperature and CO 2/N 2 concentration ratios with those calculated using an opposed-flow flame code with detailed chemistry and molecular transport submodels. For partially-premixed flames, with the rich side premixing level beyond the stable premixed flame limit, the calculations overestimate the distance between the premixed and the non-premixed flamefronts. Consequently, the calculated temperatures near the rich, premixed flame are higher than those measured. Accurate prediction of the distance between the premixed and the non-premixed flames provides an interesting challenge for future computations.

Development of an optical parametric generator with pulsed dye amplification for high-resolution laser spectroscopy

An injection-seeded optical parametric generator (OPG), coupled with three pulsed dye amplification (PDA) stages, was shown to produce tunable, narrow linewidth laser radiation. The OPG was composed of a pair of beta barium borate (β-BBO) crystals and pumped by the third harmonic (355 nm) output of a seeded Nd:YAG laser. The OPG was injection-seeded at the idler wavelength (824 nm) using an external cavity diode laser (ECDL) with a mode-hop-free tuning range of 20 GHz. Using the PDA stages, the OPG output signal (624 nm) was amplified to 19 mJ/pulse, while maintaining a spectral linewidth of approximately 160 MHz at full-width-half-maximum (FWHM) which was within a factor of 2 of the Fourier limit. A system of lenses and apertures was used to minimize amplified spontaneous emission (ASE) in the PDA stages. Using the OPG/PDA system, two-photon laser-induced fluorescence measurements of atomic oxygen were performed by sum-frequency-mixing the 624-nm beam with the third harmonic output of the seeded Nd:YAG laser to generate approximately 1 mJ/pulse of ultraviolet radiation near 226 nm. Voigt line shapes were found to be in good agreement with oxygen atom spectra in atmospheric-pressure, laminar, counter-flow flames; the magnitude of Doppler and collisional broadening was approximately the same. The measured O-atom concentration profile was found to compare well with that calculated using an opposed-flow flame code.

OH and CH luminescence in opposed flow methane oxy-flames

Emission spectroscopy is a 2-D nonintrusive diagnostic technique that offers spatially resolved data for combustion optimization and control. The UV and visible chemiluminescence of the excited radicals CH( A 2 Δ , B 2 Σ − ) and OH( A 2 Σ + ) is studied experimentally and numerically in opposed-flow diffusion flames of methane and oxygen-enriched air. The oxidized oxygen content is varied from 21 to 100% while the range of the studied strain rates spans from 20 to 40 s −1. The spectrally resolved imaging is obtained by two different methods: scattering through a grating monochromator and interposition of interference filters along the optical path. Absolute measured chemiluminescence intensities, coupled with a numerical model based on the opposed flow flame code, are used to evaluate the chemical kinetics of the excited species. The predictions of the selected model are in good agreement with the experimental data over the range of the studied flame conditions.

LIF MEASUREMENTS AND CHEMICAL KINETIC ANALYSIS OF NITRIC OXIDE FORMATION IN HIGH-PRESSURE COUNTERFLOW PARTIALLY PREMIXED AND NONPREMIXED FLAMES

We report quantitative, spatially resolved, linear laser-induced fluorescence (LIF) measurements of nitric oxide concentration ([NO]) in laminar, methane/air counterflow partially premixed and nonpremixed flames at six pressures up to 15 atm using excitation near 226.03 nm in the γ(0,0) band of NO. For partially premixed flames, fuel-side equivalence ratios (φB) of 1.45, 1.6, and 2.0 are studied at a global strain rate of 20 s−1. For nonpremixed flames, a complete set of NO measurements at global strain rates of 20 s−1, 30 s−1, and 40 s−1 is presented to supplement previously reported data in such flames. The quantitative NO measurements are compared with predictions from an opposed-flow flame code utilizing two GRI chemical kinetic mechanisms (versions 2.11 and 3.0). The effect of radiative heat loss on NO predictions is assessed by using a modified version of the code that considers radiation in the optically thin limit. The linear LIF measurements of [NO] are corrected for variations in the electronic quenching rate coefficient by using major species and temperature profiles generated by the opposed-flow flame code plus quenching cross sections for NO available from the literature. A pathway analysis provides the relative contribution of different NO formation mechanisms to the total amount of NO produced at various pressures. Quantitative reaction path diagrams are used to investigate pictorially species interactions during NO formation. Finally, we identify key reactions controlling NO concentrations in counterflow partially premixed and nonpremixed flames by using a sensitivity analysis. For the nonpremixed flames, NO measurements at pressures of 2–5 atm disagree substantially with predictions from either GRI mechanism. On the other hand, both mechanisms predict observed NO concentrations reasonably well beyond 6 atm. In general, for our nonpremixed flames, NO formation is dominated by the prompt route at lower pressures with an increasing contribution from the N2O pathway at higher pressures. For partially premixed flames, GRI 3.0 mimics the basic quantitative trends found in our LIF measurements at 1–15 atm. The kinetic analysis indicates that the prompt mechanism dominates at lower pressures, whereas the thermal and N2O pathways become more important at higher pressures. At any given pressure, a reduction in partial premixing results in kinetic behavior approaching that of the nonpremixed flames.

LIF measurements and chemical kinetic analysis of methylidyne formation in high-pressure counter-flow partially premixed and non-premixed flames

We report quantitative, spatially resolved, linear laser-induced fluorescence (LIF) measurements of methylidyne concentration ([CH]) in laminar, methane–air, counter-flow partially premixed and non-premixed flames using excitation near 431.5 nm in the A–X (0,0) band. For partially premixed flames, fuel-side equivalence ratios (ϕB) of 1.45, 1.6 and 2.0 are studied at pressures of 1, 3, 6, 9 and 12 atm. For non-premixed flames, the fuel-side mixture consists of 25% CH4 and 75% N2; measurements are obtained at pressures of 1, 2, 3, 4, 5, 6, 9 and 12 atm. The quantitative CH measurements are compared with predictions from an opposed-flow flame code utilizing two GRI chemical kinetic mechanisms (versions 2.11 and 3.0). LIF measurements of [CH] are corrected for variations in the quenching rate coefficient by using major species concentrations and temperatures generated by the code along with suitable quenching cross sections for CH available from the literature. A pathway analysis provides relative contributions from important elementary reactions to the total amount of CH produced at various pressures. Key reactions controlling peak CH concentrations are also identified by using a sensitivity analysis. For the partially premixed flames, measured CH profiles are reproduced reasonably well by GRI 3.0, although some quantitative disagreement exists at all pressures. Two CH radical peaks are observed for ϕB=1.45 and ϕB=1.6 at pressures above 3 atm. Peak CH concentrations for the non-premixed flames are significantly underpredicted by GRI 3.0. The latter agrees with previously reported NO concentrations, which are also underpredicted in these same high-pressure counter-flow diffusion flames.

Measurements of absolute CH concentrations by cavity ring-down spectroscopy and linear laser-induced fluorescence in laminar, counterflow partially premixed and nonpremixed flames at atmospheric pressure.

We report quantitative, spatially resolved measurements of methylidyne concentration ([CH]) in laminar, counterflow partially premixed and nonpremixed flames at atmospheric pressure by using both cavity ring-down spectroscopy (CRDS) and linear laser-induced fluorescence (LIF) in the A-X (0, 0) band. Three partially premixed (phiB = 1.45, 1.6, 2.0) flames plus a single nonpremixed methane-air flame are investigated at a global strain rate of 20 s(-1). These quantitative measurements are compared with predictions from an opposed-flow flame code when utilizing two GRI chemical kinetic mechanisms (versions 2.11 and 3.0). The LIF measurements of [CH] are corrected for variations in the electronic quenching rate coefficient by using predicted major species concentrations and temperatures along with quenching cross sections for CH that are available in the literature. The peak CH concentration obtained by CRDS is used to calibrate the quenching-corrected LIF measurements. Excellent agreement is obtained between CH concentration profiles measured by using the CRDS and LIF techniques. The spatial location of the CH layer is very well predicted by GRI 3.0; moreover, the measured and predicted CH concentrations are in good agreement for all the flames of this study.

QUANTITATIVE LASER-INDUCED FLUORESCENCE MEASUREMENTS AND MODELING OF NITRIC OXIDE IN HIGH-PRESSURE (6–15 ATM) COUNTERFLOW DIFFUSION FLAMES

Laser-induced fluorescence (LIF) measurements of NO concentration ([NO]) have been obtained along the centerline of methane–air counterflow diffusion flames at 6 to 15 atm. This study is an extension of our previous work involving measurements of [NO] in similar flames at two to five atm, wherein we had used a counterflow premixed flame for calibration. For the flames studied here, a method based on computed overlap fractions is developed to calibrate [NO] measurements at higher pressures. The linear LIF measurements of [NO], which are corrected for variations in the electronic quenching rate coefficient, are compared with numerical predictions from an opposed-flow flame code utilizing two Gas Research Institute (GRI) chemical kinetic mechanisms (versions 2.11 and 3.0). The effect of radiative heat loss on code predictions is accounted for by using an optically thin radiation model. The revised GRI mechanism (version 3.0) offers a significant improvement in prompt-NO predictions for these flames compared to the older version (2.11), especially at pressures below eight atm. However, a consistent discrepancy remains in the comparisons, particularly at peak NO locations for pressures lower than six atm. The measurements display a continuing trend of decreasing NO concentration with increasing pressure at 6–15 atm as expected for flames dominated by prompt NO. The discrepancy between measurements and predictions decreases with rising pressure so that the revised GRI mechanism predicts [NO] with reasonable accuracy at pressures above six atm.

Laser-induced fluorescence measurements and modeling of nitric oxide in counterflow partially premixed flames

Quantitative measurements of NO concentration ([NO]) have been obtained along the centerline of atmospheric pressure, methane–air, counterflow partially premixed flames using the laser-induced fluorescence (LIF) technique. The effect of partial premixing was studied by investigating flames with fuel-side equivalence ratios (φ B) of 1.45, 1.6, 1.8, and 2.0 at a constant global strain rate near 20 s −1. Linear LIF measurements of [NO] are corrected for variations in the electronic quenching rate coefficient by using major species profiles generated by an opposed-flow flame code and quenching cross-sections for NO available from the literature. Corrected linear LIF measurements of [NO] and temperatures measured using thin filament pyrometry are compared with numerical predictions from the opposed-flow flame code by utilizing the GRI (version 2.11) mechanism for the NO kinetics. The effect of radiative heat loss on code predictions is accounted for by using an optically thin radiation model. Reasonably good agreement exists between LIF [NO] measurements and predictions in all flames. In particular, all predictions fall within 10% of measurements at peak [NO] locations. Spatial separation was observed between regions where prompt-NO and thermal-NO dominate in the φ B = 1.45 flame. A previously modified rate coefficient for the prompt-NO initiation reaction improved agreement between predictions and measurements in the region dominated by prompt NO.

Laser-Induced Fluorescence Measurements and Modeling of Nitric Oxide in High-Pressure Counterflow Diffusion Flames

Quantitative laser-induced fluorescence (LIF) measurements of NO concentration ([NO]) have been obtained along the centerline of prompt-NO dominated, methane-air counterflow diffusion flames at two to five atm, Global strain rates of 20, 30 and 40 s−1 were investigated at each pressure, with the addition of a 15 s−1 case at three and four atm. Linear LIF measurements of [NO] are corrected for variations in the electronic quenching rate coefficient by using major species profiles generated by an opposed-flow flame code and quenching cross-sections for NO available from the literature. Corrected linear LIF measurements of [NO] are compared with numerical predictions from the opposed-flow flame code by utilizing the GRI (version 2.11) mechanism for the NO kinetics. The effect of radiative heat loss on code predictions is accounted for by using an optically thin radiation model. A modest decrease in predicted temperature owing to radiative heat loss causes a significant decrease in predicted [NO], indicating the temperature sensitivity of the prompt-NO kinetics. Comparisons between [NO] measurements and predictions show that the GRI mechanism underpredicts prompt-NO by a factor of two to three at all pressures. The underprediction peaks at 2 to 3 atm, and decreases with pressure from 3 to 5 atm. Although the GRI mechanism does not display this trend, predictions with a modified rate coefficient for the prompt-NO initiation reaction give qualitative agreement with the experimentally observed variation. However, modifying the prompt-NO initiation reaction is not sufficient to account for the differences between measurements and predictions, thus indicating a need for refinement of the CH chemistry.

00/01512 Improvement of low load combustion stability for advanced low NO x burner

Local production rates of NO have been derived from experiments in vertical flames on a burner commonly used in central heating equipment. The fuel-air equivalence ratio of around 1.9 caused a double-flame structure to develop above the burner, a phenomenon reflected by the temperature and OH concentration fields. The production rate of NO was a maximum in regions of low temperature and low OH concentration, located in the inner premixed flame and prior to the higher temperatures and OH concentrations of the outer diffusion flame. In addition, the peak rate of disappearance of NO was at the core of the premixed flame. These experimental results supported the assumption that the main mechanism of NO formation was the Fenimore Prompt route; also, an important internal “reburn” mechanism was operating. The opposed flow flame code was used here in the largest part of the simulations with the GRI 2.11 mechanism and with two additional mechanisms for a particular condition of flow and fuel-air equivalence ratio. The code was applied for fuel streams containing methane in air and oxidizer streams containing air only; the aim was to draw analogies with the experimental two-dimensional flame. Results from local rates of production and sensitivity analyses and quantitative reaction path diagrams (QRPD) for the nitrogenous species are presented for conditions representative of domestic appliances (cookers, space and water heaters, and central heating units) fired with natural gas. The co-flow experimental flame and the counter-flow theoretical flame at similar strain rates and fuel richness showed significant similarities.

Laser-induced fluorescence measurements and modeling of nitric oxide in methane–air and ethane–air counterflow diffusion flames

Quantitative laser-induced fluorescence (LIF) measurements of nitric oxide concentrations [NO] have been obtained along the centerline in atmospheric pressure methane–air and ethane–air counterflow diffusion flames. These flames are highly diluted to avoid both soot formation and the influence of radiative heat losses on NO formation, thereby ensuring NO production mostly via the prompt mechanism. Linear LIF measurements of [NO] are corrected for variations in the electronic quenching rate coefficient by using major species profiles generated by an opposed-flow flame code and quenching cross-sections for NO available from the literature. Temperature measurements are also made in the methane–air counterflow diffusion flames by using thin SiC filament pyrometry. The excellent agreement between temperature measurements and model predictions verifies the efficacy of a new calibration method developed for thin filament pyrometry. Predictions using the GRI mechanism consistently underpredict peak [NO] in all flames. This result indicates a need for refinement of both the prompt-NO and CH kinetics, especially the rate coefficient for the prompt-NO initiation reaction. A modified rate coefficient proposed for the prompt-NO initiation reaction significantly improves agreement between modeling and measurements in both the methane–air and ethane–air counterflow diffusion flames. The remaining discrepancy in some flames can be attributed to a lack of refinement in the CH chemistry. Overall, the modified rate coefficient proposed here seems to be a good choice over a wide range of strain rates for both methane and ethane fuels.

NO x Mechanisms in Rich Methane-Air Flames

Local production rates of NO have been derived from experiments in vertical flames on a burner commonly used in central heating equipment. The fuel-air equivalence ratio of around 1.9 caused a double-flame structure to develop above the burner, a phenomenon reflected by the temperature and OH concentration fields. The production rate of NO was a maximum in regions of low temperature and low OH concentration, located in the inner premixed flame and prior to the higher temperatures and OH concentrations of the outer diffusion flame. In addition, the peak rate of disappearance of NO was at the core of the premixed flame. These experimental results supported the assumption that the main mechanism of NO formation was the Fenimore Prompt route; also, an important internal “reburn” mechanism was operating. The opposed flow flame code was used here in the largest part of the simulations with the GRI 2.11 mechanism and with two additional mechanisms for a particular condition of flow and fuel-air equivalence ratio. The code was applied for fuel streams containing methane in air and oxidizer streams containing air only; the aim was to draw analogies with the experimental two-dimensional flame. Results from local rates of production and sensitivity analyses and quantitative reaction path diagrams (QRPD) for the nitrogenous species are presented for conditions representative of domestic appliances (cookers, space and water heaters, and central heating units) fired with natural gas. The co-flow experimental flame and the counter-flow theoretical flame at similar strain rates and fuel richness showed significant similarities.