An experimental study of well-defined turbulent nonpremixed spray flames

doi:10.1016/S0010-2180(99)00115-7

Combustion and Flame

Volume 121, Issues 1–2, April 2000, Pages 1-23

https://doi.org/10.1016/S0010-2180(99)00115-7 Get rights and content

Abstract

An experiment system was designed for the study of well-defined turbulent nonpremixed spray flames. Particular emphasis was placed on minimizing the influence of the injector design and on maximizing turbulence within the spray flames. A comprehensive description of the structure of such flames was obtained by applying a variety of complementary diagnostic techniques, including: broadband chemiluminescence imaging, CH∗ emission imaging, phase Doppler interferometric techniques, and spontaneous Raman spectroscopy. Two methanol spray flames were examined in detail, with Reynolds number ranging from 2.1 × 10⁴ to 2.8 × 10⁴. Flame appearance and detailed measurements confirmed the occurrence of group combustion. Near the burner mouth, a dense column of drops enveloped by a common flame was observed. Further up, large corrugated structures were visualized which eventually developed into separate “islands.” A significant fraction of the spray escaped unburned, which implies that droplet evaporation is slow in this configuration. Detailed scanning of the flames provided an extensive database of average and fluctuating components of gas velocity and temperature, as well as spray and droplet size–classified properties. Key conclusions from such measurements include: the evidence of two-way coupling between the two phases along the centerline near the burner mouth; a velocity acceleration in the densest areas of the spray flames, as a result of momentum addition through vaporization, followed by deceleration farther downstream as the jet spreading predominates; and the droplet inertial behavior, especially for the large size classes, as confirmed by estimates of some relevant Stokes numbers. The average flame height was found to correlate with an overall equivalence ratio and with the initial concentration of droplets at the burner mouth.

Introduction

Spray combustion is an important aspect of combustion technology. Its most common manifestation is as nonpremixed flames, which have been investigated experimentally with advanced laser diagnostic techniques either in laminar environments 1, 2 or in more realistic and complex turbulent ones [e.g., 3, 4, 5, 6, 7, 8, 9]. If the latter are to mimic practical spray flames, one major difficulty associated with the rich variety of practical systems arises in their design. Different spray injectors, often tailored to specific combustor geometries, produce different patterns. The resulting atomization “signature” on the droplet initial conditions usually controls the spray flame development and morphology, especially in the case of inertial droplets. This variability is necessary for practical purposes, but may hinder a fundamental understanding of spray flames, because of the complex relationship between injector design, droplet initial conditions, and the atomization process. As a result, there is no “universal” spray flame. Yet certain principles or features are common to most practical realizations. For example, aspects of group combustion models of spray burning 10, 11 may apply, to a different extent, to portions of all spray flames. Another broad classification can be based on spray flame diluteness. Most practical ones can be often dense, while laboratory configurations tend to be dilute. The behavior of the two can be quite different, since in the former case a “two-way” coupling between the two phases is expected and the presence of the droplets influences the behavior of the gas.

Advanced laser diagnostic techniques that are crucial to progress in such challenging environments have been introduced at a slower pace by comparison with gaseous turbulent flames, principally because of difficulties associated with the presence of the dispersed phase. In particular, techniques for the physical characterization of the system (e.g, droplet size and velocity statistics) are well-established. They overwhelmingly rely on phase Doppler anemometry. Detailed measurements of this type have been performed for a number of years in combustion environments, starting with [12]. On the other hand, spectroscopic studies for the identification of chemical species have been slow to emerge, often limited to feasibility studies of laser-induced fluorescence 13, 14, 15 and spontaneous Raman spectroscopy 16, 17. Perhaps the most comprehensive investigation among the latter is a characterization of the near field structure of a practical spray flame by a combination of coherent anti-Stokes Raman spectroscopy (CARS) and phase Doppler techniques [18].

In view of the difficulties of identifying a paradigm spray flame, our approach in this investigation had the goal of studying selectively aspects of spray combustion that are of general, if not universal, nature in some regions of practical sprays. The resulting design was guided by the need to minimize the importance of initial conditions and, consequently, the influence of the specific injector. Droplet concentrations were chosen to establish a group combustion of drops in a convective-diffusive boundary layer flow. Diluteness effects were also accounted for, with large drop concentrations prevailing close to the injector and leading to strong coupling between the two phases, and with one-way coupling dominating in the far field. In addition to phase Doppler techniques, Raman scattering was extensively used to map the temperature field. As a result of this comprehensive effort, a picture of a simplified turbulent spray flame emerges and an extensive database is established for subsequent computational modeling. The present article focuses on the phenomenology of such a flame, the description of its structure, and the presentation of an extensive database. Self-similarity and scaling were addressed quantitatively in a companion article [19].

Section snippets

Burner design criteria

The design of a “canonical” turbulent spray flame was a first requirement of this investigation. It was guided by two principles: maximizing the turbulence level, to offset the inevitable relaminarization of the high combustion temperatures; and minimizing the effect of the drop initial conditions. These goals were realized with the burner sketched in Fig. 1. The liquid fuel was injected using a commercial ultrasonic nebulizer (Sonotek) that can disperse modest liquid flow rates, while

Chemiluminescence imaging

A few images (150-μs exposure time) of the visible chemiluminescence from a typical spray flame are shown in Fig. 5. The ruler on the right is in burner diameter units. The pictures show the rich morphology of these flames and confirm the realization of the burner design objectives. Evident in the sequence, particularly close to the burner exit, is an inner dark region, which can be best seen in a medium-exposure (1 ms) image shown in the inset. Flame corrugations are less pronounced close to

Conclusions

A comprehensive experimental study was conducted in simplified turbulent spray flames using phase Doppler interferometric techniques, spontaneous Raman spectroscopy, and chemiluminescence imaging. The role of the initial conditions was deliberately minimized by using an ultrasonic nebulizer for the dispersion of the liquid phase, thereby ensuring that the velocity slip between the two phases was small. Two methanol flames were studied in detail, corresponding to Reynolds numbers of 2.1 × 10⁴

Acknowledgements

The support of NASA, under the Microgravity Science and Applications Program, Grants NAG3-1259 and NAG3-1688, and of NSF, through a NSF Young Investigator Award and Equipment Grant CTS-9112601, is gratefully acknowledged.

References (36)

G. Chen et al.
Combust. Flame
(1997)
D. Stepowski et al.
Combust. Flame
(1994)
R. Aftel et al.
A.N. Karpetis et al.
H.H. Chiu et al.
F. Takahashi et al.
J.B. Greenberg
Combust. Flame
(1989)
H.A. Becker et al.
Combust. Flame
(1978)
G. Chen et al.
Combust. Sci. Technol.
(1996)
V.G. McDonnel et al.
Atomiz. Sprays
(1993)

C.F. Edwards et al.

Y. Hardalupas et al.

Combust. Sci. Technol.

(1994)

Karpetis, A. N., and Gomez, A., (1996) 9th Annual Conference on Liquid Atomization and Spray Systems (ILASS-Americas...

Labowsky, M., and Rosner, D. E., in Evaporation-Combustion of Fuels, Adv. Chem. Ser., No. 166, (J. T. Zung, Ed.),...

McDonell, V. G., and Samuelsen, G. S., Twenty-Second Symposium (International) on Combustion, Pittsburgh, 1988, p....

M.G. Allen et al.

Upshulte, B. L., Allen, M. G., and Mcmanus, K. R., Twenty-Sixth Symposium (International) on Combustion, Pittsburgh,...

A. Cessou et al.

Combust. Sci. Technol.

(1996)

Cited by (55)

Are the available data from laboratory spray burners suitable for CFD modelling validations? A review
2022, Energy Conversion and Management: X
A systematic compilation and evaluation of turbulent spray flame combustion data acquired from lab-scale spray-burners and CFD models validated with these are presented. Previous review studies have been limited to presentations of the progress with validation targets and modelling methods. This work expands these by presenting the details of lab-scale spray-burners and the characteristics of their datasets, the computational methodologies developed to capture the turbulent spray combustion processes acquired from the burners and their validation with the dataset. Additionally, by critically evaluating the spray and inflow boundary conditions, the reacting and non-reacting features and the main flame characteristics and data, direct correlations between the effectiveness of the dataset from the lab-scale spray burners and the outcomes from the computational methodologies using them for validation are clearly elucidated. The evaluation indicates that the turbulence modelling and validation of the swirling non-reacting flows before the application of the combustion models are important. A better representation of the multi-mode nature and the flame-turbulence-chemistry interactions in the combustion models are required. Also, a clearer experimental presentation of the flame structure and modes would help with the development of the modelling techniques. To plug these research needs, this paper reinforces the importance of developing improved numerical techniques that can be validated by selecting an appropriate lab-scale spray-burner with a robust database. Together these can be used to improve the development of and reduce emissions from combustors burning future green fuels.
Spontaneous Raman-LIF-CO-OH measurements of species concentration in turbulent spray flames
2021, Proceedings of the Combustion Institute
Citation Excerpt :
The difficulty lies in the weak Raman signals which get obliterated by intense interferences arising from several sources, including breakdown of the suspended particles or droplets. Earlier attempts to apply conventional Raman scattering to turbulent spray flames are limited and include measurements of both Stokes and anti-Stokes Raman scattering from nitrogen to obtain the gas temperature [20,21]. A similar approach, but with the addition of Raman imaging of fuel, has been employed in the non-reacting region of sprays, injected into an engine to yield equivalence ratio and temperature [22].
This paper presents new measurements of species concentrations, temperature and mixture fraction in selected regions of a turbulent ethanol spray flame. The line-Raman–LIF–COOH setup developed at the Sandia's Combustion Research Facility is utilised to probe regions of a spray flame where laser breakdown of liquid droplets is avoided and the remaining interferences can be corrected. The spray flame is stabilised on the piloted Sydney needle spray burner, where axial translation of the liquid injecting needle in the air-blast stream can transition the spray from dilute to dense. The solution to obtaining successful measurements is found to be multifaceted and includes: the appropriate selection of flame conditions; high sensitivity of the Raman detection system permitting reduced laser energies; development of a pre-processing algorithm to reject strong droplet interferences; and application of the hybrid matrix inversion method combined with wavelet denoising to account for interference corrections and noise at the very low signal levels obtained. Unique and necessary for the successful measurements reported in this paper, a pre-processing algorithm is outlined that removes data points corrupted with strong interferences from droplets. These interferences arise from a range of sources, but the most intense are due to the laser interaction with surrounding mist or liquid fragments, such that measurements near the jet centreline are corrupted and hence discarded. Reliable measurements of mixture fraction, temperature obtained from the sum of the species number densities, and species mole fractions are reported for regions in the flames sufficiently far from the centreline. The paper demonstrates the feasibility of the judicious use of Raman scattering in turbulent spray flames, the results of which will be extremely useful for validating numerical simulations.
Investigation of fuel dilution in ethanol spray MILD combustion
2019, Applied Thermal Engineering
MILD combustion is a promising technique for low heating value fuels. In this paper, spray MILD combustion of diluted liquid fuel is numerically studied. The modelling framework is based on the RANS approach, the EDC combustion model with a skeletal chemical mechanism, and the Lagrangian tracking of droplets. The validated numerical model is used to simulate the combustion of ethanol diluted with water is studied for different oxidizer temperatures and O₂ concentrations under MILD condition. Results indicate that a low-level fuel dilution up to 5 percent has tendency toward expansion of MILD region, but higher degrees fuel dilution within the range of 5–20 percent results in a reduction of MILD combustion regions due to domination of physical effect contribution of addition of water to ethanol fuel compared to its chemical effect. Fuel water addition also leads to an increase in OH concentration due to H₂O dissociation through O + H₂O → 2OH, which assist the oxidation of fuel driven species, CO in particular. Changing the oxidizer stream temperature can lead to considerable changes in the flame structure: the typical flame shape is converted to a bifurcation style by reducing hot co-flow stream temperature from 1400 K to 1200 K. Moreover, for liquid fuel dilution, preheating and inlet air dilution do not lead to monotonic strengthening the MILD zone in spite of gaseous fuels.
CO Imaging in piloted liquid-spray flames using femtosecond two-photon LIF
2019, Proceedings of the Combustion Institute
Citation Excerpt :
However, dilution effects are present in the far field from the spray exit of group combustion. Karpetis and Gomez [5] measured droplet number densities in turbulent non-premixed flames and stated that large droplet densities were achieved at the centerline close to the exit while the densities were reduced significantly along the axial position due to the dispersion and evaporation of the jet spray. Previous studies of spray combustion have been mainly focused on the characterization of various sprays using laser-based techniques [6–8].
Liquid-spray flames are encountered in many practical combustion devices such as gasoline direct injection and diesel engines, gas turbine combustors as well as industrial furnaces. As opposed to gaseous fuels, additional phase-change steps present in liquid sprays not only complicate the overall combustion process, but also make in-situ, laser-based combustion diagnostics challenging. In particular, the formation of carbon monoxide (CO) due to incomplete fuel-air mixing and partial oxidation becomes a major challenge. In this study, we apply femtosecond, two-photon laser-induced fluorescence (fs-TPLIF) to measure CO concentration in piloted liquid-spray flames, taking into account possible signal interferences in the 230.1-nm, B¹Σ⁺←X¹Σ⁺ excitation scheme. A modified, flat-flame McKenna burner fitted with a direct-injection high-efficiency nebulizer (DIHEN) was used to produce piloted liquid-methanol spray flames. Although single-laser-shot OH-PLIF images show the presence of strong turbulent interactions in the core region, shot-averaged OH-PLIF images indicate that near the nozzle-exit region, the primary reaction takes place in an annular region around the droplet cloud, in general. A detailed spectroscopic study reveals that the signal interference at 460 nm originating from the second-order scattering of the excitation laser, which becomes approximately an order of magnitude stronger than CO fluorescence spectral lines near the nozzle exit region. The specific spectral filtering scheme introduced in our recent work is proved to be capable of suppressing interferences primarily originating from C₂ Swan-band emissions. Two-dimensional CO maps along with OH-PLIF flame structure data provide key insights into the CO formation in piloted liquid-spray flames, while providing critical validation datasets for advanced computational models.
Experimental investigation of turbulent flames in uniform dispersions of ethanol droplets
2017, Combustion and Flame
Citation Excerpt :
In comparison to the field of turbulent premixed flames where a larger collection of data measuring parameters pertinent to flame structure and propagation (such as burning rates, displacement speeds, surface density, curvature statistics) are widely available [17], such information is not readily available for turbulent spray flames [18]. Additionally, the majority of the detailed work on flame structure and propagation in droplet-laden mixtures has focused on laminar flows, with few studies investigating turbulence effects in detail available in the literature [19–22]. The lack of data on the canonical problem of turbulent planar flame propagating in a uniform dispersion of droplets, which is fundamental to understanding spray flame stabilisation and important applications such as gas turbine relight, has been highlighted in a recent review article [23].
A turbulent flame in an ethanol droplet-laden uniform mixture is investigated at overall equivalence ratios (ϕ_ov) of 0.62, 0.72 and 0.82, using a piloted Bunsen burner. Imaging of OH* chemiluminescence and simultaneous imaging of OH PLIF and Mie scattering, both at 5 kHz, and imaging of CH₂O-fuel PLIF at 5 Hz, were used to obtain instantaneous and time-averaged images, temporal sequences and 2-D estimates of flame surface density and curvature. 1-D PDA and LDA measurements were used to obtain droplet size and velocity statistics. At ϕ_ov = 0.62, the flame takes a cylindrical shape, and changes to a cone shape with increasing fuel loading to obtain higher ϕ_ov. Larger droplets are generally observed to have lower average and RMS axial velocities than smaller droplets. Profiles of droplet size distributions indicate a decreasing droplet number density downstream together with a shift to larger droplet diameters. The flame structure is observed to be relatively smooth at locations near the burner exit, and becomes more contorted with distance downstream. In general, droplets are observed to coincide with low-to-intermediate regions of OH. Occasionally, droplets appear to penetrate the flame front, and are detected in regions of intermediate-to-high OH. This occurs particularly at the downstream locations where the flame closes across the jet, with no significant averaged droplet penetration observed past 2 mm in the direction normal to the flame front. Measurements show a gradual reduction in flame surface density and higher flame front curvature with both distance downstream and increasing fuel loading. Estimates of the average droplet evaporation rate increase with both distance downstream and ϕ_ov, as droplets appear in higher mean progress variable regions. The measurements reported here are useful for model validation of flame propagation in dilute sprays.
Characterization of atomization and combustion in moderately dense turbulent spray flames
2015, Combustion and Flame
This paper presents a study of spray jets and flames where the liquid fuel loading in the carrier air is such that there are substantial liquid–liquid interactions close to the exit plane. The burner consists of an air-blast atomizer located within a wide hot co-flow used to stabilize turbulent auto-igniting spray flames. Laser/phase Doppler anemometry, microscopic backlit imaging coupled with advanced image processing, broadband chemiluminescence and OH-PLIF imaging are utilized. A key focus here is on a novel characterization of the spray boundary condition in terms of non-spherical shapes of fluid fragments. Three different classes of shapes, namely: ligaments, droplets, and large, generally irregular objects are examined. Statistics for each of these fragments are computed in a range of sprays and it is found that their size and probability of occurence depends on the initial Weber number and fuel/air mass ratio (F/A). The change in chemiluminescence emission that is measured as a function of the F/A ratio trends in the same manner as the change in the size of the largest non-spherical objects in the spray. In addition, it is found that changes in the growth of reaction zone width occur as estimated from OH images of laser induced fluorescence, and this is also controlled by the F/A ratio. Therefore, the spray structure at the exit plane may partly dictate the downstream flame characteristics; and this could be largely due to the presence of non-spherical fragments and hence the different rates of atomization and vapourization amongst the different sprays. The burner, as well as the classification of the spray introduced here, while complex, can form a platform for the improvement of models for moderately dense reacting sprays.

View all citing articles on Scopus

View full text

Original ArticlesAn experimental study of well-defined turbulent nonpremixed spray flames

Abstract

Introduction

Section snippets

Burner design criteria

Chemiluminescence imaging

Conclusions

Acknowledgements

Combust. Flame

Combust. Flame

Combust. Flame

Combust. Flame

Combust. Sci. Technol.

Atomiz. Sprays

Combust. Sci. Technol.

Combust. Sci. Technol.

Original Articles
An experimental study of well-defined turbulent nonpremixed spray flames