Jeffrey M. Donbar

Mixing and Combustion Studies Using Cavity-Based Flameholders in a Supersonic Flow

An experimental investigation of the mixing and combustion processes that occur in and around a cavity-based flameholder in a supersonic flow is reported. Cavity-based flameholders are commonly found in hydrocarbon-fueledscramjet combustors; however, detailed information concerning the behavior of these devices, their optimal shape and fueling strategies, combustion stability, and interactions with disturbances in the main airflow (i.e., shock trains or shock-boundary layer interactions) is largely unavailable in the existing literature. This work is part of an ongoing research program aimed at providing information to help fill these voids and improve the overall understanding of cavities for use as scramjet flameholders.

Scalar and velocity field measurements in a lifted CH4–air diffusion flame

Experiments have been performed to investigate the leading edge of a lifted jet diffusion flame. The first portion of this study is a simultaneous particle image velocimetry (PIV) and planar laser-induced fluorescence (PLIF) investigation of a lifted methane flame. The simultaneous technique is an approach for establishing the 2-D velocity field in conjunction with the flame front location indicated by laser-induced fluorescence from CH radicals within the reaction zone. The results show that the lifted flame stabilizes in a region of relatively low incoming gas velocity. Furthermore, the radial movement of large-scale vortices appears to play a crucial role in local flame extinction. The second set of experiments involves a simultaneous CH and OH PLIF investigation of the same lifted flame. The relative positions of the two radical fields have remarkable agreement. The CH profile is indicative of the fuel-rich region of the reaction zone and closely follows the inner edge of the OH profile. Furthermore, the OH zone is more than three times as thick as the CH zone, and the structures in both images support the radial motion of vortices established by the joint PIV/CH-PLIF measurements.

Reaction zone structure in turbulent nonpremixed jet flames—from CH-OH PLIF images

Abstract It is shown that simultaneous images of the CH and OH concentration fields can be obtained throughout a high-Reynolds-number (18,600) turbulent nonpremixed, nonsooting jet flame, and that the CH-OH boundary is a useful marker of the instantaneous stoichiometric contour. Previous CH-OH imaging was confined to the flame base. The structure of the fuel-decomposition zone—identified by the CH images—includes the following regions: those with high-curvature cusps; those with low CH concentration; and those where the flame “pinches” due to oxidizer being entrained to the centerline. It is found that the reaction zone that is associated with fuel decomposition (i.e., the CH layer) remains thin and rarely exceeds 1 mm, even near the tip of the high-Reynolds-number flame. CH layers in the turbulent flame are not thicker than the CH layers in the laminar jet flame at the same x / d location. In fact, CH layer thickness is relatively insensitive to Reynolds number and the level of turbulence. This implies that turbulence does not broaden the CH reaction zone, and that flamelet concepts are justified in modeling the reaction zone associated with fuel decomposition. The CH layers become thicker in the streamwise ( x ) direction, which is expected because scalar gradients and the dissipation rate are expected to decrease in the streamwise direction. Imaging the CH layer makes it possible to measure the flame surface density (Σ), which has a typical value of 0.2 mm −1 . Surface density is shown to be related to the turbulent brush thickness and the degree of wrinkling.

Strain rates measured along the wrinkled flame contour within turbulent non-premixed jet flames

Abstract The thin, wrinkled CH reaction layers within moderate- (Re = 9,100) and high- (18,600) Reynolds-number turbulent non-premixed jet flames were identified by using planar laser-induced fluorescence, and the in-plane strain rates on these reaction layers were measured using simultaneous Particle Imaging Velocimetry (PIV). The PIV diagnostics resolved the Taylor scale; the strain-limited diffusion length scale was fully resolved for half the cases studied and nearly resolved for the others. In the high-Reynolds-number jet, instantaneous strain rates on the flame surface are highly intermittent, with peak values exceeding 10,000 s −1 . Mean strain rates, conditioned on the CH-peak contour, are relatively constant (150 s −1 ) in the Re = 9100 flame and increase (650–1700 s −1 ) with axial location in the Re = 18,600 flame, resulting from the flame wrinkling process. The CH-layer thickness does not appear to respond in amplitude or in phase with the strain field, indicating that quasi-steady conditions do not occur. The strain field apparently oscillates at frequencies as high as 5–10 kHz—which is the inverse of the crossing time of integral-scale eddies—perhaps because only the low-frequency component of strain effectively acts on the flame. Mean axial velocities, conditioned on the CH-peak contour, were found to remain constant from the flame base to tip and to approximately equal the product of the stoichiometric mixture fraction and the fuel-exit velocity, in agreement with prediction.

Simultaneous Rayleigh imaging and CH-PLIF measurements in a lifted jet diffusion flame

Abstract Simultaneous Rayleigh scattering and CH planar laser-induced fluorescence (PLIF) measurements near the stabilization region of a lifted methane–air diffusion flame are presented. The goals of this investigation are to establish flow patterns responsible for complete breaks in the CH profile that indicate local flame extinction and evaluate the stabilization mechanisms over a range of flow conditions. Considerable attention has been given to vortex–flame interactions as a primary extinction mechanism of turbulent diffusion flames. The existence of holes in the flame zone is thought to result from the radial penetration of the flame by vortices from the internal fuel jet. In this investigation, Rayleigh scattering is used as a qualitative indication of gas temperature, thereby providing valuable information about the fluid near regions of local extinction, as indicated by well-defined breaks in the CH layer. The extent of premixedness in the region upstream from the CH structure is also assessed from the Rayleigh signal level. Furthermore, the roles of premixedness in flame stabilization, the nature of the leading edge, and lift-off height oscillation are discussed.

Newly Developed Direct-Connect High-Enthalpy Supersonic Combustion Research Facility

Anew continuous-e ow,direct-connect,high-enthalpy, supersonic combustion researchfacility isdescribed. This test facility provides combustor inlet e ow conditions corresponding to e ight Mach numbers between 3.5 and 7, at dynamic pressures up to 95.8 kPa. Most of the major components of the new facility are water cooled (including the vitiated heater, the instrumentation and transition sections, and the facility nozzle and isolators ); the current exception is the variable-geometry heat-sink combustor. A variety of conventional and advanced instrumentation, including a steam calorimeter and a thrust stand, exists for accurate documentation of combustor inlet and exit conditions and performance parameters. In a recent calibration effort, pitot pressure surveys, total temperature surveys, and wall static pressure distributions were obtained for a wide range of inlet conditions using Mach 1.8 and 2.2 facility nozzles. In addition, three-dimensional numerical simulations of each test case were completed. Results from thecomputations compare favorably with experimental results for all cases and yield estimates of the integral boundary-layer properties at the isolator exit.

Observations on the leading edge in lifted flame stabilization

The objective of this paper is to report some of the first experimental evidence for the “leading edge” flame as the stabilization mechanism in lifted jet diffusion flames [1–5]. CH fluorescence has been used to indicate the flame front location (i.e., region of chemical reaction) and thereby characterize features of the stabilization region [5, 6]. The “leading edge” flame phenomenon reported within refers to the outward-extending branch of CH fluorescence at the base of the streamwise CH zones. Whether the “leading edge” flame is a special case of the more general triple flame is a question which remains unanswered. It is evident from previous computational studies [7, 8] that the triple flame, when interacting with a vortex or pair of vortices, can take on characteristics of the “leading edge” flames introduced in the present study. Veynante et al. [8] illustrate the contortion of the premixed branches of the triple flame by the flowfield where the premixed branches are swept into the trailing diffusion flame. These simulated triple flame/vortex interactions are consistent with the results of this study which show a trailing diffusion flame and the leading edge reaction zone structure.The test conditions and measurement locations for this investigation are shown in Fig. 1. Download high-res image (204KB) Download full-size image Fig. 1. Test conditions and measurement locations for (a) the lowest flow rate and (b) the highest flow rate. The intermediate flow rate (not shown) corresponds to a methane velocity of 21.2 m/s while the bottom of the image region is 37.4 mm from the jet exit and includes both sides of the lifted flame. The axisymmetric burner consists of a 5-mm inner diameter fuel jet surrounded by a 150-mm i.d. coflow tube. Methane is delivered through the fuel jet, while low-speed air (∼0.15 m/s) passes through the coflow annulus. The stabilization regions of three lifted flames are investigated by varying the methane and air flow rates and adjusting the burner position accordingly so that the image region includes the leading edge of the reaction zone. The methane exit velocities are 15.8, 21.2, and 27.5 m/s, corresponding to jet Reynolds numbers of 4800, 6400, and 8300, respectively. Both sides of the lifted flame are imaged during the two lower flow rates (Fig. 1a) while the turbulent fluctuations and wider stabilization region resulting from the highest flow rate limit this case to one side of the flame (Fig. 1b). The CH planar laser-induced fluorescence (CH-PLIF) technique has been described elsewhere [5, 6]. The setup includes a Nd:YAG-pumped dye laser which excites the Q1(7.5) transition of the B2Σ−–X2π(0,0) band of CH at λ = 390 nm. Fluorescence from the A–X(1,1), (0,0) and B-X(0,1) bands between λ = 420 and 440 nm is recorded. This approach has resulted in acceptable CH signal levels and excellent image quality (i.e., spatial resolution and contrast), which the authors find superior to the 431.5 nm laser excitation employed in earlier studies [9, 10]. The authors believe the excellent resolution resulting from the signal levels and laser sheet characteristics in this study are extremely important in uncovering the leading edge premixed branch, which is generally weaker in signal level than the trailing diffusion flame. It is likely that these stabilizing leading edge flame observations are not reported in studies with less spatial resolution or are possibly not at all detectable due to limitations in the specific CH excitation/detection scheme. Several diagnostic studies involving lifted flames present the lifted flame structure as a continuous flame surface, similar to a distorted cylindrical object, emanating from a ring-shaped structure where the flame is stabilized [4, 5, 11]. Most previous work, however, does not give experimental evidence of the mechanism of lifted flame stabilization. Figure 2 Download high-res image (3MB) Download full-size image Fig. 2. CH-PLIF images illustrating the leading edge phenomenon where the CH zone extends outward at the stabilization point. (a)–(c) are from the lowest Re = 4800 flow condition (Fig. 1a); (d)–(h) are from the intermediate Re = 6400 flow condition; (i)–(l) are from the highest Re = 8300 flow condition and only include the right side of the flame (Fig. 1b). consists of several instantaneous 35.1 mm × 23.4 mm CH-PLIF images which provide such evidence. The images clearly show a continuous vertical distribution of CH which represents the primary diffusion flame reported in many previous studies. In addition to the vertical trailing diffusion flame, a structure is witnessed near the flame base which curls toward the outside, or fuel-lean, portion of the reaction zone. In comparison to ideal, laminar tribrachial structures, evidence of both rich and lean branches of premixed flame is not present, only the one branch extending outward near the jet edge. However, the rich branch on the fuel side of the diffusion flame may be overlapped into the diffusion flame by the flowfield as illustrated by Veynante et al. [8]. It is believed that the branch in the CH zone is a leading edge flame, stabilized by opposing the flow in the relatively low-speed region (∼1.0 m/s) near the outside edge of the jet. This branch of CH is not obviously present in all of the data; it only appears in approximately 30% of the images. The authors reason that the leading edge phenomenon may not be present around all 360 degrees of the stabilizing “ring,” but only in a portion of the flame sufficient to generate enough thermal energy to stabilize the flame globally. The reader must keep in mind that while the laser imaging techniques allow one to investigate the flame in detail locally, global behavior, most notably out-of-sheet activity, may be dominant at any given instant. With the flow inherently three-dimensional and time-dependent [11, 12], sheet imaging techniques often provide data to support a theory, but rarely provide definitive evidence. Since the measurements only investigate the flowfield in one plane, it is feasible that the leading edge structure could be present outside the measurement slice during the instances when no premixed CH branch is witnessed. In addition, recent cross-sectional images of lifted flames near the stabilization zone clearly render three-dimensional lobed structures that are consistent with this theory [11, 12]. The extent of mixing and the entrainment of ambient air into the fuel is of central importance to this problem. Based on comparisons with mixture fraction images presented by Starner et al. [13], which illustrate that the portion at the base of a lifted methane flame has a flammable composition, the authors are confident that the leading edge flame lies in a flammable mixture fraction region. Furthermore, fluctuations in the axial location of the leading edge, along with its orientation relative to the trailing diffusion flame, imply axial propagation into the unburned gas region. These observations imply that the physics of flame stabilization is likely a combination of multiple mechanisms based on premixedness, strain rate considerations [6, 14], and flame propagation into nonhomogeneous flowfields with flow separation, scalar gradients, and a range of mixture fractions [13].

Planar imaging of CH, OH, and velocity in turbulent non-premixed jet flames

Simultaneous planar laser-induced fluorescence of the CH and OH radicals and two-dimensional particleimage velocimetry were used to investigate the structure of turbulent non-premixed methane/nitrogen jet flames ( Re jet =18,600) in an oxygen coflow. The motivation for this study is to investigate the relationship among regions of high CH/OH concentration and kinematic quantities such as vorticity, strain rate, and dilatation. The results show that in the lower part of the flame, the direction of the two-dimensional principal compressive strain axis exhibits a preferred orientation of about 45° with respect to the flow direction, whereas near the flame tip, the strain exhibits a more random orientation. Furthermore, CH structures are more likely to align orthogonal to the principal compressive strain axis in the downstream half of the flame. Probability density functions (PDFs) show that the most probable value of vorticity on CH structures is about Δ U /δ (where Δ U is the difference between the jet centerline and coflow velocities and δ is the full width at half-maximum of velocity profile), but near zero on the OH structures. Furthermore, joint PDFs of strain and dilatation show that CH structures are more likely to be associated with positive dilatation than are OH structures. These results are consistent with previous studies that have shown that jet flame kinematics are substantially affected by heat release and further show that these effects are more closely correlated with zones of high CH concentration than with zones of high OH concentration.

Axisymmetric jet shear-layer excitation by laser energy and electric arc discharges

Two energy deposition methods (electric arcing and laser-induced optical breakdown) were used to force and control compressible mixing layers of axisymmetric jets. The effects of energy-deposition forcing methods have been experimentally investigated with schlieren imaging, particle image velocimetry, product formation flow visualizations, and high-frequency pressure measurements. Large-scale structures were forced in perfectly expanded jets with nozzle-exit Mach numbers of 1,38,1.5, and 2.0, utilizing single pulse-laser energy deposition focused at the nozzle exit. Structures were successfully forced over a range of convective Mach numbers from 0.63 to 0.85 using laser pulse energies from 5 to 40 mJ. The large-scale structure forced by laser perturbation in the Mach 1.38 jet was characterized with detailed measurements of the velocity and vorticity fields and the fluctuating pressure history

On scalar dissipation and partially premixed flame propagation

Measurements of the scalar dissipation rate in the region immediately upstream of a lifted jet flame are presented. The scalar dissipation is determined in this isothermal region from a planar measurement of a two-dimensional conserved scalar (jet fluid) using laser Rayleigh scattering. Fields of the scalar dissipation rate are presented in addition to tabulated values for three different liftoff heights ( Re d =4800, 6400, and 8300). Scalar dissipation rates do not reach levels thought to cause extinction of the leading edge based on comparison with extinction data for counterflow diffusion flames. Additionally, results are presented on the axial flame propagation velocities relative to the jet flow. The data indicate that over the three flow conditions, the flame velocity relative to the flow is approximately constant during the case of a quasi-stationary lifted flame. In light of these findings, it is suggested that concepts involving partially premixed flame propagation, rather than those of critical scalar dissipation rate, are central to modern lifted flame stabilization models.