Preface 1.General Considerations 2.Basic Processes in Atomization 3.Drop Size Distributions of Sprays 4.Atomizers 5.Flow in Atomizers 6.Atomizer Performance 7.External Spray Charcteristics 8.Drop Evaporation 9.Drop Sizing Methods Index

Atomization and Sprays

Dynamics and stability of lean-premixed swirl-stabilized combustion

We present a practical implementation of calibration-free wavelength-modulation spectroscopy with second harmonic detection (WMS-2f) for measurements of gas temperature and concentration in harsh environments. The method is applicable to measurements using lasers with synchronous wavelength and intensity modulation (such as injection current-tuned diode lasers). The key factors that enable measurements without the on-site calibration normally associated with WMS are (1) normalization of the WMS-2f signal by the first harmonic (1f) signal to account for laser intensity, and (2) the inclusion of laser-specific tuning characteristics in the spectral-absorption model that is used to compare with measured 1f-normalized, WMS-2f signals to infer gas properties. The uncertainties associated with the calibration-free WMS method are discussed, with particular emphasis on the influence of pressure and optical depth on the WMS signals. Many of these uncertainties are also applicable to calibrated WMS measurements. An example experimental setup that combines six tunable diode laser sources between 1.3 and 2.0 mum into one probe beam for measurements of temperature, H(2)O, and CO(2) is shown. A hybrid combination of wavelength and frequency demultiplexing is used to distinguish among the laser signals, and the optimal set of laser-modulation waveforms is presented. The system is demonstrated in the harsh environment of a ground-test scramjet combustor. A comparison of direct absorption and 1f-normalized, WMS-2f shows a factor of 4 increase in signal-to-noise ratio with the WMS technique for measurements of CO(2) in the supersonic flow. Multidimensional computational fluid-dynamics (CFD) calculations are compared with measurements of temperature and H(2)O using a simple method that accounts for the influence of line-of-sight (LOS) nonuniformity on the absorption measurements. The comparisons show the ability of the LOS calibration-free technique to gain useful information about multidimensional CFD models.

Wavelength-modulation spectroscopy for measurements of gas temperature and concentration in harsh environments

DOI: 10.2514/1.40966 The dynamics of unstart of a floor-mounted inlet/isolator model in a Mach 5 flow are investigated experimentally. The inlet section contains a 6-deg compression ramp, and the isolator is a rectangular straight duct that is 25.4-mm high by 50.8-mm wide by 242.3-mm long. Measurements made include 8-kHz schlieren imaging and simultaneous fast-response wall pressures along the length of the inlet/isolator. Unstart is initiated by deflecting a flap at the downstream end of the isolator. The shock system, induced by unstart, initially propagates upstream through the isolator at a velocity of about 35 m=s (in the laboratory frame of reference), then decelerates to about 20 m=s near the isolator entrance, and then accelerates to a velocity of about 74 m=s within the inlet. Throughout the isolator, unstart is seen to be strongly associated with boundary-layer separation. Once the inlet has unstarted, a highamplitudeoscillatory(periodic)unstarted flowensues,forwhichtheoscillationfrequencyisabout124Hz.However, under some conditions, an 84-Hz oscillatory unstarted flow mode, with lower pressure fluctuations, is observed. Under other conditions, a nonoscillatory unstarted flow, with much lower pressure fluctuations, is observed.

Experimental Investigation of Unstart in an Inlet/Isolator Model in Mach 5 Flow

Combustion modes of hydrogen jet combustion in a cavity-based supersonic combustor

Flame structures and operating limits of an ethylene-fueled recessed cavity flameholder were investigated both experimentally and numerically, using a newly developed AFRL research scramjet flowpath at Wright-Patterson Air Force Base. Flush-wall low-angled injectors were used as main fuel injectors. The recessed cavity features an array of fueling ports on the aft ramp for direct cavity fueling. The cavity operating conditions include 1) direct cavity fueling, 2) direct cavity fueling with back pressurization, and 3) fueling from main injectors with and without direct cavity fueling. With direct cavity fueling, significant variation in the shape and spatial distribution of the cavity flame was observed at various fuel flow rates with and without back pressurization. It was found that both lean ignition and blowout limits increase with the characteristic air flow rate. The lean blowout limit is decreased toward a lower value as the shock train is pushed toward upstream. With fueling from main injectors, the flame is mainly distributed within the body wall corners for the present flowpath. The rich blowout limit for a cavity fueled with both main and cavity fuel is lower than for the case with cavity fuel alone, due to main fuel entrainment from the low-angle injectors. Qualitative composition analysis indicates that the gas mixture inside the cavity mainly contains combustion products and is relatively rich with main fuel only. Consequently, additional fuel injection into the cavity increases the probability of blowing out the entire flame by disabling the flame holding capability of the recessed cavity for the present flowpath and injector designs. The rich blowout limit with main fuel injection was found to increase with the body-side fuel flow rate. Merging of fuel plumes injected from upstream injectors creates an aerodynamic blockage for air entrainment into the cavity and, consequently, reduces the rich blowout limit. NOMENCLATURE

Flame Characteristics and Fuel Entrainment Inside a Cavity Flame Holder in a Scramjet Combustor

The structures of sonic ethylene jets delivered from orifices of three different diameters and two injection angles (30 and 90deg) into aMach 2 supersonic crossflowwere studied experimentally. The ratio of the cross-sectional areas of the largest and smallest injectors is 25:1. Time-averaged spontaneous vibrational Raman scattering was used to quantify injectant concentrations by constructing two-dimensional spanwise concentration images from the onedimensional linewise Raman scattering images. Based on the present data set, new penetration height correlations were developed to treat cases with injection angles of both 30 and 90 deg. Excluding the influence of wall boundary layer, the present measurements show that the properties of fuel plume structures, such as shape, size, and concentration profiles, are scalable with the injector size. Themeasured ethylene concentrations were also compared with predictions from the revised jet penetration code, which was calibrated primarily with hydrogen and helium. Discrepancieswere observedbetween themeasurements and the jet penetration code predictions for the structures of ethylene fuel plumes. The experimental data generated from the present study can be used to validate the numerical simulations.

Raman Scattering Measurements of Gaseous Ethylene Jets in a Mach 2 Supersonic Crossflow

The structures of water jets injected into an M=1.94 crossflow were studied experimentally. Two plain orifice nozzles with L/d0 of 20 and orifice diameters of 0.5 and 1.0 mm were tested. Liquid injectors were flush mounted on the bottom plate of the wind tunnel to provide normal injection. Wide ranges of test conditions for jet-to-air momentum flux ratios, aeration levels, and freestream air velocity were tested. A two-component phase Doppler particle analyzer (PDPA) was utilized for the measurement of droplet and spray plume properties along the centerline and across the half plane of spray plumes at various freestream locations. Based on the PDPA measurements, correlations for the penetration heights of pure- and aerated-liquid jets were developed. It was found that once the jet is aerated, the penetration height and cross-sectional area of the spray plume increase dramatically to create a more uniformly distributed spray plume for injected liquid. The atomization processes of pure- and aerated-liquid jets are completed at x/d0<100 at the M=1.94 crossflow, due to the strong action of the supersonic freestream air. The flux- averaged SMD is fairly constant for x/d0≥100 and is on the order of 10 µm for both pure- and aerated-liquid jets. Centerline distribution profiles of droplet and spray plume properties in the freestream direction can be normalized by the penetration height of each spray to obtain universal curves for both pure- and aerated- liquid jets in regions where the liquid atomization process is complete. The normalized distribution profiles for droplet size and x-component droplet velocity exhibit S and mirrored-S shapes, respectively. These S-type distribution profiles are caused by the presence of the bottom floor. These universal curves can potentially be used for the modeling of the far-field structure of liquid jets in supersonic crossflows. NOMENCLATURE

Structures of water jets in a mach 1.94 supersonic crossflow

*† ‡ § ¶ This paper deals with the thermoacoustic instability and ensuing flow oscillation in a scramjet engine, a phenomenon commonly known as combustion instability. The analysis is based on a quasi-one-dimensional treatment of unsteady flow motion, which simulates the main features of the oscillatory flowfields in both the isolator and combustor. The model also accommodates the response of local heat release to acoustic excitation. The calculated oscillation frequency agrees well with the measured values of around 350 Hz. A companion analytical analysis is also established to help explore the underlying mechanisms responsible for driving and sustaining thermoacoustic flow instabilities. In particular, the interactions between the unsteady heat release, fuel injection and mixing, and shock response are examined. Their influence on the acoustic oscillation characteristics is identified.

Thermoacoustic Flow Instability in a Scramjet Combustor

Shock train structures inside constant-area isolators were investigated experimentally and numerically. Two heat- sink isolators, one with a rectangular configuration and one with a round configuration, and one ceramic-paneled rectangular isolator with an identical cross-sectional area a length of 25" were back pressurized using a throttle valve. Both heat-sink isolators were coated with a thermal barrier, which has a relatively rough surface. Pressure profiles inside the shock train and flow properties at the isolator exit plane were measured, using wall static pressures, Pitot pressures, and stagnation temperatures, at various back pressures, in Mach 1.8 and 2.2 flows. It was found that the round isolator can sustain a higher back pressure before unstarting the flow than the rectangular isolator. The rectangular isolator with smooth ceramic panels exhibits better shock holding capability than the heat- sink rectangular isolator. The shock train location is very sensitive to back pressure variation when the shock is stationed near the isolator entrance. As the shock train is pushed toward the isolator entrance by increasing back pressure, much more uniform distributions of flow properties are created at the isolator exit plane, where the centerline flow Mach number decreases linearly with the pressure ratio along the entire isolator length. The parameter group M0 2 ((Ps-Ps´)/P0)/(Pb/P1) was found to be able to bring the spatially shifted pressure profiles of the shock trains into a narrowly-distributed similarity profile with minor dependences on isolator configuration and flow Mach number. NOMENCLATURE D = characteristic length L1.0 = shock train length M = Mach number M0 = nominal Mach number for facility nozzle P = pressure Ppitot = Pitot pressure PR = pressure ration across the isolator, Pb/P1 Reθ = Reynolds number based on boundary layer momentum thickness T = temperature x = free stream direction xs = shock train leading edge x' = freestream distance from shock leading edge, x'=x-xs y = transverse direction θ = boundary layer momentum thickness Superscript ' = condition at shock leading edge

Kuo-Cheng Lin

Papers

Flame Characteristics and Fuel Entrainment Inside a Cavity Flame Holder in a Scramjet Combustor

Raman Scattering Measurements of Gaseous Ethylene Jets in a Mach 2 Supersonic Crossflow

Structures of water jets in a mach 1.94 supersonic crossflow

Thermoacoustic Flow Instability in a Scramjet Combustor

Characterization of Shock Train Structures inside Constant-Area Isolators of Model Scramjet Combustors