Search

Combustors with fuel-spray atomizers are particularly susceptible to a low-frequency oscillation at idle and subidle conditions. For aeroengine combustors, the frequency of this oscillation is typically in the range 50-120 Hz and is commonly called 'rumble.' The mechanism involves interaction between the plenum around the burner and the combustion chamber. Pressure variations in the plenum or the combustor alter the inlet air and fuel spray characteristics, thereby changing the rate of combustion. This in turn leads to local 'hot spots' which generate pressure oscillations as they convect through the downstream nozzle. In order to eliminate the combustion oscillations, it is essential to determine which fuel atomizers are particularly likely to lead to instability by quantifying their sensitivity to flow perturbations. This can be done by identifying the system through understanding the transfer function, which represents the relationship between the unsteadiness of combustion and the inlet fuel and air. In the present work, various types of signals are applied to produce a small change in the inlet fuel and air flow rates, the response in the rate of heat release caused downstream was calculated and stored for subsequent analysis. Afterwards, the system transfer function is calculated by determining the coefficients of an IIR filter (Infinite Impulse Response)for which the output signal is the downstream heat release rate and the input signal is the inlet flow rate. The required transfer function then follows from the Fourier transform of this relationship. The resulting transfer functions are compared with those obtained by the forced harmonic oscillations at a fixed given frequency. Suitably chosen input signals accurately recover the results for harmonic forcing at a single frequency, but also give detailed information about the combustor response over a wide frequency range. There are two distinct forms to the low-frequency quasisteady response. In the primary zone, the rate of combustion is influenced by the turbulence and is enhanced when the inlet air velocity is large. Near the edge of combustion zone, the rate of combustion depends on the mixture fraction and is high when the mixture fraction is close to the stoichiometric value. This generates 'hot spots' which convect into the dilution zone. At higher frequencies, the combustion lags this quasi-steady response through simple lag-laws and the relevant time delays have been identified. [DOI: 10.1115/1.1806451]

Byline: A. V. Voloshin (1), A. Kh. Gil'mutdinov (1), Yu. A. Zakharov (1), A. A. Sevast'yanov (1) Abstract: Shadow spectral filming is used to study the spatiotemporal dynamics of the formation and dissipation of vapors of potassium sulfate and aluminum and indium nitrates in a transversely heated graphite furnace atomizer. The integrated platform, the Pd--Mg modifier, the internal flow of the sheath gas, the magnetic field of the nonselective background corrector, and the diffusion of oxygen from the ambient air are responsible for specific nonuniformities in the spatial structure of the vapor cloud. The nonuniformities result to a large extent from the transverse nonisothermal conditions of the graphite furnace. The explosive splashes of aluminum vapors interfering with the analysis are generated at the cold side walls of the graphite furnace, where the vapors are condensed in the process, rather than on the particles of the dry residue of the sample. The beating of the magnetic-field induction inside the graphite furnace atomizer caused by the overlapping of the fields of the heating current (50 Hz) and the background corrector (54 Hz) results in low-frequency (4--5 Hz) oscillations of the spatial position of the vapor cloud. The matrix modifier can stimulate these oscillations. Author Affiliation: (1) Physics Faculty, Kazan State University, Kremlevskaya ul. 18, Kazan, 420008, Russia Article History: Registration Date: 06/10/2004

Most types of combustion-driven devices experience combustion instabilities. For aeroengine combustors, the frequency of this oscillation is typically in the range 60-120 Hz and is commonly called 'rumble.' The rumble oscillations involve coupling between the air and fuel supplies and unsteady flow in the combustor. Essentially pressure fluctuations alter the inlet fuel and air, thereby changing the rate of combustion, which at certain frequencies further enhances the pressure perturbation and so leads to self-excited oscillations. The large residence time of the liquid fuel droplets, at idle and subidle conditions, means that liquid and gaseous phases must both be considered. In the present work, we use a numerical model to investigate the forced unsteady combustion due to specified time-dependent variations in the fuel and air supplies. Harmonic variations in inlet air and fuel flows have been considered and the resulting unsteady combustion calculated. The influence of droplet size distribution has also been investigated. The calculations provide insight into the interaction between atomization, unsteady combustion, and flow oscillations. [DOI: 10.1115/1.1396841]

Planar laser scattering (PLS) and planar laser-induced fluorescence (PLIF) techniques are currently being used for rapid characterization of fuel sprays associated with gas turbine atomizers, diesel injectors, and automotive fuel injectors. These techniques can be used for qualitative, quantitative, and rapid measurement of fuel mass, spray geometry, and Sauter mean diameters in various sprays. The spatial distribution of the fuel mass can be inferred directly from the PLIF image, and the Sauter mean diameter can be measured by simultaneously recording the PLIF and PLS images and then rationing the two. A spray characterization system incorporating the PLS and/or PLIF techniques has been loosely termed an optical patternator, and in this study, it has been used to characterize both steady and pulsed sprays. The results obtained with the optical patternator have been directly validated using a Doppler particle analyzer (PDPA).

The relationship between atomizing gas molecular weight and spray mean drop size, Rosin-Rammler distribution parameter, and number averaged drop velocity is reported for a low mass flowrate effervescent atomizer-produced spray. Experimental data at lower gas-liquid ratios (GLR's) demonstrate that an increase in the molecular weight of the atomizing gas increases mean drop size and decreases number averaged drop velocity. The increase in mean drop size is attributed to an increase in the thickness of the liquid annulus at the nozzle exit and a subsequent increase in the diameter of ligaments formed there. The decrease in number averaged velocity results from a decrease in jet momentum rate. A model developed to explain the atomization process indicates that the gas flow is choked at higher GLR's.

Experimental and numerical investigations are made on the behavior of droplets in a hollow-cone spray paying attention to the liquid sheet formed at the orifice of pressure-swirl atomizer. Simultaneous measurements of droplet sizes and velocities are made by phase-Doppler technique and numerical simulations are carried out based on the transient Eulerian equations for the gas and the Lagrangian equation for the droplets, taking account of the liquid sheet formed at the atomizer orifice. It is shown that the simulation gives good predictions by incorporating the existence of the liquid sheet. The predicted results indicate that the movement of the liquid sheet induces a strong air stream which acts as a strong side wind against the droplets immediately after breakup. This air stream selectively transports small droplets toward the central region and plays an essential role in the classification of droplets by size. Accordingly, the existence of the liquid sheet is significant for the characteristics of droplet dispersion and it should not be neglected in the prediction of hollow-cone spray flows. In addition, the shape of the liquid sheet is theoretically computed based on the simplified equations of motion. The comparison between the theoretical computation and the experimental result suggests that the surface tension of liquid is predominant in determining the shape of the liquid sheet.

The absorbance of Cu atoms sputtered from a Cu cathode in a laboratory-modified radio frequency glow discharge atomizer for atomic absorption spectrometry was measured at different regions within the atomizer chamber for different discharge parameters. The results show that not only the discharge gas flow rate and pressure but also the applied rf power and the degree of extension of the negative glow have significant effects on the atomization and on the distribution of sputtered atoms within the atomizer chamber. The influence of various diameters and thicknesses of the sampling orifice (anode) on the plasma atomization characteristics and the sample weight loss rates has been examined. By increasing the diameter of the sampling orifice, a larger area of the cathode surface (Cu) was allowed to be sputtered, thereby generating a larger number of atoms (higher absorbance) in the analysis volume, where the sputtered atoms were studied by atomic absorption spectrometry. The absorbance of Cu atoms was inversely related to the thickness of the sampling orifice disk because the distance of the investigated region of the atomizer chamber from the cathode surface increases with increasing thickness of the sampling orifice disk. The rate of sample weight loss was inversely related to the diameter of the sampling orifice at orifice diameters higher than or equal to 4 mm. The rate of sample weight loss as a function of the thickness of the sampling orifice disk indicated that the transport efficiency of the sputtered atoms into the atomizer chamber was higher when a sampling disk having a thickness of 2 mm was used. Changing the location of the discharge gas inlet port resulted in a large increase in the atomic absorption signal. A larger increase in the atomic absorption signal of the sputtered Cu atoms was observed for the modified rf atomizer having the discharge gas inlet port on the body of the sampling orifice disk than for the unmodified rf atomizer having the discharge gas inlet port on the top of the atomizer chamber.

A single 90 [degrees] off-axis ellipsoidal mirror fragment was used in a dispersive detection system for electrothermal atomization laser-excited atomic fluorescence spectrometry. The performance of the new optical arrangement was compared with those of optical arrangements that employed a plane mirror in combination with biconvex or plano-convex lenses. All the optical arrangements collected fluorescence in a scheme called front surface illustration. BEAM-4, an optical ray tracing program, was used for calculations of spatial ray distributions and optical collection efficiency for the various optical configurations. Experimentally, the best collection efficiency was obtained by use of the ellipsoidal mirror, in qualitative agreement with simulations done by use of the BEAM-4 software. The best detection limit for cobalt with the new optical arrangement was 20 fg, which was a factor of 5 better than that obtained with conventional optical arrangements with otherwise the same instrumentation. The signal-to-background ratio and the fluorescence collection efficiency were also studied as a function of position of the optical components for the various optical arrangements. For both cobalt and phosphorus, the signal-to-background ratio with the new optical arrangement remained stable within 10-20% during [+ or -]8 mm shifts in the position of the detection system from the focal plane of the optics. Overall, the new optical arrangement offered high collection efficiency, excellent sensitivity, and facile optical alignment due to efficient spatial separation between the fluorescence signal and the background radiation. The advantages of the new optical arrangement were particularly important during measurements in the presence of high levels of blackbody radiation.

A new approach is described for processing transient data from electrothermal atomizers used in atomic absorption spectroscopy. The transient responses are first integrated and then a pseudo-first-order model is fit to the time-dependent integrals in order to predict the response that would be measured if the atomization process were monitored to completion. The principal advantage expected and observed for the new approach is its ability to reduce effects of variables such as atomization temperature. For all elements studied (Cr, Mn, K, Yb, Fe), the new predictive approach is shown to be virtually independent of temperature in the range from 2200 to 2600[degrees]C. The predictive approach exhibited lower temperature coefficients than either the peak-height or peakarea options for all elements examined. For the more volatile elements (Mn, K, Yb, Fe), the improvement ratio at 2400[degrees]C of the predictive approach relative to the others ranged from 1.4 to 8.2. For chromium at 2400[degrees]C, the temperature coefffcient of the predictive method was approximately 10- and 30-fold lower than those for the peak-area and peak-height options, respectively.

Liquid sheet break-up in coflowing shear flow is the mean by which liquids are atomized in practical injectors for gas turbine combustors. The present study explores experimentally the mechanisms of liquid sheet instabilities and spray formation. Experiments are conducted on four airblast geometries. A high-speed video camera associated with an image processing unit was used to study the liquid sheet instabilities. A microphone and a frequency analyzer were used to track the disintegration frequency. Instability amplitude and disintegration length of the liquid sheet were measured. A two-component Phase Doppler Particle Analyzer was used to characterize the resultant spray. The spatial distribution of the particle size is influenced by the swirling flow field. These experimental results will be used to assess models of fuel sheet instabilities and disintegration.

Measurements of circumferential liquid distribution are carried out over ranges of injection pressure from 0.34 to 1.72 MPa (50 to 250 psi) using five different simplex nozzles to achieve variations in the discharge orifice length/diameter ratio from 0.5 to 4.0. Two additional simplex nozzles of the same flow number are also examined in order to ascertain the effect of variations in the number of swirl chamber feed slots on circumferential liquid distribution. Mixtures of water and glycerine are used to provide a twelve to one variation in liquid viscosity. The results obtained show that spray uniformity improves markedly at higher injection pressures. Increase in liquid viscosity also has a beneficial effect on spray uniformity. The most uniform circumferential liquid distributions are obtained with nozzles having a discharge orifice length/diameter ratio of between 1 1/2 and 2.

In order to formulate a common approach that could provide the spray parameters of airblast atomizers, various processes of liquid preparation, breakup, and secondary atomization have been included in a semi-analytical calculation procedure. The air velocity components in the atomizer flow field are provided by mathematical expressions, and the spray droplets are considered to form at ligament breakup through a disturbance wave growth concept. The validation of the developed approach included the application to six atomizers that significantly varied in concept, design, and size. They represented both prefilming and plain-jet types, and the data utilized in the present effort were obtained with six different liquids. Satisfactory agreement between the measurements and the predictions has been achieved under wide ranges of air/fuel ratio and air pressure drop for various test liquids. The results of this investigation indicate the potential of using such an approach in the early phases of airblast atomizer design, and may be followed by more detailed calculations using analytical tools., The characteristics of sprays produced by airblast atomizers are studied through calculations from a derived semi-analytical approach, and experiments performed on six varying types of atomizers. Results show that spray quality depends not on the kind of atomizer used, but on the way the liquid is injected into the air. Theoretical and experimental results show good agreement. The calculation approach may also be applied in studying other common liquids for atomizer design.

The spray cone angles of seven simplex nozzles are investigated using three liquids of varying viscosities. Results show that the spray cone angle increases with an increase in injection pressure and decreases with an increase in the liquid viscosity and length/diameter ratio of the final discharge orifice. Variations in the number of swirl chamber feed slots had minimal effect on the spray cone angle.

For some prefilming airblast atomizers, the manner in which the atomizing air impinges on the liquid sheet prevents the wave formation that usually precedes the breakup of a liquid sheet into drops. When this occurs, the liquid shatters into various drop sizes. A study is conducted to show that these two types of atomization is dependent on the angle in which the atomizing air impinges on the liquid sheet. A formulation for mean drop size is also derived.

Phase Doppler measurements of size, velocity, liquid flux, and average mass fractions were obtained in sprays produced by three identical coaxial airblast atomizers, with their axes placed in a triangular arrangement at distances of two air jet diameters from each other; the arrangement simulates the spray interaction in the preburner of the space shuttle main engine with water and air respectively replacing the liquid oxygen and hydrogen of the preburner sprays. Each nozzle comprised a liquid jet with exit diameter of 2.3 mm centred in a 8.95 mm diameter air stream. Two liquid flowrates were examined, while the air flowrate was kept constant, resulting in Weber number at the exit of the nozzle around 1100, air-to-liquid momentum ratio 8.6 and 38, velocity ratio 24 and 51, mass flowrate ratio 0.35 and 0.75, liquid jet Reynolds number 10,000 and 21,000 and air jet Reynolds number around 108,000. The air flow characteristics were compared to the flow without liquid injection. Up to 10 air jet diameters from the nozzle exit, individual spray characteristics dominated and maximum Sauter mean diameters, typically around 150 [[micro]meter], and liquid flux were observed on the geometrical axes of the nozzles. Spray merging was strong in the region between the nozzle axes, where the Sauter mean diameter reduced and the liquid flux and the mean and rms of the fluctuations of the axial velocity of the droplets and the air flow increased relative to the single spray. Downstream of 25 air jet diameters from the nozzle exit, the multiple sprays merged to a single spray-like flow produced by a nozzle located at the centre of the triangular region between the nozzle axes. Reduction of the liquid flowrate by 50 percent, improved atomization by 25 percent, shortened the axial distance from the nozzles where the individual spray characteristics disappeared by 30 percent and increased the airflow turbulence by 20 percent. Droplet coalescence was negligible for high liquid flowrates, but for reduced liquid flowrates coalescence became important and the Sauter mean diameter increased with the axial distance from the exit by around 15 percent. Spray merging increased the air flow turbulence and the local mass fraction distribution of the air in the region between the nozzle axes by around 50 and 40 percent respectively relative to the single sprays, resulting in a fuel rich region with increased gas flow turbulence which may influence the ignition process in the preburner of the space shuttle main engine., The paper regarding the spray characteristics of multiple coaxial airblast atomizers similar to the equipment used by space shuttle engines are important due to various reasons. It provides combustion engineers with information needed to enhance the performance of space shuttle engines. Moreover, it opens up new avenues for analyzing concepts regarding atomization.