A review and comparison of initiation of combustion processes by conventional electric spark or thermal means with laser sources is presented. A description of the fundamentals of ignition processes is used as basis for interpretation of experimental and theoretical studies of laser ignition. It is shown that many features of laser and conventional ignition can be understood on the basis of simple thermal concepts, particularly when the effects of thermal or radical losses are considered. It is proposed that the main advantages of laser sources is likely to be in the timing and placement of ignition rather than the inherent energy requirements. Potential applications to combustion systems of practical importance, e.g. high-speed propulsion systems, are discussed and instructive laboratory-scale experiments are suggested.

One of the main purposes of the present study is to explore the use of laser radiation to initiate and support combustion of coals. Within this context, some experiments were conducted to study the interaction of the Nd-Yag laser radiation with four different coals. Laser intensities ranging from 0.5 X 10³ to 1.5 X 10⁴ W/cm² at 1.064 micrometers wavelength and the pulse duration of 5 ms were used. For laser intensities less than 800 W/cm², no ignition was observed for all coals. For laser intensities above this value, two ignition mechanisms were observed: the surface ignition followed by the gas phase ignition when Wyoming subbituminous, Indian lignite and North Dakota lignite coals were used. For the same range of the laser intensities, however, only the gas phase ignition was observed when Pittsburgh bituminous coal was used. It was also noted that a significant amount of the external laser radiations were absorbed by the pyrolysis products during the early stages of the ignition period.

The unique chemistry of methane combustion, including strong C-H bond energy, leads to difficulties in use of natural gas as an engine fuel. Problems include low combustion efficiency, knocking, unreliable ignition (misfiring), and NO_x emission. It is well established that improvement of the above-mentioned combustion phenomena requires the presence of high concentration of chain-initiating and chain-branching reactive radicals. This project explores a novel approach called Infrared Multiphoton Dissociation (IRMPD), for producing reactive radicals. IRMPD involves the absorption of multiple infrared photons by target molecules within the duration of an infrared laser pulse, leading to formation of high energy species. This study is an exploration of the applicability of IRMPD to natural gas excitation, and subsequent enhancement of combustion. IRMPD is demonstrated to be a feasible concept for natural gas engine ignition and for combustion enhancement through reduction in ignition delay time. The next stage of development would be to identify and implement a prototype laser and optical hardware for single-cylinder engine tests.

Spherical detonations of C₂H₂/O₂/N₂ mixtures in an open flow system (initially at 1 atmosphere) and planar detonations of C₂H₂/O₂ and H₂/O₂/C₂H₂ mixtures in an enclosed tube are successfully initiated by use of an ArF laser at 193 nm. The required critical energy for the initiation of spherical detonations is found to be relatively low: approximately 12 +/- 2 mJ for a 40% C₂H₂ in C₂H₂/O₂ mixtures. This small critical energy may be attributed to a relatively strong absorption of C₂H₂ at 193 nm, and possible enhancement by the photodissociation products of C₂H and H. The initiation appears to be accomplished without overdriving the mixtures through a blast wave. The critical energy, delay time, detonation velocity and pressures are measured as functions of stoichiometric mixture ratio, initial pressure and incident laser energy, for both spherical and planar detonations.

Successful operation of future supersonic transport aircraft systems will require significant advances in combustion technology. The conditions of supersonic flight and high altitude impose severe strains on ignition reliability, stability, and overall combustion efficiency. Of particular interest is the attainment of combustion enhancement and relight at flight envelopes with low ambient temperature and pressure. This research program examines the feasibility of obtaining reliable relight of Jet A and other fuels by laser energy sources. Excimer laser pulses (193 and 248 nm) are used to photochemically dissociate fuel, oxidizer, or sensitizer molecules to produce chain-branching radicals, which initiate combustion.

A laser ignition system is proposed for the Combustion Experiment Module on an orbiting spacecraft. The results of a design study are given using the scheduled 'Flame Ball Experiment' as the design guidelines. Three laser ignition mechanisms and wavelengths are evaluated. A prototype laser is chosen and its specifications are given, followed by consideration of the beam optical arrangement, the ignition power requirement, the laser ignition system weight, size, reliability, and laser cooling and power consumption. Electromagnetic interference to the onboard electronics caused by the laser ignition process is discussed. Finally, ground tests are suggested.

New laser-based diagnostic techniques, developed primarily within the combustion community, offer considerable promise for measurements in reactive gaseous flows. In this paper we overview three diagnostic methods under development at Stanford University: spectrally resolved line-of-sight absorption (LOSA), conducted with wavelength-modulated semiconductor diode and ring dye laser sources; spectrally resolved single-point laser-induced fluorescence (SP LIF), conducted with a rapid-tuning ring dye laser; and planar laser-induced fluorescence (PLIF), conducted with a tunable pulsed dye laser source and intensified CCD array camera. These methods have unique capabilities for nonintrusive measurements of flow- field properties such as temperature, species concentration, velocity, density and pressure, as well as quantities derived from these properties such as mass flux (product of velocity and mass density). Species monitored in current studies include NO, OH, O₂ and H₂O.

Localized velocity, temperature, and species concentration measurements in rocket flow fields are needed to evaluate predictive computational fluid dynamics (CFD) codes and identify causes of poor rocket performance. Velocity, temperature, and total number density information have been successfully extracted from spectrally resolved Rayleigh scattering in the plume of small hydrogen/oxygen rockets. Light from a narrow band laser is scattered from the moving molecules with a Doppler shifted frequency. Two components of the velocity can be extracted by observing the scattered light from two directions. Thermal broadening of the scattered light provides a measure of the temperature, while the integrated scattering intensity is proportional to the number density. Spontaneous Raman scattering has been used to measure temperature and species concentration in similar plumes. Light from a dye laser is scattered by molecules in the rocket plume. Raman spectra scattered from major species are resolved by observing the inelastically scattered light with a linear array mounted to a spectrometer. Temperature and oxygen concentrations have been extracted by fitting a model function to the measured Raman spectrum. Results of measurements on small rockets mounted inside a high altitude chamber using both diagnostic techniques are reported.

Two-dimensional laser induced fluorescence of OH radicals is used to image flames in an optically accessible internal combustion engine. The two-dimensional plane is defined by forming the input XeCl excimer laser into a vertical sheet at the center of the combustion chamber. Nonconsecutive chemiluminescence and laser induced fluorescence images were observed at different delay times after the spark. The chemiluminescence images correlated well with the laser induced fluorescence images. A highly wrinkled turbulent flame in a tumble flow field as well as large cyclic variations in flame growth were observed in laser induced fluorescence images.

Velocity and dropsize measurements are reported for a swirl-stabilized, combusting spray. For the gas phase, three components of mean and fluctuating velocity are reported. For the droplets, three components of mean and fluctuating velocity, diameter, and number flux are reported. The liquid fuel utilized for all the tests was heptane. The fuel was injected using an air-assist atomizer. The combustor configuration consisted of a center-mounted, air-assist atomizer surrounded by a coflowing air stream. Both the coflow and the atomizing air streams were passed through 45 degree swirlers. The swirl was imparted to both streams in the same direction. The combustion occurred unconfined in stagnant surroundings. The nonintrusive measurements were obtained using a two-component phase/Doppler particle analyzer. The laser-based instrument measured two components of velocity as well as droplet size at a particular point. Gas phase measurements were obtained by seeding the air streams with nominal 1 micron size aluminum-oxide particles and using the measured velocity from that size to represent the gas phase velocity. The atomizing air, coflow air and ambient surroundings were all seeded with the aluminum-oxide particles to prevent biasing. Measurements are reported at an axial distance of 5 mm from the nozzle. Isothermal single- phase gas velocities are also reported for comparison with the combusting case.

The scope of this study is to obtain realistic comparisons between numerical calculations and laser Doppler velocimetry measurements of the axial velocity profiles for a large number of strained laminar premixed propane-air double flames. Stationary conservation equations for strained flames have been derived including one-parameter (strain rate) and two-parameter (radial pressure gradient and flow divergence) formulations. The experimental study consists in measuring the axial velocity across the flame by means of LDV system for stoichiometric, lean, rich and near extinction limit flames.

Measurements of degenerate four-wave mixing (DFWM) from NO₂ are presented as a function of buffer gas pressure and laser power. The signal strength first decreases and then increases with increasing buffer gas pressure. This dependence on buffer gas pressure is apparently due to the onset of thermal gratings. In the regime where thermal gratings are the dominant source of DFWM signals, the DFWM signal intensity increases as the square of the laser power until saturation.

A power spectral density estimation was performed to detect characteristic velocity fluctuations in the surrounding air flow field of a air-assisted spray flame. Measurements were carried out using randomly spaced in time velocity data obtained from a phase Doppler instrument. A fast algorithm enabled large numbers of data points to be analyzed in a relatively short time. An unbiased estimate was made of the power spectral density that considered the variable measurement times of individual realizations. The measured droplet arrival times from the phase Doppler particle sizer were used to determine the appropriateness of the Poisson distribution in describing droplet arrival times. The results indicate that the variance in droplet arrival statistics is larger than would be expected for a Poisson process. In addition, several positions in the spray have bimodal distributions which limits the utility of the Poisson process.

Rayleigh Scattering (RS) and Planar Laser Induced Predissociative Fluorescence (PLIPF) are used to obtain 2-D images of total and of state specific [i.e., of definite rotational levels of OH (v' equals 0) and O₂ (v' equals 2, 3, 6, or 7)] densities of the burned and unburned zones inside a combustion bomb. A tunable excimer laser, operated in either the 193 nm- or 248 nm-range, is used. Our bomb is about the size of an automobile engine cylinder. As each volume element of the fuel mixture burns, it forms high-T products with large specific volumes. Because the thermodynamics are well known, the bomb may serve as a reference device for diagnostics for high temperature species via pulsed-laser methods. The total densities and the mole fractions of the various constituents can be accurately calculated. Work has been done with various compositions of hydrogen-air and propane-air fuel mixtures.

Optical scattering by fuel droplets of spray jets can provide information about droplet evaporation in a combustion chamber. Evaporating fuel droplets are often spatially inhomogeneous and non-spherical. Spatial inhomogeneities in droplets can occur in the imaginary or in the real part of the index of refraction. We review some of the theoretical progress to date in modeling scattering by non-spheres and by inhomogeneous droplets. These models help investigate the effects of shape and refractive index perturbations on morphology dependent resonances of droplets. These investigations may lead to the use of such effects as tools for droplet diagnostics.

Combustion processes may involve a variety of particles in the micrometer size range. Such microparticles include coal macerals, and the processes can generate particulate matter via the production of flyash, soot, or condensables. In addition, sorbent microparticles can be added to a combustor such as the fluidized bed coal combustor to remove hydrogen sulfide and/or sulfur dioxide, the so-called dry-scrubbing process of desulfurization. In this study, inelastic light scattering and infrared laser heating techniques were developed and used to follow chemical changes in single microparticles levitated electrodynamically. The reactions between CaO and CuO sorbent particles with SO₂ were explored, coal macerals were characterized by inelastic scattering, and the effects of heating on a black carbonaceous microparticle were examined. Raman spectra obtained for CaO/Ca(OH)₂ particles levitated in a stream of oxygen and SO₂ show the formation of CaSO₃ and CaSO₄ upon heating. The uptake of water by the sorbent leads to fluorescence, which can mask the Raman spectrum. Single vitrinite and liptinite macerals are shown to have significantly different spectra, the latter being dominated by fluorescence. Heating and inelastic scattering measurements performed using fructose microspheres as a model carbonaceous blackbody show large changes in the particle due to pyrolysis. The apparatus, procedures and experimental results are presented together with a discussion of the problems and limitations associated with laser heating of levitated microparticles.

Single micron-sized spherical particles of carbon and aluminum oxide are suspended in an electrodynamic levitator equipped with a high vacuum system. A CO₂ laser is used to heat the particles while their infrared emission signals are monitored from 1.3 - 5.5 micrometers . At atmospheric pressure, particle cooling from air conduction is quite rapid, and the IR signal from a particle follows the 5 millisecond laser pulse. When the levitator is evacuated to pressures ranging from 1 X 10^-4 to 1 X 10^-5 torr, the efficiency of molecular impact cooling is greatly reduced, and radiative cooling becomes dominant. Under these conditions the particle temperature and emissivity can be found by theoretical fits of the radiative decay curves. For the carbon samples, our analysis has yielded emissivity values in good agreement with those calculated from Mie theory. One interesting feature of this agreement is the determination of emissivities greater than one for the smaller carbon particles. The emissivities found from the thermal decays of Al₂O₃ micro- particles are much larger than those predicted using Mie theory with bulk optical constants. These anomalously high emissivities for the alumina particles may be caused by surface contaminants.

The evaporated gas behind each flowing droplet affects the evaporation rate and drag of trailing droplets. For interacting droplets, we present a diagnostic technique that is capable of measuring evaporation-related droplet radius changes and drag-related flow velocity changes. When irradiated by a pump-laser beam, each dye-containing droplet acts as a laser, emitting at discrete wavelengths that correspond to morphology-dependent resonances of a sphere. Small wavelength shifts in the lasing spectra from each droplet are related to its radius change and hence, the decrease of the droplet volume. For an isolated droplet-stream segment, the evaporation rates of trailing droplets behind a lead droplet are determined.

Light can be trapped inside droplets by total internal reflection, and resonant conditions are known to exist for particular values of the radiation wavelength, droplet size, and refractive index. Such resonances have been observed previously in fluorescence from dye-doped droplets, and their presence or absence is controlled by the amount of droplet absorption at the resonance wavelength. These resonance peaks can be regarded as very sensitive absorption indicators, providing both quantitative and spectral information. Recent efforts have shown that the relative strength of resonance peaks in fluorescent emission can be modeled to give excellent agreement with known concentrations of rhodamine dye dissolved in ethanol droplets both for self-absorption and for absorption due to nigrosin additive, with sensitivities approaching 0.001 cm^-1. Using the model developed from these experiments, additional work with bromo-cresol green dye additives shows diagnostic application of the 'missing resonance spectroscopy' could provide new experimental information as a droplet concentration or pH sensor.

An electrodynamic levitator-trap (Paul trap) is used at line frequency (i.e. 60 Hz) as a sample cell for the long term microphotography of a microparticle in air at STP. Images are obtained in a trap modified to eliminate stray static fields at its AC 'null' point. Resolution in these long term images is found to be limited principally by stochastic thermal fluctuations and optical diffraction. A stochastic differential equation constructed for describing the particle's motion is found to be in good agreement with imaging experiments. This model provides an optimal limit to which a particle may be localized by increasing the drive potential, and indicates that this limit is a function principally of particle size and temperature. Images taken in fluorescence from a glycerol particle containing a probable surfactant are presented. Polarization resolution of these images clearly shows segregation of the molecule to the surface and identifies the orientation of the molecular emission moment in relation to the surface normal.

Resonance structure in elastic scattering (at approximately 90 deg from the direction of forward scattering) has been measured for evaporating micron-sized glycerol droplets suspended in an electrodynamic trap. Seeding the droplets with polystyrene latex particles having diameter 30 nm <EQ d <EQ 105 nm broadens and attenuates the highest Q (Q approximately equals 10⁴-10⁵) resonances. Further, regardless of whether resonance conditions are satisfied, the addition of latex particles causes fluctuations of the elastic scattering with amplitude up to approximately equals 10% of the total signal.

Stimulated Raman scattering (SRS) in laser-irradiated microdroplets is suppressed by the addition of nm-sized latex particles. The microdroplets consist of either pure ethanol or a solution of Rhodamine 6G dye in ethanol, seeded with latex particles having diameters 50 < d < 500 nm. SRS emission occurs at droplet morphology-dependent-resonances (MDR's) following either direct pumping by the incident 532 nm laser, or indirectly whereby the pump laser first initiates dye lasing which in turn pumps SRS. It is noteworthy that no Raman emission is observed from the corresponding bulk sample. For large latex, we observe SRS suppression at a near-coincident threshold concentration independent of the presence of dye; whereas, for small latex, adding dye reduces the threshold concentration by more than an order of magnitude. These findings are consistent with the interpretation that, for large latex, approximately 1 particle much occupy the MDR mode volume at threshold while for small latex, the addition of particles facilitates Forster-assisted annihilation of both 532 nm- and dye lasing MDR-pump photons.

A numerical technique is presented for the study of drops and particles in three-dimensional flows. Solutions have been obtained for uniform flow past drops, and for linear shear flows past particles and drops. The results indicate that, at low Reynolds (Re) and finite Weber (We) numbers, the drops deform to spherical cap shapes; as the Reynolds number is increased, the shapes become ellipsoidal. Coincident with the ellipsoidal shapes is the appearance of a detached recirculating eddy on the downstream side of the drop. For non-zero shear rates, the lift coefficient is approximately constant over a wide range of intermediate Reynolds numbers. At low Reynolds numbers and shear rates, the lift coefficient varies as Re^-1/2 for constant shear rate.

Most models involving absorption and scattering of light by particulate matter have relied on theories that treat the scatterers as homogeneous spheres or as spherical cores centered in spherical hosts. Aggregation and scavenging processes are commonly encountered in the study of aerosol particles, frequently resulting in situations where such spherically symmetric systems are unrealistic. A theoretical study of the absorption properties of carbon grains entrained eccentrically as spherical inhomogeneities in otherwise homogeneous droplets is presented. Although the focus of the calculations discussed herein is on single inclusions, a rigorous theoretical treatment is outlined for host spheres possessing multiple spherical inclusions.

A computer program is being developed for the theoretical analysis of the propagation of a laser pulse optically focused within an aerosol spray. The computer program can be applied, for example, to analyze laser ignition arrangements where a focused laser pulse would be used to ignite a liquid aerosol fuel spray. Laser light scattering and absorption of the individual aerosol droplets are evaluated using electromagnetic Lorenz-Mie theory. Initially, beam propagation is being modeled using a simple modified paraxial theory. Arbitrary input parameters to the computer program describing the optical/laser/aerosol spray arrangement include the liquid volume fraction, average droplet size, droplet size distribution, laser wavelength, laser pulse energy, laser pulse duration, lens focal length, beam diameter incident on the lens, and the choice of aerosol liquid and surrounding gaseous medium (through arbitrary inputted values of the thermodynamic and optical properties of the aerosol liquid and the gaseous medium). The output of the computer program includes, as a function of spatial position along the laser propagation axis within the spray, the laser pulse intensity and energy, the overall volumetric absorption of laser energy by the aerosol liquid and by the gaseous medium, and the overall average temperature rise of the aerosol liquid and of the gaseous medium.

Light scattering by radially inhomogeneous fuel droplets has been calculated using both geometrical optics (GO) and the exact separation of variables (SV) solutions. The refractive index profiles of the fuel droplets were those calculated by Kneer et al. The GO and SV solutions agree very well in the forward direction (for scattering angles between 30 and 60 degrees), and less well in the backward direction (for scattering angles between 140 and 170 degrees). Both amplitudes and phases of the scattered light are compared. The agreement in the backward direction is much better for 40 micrometers diameter droplets than for 20 micrometers diameter droplets.

The problem of scattering and absorption of electromagnetic radiation by particles can be solved analytically for only the simplest case, but established numerical methods allow a straightforward extension to particles with arbitrary homogeneities, arbitrary shapes, and nonlinear response. In this paper a recently developed frequency domain method involving CFD techniques is reviewed and applied to the problem of a dielectric coated sphere of arbitrary size parameter. Numerical results are in good agreement with analytical solutions. The problem is solved via the Debye amplitude formulation and then the field observables such as the heat generation are calculated. Results obtained suggest that finite element methods have promise for analytically intractable scattering/absorption problems.

Measurements of the scattered light intensity from individual droplets illuminated by focused Gaussian laser beams are presented. The results are compared with both the generalized Lorenz-Mie theory and classical Lorenz-Mie theory of elastic scattering. The data are shown to be consistent with the generalized theory. Besides experimental results, computations indicate that differences between the two theoretical approaches exist. It is shown that for sufficiently small beam diameter, characteristic of focused laser beam illumination, the departure of elastic scattering from that predicted by the classical Lorenz-Mie theory can be significant.

We have recently modeled third-order sum-frequency generation (TSFG) in droplets. The basic approach is similar to the model developed by Cooney and Gross for coherent anti- Stokes Raman scattering (CARS) from droplets. In this model, three generating waves interact to generate a third-order nonlinear polarization, which then radiates inside the sphere as described by the model of H. Chew et al. The intensity of the output waves at the sum frequency is proportional to the spatial overlap (amplitude and phase) of the nonlinear polarization with the output resonance of the droplet cavity mode, and to the integral of the products of the frequency dependence of the nonlinear polarization and the output resonance mode. Here we review our approach to modeling TSFG in droplets, discuss second-order sum frequency generation (SSFG) and CARS in droplets, stressing the similarities and differences among TSFG, SSFG, and CARS in droplets, and discuss the possible application of these mixing processes for fuel droplet characterization. We not that TSFG and SSFG from droplets are too weak to be useful for fuel droplet characterization, but that CARS is readily detectable from droplets and may be useful for determining the concentrations of chemical species in fuel droplets.