Douglas A. Feikema

Blowout of nonpremixed flames; Maximum coaxial air velocities achievable, with and without swirl

The present study demonstrates how to optimize parameters in order to maximize the amount of coaxial air that can be provided to a nonpremixed jet flame without causing the flame to blow out. Maximizing the coaxial air velocity is important in the effort to reduce the flame length and the oxides of nitrogen emitted from gas turbines and industrial burners, a majority of which use coaxial air. Previous measurements by the latter two authors have shown that a sixfold reduction in the NO x emission index of a jet flame is possible if sufficient coaxial air can be provided without blowing the flame out. The coaxial air shortens the flame and forces the reaction zone to overlap regions of higher gas velocity, which reduces the residence time for NO x formation. The present work concentrates on demonstrating ways to prevent flame blowout when the following two constraints are imposed: (1) the coaxial air velocities must be sufficient to shorten the flame to a specified length (in order to reduce NO x emissions) and (2) the coaxial air flow rate must be sufficient to complete combustion without the need for ambient air, which is a common practical constraint. The zero swirl case is considered first, and the effects of adding swirl are measured and directly compared. The following were systematically varied: fuel velocity, air velocity, fuel tube diameter, air tube diameter, fuel type, and swirl number. Measurements demonstrate that coaxial air alone (with zero swirl) can cause up to a twofold reduction in flame length. However, the flame is stable only if the velocity-to-diameter ratio of the fuel jet does not exceed a critical value. It is found that the addition of swirl improves the maximum-air blowout limits by as much as a factor of 6. The results identify a strain parameter, based on the ratio of air velocity to air tube diameter (U A/dA), which collapses the blowout curves for ten different conditions (burner size, swirl number) approximately to a single curve. A physical mechanism that explains the swirl flame data is presented. Swirl is believed to be beneficial because it reduces the local velocities, and thus the local strain rates, near the forward stagnation point of the recirculation vortex, where the flame is stabilized.

Images of the strained flammable layer used to study the liftoff of turbulent jet flames

Images were obtained to visualize the flammable layer upstream of a lifted turbulent, non-premixed jet flame in order to (a) quantify the degree of premixing upstream of the flame base, (b) visualize the path along which the flame base can propagate, and (c) show how scalar dissipation layers, which identify high strain regions, overlap the flammable layer. It is found that a significant amount of premixing occurs at x/d =5 that creates a continuous, contorted flammable layer, appearing as thin as 70 microns or as thick as 3 mm, depending on the local strain. The structure of many larger wrinkles in the flammable layer is consistent with the structure predicted by the flame-vortex simulation of Ashurst and Williams: a vortex that moves radially outward tends to cause extensive strain and a thinning of the layer at the largest radial locations. Upstream of the vortex, compressive strain causes a local broadening of the flammable layer, as predicted. Where the dissipation layers are observed to overlap the flammable layer, the dissipation rate is larger than that required to extinguish either a premixed or a non-premixed strained flame. The dissipation layers have a distinet structure in the jet near field—they are aligned at approximately 45° to the jet axis, unlike the isotropically aligned dissipation layers downstream.

Quantitative rainbow schlieren deflectometry as a temperature diagnostic for nonsooting spherical flames

Numerical analysis and experimental results are presented to define a method for quantitatively measuring the temperature distribution of a spherical diffusion flame using rainbow schlieren deflectometry in microgravity. The method employed illustrates the necessary steps for the preliminary design of a rainbow schlieren system. The largest deflection for the normal gravity flame considered in this paper is 7.4 x 10(-4) rad, which can be accurately measured with 2 m focal-length collimating and decollimating optics. The experimental uncertainty of deflection is less than 5 x 10-(5) rad.

Analysis of the Laser Propelled Lightcraft Vehicle

Advanced propulsion research and technology require launch and space flight technologies, which can drastically reduce mission costs. Laser propulsion is a concept in which energy of a thrust producing reaction mass is supplied via beamed energy from an off-board power source. A variety of laser/beamed energy concepts were theoretically and experimentally investigated since the early 1970s. During the 1980s the Strategic Defense Initiative (SDI) research lead to the invention of the Laser Lightcraft concept. Based upon the Laser Lightcraft concept, the U.S. Air Force and NASA have jointly set out to develop technologies required for launching small payloads into Low Earth Orbit (LEO) for a cost of

Enhancement of flame blowout limits by the use of swirl

1.0M or

Images of Dissipation Layers to Quantify Mixing Within a Turbulent Jet

1000/lb to

Markstein Numbers of Negatively-Stretched Premixed Flames: Microgravity Measurements and Computations

100/lb. The near term objectives are to demonstrate technologies and capabilities essential for a future earth to orbit launch capability. Laser propulsion offers the advantages of both high thrust and good specific impulse, I(sub sp), in excess of 1000 s. Other advantages are the simplicity and reliability of the engine because of few moving parts, simpler propellant feed system, and high specific impulse. Major limitations of this approach are the laser power available, absorption and distortion of the pulsed laser beam through the atmosphere, and coupling laser power into thrust throughout the flight envelope, The objective of this paper is to assist efforts towards optimizing the performance of the laser engine. In order to accomplish this goal (1) defocusing of the primary optic was investigated using optical ray tracing and (2), time dependent calculations were conducted of the optically induced blast wave to predict pressure and temperature in the vicinity of the cowl. Defocusing of the primary parabolic reflector causes blurring and reduction in the intensity of the laser ignition site on the cowl. However, because of the caustic effect of ray-tracing optics the laser radiation still forms a well-defined ignition line on the cowl. The blast wave calculations show reasonable agreement with previously published calculations and recent detailed CFD computations.

Flame-Vortex Studies to Quantify Markstein Numbers Needed to Model Flame Extinction Limits

The blowout limits of a number of swirl stabilized flames were measured and the trends are explained by applying the concepts proposed in recent flame blowout theories, which previously have been applied only to non-swirling flames. It is shown that swirl flame blowout limits can be compared to wellknown limits for non-swirling simple diffusion flames by using the proper nondimensional parameter, i.e., the inverse Damkohler number ( U F ~ d F ) ~ ( S ~ 2 / a ) . Unlike most previous work. four parameters were systematically varied: the fuel tube diameter (dF), the fuel type and thus reaction rate, which is related to the maximum laminar burning velocity (SL), the coaxial air velocity (UA), and the swirl number. Results show that the maximum fuel velocity (UF) and thus the maximum heat release rate for a swirl flame is as much as four times larger than that for a non-swirling flame. Blowout velocity (UF) increases with burner size (dFj and laminar burning velocity squared (SL2); this is similar to non-swirling flames except that a new parameter that includes swirl number must be added. The major reason why swirl increases the stability of a flame is because of a flame-vortex interaction. The toroidal recirculation vortex reduces the centerline velocity below that of a non-swirling flame and the analysis shows that this is strongly stabilizing.