Effects of positive flame stretch on the laminar burning velocities of hydrocarbon/air mixtures were studied experimentally using outwardly propagating spherical flames. The test conditions included propane, methane, ethane, and ethylene-air flames at various fuel-equivalence ratios and normal temperature and pressure. Karlovitz numbers generally were less than 0.3 so that the flames were remote from quenching conditions. Within this range, the ratio of the unstretched (plane flames) to stretched laminar burning velocities varied linearly with Karlovitz numbers, yielding Markstein numbers that were independent of Karlovitz numbers for a particular reactant mixture. In addition, Markstein numbers varied in a roughly linear manner with fuel-equivalence ratios over the range of the measurements, which were somewhat removed from flammability limits where behavior might differ. Effects of stretch were substantial: Markstein numbers varied from -2.5 to 7.2, yielding corresponding laminar burning velocity variations of 0.4-2,7 times the value for an unstretched (plane) flame over the test range. The ranges of fuel-equivalence ratios for unstable preferential-diffusion conditions (negative Markstein numbers) were as follows: propane, greater than 1.44; methane, less than 0.74; ethane, greater than 1.68; and ethylene, greater than 1.95. Fuel-equivalence ratios for maximum flame temperatures and laminar burning velocities are near unity for the present flames; therefore, neutral preferential-diffusion conditions are shifted toward fuel-equivalence ratios on the unstable side of unity, in qualitative agreement with recent approximate theories treating the effects of stretch on laminar premixed flames.

Effects of positive flame stretch on laminar burning velocities, and conditions for transition to unstable flames, were studied experimentally for freely propagating spherical flames at both stable and unstable preferential-diffusion conditions. The data base involved new measurements for H2/O2/N 2 mixtures at values of flame stretch up to 7600 s-1, and existing measurements for C3Hs/O2/N 2 mixtures at values of flame stretch up to 900 s-1. Laminar burning velocities varied linearly with increasing Karlovitz numbers-either decreasing or increasing at stable or unstable preferential-diffusion conditions-yielding Markstein numbers that primarily varied with the fuel-equivalence ratio. Neutral preferential-diffusion conditions, however, were shifted toward the unstable side of the maximum laminar burning velocity condition that the simplest preferential-diffusion theories associate with neutral stability. All flames exhibited transition to unstable flames: unstable preferential-diffusion conditions yielded early transition to irregular flame surfaces, and stable preferential-diffusion conditions yielded delayed transition to cellular flames by hydrodynamic instability. Conditions for hydrodynamic instability transitions for H2/O2/N 2 mixtures were consistent with an earlier correlation due to Groff for propane/air flames, based on the predictions of Istratov and Librovich.

An experimental study of primary breakup of turbulent liquids in gadliquid mixing layers is described. The experiments involved mixing layers along large liquid jets (3.6.6.4 and 9.5 mm dia.) injected at various velocities into still air at atmospheric pressure with fully-developed turbulent pipe flow at the jet exit. Liquids studied included water, glycerol (42% glycerin by mass) and n-heptane. Pulsed shadowgraph photography and holography were used to find conditions where turbulent primary breakup was initiated and drop sizes and velocities after primary breakup. Drop sizes after primary breakup satisfied Simmons' universal root normal distribution and can be characterized solely by their SMD. Mass weighted mean streamwise and crosstream drop velocities after primary breakup were comparable to mean streamwise and crossweam rms fluctuating velocities in the liquid, respectively, with effects of mean velocity distributions in the jet passage reflected by somewhat lower streamwise drop velocities near the jet exit. Conditions for the initiation of turbulent primary breakup and the variation of SMD with distance from the jet exit were correlated reasonably well by a phenomenological analyses considering effects of surface tension and liquid turbulence properties alone.

An experimental study of the structure of round buoyant turbulent plumes was carried out, emphasizing conditions in the fully developed (self-preserving) portion of the flow. Plume conditions were simulated using dense gas sources (carbon) dioxide and sulfur hexafluoride) in a still air environment. Mean and fluctuating mixture fraction properties were measured using single-and two-point laser-induced iodine fluorescence. The present measurements extended farther from the source (up to 151 source diameters) than most earlier measurements (up to 62 source diameters) and indicated that self-preserving turbulent plumes are narrower, with larger mean and fluctuating mixture fractions (when appropriately scaled) near the axis, than previously thought. Other mixture fraction measurements reported include probability density functions, temporal power spectra, radial spatial correlations and temporal and spatial integral scales.

An experimental study of the structure of round buoyant turbulent plumes was carried out, limited to conditions within the fully developed (self-preserving) portion of the flow. Plume conditions were simulated using dense gas sources (carbon dioxide and sulfur hexafluoride) in a still air environment. Velocity statistics were measured using laser velocimetry in order to supplement earlier measurements of mixture fraction statistics using laser-induced iodine fluorescence. Similar to the earlier observations of mixture fraction statistics, self-preserving behavior was observed for velocity statistics over the present test range (87–151 source diameters and 12–43 Morton length scales from the source), which was farther from the source than most earlier measurements. Additionally, the new measurements indicated that self-preserving plumes are narrower, with larger mean streamwise velocities near the axis (when appropriately scaled) and with smaller entrainment rates, than previously thought. Velocity statistics reported include mean and fluctuating velocities, temporal power spectra, temporal and spatial integral scales, and Reynolds stresses.

An experimental study of the structure of round buoyant turbulent plumes was carried out, limited to conditions in the self-preserving portion of the flow. Plume conditions were simulated using dense gas sources (carbon dioxide and sulfur hexafluoride) in a still and unstratified air environment. Velocity/mixture-fraction statistics, and other higher-order turbulence quantities, were measured using laser velocimetry and laser-induced fluorescence. Similar to earlier observations of these plumes, self-preserving behavior of all properties was observed for the present test range, which involved streamwise distances of 87–151 source diameters and 12–43 Morton length scales from the source. Streamwise turbulent fluxes of mass and momentum exhibited countergradient diffusion near the edge of the flow, although the much more significant radial fluxes of these properties satisfied gradient diffusion in the normal manner. The turbulent Prandtl/Schmidt number, the ratio of time scales characterizing velocity and mixture function fluctuations and the coefficient of the radial gradient diffusion approximation for Reynolds stress, all exhibited significant variations across the flow rather than remaining constant as prescribed by simple turbulence models. Fourth moments of velocity and velocity/mixture fraction fluctuations generally satisfied the quasi-Gaussian approximation. Consideration of budgets of turbulence quantities provided information about kinetic energy and scalar variance dissipation rates, and also indicated that the source of large mixture fraction fluctuations near the axis of these flows involves interactions between large streamwise turbulent mass fluxes and the rapid decay of mean mixture fractions in the streamwise direction.

No abstract

No abstract