| OCR Text |
Show number flames because of the greater sensitivity of the photographic film to the yellow flickering part of the flame relative to the blue cone. NOx emission indices were calulated as shown in Eqn. 2, where the molecular weight of NOx is assumed to be that of N02, following past practices (Turns & Myhr, 1991; Turns et al., 1993): MN0 2 XNOx .103 (12.01 + 1.008Y)(XC02 - XC02._ +XCO) (2) where Xi is the mole fraction of species i, 00 signifies the background concentration, and y is the molar ratio of hydrogen to carbon in the fuel. The calculation of CO emission indices has been described by Turns & Bandaru (1993). Test Conditions Tests were conducted using methane (> 97% purity) and propane (> 96% purity) fuels. Motor-driven nozzles having 10-mm or 3-mm exit diameters were employed. Nozzles with deflection angles, $, of 30°,45°, and 60° were employed with the 10-mm nozzles, while a single deflection angle ($ = 45°) was used with the 3-mm nozzle. The motor-driven mechanical nozzle has a maximum rotational speed of 7,000 rpm, which limits the range of operating conditions. Table 1 shows test conditions and the maximum achievable Strouhal and Reynolds numbers. RESULTS AND DISCUSSION Visual Observations Cold flow measurements (Schneider et al., 1992, 1995) have shown that it is possible to obtain a wide range of turbulent flowfields by varying Stp in a precessing jet flow. That the nature of the flowfield is important to flame structure is apparent from the visual appearance of the flames at different conditions, as illustrated in Figs. 3 and 4. At low precession Strouhal numbers, Stp ~ 0.005, the flame does not radiate strongly in the visible spectrum. The flame resembles an inverted cone, whose half-angle corresponds to the deflection angle of the nozzle; the flame is blue in color and is highly turbulent (Fig. 3). Under some conditions in this regime, weak yellow flamelets, 6 |
