OCR Text |
Show ~n Fig. 3. As will be shown in the next figure, this is a region where the flow is drawn radially mward by entrainment from the outer edge of the main air jet. Located along the centerline is the fourth major region of the flow-the central recirculation zone. This region shows a strong reverse flow with moderate fluctuations. Note that the recirculation extends well into the droplet biased region and comes to an end near the mean stagnation point at z = 25 mm. In order for this flow to come to a mean stagnation point, the re~erse flow from this region must be entrained at the inner edge of the fuel jet. This is the pnmary mechanism of stabilization in swirling flames: hot product gases are convected from downstream along the centerline of the flow to be entrained by, and to inflame, the fresh mixture issuing from the nozzle. ~t is interesting that a negative mean velocity is measured in the internal recirculation zone even In ~e droplet biased region. One might expect that the droplets in this region had flown off of the Inner edge of the spray sheath and would therefore exhibit a positive axial velocity. The widths of the distributions at z = 50 mm are large enough relative to the mean that the tails of the distrit;>~tions do extend to positive velocities. The question that remains is whether the positive veloclues measured are due to droplets penetrating the reverse flow of the gas phase or whether the gas flow (with embedded droplets) alternates between flow in the forward and reverse direction at some time interval. As shown in Fig. 3, a luminous flame can be observed in this region. This flame can be seen to bob up and down at a frequency of order I-10Hz. If this luminous zone is interpreted as marking the penetration depth of the internal recirculation zone, then this bobbing would lend support to the "flow alternating in time" interpretation of the data. The fifth region of the flow is the wake region fonned by the bluff-body action of the nozzle between the fuel jet and the main air jet This region is interesting only in that it is responsible for the occurrence of intennittent reverse flow along the mixing boundary between the air and fuel jets. Note that although the data at z = 25 mm, r = 30 mm shows a slight positive mean velocity, the distribution extends well into negative velocities. The same is true at z = 50 mm and 75 mm. It is this intermittent reverse flow that accounts for the swirling flamelets which appear on the outer edge of the spray sheath, as mentioned above in connection with Fig. 3. It is interesting to note that under other operating conditions (Edwards, 1988) these flamelets (and presumably this mixing layer reverse flow) are the only visible means of stabilization. The last major region of the flow is just outside of the internal recirculation zone where the fuel spray and main air jets no longer exhibit distinct axial velocities (z = 125 mm, r = 70 mm). In this region the two jets have achieved a sufficient degree of mixing that combustion intensity (as observed by luminosity) increases dramatically. Note in the figure that the side-by-side narrow range of fluctuations of the air jet and broad fluctuations of the fuel jet have been replaced by a wide region where the fluctuations are of the same order as the mean. Figure 6 shows the radial velocity component of the flow. Velocity vectors emanating from the centerline outward are shown as positive (data above the baseline) while those pointed inward are shown as negative (data below the baseline). Data for the fuel jet and main air regions appear very similar to the axial velocity data-in both cases the flow is predominantly in the positive direction and fluctuations are largest in the fuel jet. The radial velocity entrance profile shows that the flow is purely axial (in the r-z plane) across most of the throat until just adjacent to the nozzle. Here entrainment by the fuel jet induces a radially inward velocity. The external recirculation region (outside of the main air jet) shows a slight negative mean radial velocity with moderate fluctuations. This is the region where the axial velocity was zero, leading to the conclusion that the flow is being slowly drawn in the radial direction by entrainment at the outer edge of the main air jet. - 7 - |