OCR Text |
Show Flowfield Structure Figure 4 illustrates the method of presentation of the data showing the axial velocity component as a function of radius at seven axial stations in the flame. The axial stations are located at 25 mm intervals starting at the entrance plane of the furnace (z = 0 mm). Data were obtained as a function of radial position and are shown by the diamonds in the figure. Data are only presented for a single angular coordinate and have been mirrored about the furnace centerline to aid in comparison with Fig. 3. Data taken across the full diameter of the region indicates that some asymmetry does exist, but it is not sufficient to alter the physical structure depicted in the figures. The magnitude of the velocity at each axial station is indicated by the displacement of the data from a baseline at each station (the horizontal lines). The axis to the left of the figure shows the magnitude of velocity which corresponds to the height of the adjacent vertical bar. In this figure, for example, an axial velocity of 20 mls is shown by displacement of the data by 20 mm (in z) from its baseline. Cubic spline fits have been made through the data as shown by the solid lines in the figure. In subsequent figures, only the data fits will be shown since inclusion of the individual data points quickly obscures the more complex plots. The correspondence between the fits and the data shown in Figure 4 is typical of the quality of the fits for the remaining data. The hatched region in this and subsequent figures indicates those regions of the flow where the velocities are biased by the presence of a high number density of droplets. As shown by Durst and Ruck (1987), the conventional laser doppler velocimeter is optimized for detection over a relatively small range of particle sizes. In the absence of significant sources of noise or laser beam extinction, the laser-LOY -seed system can be optimally coordinated to provide data from those seed particles sized to follow the flow. This can be accomplished because of the nearlymonotonic relationships of scattering intensity and signal visibility with particle size (resonances neglected) when collection optics, laser intensity and background noise are constant. In the present case these conditions are not met: laser intensity varies as a result of beam extinction by soot, the coincidence of the beams varies at the imaged probe volume as a result of beam steering, and the flame luminosity and scattering of the laser beams into the collection optics (by soot) cause sizable variations in the background noise level of the detected signals. The result is that even with a system tuned for small particles that follow the flow, larger particles which are ballistic relative to the flow may still be validated. For this reason the data in the spray region is labeled as "biased" and indicates some blending of validations from particles which do and do not follow the gas phase flow. As such, this data represents neither the gas phase velocity nor the true mean spray velocity, but is a mixture of both. Data from this region are presented only because they provide some indication as to the velocity characteristics of the two-phase flow in this region. Spatial extent of the biased region was determined by measuring the data validation rate without seeding the flow. Any region where the data rate without seeding exceeded 100Hz was considered to be biased. Since the minimum seeded data rate used in this study was 1 kHz, this means that in order for a region to be considered unbiased, there must be at least one order of magnitude more validations from seed particles than droplets. In fact, the percentage of validations due to seed particles is probably higher since many of the validations in the nonseeded case were in fact due to seed particles which had remained in the furnace from previously seeded operation. This was confirmed by observation of a significant data rate in the hot furnace gases long after the spray had be shut off. It should be noted that during operation in one of these biased regions the LOY system tuning was not altered from that used in the seed-only parts of the flow (Le. the system was kept optimized for small particles). This resulted in a very high rejection rate by virtue of the amplitude limit setting when operating in the spray. - 5 - |