OCR Text |
Show is provided by Figure 13 which shows the variation of the cumulative number density above 3 microns as a function of reactor position or increased residence time for all slurries. (Although the syringe drive used to feed the slurries is set at a given feedrate of 1 ml/min, we found that the total mass flowrate exiting the slot varied by up to a factor of 2 as measured by the vacuum sample system.) For a better comparison between the slurries we have normalized the total countrate for each slurry to correspond to the mass feedrate of slurry #2. In general, the total number density is decreasing with increased residence time. If we assume that particle diffusion is negligible, this is a significant result since it shows that the particles are not fragmenting as they heat up and react in the high temperature flame region. In contrast to a total number density decline and independent of whether particle diffusion is important or not, we consistently observe an increase in the largest particle number density due to particle swelling. Particle densities are sufficiently dilute that particle growth by collision is improbable. To better quantify the number density decrease, we determined the effective slot width at various axial positions within the reactor by measuring the total particle count rate above 3 microns at the center of the slot, and then determined the width at which the particle count rate was half the peak value. If particle diffusion were significant, the halfwidth would increase with increasing reactor position. Surprisingly, the number count half-width varied by less than 25% up to 8 cm height (35 milliseconds) in the reactor and, in fact, decreased in the mid-region of the reactor. As shown in Figure 13, the total particle number density above 3 microns at the center of the slot decreases by a factor of 5-10 at 30 milliseconds. Thus, we conclude that the drop in number density with reactor height is due primarily to particle size reduction by reaction, and that particle fragmentation is negligible for the conditions studied here. Particle volume frequency distributions shown in Figure 14 for slurry #1 give further information on the physical behavior of the slurry particles. The volume frequency distributions show a similar behavior with increased residence time as described for the corresponding number distributions of Figure 9. Note that in contrast to the number distributions, volume weighting reverses the emphasis to the largest particles in the system. Instruments based on volume measurements alone will not provide accurate measurements of low percentage-volume particles even though the number density may be high. In practice, information on both number and volume is required to provide a complete picture of size distribution effects. These volume distributions were obtained by assuming spherical particles and integrating the volume-weighted number distributions. Assuming that the slurry particles have a density of 1.2 g/cc, and that the flux of particles across the reactor slot is uniform, a total mass flow of slurry particles can be obtained from the optical measurements. For cold flow conditions, we have used a vacuum-filter system to collect the total flow of particles exiting the reactor slot for a known period of time to obtain an independent measure of the total mass flowrate. Comparison of the results for various sets of conditions are shown in Table II and indicate satisfactory agreement between the two measurements. The mass measurements are a very sensitive measure of both the size and number |