DROP DISTRIBUTION EFFECTS ON PLANAR LASER IMAGING OF SPRAYS

Publisher Website
Google Scholar
Abstract
A study of the effects of drop distribution on the interpretation of light-scattering polarization ratio measurements for characterizing mean drop diameter is presented. The work focuses on the appropriateness of logarithmic-normal, Rosin-Rammler, Nukiyama-Tanasawa, and upper-limit distributions in representing the drop size distribution. Comparisons between the individual drop distributions are made on the basis of fits to experimental measurements made in a pressure-atomized water spray using the Phase Doppler Particle Analyzer (PDPA). The results from the drop size distribution comparisons indicate that care must be exercised when fitting number or volume distributions, since the diameter-weighting differences in fitting diameter number or volume distributions can significantly alter the best-fit results. In general, good fits to either diameter number or volume distributions can be obtained with the previously mentioned distributions. However, complementary transformations between the diameter number and volume distributions generally resulted in unacceptable fits. Furthermore, number density predictions were typically less satisfactory for the Rosin-Rammler distribution. Both the logarithmic-normal and Nukiyama-Tanasawa distributions were applied to the polarization ratio technique analysis for measurement of the Sauter mean diameter. Differences observed between the PDPA and polarization ratio results could not be resolved through more accurate representation of drop size distribution. In fact, direct Mie scattering calculations using the measured size distributions showed discrepancies with the observed polarization ratio measurements from the experiment. These observations eliminate size distribution as the only source of inaccuracy in the polarization ratio technique and point to examination of other light-scattering-related effects such as technique size biases.

This publication has 0 references indexed in Scilit: