Impact of polarization and detection angle on Mie scattering signals for planar droplet sizing

The planar droplet sizing (PDS) method for determining the droplet diameter from the ratio of laser-induced fluorescence (LIF) and Mie scattering signal (Mie) has been extensively applied for right-angle scattering detection. The present work consists of a parametrical study of the Mie signal dependency on the scattering angle and polarization, as well as further experimental investigations in single droplets to ultimately improve the intensity fluctuation characteristics of the signal through a favorable experimental arrangement. Finally, we investigated how these findings impact droplet sizing in a spray on a shot-to-shot basis, as well as averaged across a dataset, via microscopic and macroscopic spray measurements. In the parametrical study, we found that horizontally polarized incident light results in a higher degree of Mie signal fluctuation than vertically polarized incident light at a scattering angle of \(\Theta ={90}{^\circ }\) $\mathrm{\Theta }=90_{}^{\circ }$ , whereas the reduction to \(\Theta = {60}{^\circ }\) $\mathrm{\Theta }=60_{}^{\circ }$ resulted in the exact opposite observation. Furthermore, the reduction from \(\Theta = {90}{^\circ }\) $\mathrm{\Theta }=90_{}^{\circ }$ to \(\Theta = {60}{^\circ }\) $\mathrm{\Theta }=60_{}^{\circ }$ resulted in an overall improvement to the \(D^2\) $D^2$ -dependency and signal fluctuation of the Mie signal. We observed the effects shown in the parametric study in the experimental droplet measurements. For instance, when observing single droplets, the change from \(\Theta = {90}{^\circ }\) $\mathrm{\Theta }=90_{}^{\circ }$ to \(\Theta = {60}{^\circ }\) $\mathrm{\Theta }=60_{}^{\circ }$ resulted in a reduction in signal fluctuation by 46.1% and 17.6% for horizontally and vertically polarized incident light, respectively. An implementation into a spray setup yielded reduced shot-to-shot sizing deviations and closer adherence to the microscopically measured diameter when averaged across the dataset.