Abstract
<div class="section abstract"><div class="htmlview paragraph">The distribution of spray droplet sizes plays a pivotal role in internal combustion engines, directly affecting fuel-air mixing, evaporation, and combustion. To gain a precise understanding of droplet size distribution in a two-dimensional space, non-intrusive optical diagnostics emerge as a highly effective method. In the current investigation, two-dimensional (2D) diesel spray droplet sizes mapping using a simultaneous combination of planar laser-induced fluorescence (PLIF) and Mie-scattering techniques is introduced. The assessment of droplet diameter relies on the interplay between fluorescent and scattered light intensities which correspond the light based on volumetric droplets and surface area of the droplets. This calculation is made possible through the LIF/Mie technique. However, traditional LIF/Mie methods are plagued by inaccuracies arising from multiple light scattering. To overcome this challenge and to attain higher accuracy than conventional LIF/Mie technique, we introduce a sparsity deconvolution approach to eliminate unwanted light interference on both LIF and Mie images. The core concept of sparsity deconvolution is to reduce disturbances caused by multiple scattering and offer sharp and finely detailed images for LIF/Mie ratio estimation. To enhance spatial sharpness and remove the undesired scattering light, an iterative Richardson–Lucy (RL) and Land Weber (LW) filters are introduced for image deconvolution. The results reveal that RL deconvolution is particularly well-suited for the intricate task of deconvolving complex liquid sprays, producing sharper and finer detailed droplet images. Additionally, the further calibration of 2D droplet size mapping based on microscopic method is implemented to approximate the linear fitting curve of dependence between macro LIF/Mie ratio and droplet diameter. This comprehensive approach advances the understanding of the critical role played by droplet size distribution under engine-like conditions.</div></div>