An optical soot sizing method without prior morphology parameters

Abstract

Great concern has been focused on the particle size distribution (PSD) of soot as a crucial factor of the physical and chemical properties. Optical methods are widely used to analyze the PSD of soot due to their non-invasive and real-time measurement, where the morphology parameters are essential for retrieving the PSD from optical signals. However, these parameters are difficult to obtain in field measurement. In this work, we propose a new PSD sensing method for soot without prior knowledge of the morphology parameters based on the light-scattering angular spectrum using a hybrid inverse algorithm. The light scattering angular spectrum is the intensity distribution of scattering light at different observation angles for simultaneously characterizing the PSD and the morphological parameters, which are decoupled by a hybrid inverse algorithm consisting of Tikhonov regularization and adaptive weighted particle swarm optimization (AW-PSO). A compacted and lens-free prototype sensor is designed to verify our method. The calibration experiments by spherical aerosols show that our prototype sensor provides high-precision light scattering angular spectrum measurement with maximum relative error (RE) of 14.78 %. In the tests of soot, the maximum Kullback-Leibler Divergence (DKL) of PSD is 0.28 in comparison with the reference volume equivalent PSD. The sensing method proposed in this paper provides volume equivalent PSD measurement for soot without requirement of the morphology parameters, and our prototype sensor shows great potential for soot and other non-spherical aerosol analysis in routine field measurements outside the laboratory.