Quantifying extinction imaging of fuel sprays considering scattering errors

Abstract

In this work, we use the measurement technique of high-speed Diffuse Back Illumination Extinction Imaging (DBI-EI) to obtain quantitative information in the form of projected liquid volume (PLV) in a highly transient GDI process. For the DBI-EI setup we use a LED-Panel as the light source, which fulfills diffuse back illumination extinction imaging criteria. Measurements were carried out in a constant volume chamber, allowing easy optical access, and enabling measurements at real world ambient engine conditions. For the experiments, we use an Engine Combustion Network (ECN) Spray G injector and measure the sprays at ECN conditions. Moreover, we mount the injector in a motorized rotational system, enabling measurements of the sprays at precisely defined angles of observation. The DBI-EI technique requires a light source radiating uniformly in a certain range of an angle. Because of the diffuse radiation, an error in the quantification of the liquid phase results from the detection of multiple and forward scattered photons. This leads to an underestimation of the optical depth ([Formula: see text]), which further results in a false calculation of the projected liquid volume. Therefore, we must assume that DBI-EI results are wrong. To enable the use of DBI-EI in all spray regions independent of the measurement setup, we present a simulation-based method, which is correcting the [Formula: see text] for scattering effects. Results show, that the measured [Formula: see text] of the experimental setup, which we used in this work, is underestimated by at least a factor of 2.2. This factor increases with increasing spray densities. We can use the corresponding corrected PLV data to reconstruct three-dimensional data of the liquid volume fraction with the tomographic method filtered back projection. Thus, we obtain time and spatial resolved quantitative spray information, with an approach to correct undesired scattering effects, while keeping the experimental effort low.