Determination of spray droplet size by wavelet analysis of interferometric images

Abstract

The problem of size droplet determining by the Interferometric Particle Imaging (IPI) in a spray of an optically transparent liquid is considered. The measurements were carried out using the example of a fuel injector flame. A new method for analyzing the IPI image of a droplet using a continuous wavelet transform is proposed. The spatial distribution of wavelet coefficients is used to introduce an integral measure similar to the spectral density of the signal. As a reliable estimate of the droplet size, we considered the results by the Glare Point Technique (GPT), which were applied simultaneously. A comparative analysis of the sizes obtained using GPT and IPI showed that the application of the wavelet transform to IPI images, in comparison with the Fourier method, significantly reduces the frequency of occurrence of a systematic error in determining the size, which is most specific for large droplets. It was found that the false maxima in the Fourier spectrum causing these errors are due to the appearance of distortions at the boundary of the interferometric image of the droplet. The wavelet transform allows one to separate these distortions from the main maximum in the vicinity of the droplet center.