Drop size distribution measured by imaging: determination of the measurement volume by the calibration of the point spread function

Abstract

Measurement of drop size distributions in a spray depends on the definition of the control volume for drop counting. For image-based techniques, this implies the definition of a depth-of-field (DOF) criterion. A sizing procedure based on an imaging model and associated with a calibration procedure is presented. Relations between image parameters and object properties are used to provide a measure of the size of the droplets, whatever the distance from the in-focus plane. A DOF criterion independent of the size of the drops and based on the determination of the width of the point spread function (PSF) is proposed. It allows to extend the measurement volume to defocused droplets and, due to the calibration of the PSF, to clearly define the depth of the measurement volume. Calibrated opaque discs, calibrated pinholes and an optical edge are used for this calibration. A comparison of the technique with a phase Doppler particle analyser and a laser diffraction granulometer is performed on an application to an industrial spray. Good agreement is found between the techniques when particular care is given to the sampling of droplets. The determination of the measurement volume is used to determine the drop concentration in the spray and the maximum drop concentration that imaging can support.

Experimental characterisation of sprays resulting from impacts of liquid-containing projectiles

“…An advanced Spoke filter (from Minor & Sklansky 1981) was developed to locate the circular-like droplets based on the greyscale gradient directions along the droplet edge pixels. The region of each droplet image is analysed with a point-spread function approach presented by Fdida & Blaisot 2010. The detected in-focus droplets are tracked over time with an iterative version of the bestestimate tracking algorithm (Ouellette & Bodenschatz 2006), which finds the best match for each droplet in three consecutive image frames.…”

An improved image processing method for particle characterization by shadowgraphy

“…Correction to these out-of-focus particles is not included in this study. Fdida and Blaisot proposed a method to correct the out-of-focus particles or to define a depth-of-field thickness where all particles are focused, based on the point spread function [8]. These methods will be implemented and investigated in the next stage of this project.…”

“…The factor of 0.61 is proposed by [8]. This value is the one which allows obtaining, based on point spread function, a polynomial relationship between contrast of particle image and measured particle diameter.…”

Flow characteristics of a large-size pressure-atomized spray using DTV

“…In several recent papers, DTV calibration is often performed with a standard calibration plate ( Nishino et al (20 0 0) ; Fdida and Blaisot (2010) ; Castanet et al (2013) ; Sinha et al (2015) ) and validated on a mono-dispersed droplet generator ( Anand et al, 2012 ). In our work, this initial calibration step is performed with calibrated glass beads.…”