Precision in Spray Dynamics: Combining Advanced Imaging and Phase Doppler Anemometry
Zuhaib Nissar, Oyuna Rybdylova, Guillaume De Sercey, Steven M Begg
University of Brighton, Brighton, United Kingdom
DOI:
Atomisation in sprays exhibits a stochastic nature, yet it underpins the functionality of technologies reliant on sprays. To accommodate this inherent randomness, it is important to assess all possible outcomes of droplet sizes and velocities at each location, which is crucial for advancing spray-based applications. Phase Doppler Anemometry (PDA) is typically employed to measure these properties, though its use is largely restricted to the far-field, dilute region of a spray where droplets are generally smaller and spherical compared to those in the breakup zone. A new approach involving high-resolution, high-speed, microscopic imaging was developed to complement PDA measurements. Firstly, in this method, a composite image of the spray was constructed by translating the small field of view across the entire spray region in steps, resulting in a set of non-overlapping images. Along the boundaries, the artefacts due to composition were taken care of by utilising four grids, each offset vertically and/or horizontally by half the height and/or width of the field of view to reconstruct data along the ’stitches’. Secondly, the image processing algorithm enabled identification of perforations in the liquid sheet, breakup region and local droplet size and velocity distributions in the spray. In total, 1,200 composite images were produced for each grid at a frame rate of 84 kHz. Droplet sizes greater than 10 μm could be measured, with PDA extending the measurable range down to 1-2 μm. In addition, particle image velocimetry was performed on temporally resolved composite images to derive velocities in the field. The characteristics of a steady-state flat fan water spray at a flow rate of 5 kg/h were analysed using both methodologies. The findings indicated a strong qualitative agreement between the techniques, showing consistent trends, though some quantitative discrepancies were noted. The velocity measurements varied by up to 20% between the two methods. These variations may stem from the exclusion of non-spherical droplets in PDA or from droplets within the depth of field in the images, which move slowly and lead to bias during image analysis.
