Image analysis of liquid breakup in a flat fan spray for the development of a fully Lagrangian modelling approach

The current work is focused on the development of a suitable methodology to determine droplet distribution data in a planar, flat fan spray atomisation process. The aim is to support the development of a new simulation framework, based upon the Fully Lagrangian Approach (FLA) [1], that has been generalised to polydisperse evaporating droplets (gFLA) [2]. The FLA has advantages of both Lagrangian droplet tracking and continuous formulations for the admixture [1, 3]. Validation of this modelling approach requires the identification of transient droplet size distributions, unlike more routinely used techniques, that measure size distributions at a single point location, over the entire period.
In the experiments, a commercial flat fan nozzle is used to inject water under steady flow conditions via low pressure pump. The atomisation is assumed to occur such that the liquid sheet breaks up into droplets that are contained within a 2-D spray plane, with minimal out of plane deviation. The optical arrangement consists of a pulsating LED for backlit illumination while the spray is captured using a high-speed camera fitted with a long-distance, microscopic lens. The experimental setup was used to perform 2-D spray analysis using an in-house image processing code to differentiate between structures in the spray and to identify the transition region where the flow transforms from a continuous liquid film into ligaments and individual droplets. More specifically, image processing criteria were applied in order to determine the size and position of individual droplets in the downstream atomisation region in consecutive image sequences across all time periods. A statistical approach is proposed to obtain the droplet distribution in different regions of the spray.