Interphase coupling in polydisperse evaporating sprays using the generalised fully Lagrangian approach

A novel methodology is presented for the modelling of polydisperse evaporating droplet flows by combining the generalised Fully Lagrangian Approach (gFLA) with the statistical learning technique of kernel regression. This procedure is shown to retain the detail of the structures inherent in the droplet concentration field for such systems, whilst offering a reduction in computational cost due to the lower droplet seeding required by the gFLA in comparison to conventional methods. The developed approach is further applied to the scenario of tracking the droplet vapour field arising due to droplet evap oration, with the interphase coupling source terms being constructed using the droplet probability density provided by the gFLA.