Development and Validation of an Operational Numerical Model for the Simulation of the Aerial Drop of Firefighting Products

Aerial firefighting plays an important role in the protection of human lives and patrimony, particularly in situations requiring a rapid intervention, such as emerging fires, inaccessible mountainous areas, or highly risk areas. The efficiency of aerial means is, however, extremely dependent on fire characteristics, atmospheric conditions and pilot expertise. The current work describes the development and validation of the Aerial Dropping Model (ADM), intended for the simulation of the aerial drop of firefighting products. ADM simulates the vegetative canopy-induced wind flow, under varying atmospheric stability conditions. Vertical turbulent fluxes are calculated through a set of modified flux-profile relationships valid in the atmospheric roughness sublayer. The efflux of liquid from the aircraft tank is calculated based on tank geometry and door-opening rate. The size of droplets formed by Rayleigh-Taylor and Kelvin-Helmholtz instabilities acting on the surface of the jet is estimated applying the linear stability theory, while the size distribution and velocity of the droplets formed by secondary breakup (bag and shear breakup regimes) is based on experimental correlations dependent on the Weber and Ohnesorge dimensionless numbers. A Lagrangian approach is applied to the simulation of the spray cloud deposition, during which dynamical drag laws allow to account for droplet shape deformation. The process ends with the canopy retention of the spray cloud. Main outputs are the spatial ground distribution of concentrations, and the line length and area per coverage level. The validation process included the statistical comparison of ADM outputs against a set of real scale drop tests conducted with a conventional and a constant flow delivery systems. From the investigation of model performance, good accuracy was obtained for the wide range of input conditions tested. Ground pattern shape shows the features observed in measured data. The average normalised mean squared error for the estimation of line lengths is 0.01 and 0.03 for the prediction of areas occupied per coverage level. Due to its operational characteristics this tool is primarily indicated for testing the effectiveness of new firefighting chemicals or delivery systems, complementing or substituting the data obtained from exhaustive real scale drop tests. It can potentially be used as a demonstration tool in firefighters’ training activities.