Modelling of Sprinkler Sprays using Deterministic and Stochastic Approaches

Modelling sprinkler atomization is a challenging task, due to the stochastic nature of the liquid sheet breakup process. The present study aims to improve prediction of the initial sprinkler spray characteristics through exploring different physics based modelling approaches. The sub-models for film flow and sheet trajectory adopted in the development of the fire sprinkler spray models are reviewed. Three new deterministic approaches for sprinkler atomization have been proposed by employing an existing film sub-model and a detailed water sheet trajectory sub-model which has never been used for fire sprinkler applications. The numerical integration of the sub-models with sheet instability theory has been explained in detail. The developed methods simulate the orthogonal impingement of water jet to a deflecting disk, with the potential to be adapted for tilted deflectors. A comparative analysis is carried out between the three introduced methods and a reference model in terms of their predictions for droplet median diameter and initial droplet location for a range of ambient temperatures and water injection pressures. The developed methodologies have been further expanded by incorporating random behaviour to the spray formation procedure. The stochastically predicted mean velocity and volume median diameter have been compared against robust experimental data and empirical correlations. The improvements obtained by the developed methodologies have been promising.