A quantitative analysis of atomization mechanisms from high-fidelity simulations

Recent advances in computational efficiency and numerical methods have allowed researchers to simulate atomizing liquid systems. These simulations have yet to result in a significantly deeper understanding of the underlying processes of atomization. The primary obstacle limiting simulations' usefulness is the massive size of resultant datasets, which contain millions to billions of spatial cells per timestep and can fill hundreds of terabytes of storage. The present work focuses on utilizing a methodology to extract relevant information from simulations, as they run, to produce easily accessible databases full of novel, statistically relevant information. This methodology will be applied to two simulations. First, a diesel-type high-pressure round jet injected into quiescent air. And second, an air-blast atomizing round jet. Novel statistics describing the local flow fields and droplet characteristics from breakup events throughout these simulations will be processed and compared. Special consideration will be given to comparing the two jets for an analysis of the effects of aerodynamic stresses on liquid breakup. This work will result in a foundation for future higher-resolution studies of jets and the development of new physics-based reduced-order atomization models.