Multiscale simulation of swirling atomization using a hybrid conservative level set/Lagaragian particle tracking method

Abstract

Atomization is complex for intricate interfaces and multiple length scales, leading to a fundamental challenge to accurately describe its characteristics. To address this issue, we develop a hybrid conservative level Set (LS)/Lagrangian Particle Tracking (LPT) method that efficiently models both the liquid films and small droplets. The simulation of a settling droplet is utilized to validate the accuracy of the hybrid method, with results showing excellent agreement with theoretical predictions and the pure LS method. Additionally, the hybrid method effectively detects adhered structures and facilitates droplet conversion under specific criteria in parallel processing. Then the method is applied to studying swirling atomization characteristics. Despite minor discrepancies with the pure LS method, the hybrid method significantly reduces computational costs by 80% in this realistic scenario. With the effective method, we explore the influence of inlet conditions on atomization. Turbulent inflow and increased swirl intensity reduce the peak value of the volume probability density function, suggesting these factors promote more uniform droplet formation and enhance atomization. Vortex structure analysis reveals that swirl and Kelvin-Helmholtz instability dominate in high-velocity regions. The enstrophy, a key turbulence metric, increases with higher turbulence and swirl intensity. Enstrophy transport budget analysis highlights vortex stretching as the dominant term, with vorticity vectors consistently aligning with the intermediate principal strain rate across cases.