A comprehensive theoretical model for the primary breakup of liquid jets in subsonic crossflow was developed. The model theoretically analyzed the jet deformation process, mass stripping process, and the influence of several critical forces and consequently provided highly accurate predictions of the jet trajectory. Deformation of the liquid jet cross section was considered as a two-stage process based on the physical characteristics, including the spring-mass analogy deformation and the mass stripping induced deformation. The mass stripping process was modeled as an exponential function of time based on experimental findings for liquid jets and droplets. Balance of critical forces acting on the jet were analyzed, both along the gas and jet flow directions, which included aerodynamic drag, viscous force, surface tension, and gravitation. The model provided precise prediction to the jet trajectory against experimental data without any initial jet velocity assumption across a wide range of gaseous Weber numbers and gas to liquid momentum ratios. In addition, quantitative effects of viscous force, surface tension force, and aerodynamic drag on jet trajectory were fully investigated based on the new model, which provided more insight into jet breakup characteristics and the effects of fuel properties on jet trajectory and deformation. Furthermore, the three-dimensional structure of the jet was reconstructed through the present model, which matched well against numerical results. Importantly, the current mathematical primary breakup model could be integrated with Lagrangian methods, obtaining more detailed vortex structures and accurate droplet dispersion with reduced computational time.
The transformation criterion is critical for the coupled volume of fluid and lagrangian particle tracking (VOF-LPT) method for jet breakup and atomization simulations, which determines the conversion from VOF droplets to Lagrangian particles. It is usually developed based on experience, varies among researchers (e.g., different maximum transformation diameter), and consequently leads to considerable computational errors. To mitigate such errors in simulations of liquid jets in crossflow (LJCF), an optimized transformation criterion of the VOF-LPT method was developed and validated in this study. First, a comprehensive parametric study of a two-dimensional LJCF case was conducted to investigate the primary breakup process. It was found that liquid column deformation was more sensitive to mesh resolution than to interface capturing methods, and a minimum of 32 cells across the jet diameter was essential to accurately capture the periodic deformation characteristics. Second, the secondary breakup process was studied through detailed three-dimensional simulations of droplets in crossflow. An optimized transformation criterion was developed based on quantitative error analysis of the droplet displacement of VOF and LPT methods. It was established based on the local droplet Weber number and local mesh resolution, and without any coefficient that needs calibration. Finally, the proposed criterion was validated using a three-dimensional LJCF case, demonstrating a remarkable improvement of over 30% in droplet size prediction accuracy compared to conventional implementations based on the maximum diameter criterion.
This study investigated the influence of swirl number (Sn = 0–3.9) on the flow characteristics and atomization behavior of coaxial liquid jets. Velocity and vorticity fields, jet breakup processes, and droplet dynamic characteristics were analyzed within a wide range of swirl numbers. Results showed that geometric swirl numbers agreed well with experimental and numerical swirl numbers under weak swirl conditions, but were significantly higher at ultra-strong swirl conditions due to flow separation and angular momentum loss. The central toroidal recirculation zone (CTRZ) was small at weak swirl condition, but expanded significantly with increasing swirl numbers due to enhanced centrifugal force. The flow field evolved from a single-vortex feature (at weak swirl) to dual- and multi-vortex features (at strong and ultra-strong swirl), resulting in more homogeneously distributed vorticity. After introducing the liquid jet, flow stability decreased significantly as the swirl number increased. Under weak swirl condition, the CTRZ remained suppressed and the flow structure was relatively stable. Under ultra-strong swirl condition, the CTRZ expanded dramatically and flow instability intensified. Furthermore, as the swirl number increased, both the normalized breakup location and timing of the jet were delayed. However, atomization was improved with increasing swirl number, and the droplet dynamic characteristics transitioned from jet-dominated to swirl-dominated.