A Laval nozzle is a pivotal component of supersonic separation technology for dehydration, but its performance is constrained by low separation efficiency due to intractable conflicts between droplet nucleation and growth rates. To address this issue, this work for the first time applied the Foelsch-MOC method to design an expansion section with a first-order differentiable and second-order continuous profile. A numerical model for pure water vapor condensation based on the ideal gas equation of state (EOS) was then established to optimize nozzle structures for droplet growth. Major findings revealed that the droplet radius of the Foelsch-MOC nozzle was 30.5% higher than that of the conical nozzle, while its outlet liquid mass fraction was 5.8% lower. The liquefaction capacity was found to increase first and then decrease with the expansion coefficient, peaking at 0.5. Moreover, increasing the axis shift from 9.3 mm to 25 mm promoted the outlet liquid mass fraction by 1.2% and decreased the droplet radius by 9.3%, thereby improving liquefaction capacity. These results highlight the droplet growth superiority of the novel expansion profile, providing a valuable guideline for designing nozzles that simultaneously ensure liquefaction efficiency and facilitate droplet development.
To achieve efficient bubble refinement under high gas–liquid ratio (GLR) conditions in industrial applications, this study investigates the effects of Venturi geometry parameters and operating conditions on bubble size distribution under high GLR conditions. Experiments were conducted with throat velocities ranging from 4.34 to 13.02 m/s and GLRs from 10% to 60%, examining the effects of throat diameters (4 mm and 8 mm) and divergent angles (7.5°, 10°, and 12.5°). A novel baffled Venturi bubble generator was designed by maintaining constant throat flow area and principal parameters. Results showed that Venturi tubes with an 8 mm throat diameter produced smaller bubble sizes at higher liquid flow rates. In contrast, excessively large divergent angles produced unfavorable larger bubble sizes due to higher GLR and flow separation. Based on experimental data (Reynolds number range: 12,163–106,084), a comprehensive bubble size prediction model was established, which simultaneously considers the effects of throat diameter and divergent angle. Increasing the liquid flow rate and reducing the GLR effectively minimize bubble size, while the newly developed bubble generator demonstrates superior performance compared to conventional Venturi designs. These findings provide practical references for the design and industrial application of Venturi bubble generators under high GLR conditions.