Improved CFD modeling of radial void fraction distribution in flow boiling: A novel approach for wide-ranging operating conditions

Abstract

Accurate prediction of void fraction distribution in flow boiling is of great significance for the thermal–hydraulic analysis and safety of nuclear power plants. Computational Fluid Dynamics (CFD) methods have consistently proven to be effective for boiling simulation. Due to the complexity of boiling phenomena, the prediction of radial void fraction distribution in tubes tends to be challenging, thus hindering a clear understanding of the underlying mechanism. In this work, with the aim of achieving the void fraction prediction over a wide range, the DEBORA experiment operated with R12 was simulated within the pressure range of 1.46–3 MPa and the mass flux range of 1000–5000 kg·m⁻²·s⁻¹. The closures of the two-phase equation were compared and the applicability of R12 was discussed. Meanwhile, sensitivity analyses were conducted on the models for bubble departure diameter, lift force, and turbulent dispersion force. The Kocamustafaogullari model was selected for calculating bubble departure diameter. According to the cross-sectional average void fraction, the appropriate selection criteria for the effective turbulent Prandtl number model was recommended. An abnormal upward trend of void fraction near the wall was observed for G ≥ 3000 kg·m⁻²·s⁻¹ due to the small or even negative relative velocity rather than the lift force coefficient model. A new lift force coefficient model relevant to pressure, mass flux, and void fraction has been proposed. The new CFD model combination demonstrated a good agreement with the experiment, providing an effective tool for analyzing the radial void fraction distribution.