A generalized non-equilibrium heat transfer model for annular flow boiling of binary zeotropic mixtures

Zeotropic refrigerant mixtures are promising alternatives to high–global warming potential fluids due to their thermodynamic flexibility and temperature glide, which enables improved thermal matching in heat exchangers. However, the non-isothermal phase change and associated mass transfer effects introduce strong non-equilibrium behavior that challenges accurate prediction of flow boiling heat transfer. In this study, a generalized non-equilibrium heat transfer model is developed for annular flow boiling of binary zeotropic mixtures based on film theory. The model explicitly resolves coupled heat and mass transfer across the liquid film, vapor–liquid interface, and vapor core, accounting for interfacial temperature variation, axial and radial mass diffusion resistance, and species segregation. An iterative solution strategy is employed to simultaneously determine the interfacial temperature and heat flux by enforcing energy balance across all phases. A key advantage of the proposed framework is its flexibility, allowing integration of multiple liquid-film flow boiling correlations to optimize predictive performance for different mixtures and operating conditions. The model is validated against 1139 experimental data points covering 28 binary refrigerant pairs, a wide range of operating conditions, and temperature glides up to 35.8 °C, and it has achieved 81% of predictions within ±30% deviation. Overall, the proposed non-equilibrium framework consistently outperforms eleven existing flow boiling correlations, demonstrating improved robustness and broad applicability for modeling evaporation in zeotropic mixtures and supporting the design of advanced refrigeration and heat pump systems. • A generalized non-equilibrium model is developed for binary zeotropic evaporation. • Interfacial temperature and concentration gradients are included. • Validated against 1139 data points from various refrigerant pairs. • Allows tailored predictions for specific refrigerant mixtures. • Outperforms eleven existing correlations with 81% accuracy within ±30%.