Modeling non-equilibrium water-ethanol condensation: Ideal and non-ideal approaches

Accurate prediction of non-equilibrium wall-film condensation of water-ethanol mixtures in Baijiu distillation condensers is essential for reliable modeling of heat and mass transfer. Yet, it remains challenging because of strong coupling among vapor partial pressures, saturation temperatures, and non- ideal phase behavior. Yet, it remains challenging because of strong coupling among vapor partial pressures, saturation temperatures, and non-ideal phase behavior. The conventional Lee model assumes a constant saturation temperature, limiting its ability to capture these coupled effects and leading to errors in predicting the condensation driving force. To address this limitation, this study develops an improved framework for simulating condensation of a water-ethanol mixture by reformulating the saturation temperature as the key phase-change driver. Three Lee-model-based approaches are proposed, including a pure-component model, an ideal-solution model, and a non-ideal solution model. Comprehensive comparative analyses are then conducted to reveal their respective prediction performances under different ethanol concentrations. The pure-component model estimates the saturation temperature from individual partial pressures, without accounting for interactions. The ideal-solution model applies Raoult's law combined with Dalton's law to describe vapor-liquid equilibrium. The non-ideal model introduces an experimentally fitted relationship between the saturation temperature and liquid-phase composition, based on atmospheric VLE data, to account for non-ideal effects. To improve numerical robustness, a liquid-phase fraction threshold is introduced to separate gas-dominated and liquid-significant regimes, avoiding breakdown of Raoult's law at low concentrations and ensuring a smooth model transition. This work adopts two-dimensional CFD numerical simulation combined with the Volume of Fluid (VOF) multiphase model and self-compiled UDF-modified Lee phase-change model, and an independent experimental test rig is self-developed to verify the reliability of the proposed models. Results show that the threshold approach stabilizes computation and improves temperature-field prediction. Non-ideal effects are most significant at intermediate compositions ( x = 0.3–0.7), where model deviations are largest. Near azeotropic conditions, only the non-ideal model accurately captures component enrichment, while simpler models show clear discrepancies. Overall, the pure-component model shows the largest deviation across all conditions, whereas the ideal-solution model approaches the non-ideal solution behavior at higher concentrations ( x ≥ 0.7). This work defines the applicability limits of the three proposed condensation models and provides a robust framework for multicomponent condensation, with its reliability confirmed by independent experimental tests on wall temperature and outlet condensate composition.