Effect of turbulence models on loss prediction in steam turbine cascades considering non-equilibrium condensation using OpenFOAM

In a humid steam environment, the unbalanced condensation phenomenon within the steam turbine blade channels significantly alters the flow field structure and leads to thermodynamic losses; accurately predicting this process is crucial for the aerodynamic performance analysis of turbine machinery. Utilizing the high scalability of the open-source computational platform OpenFOAM, this study developed a transonic humid steam solver that combines multiple advanced turbulence models, and selected nine representative RANS models to systematically evaluate their ability to predict unbalanced condensation flow in two-dimensional turbine blades. The results show that the k-ωSSTLM transition model performs the best. This model, by explicitly describing the development of laminar, transition, and turbulent regions in the boundary layer, can more accurately depict the flow and heat transfer characteristics in the near-wall region. The predicted surface pressure distribution and shock angle are most consistent with the experimental data, and reasonably reveal the influence of the transition process on local supercooling and droplet growth behavior. In terms of loss prediction, this model accurately captures the low friction characteristics of the laminar region near the leading edge, thereby avoiding the inherent problem of excessive losses in the traditional “fully turbulent" assumption. Future work will incorporate the dehumidification mechanism and extend the research to three-dimensional, full-stage turbine simulation and optimization studies.