Hybrid turbulence modeling approach for flat-plate turbulent boundary layers: numerical simulation and experimental validation

To further improve the accuracy of numerical simulations for turbulent boundary layer flows, this study introduces a hybrid turbulence model for turbulent boundary layer flows. This model is based on the Partial Averaged Fluctuation Velocity (PAFV) and k-ω turbulence models and features only one empirical coefficient. Leveraging the OpenFOAM open-source software platform, the development of the proposed hybrid turbulence model is carried out. Numerical simulations are performed for turbulent boundary-layer flow over a flat plate with zero pressure gradient. This analysis yields insights into the mean velocity, PAFV distributions, and allows for a comparative study of calculated and experimental values pertaining to the logarithmic law, velocity defect law, friction coefficient, and turbulent kinetic energy of the turbulent boundary layer. In comparison to traditional models, this model offers a more comprehensive velocity distribution near the viscous sublayer, with the predicted velocity being approximately 6% higher than values from conventional models. Numerical simulations present explicit distributions of the PAFV vector within the flat-plate turbulent boundary layer, indicating the PAFV’s utility as a physically interpretable turbulence velocity scale. The computed wall friction coefficient C_f deviates by less than 1.59% from theoretical correlations under high-Reynolds-number conditions. Furthermore, the predicted turbulent kinetic energy and Eddy Kinematic Viscosity exhibit consistent distribution trends with established theoretical analyses. Experimental measurements of turbulent fluctuation velocity suggest that the measured fluctuation intensity is roughly twice the predicted PAFV magnitude, thereby supporting the use of the empirical coefficient C ≈ 2.0. These findings quantitatively validate the efficacy of the proposed PAFVH-ω model in forecasting flat-plate turbulent boundary layer flows and introduce a fresh framework for incorporating first-order fluctuation data into RANS-based turbulence modeling.