A perspective on turbulence models for aerodynamic flows

Turbulence modelling options are discussed in the context of steady aerodynamic flows. After a brief overview of popular turbulence models, four criteria are presented that should be satisfied in order to conclusively evaluate a turbulence model with respect to its ability to predict a specific flow. Many past studies do not meet these criteria. This is followed by some sample results for several turbulence models, including one-equation, two-equation and algebraic Reynolds stress models. The three main conclusions are as follows. First, more combined experimental–numerical studies are needed that meet the four criteria for assessment of turbulence models. Second, of the models studied, the Spalart-Allmaras model provides the most accurate results for the high-lift flows examined. Finally, the most significant factor limiting our present ability to predict many aerodynamic flows accurately is our inability to reliably predict laminar-turbulent transition. Until this issue is addressed, the benefits of an improved turbulence model will be limited.