Aerodynamic Robustness of RANS Turbulence Models Under Parametric Uncertainty

Abstract

This study proposes an uncertainty-aware framework for evaluating turbulence models applied to the NACA 0012 airfoil. Validated two-dimensional CFD simulations with a hybrid mesh are performed and compared with experimental data. Unlike conventional deterministic approaches, the analysis incorporates a Six Sigma-based probabilistic methodology to quantify the impact of inlet turbulence variability on aerodynamic coefficients. Latin Hypercube Sampling is used to efficiently explore uncertainties in turbulence intensity and length scale. The results reveal distinct sensitivity patterns across turbulence models, highlighting differences in their response to input variability. In particular, the Spalart-Allmaras model exhibits the most stable behavior, with reduced variability in aerodynamic predictions. This work contributes a structured approach that combines performance assessment and robustness analysis, providing practical support for turbulence model selection under uncertain flow conditions.