Higher-order RANS turbulence models for separated flows (RANS)

Higher-order Reynolds-averaged Navier-Stokes (RANS) models are developed to overcome the shortcomings of second-moment RANS models in predicting separated flows.  These differential transport models hold promise in predicting reversal of turbulent transport typical in adverse pressure gradient flows.  Third- and fourth-order moment transport models are developed and used to simulate canonical boundary layer and two-dimensional separated flows.

Existing RANS turbulence models over-predict the extent of separation, making them unfit for off-design aerodynamic studies.  Second-moment closures (SMC) have failed to improve flow separation predictions over simpler two-equation models.  Failure may be in part due to inability of algebraic models for turbulent transport to predict transport reversal typical in decelerating flows.  To overcome this shortcoming, we represent the turbulent transport term using a differential model.  The differential transport model leads to closing the RANS system equations either at the fourth- or fifth-order moment level.    Models are constructed based on a-priori studies on existing moment budget term models.  The system of about 10-14 turbulence equations are solved iteratively.