Reynolds shear stress modeling in turbulent boundary layers subject to very strong favorable pressure gradient

• Published in: Computers & fluids Vol. 202; p. 104494
• Main Authors: Saltar, German, Araya, Guillermo
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Ltd 30.04.2020; Elsevier BV
• Subjects: Acceleration; Coefficient of friction; Computational fluid dynamics; Computer simulation; Direct numerical simulation; DNS; Kinetic energy; Laminarescent; Numerical prediction; Quasi-laminarization; RANS; Reynolds averaged Navier-Stokes method; Shear stress; Skin friction; Stresses; Temporal resolution; Turbulence modeling; Turbulence models; Turbulent boundary layer; DNS; RANS; Quasi-laminarization; Turbulence modeling; Laminarescent
• ISSN: 0045-7930; 1879-0747
• DOI: 10.1016/j.compfluid.2020.104494

•Assessment of the SA and SST models in very strong FPG based on a DNS database.•The SA model showed quick spatial adaptation to turbulent inflow conditions.•SA and SST were able to properly predict the freestream and friction velocity.•However, they fail to reproduce the decreasing behavior of Cf.•Both models overpredicted production of TKE in very strong FPG flows. Recent numerical predictions of turbulent boundary layers subject to very strong Favorable Pressure Gradient (FPG) with high spatial/temporal resolution, i.e. Direct Numerical Simulation (DNS), have shown a meaningful weakening of the Reynolds shear stresses with a lengthy logarithmic behavior [1,2]. In the present study, assessment of the Shear Stress Transport and Spalart-Allmaras turbulence models (henceforth SST and SA, respectively) in Reynolds-averaged Navier-Stokes (RANS) simulations is performed. The main objective is to evaluate the ability of popular turbulence models in capturing the characteristic features present during the quasi-laminarization phenomenon in highly accelerating turbulent boundary layers. A favorable pressure gradient is prescribed by a top converging surface (sink flow) with an approximately constant acceleration parameter of K=4.0×10−6. Validation of RANS results is carried out by means of a large DNS dataset [1]. Generally speaking, the SA turbulence model has demonstrated the best compromise between accuracy and quick adaptation to the turbulent inflow conditions. Turbulence models properly captured the increasing trend of the freestream and friction velocity in highly accelerated flows; however, they fail to reproduce the decreasing behavior of the skin friction coefficient, which is typical in early stages of the quasi-laminarization process. Both models have shown deficient predictions of the decreasing and logarithmic behavior of Reynolds shear stresses as well as significantly overpredicted the production of Turbulent Kinetic Energy (TKE) in turbulent boundary layers subject to very strong FPG.