Influence of Surface Roughness on the Flat-Plate Boundary Layer Transition Under a High-Lift Airfoil Pressure Gradient and Low Freestream Turbulence

Effects of surface roughness on the transition of flat-plate boundary layers under a high-lift airfoil pressure gradient with low incoming freestream turbulence level have been investigated. Time-resolved streamwise and wall-normal velocity fields with surface roughness values of Ra/C = 0.065·10−5, 4.417·10−5 and 7.428·10−5 have been measured at a fixed Reynolds number of 5.2·105 and freestream turbulence intensity of 0.2%. For the reference Smooth surface of Ra/C = 0.065·10−5, a laminar separation bubble forms from about 64% to 83% of the chord length. Displacement thickness increases downstream of separation and then decreases during the transition (reattachment), and momentum thickness increases due to the vortices shed from the separation bubble. Increasing surface roughness has little impact on the laminar boundary layer separation onset but reduces the height and length of the separation bubble and induces earlier transition. For Ra/C = 4.417·10−5, displacement thickness during transition is slightly thinner and the overall momentum deficit is slightly lower than those for Ra/C = 0.065·10−5. For Ra/C = 7.428·10−5, the separation bubble becomes hardly visible as the transition mode approaches the attached mode, and turbulent mixing by the wall-bounded turbulence becomes dominant. In addition, the portion of turbulent wetted area increases due to earlier transition, and momentum deficit increases more rapidly in the turbulent wetted area. Thus, the overall momentum deficit for Ra/C = 7.428·10−5 is larger than that for Ra/C = 0.065·10−5.

Numerical Modeling of Freestream Turbulence Decay Using Different Commercial Computational Fluid Dynamics Codes

This work models the spatial decay of freestream turbulence using three different commercial computational fluid dynamics (CFD) codes: Fluent, star-ccm+, and cfx. The two-equation shear stress transport k–ω (SST-k–ω) steady Reynolds-averaged-Navier–Stokes (RANS) model was used, within each of these three different commercial codes, and the modeling variations were analyzed. Comparison of the results from the SST-k–ω model with experiments and large eddy simulation (LES) (carried out using star-ccm+) were also made, which reveal that all the commercial CFD codes demonstrate either a higher or slower rate of spatial turbulent kinetic energy (TKE) decay. Attempts were then made to unify the resultant modeling approach between these three CFD tools, by careful manipulation of the inlet boundary conditions and subsequent fine-tuning of the SST-k–ω model constant (β∞∗). The results obtained not only displayed uniformity among the three CFD codes but also demonstrated a much better agreement to the experiments and the LES results. Thereafter, the optimized model coefficient (β∞∗) was integrated with the three-equation k–kl–ω transition model to examine its applicability in modeling a turbulent boundary layer flow over a flat plate with low incoming turbulence. The results showed good agreement with the theoretical boundary layer correlations, with correct prediction of the transition location. The findings from this study can be used as a suitable modeling method to accurately model the effects of freestream turbulence on bluff-body and boundary layer flows.