Evaluation of a Correlation-Based Transition Model and Comparison with the eN Method

Abstract

The correlation-based -Re t transition transport model has been implemented into a hybrid Reynolds-averaged Navier-Stokes solver and evaluated on various test cases. The original model formulation and recently published approaches of different authors for relevant empirical model functions as well as different transition criteria are compared. The prediction of transition with the correlation-based model was applied to a zero pressure gradient flat plate test case and some well-known one-element airfoil test cases. A comparison with the standard approach for transition prediction in the flow solver used based on the e N method was accomplished. The simulation results in terms of transition locations and skin-friction coefficient distributions, performance, and limiting factors of both transition prediction methods for the various test cases are presented and discussed.function that controls the transition length k = turbulent kinetic energy k amb = lower bound for turbulent kinetic energy in the far field M = Mach number N = critical N factor Re x = Reynolds number based on axial coordinate x Re c = Reynolds number based on momentum thickness, critical Re t = Reynolds number based on momentum thickness at transition onset Re = Reynolds number based on momentum thickness Re = Reynolds number based on vorticity Tu = local turbulence intensity Tu 0 = turbulence intensity at the far-field boundary Tu 1 = freestream turbulence intensity U = local velocity x = axial coordinate y = distance in wall coordinates = angle of attack = model constant for the shear-stress transport k-! turbulence model = model constant for the supersonic transport k-! turbulence model = intermittency = molecular viscosity t = eddy viscosity = density ! = specific turbulent dissipation rate ! amb = lower bound for turbulent dissipation rate in the far field

Numerical Investigations of Local Correlation-Based Transition Model in Hypersonic Flows

“…The runs with LCTM models match pretty good the laminar heat flux upstream the transition , however, downstream the transition region they tend to slightly over predict the heat flux regarding the one computed in a fully turbulent flow. This phenomenon was not observed in previous works 4,5,19,20 where other correlations were used. However a bump over the turbulence profile is observed in experiments.…”

Development and Application of Transition Prediction Techniques in an Unstructured CFD Code (Invited)

“…46, where the results were computed for M = 0.2, Re = 2.0×10 6 , and α = 3.5° and α = 7.5° [150]. In contrast to the results shown, for example, in [121] here the free-stream turbulence intensity (FSTI) was kept at the same numerical value in each simulation using the RSM (fully-turbulent), the γ-Re θ -SST model and γ-Re θ -RSM model and for each angle of attack similar to the situation in a wind tunnel when an experiment is carried out. In [121], where partially a good agreement was shown using the γ-Re θ -SST model, it was described that this agreement could be achieved only if the numerical FSTI value was adapted for each angle of attack.…”

“…In the TAU code the original formulation of the γ-Re θ model coupled to the k-ω SST model [77] according to [76] is available and has been applied to various configurations [119][120][121].…”

Key role of human leukocyte antigen in modulating human immunodeficiency virus progression: An overview of the possible applications

“…The γ-Re θt transition transport model has been implemented into the hybrid, three-dimensional, parallelized RANS based DLR TAU code [7][8][9]. Basic investigation and validation was performed for a flat plate test case and for certain one element airfoils such as the AS-A, AS-B and Somers NLF airfoil [10]. Different approaches for model relevant empirical functions and various transition criteria have been analyzed.…”