15-RANS transition predictions
Comparison of RANS transition model predictions on hypersonic three-dimensional forebody configurations
J.I. Cardesa1,*, G. Delattre1,**
1ONERA/DMPE, Université de Toulouse, F-31055 Toulouse, France
* jcardesa@onera.fr – ** delattre@onera.fr
As the interest for hypersonic vehicles has been growing over recent years, so has the research activity in the field of laminar-turbulent transition at hypersonic velocities due to the paramount importance of accurate drag and heat transfer prediction during the design phase.
Currently, RANS computations are still widely used for design studies, hence the need to cater for transition phenomenology within RANS codes. Modern literature offers a wide range of RANS-coupled transition models with various levels of formulation complexity. It is commonly assumed that choosing amongst such models boils down to a compromise between model predictive power and ease of implementation within existing codes – which goes hand in hand with ease of deployment and use.
In this talk, we report on our experience introducing low-overhead, low-complexity RANS-compatible transition models in our journey towards a comprehensive collection of transition models within ONERA’s multiphysics software suite CEDRE.
More specifically, three transition models that couple swiftly with Menter’s SST turbulence model by introducing a single additional scalar transport equation for intermittency are benchmarked on common hypersonic test cases.
A commonality amongst the chosen models is their reliance on local quantities to infer characteristic boundary layer length scales, hence avoiding the need to determine the boundary layer edge position and compute integrals along wall-normal lines.
Results obtained with two such models on hypersonic vehicle forebodies (a generic forebody and BOLT) are compared against experimental data from quiet wind tunnel tests.
The differences between the two models will be discussed, with emphasis on the dependence of the results on the domain inlet conditions and the tunable model constants. Where applicable, different transition mechanisms along the surface of the vehicle will be highlighted.
Fig. 1: RANS simulation results for the surface heat flux on a hypersonic forebody at Mach = 6 at quiet wind tunnel test conditions, along with a cut at the forebody end providing the Mach number distribution.
