Large-Eddy Simulation of the Bachalo-Johnson Flow, with Shock-Induced Separation

PI Name: 
Philippe Spalart
PI Email:
Allocation Program: 
Allocation Hours at ALCF: 
135 Million
Research Domain: 

Turbulence strongly controls many flows. Shock–boundary layer interactions (SBLIs) are of particular interest to aerospace companies and the gas-turbine industry, and some proposals for carbon capture and sequestration also involve supersonic flows in the compressors. This project will focus on the upper surface of a transport aircraft wing, especially in off-design conditions, such as an increased Mach number or an angle of attack. It takes as its reference Bachalo and Johnson’s 1981 experiment at the National Aeronautics and Space Administration (NASA), which has been a very useful test case for computational fluid mechanics (CFD), and Reynolds-Averaged Navier-Stokes (RANS) turbulence models. However, although the experiment featured fairly detailed measurements, there was no independent confirmation, and it yielded much less information compared to what a simulation can provide, for instance, in terms of skin friction.

Therefore, as researchers continue a worldwide effort to improve RANS models and to enhance the general understanding of complex turbulent flows, it will be very valuable to have a large eddy simulation (LES) of this 1981 experiment’s flow. The simulation will be very large because of the extended separated region and the high Reynolds (Re/m=13.6 × 106) number. It will be conducted at the correct Mach number (M=0.875) as well, with converged turbulence statistics and in good agreement with the experiment. The overall objective is to reach the state of the art for simulations of SBLIs for grids up to two billion points and to thoroughly extract findings pertaining to physics and numerical resolution.