Mixed compression inlets are common among supersonic propulsion systems. However they are susceptible to total pressure losses due to shock/boundary-layer interactions (SBLI's). Because of their importance, a workshop was held at the 48th American Institute of Aeronautics and Astronautics (AIAA) Aerospace Sciences Meeting in 2010 to gauge current computational fluid dynamics (CFD) tools abilities to predict SBLI's. One conclusion from the workshop was that the CFD consistently failed to agree with the experimental data. This thesis presents additional CFD and numerical analyses that were performed on one of the configurations presented at the workshop.;The additional analyses focused on the University of Michigan's Mach 2.75 Glass Tunnel with a semi-spanning 7.75 degree wedge while exploring key physics pertinent to modeling SBLI's. These include thermodynamic and viscous boundary conditions as well as turbulence modeling. Most of the analyses were 3D CFD simulations using the OVERFLOW flow solver. However, a quasi-1D MATLAB code was developed to interface with the National Institute of Standards and Technology (NIST) Reference Fluid Thermodynamic and Transport Properties Database (REFPROP) code to explore perfect verses non-ideal air as this feature is not supported within OVERFLOW. Further, a grid resolution study was performed on the 3D 56 million grid point grid which was shown to be nearly grid independent. Because the experimental data was obtained via particle image velocimetry (PIV), a fundamental study pertaining to the effects of PIV on post-processing data was also explored.;Results from the CFD simulations showed an improvement in agreement with experimental data with certain settings. This is especially true of the v velocity field within the streamwise data plane. Key contributions to the improvement include utilizing a laminar zone upstream of the wedge (the boundary-layer was considered transitional downstream of the nozzle throat) and the necessity of mimicking PIV particle lag for comparisons. It was also shown that the corner flow separations are highly sensitive to the turbulence model. However, the center flow region, where the experimental data was taken, was not as sensitive to the turbulence model. Results from the quasi-1D simulation showed that there was little difference between perfect and non-ideal air for the configuration presented.
展开▼
