1. The experiments discussed here were conducted in the University of Wyoming Aeronautical Laboratories (UWAL) 2’ × 2’ wind tunnel. The wind tunnel is a fan-driven, open-return design with a 0.61 × 0.61 × 1.219 m test section. Using a variable-speed motor, free-stream velocities of 10-50 m/s are possible at Reynolds’s number up to 2.5x106/m. Theinletsection of the tunnel has a honeycomb insert and three screens located just upstream of a contraction section with a 12:1ratio. Themeasuredfree-streamturbulenceabove the model was 0.3%. Wedge Model Design
2. 47-80 -60 -40 -20 0 20 40 60 80
3. In order to perform these corrections, two inviscid CFD simulations were performed for each model: one with a wind tunnel wall and another without the wall. The wall case was used to simulate the wind tunnel experimental condition, and the no-wall case was used tosimulatefreeflightcondition. Simplecomputational meshes for the simulations were generated using MATLAB. The grids varied in size for the different models and the different upper boundary (wall or no wall). The largest of the grids was 243 × 141. The CFD solver used was OVERFLOW, a Reynolds-Averaged Navier Stokes solver developed by NASA. The boundary conditions used for the simulations including the wind tunnel wall were free-stream/characteristic conditions at the inflow, outflow conditions at the exit, and adiabatic walls for the wedge model and the wind tunnel wall. For the unbounded simulation, the adiabatic wall condition was replaced with the freestream/characterisitic condition. Each of the CFD simulations required between 1500 to 2000 iterations to converge, and the typical computational time for each of these simulation on a DEC ALPHA computer was ∼30 minutes. A sample result is shown in Fig. 7 where velocity contours are shown for the cases with and without an upper wall. As is evident in the figure,