Accuracy of Aerodynamic Pre-dictions and its E ects on Supersonic Transport Design

An investigation of the aerodynamic modeling requirements for HSCT design is made. The purpose of this study is to determine the effects of including Euler and Navier-Stokes calculations for the supersonic aerodynamics and structural loading predictions in a multidisciplinary HSCT design optimization procedure. The accuracy, computational effort, and ease of implementation of the CFD analyses into the optimization process are compared with that of simpler models based on a combination of linear theory, slender body theory, and boundary layer theory. The accuracy of these simple models and grid-converged CFD solutions is quantified through comparisons with experimental data on analytic forebodies and wings. Differences in the cruise drag predictions of up to 2 counts are seen in the designs considered, which have cruise drag coefficients ranging from 70 to 85 counts. While this appears to be a small discrepancy, it has a large effect on the aircraft range and weight. A two count increase in cruise drag corresponds to a 120 n.mi. reduction in the range and a 56,000 lb increase in the take-off gross weight (TOGW) for an aircraft with a 5500 n.mi. range and a nominal TOGW of 770,000 lb. There is little difference noted in the wing loadings predicted from Euler and parabolized Navier-Stokes analyses. Linear theory loads match CFD loads well at cruise conditions, but not at high-lift conditions where crossflow shocks appear on the wing. Wing bending stresses are sensitive to small variations in the loading and spanwise center of pressure location. Even at cruise conditions where linear theory and Euler loads show good agreement, there are differences in the maximum bending stress of up to 26.5%. This sensitivity is a result of the fuel weight in the wings canceling much of the wing bending due to aerodynamic loading. Despite the large variations in the stresses and loads at high-lift conditions using the different aerodynamic models, the optimized wing bending-material weights show relatively small variations up to 1800 lb. Significant gains in the accuracy of our aerodynamic performance predictions can be realized using CFD calculations in the optimization, but the large computational requirements preclude the sole use of Euler or Navier-Stokes solutions in an efficient high-dimensional multidisciplinary aircraft optimization.