Aerodynamic Design Optimization of a Supersonic Transport Aircraft Considering Low-Speed Stability

Abstract

Designing supersonic transport (SST) aircraft requires accounting for performance and stability at high-speed and low-speed conditions. Previous work demonstrated that there is a trade-off between high-speed performance and low-speed stability. The objective of this dissertation is to use optimization to study design trends for SSTs, with a focus on low-speed stability.

I first assess the accuracy of Reynolds-averaged Navier–Stokes (RANS) and delayed detached eddy simulations (DDES) at predicting the vortex-dominated flows that are relevant for SSTs at subsonic conditions such as takeoff and landing. I compare the predicted aerodynamic coefficients with experimental data for a delta wing, which is a simplified representation of an SST wing. RANS accurately predicts vortex effects in the steady regime but is inaccurate at high angles of attack where the flow is unsteady. DDES is more reliable in the unsteady regime, but the computational cost is at least 100 times that of RANS. This motivates the RANS-based analysis and optimization in the remainder of the dissertation.

Next, I develop methods to improve the speed and accuracy of RANS solvers for the low and high Mach number flow regimes that are relevant for SST design. I propose a method for scaling the artificial dissipation terms in the Jameson–Schmidt–Turkel scheme to improve its accuracy at low Mach numbers while retaining the simplicity of the original scalar dissipation formulation. In addition, I introduce a dissipation-based continuation method for flows with shocks that improves robustness and accelerates convergence without sacrificing accuracy.

The second half of this dissertation focuses on aerodynamic design optimization of a full SST configuration. I perform RANS-based aerodynamic shape optimization to minimize drag at a supersonic cruise condition with and without a static margin constraint at a subsonic takeoff condition. The stable optimized designs use larger leading-edge flap deflections at the subsonic condition and have thicker wings. The increased wing thickness results in a 0.5% increase in supersonic drag for neutral stability and a 0.85% increase in supersonic drag for a 10% static margin. These results demonstrate that aerodynamic shape optimization is a valuable tool for designing SSTs accounting for supersonic performance and subsonic stability.

Finally, I investigate the impact of different trim surface configurations on SST design. I first use RANS-based optimization to compare the trim drag at a supersonic cruise condition for three-surface, canard, and conventional variants of an SST. The three-surface configuration has the lowest trim drag at the supersonic condition. I then construct a supersonic buildup model to study the effects of variable trim surface sizing. When the trim surface spans are included as design variables, the design for minimum supersonic drag has no horizontal stabilizer and a canard sized at 39% of the wing span. However, the optimized canard configuration is unstable at subsonic conditions. This emphasizes the need to simultaneously consider subsonic stability and supersonic performance for SST design.