Comparative optimization of hypersonic waveriders using analytical and computational methods

• Waverider optimization via lower-surface polynomials and leading edge perturbations • GPU-accelerated CFD workflow for waverider optimization and trajectory analysis • Waverider optimization using analytical models and high-fidelity CFD • Key differences revealed between analytical and CFD-based optimization • Waveriders geometries optimized using different aero-performance models, Euler CFD flow fields at Mach 6.85, 25 km altitude. This work presents a methodology for designing waverider geometries that incorporate leading edge rounding and parameterized surface perturbations, and evaluates their performance using two distinct aerodynamic prediction approaches: a reduced-order analytical/empirical model and a GPU-based Euler CFD solver, both applied with an empirical approach to estimate the viscous contributions. Geometry optimization was performed using a global pattern search algorithm, and optimal trajectories were computed using a direct collocation method to assess the impact of aerodynamic modeling fidelity on mission performance. Cross-comparisons between these aerodynamic performance prediction methods revealed that CFD-based optimization produces geometries that capture off-design flow phenomena that are absent from analytical-based optimizations. As a result, CFD-optimized waveriders demonstrated higher average lift-to-drag ratios than those derived from the analytical model. The analytical trajectories had lower angle-of-attack control and degraded in accuracy for vehicles with high pressure losses. These findings highlight the limitations of traditional analytical methods and underscore the feasibility of using modern GPU-accelerated CFD for resource-intensive tasks such as vehicle optimization and trajectory analysis.