RANS-Based Design Optimization for Sonic Boom Ground Noise Minimization

This paper presents a gradient-based aerodynamic shape optimization framework for minimizing ground-level sonic boom noise using a partitioned nearfield-farfield analysis approach. High-fidelity Reynolds-averaged Navier–Stokes simulations are performed in the nearfield using a finite-difference discretization on curvilinear overset grids, while farfield acoustic propagation is modeled with ray acoustics and the augmented Burgers equation solver. A coupled adjoint methodology provides efficient sensitivity calculations across the nearfield and farfield domains, enabling the use of ground-level noise metrics as optimization objectives. The use of curvilinear overset grids allows the off-body Mach-cone-aligned grid to remain fixed during optimization, preserving solution accuracy while deforming only the near-body grid. Three test cases demonstrate the framework’s capabilities: minimization of the ground signature of a two-dimensional diamond airfoil, minimization of the A-weighted sound exposure level for the JAXA wing-body configuration, and the inverse design of the NASA Concept 25D low-boom configuration. Results show good reduction of noise metrics, demonstrating the effectiveness of the proposed optimization methodology for low-boom supersonic aircraft design.