Integrated aerodynamic optimization of diverterless supersonic inlet using full turbulence adjoint method

Diverterless supersonic inlets (DSI) is the predominant system for supersonic aircraft below Mach 2, offering superior boundary layer displacement and stability margins. However, conventional discrete adjoint optimization frameworks relying on the constant eddy viscosity (CEV) assumption introduce significant gradient errors in turbulent flow sensitivity analysis, limiting performance refinement of DSI configurations characterized by shock-boundary layer interactions and separation phenomena. This study addresses this challenge through a novel discrete adjoint framework integrating upwind schemes and full turbulence adjoint modeling, enabling accurate gradient computation while maintaining numerical stability for high-speed flows. The method is applied to optimize an integrated DSI system encompassing compression bump, diffuser duct, and cowl lip geometries. Gradient validation against finite-difference results demonstrates improved convergence efficiency and sensitivity accuracy compared to CEV-based approaches. At Mach 1.8, the optimized DSI achieves shock-on-lip (SOL) conditions with 4.0% total pressure recovery improvement (0.914) and 64.2% reduction in distortion index (0.273). Off-design performance at Mach 0.6–0.9 maintains exceptional recovery coefficients (0.982/0.973), demonstrating multi-point operational robustness. The design simultaneously minimizes internal flow separation, enhances boundary layer control, and balances supersonic pressure recovery with sub/transonic efficiency. This work advances adjoint-based aerodynamic optimization by resolving CEV limitations in capturing viscous-inviscid interactions, establishing a pathway for next-generation DSI development.