Aerodynamic Shape Optimization of Doubly Offset Serpentine Diffuser using Response Surface Methodology

Abstract

Doubly offset serpentine diffusers have gained popularity in the compact design configurations of modern stealth fighters and UAVs with highly integrated propulsion systems into the airframe. In this research, the design space of a doubly offset serpentine diffuser is explored and the numerical optimization of its shape variables is achieved using response surface methodology to maximize total pressure recovery at the Aerodynamic Interface Plane between the engine and inlet. The stream-wise and transverse pressure gradients in the baseline diffuser are controlled using area distribution and centerline distribution equations respectively. The original geometry is perturbed using three control points distributed uniformly along the centerline and the central composite design has been used to select a pool of candidate designs. A steady-state flow solution has been achieved using governing Reynolds averaged Navier-Stokes equations applied through the general-purpose computational analysis tool ANSYS Fluent. A response surface is constructed out of the training data by fitting quadratic polynomials to the pressure recovery coefficients. The optimal diffuser design is found using a standard optimization algorithm from the response surface approximations. The optimized shape encompasses potential improvement in the total pressure recovery by 1.1% as compared to the baseline geometry. Results reveal that diffuser performance is a complex function of its geometric shape and any slight change in its shape variables may lead to significant performance degradation.