Feature Based Mesh Adaptation Applied on the AIAA PAW-6 Nozzle Jet Prediction

This paper highlights the advantages of anisotropic mesh adaptation for predicting the jet plume of an axisymmetric supersonic jet nozzle, as studied in the AIAA PAW-6 workshop. The exhaust nozzle is a crucial component of the engine because its sensitivity to net thrust output is greater compared to other engine components. Various designs exist; however, at supersonic Mach numbers, Convergent-Divergent (CD) nozzles are necessary to efficiently expand the hot gases into the atmosphere. Consequently, CD nozzles have been extensively studied over the years. During the 6th AIAA Aerodynamic Workshop, computational fluid dynamics (CFD) results using handcrafted meshes were compared with nonintrusive experimental data. The calculations required prior knowledge of flow physics by experienced engineers to produce high-quality meshes. The aim of this paper is to demonstrate the applicability of a feature-based remeshing framework that requires no prior knowledge of the flow physics. A further objective is to highlight how this framework can support turbulence-model developers in assessing the impact of model modifications. The study focuses on the off-design adiabatic case 4, also documented in the Turbulence Model Research Database, and achieves mesh–solution convergence starting from a very coarse mesh generated without any a priori information on the flow. The resulting predictions are then compared with both the PAW-6 workshop data and the available experimental measurements.