Optimization of Supersonic Jet Noise Using a Reynolds-Averaged Navier-Stokes Approach

The objective of this proposal is the development and demonstration of a cost-effective high-fidelity aeroacoustic design tool for future commercial supersonic nozzle designs and installations. Although eddy-resolvingnbsp; CFD methods for computing high-speed jet noise are available, such methods are computationally expensive and are currently deemed impractical for use in a design optimization loop. On the other hand, the prediction of turbulence generated noise using the Reynolds-Averaged Navier-Stokes (RANS) equations provides a less accurate but more cost-effective approach for practical design problems, wherein the turbulence length and time scales needed to model the local noise source terms can be extracted from the RANS turbulence model solution, as performed by the NASA JeNo code. The objective of this project is to develop a tightly coupled practical RANS-based jet-noise aeroacoustic analysis and design optimization capability which can be leveraged by government and industrial customers to better understand and design efficient propulsion systems that meet well defined and accepted noise metrics. The approach consists of developing an exact discrete adjoint method for a tightly coupled RANS-acoustic prediction method in order to provide sensitivities for gradient-based aerodynamically constrained acoustic optimizations. This capability will be demonstrated on realistic nozzle configurations including single and dual stream chevron nozzles, and marketed to government and industrial customers in the aerospace industry.