Towards Chevron Nozzle Shape Optimization Using a Hybrid CFD/CAA Method

AIAA/CEAS Aeroacoustics Conference

Abstract

In this paper, a hybrid CFD-CAA method is used to predict the jet noise from chevron nozzles, which is combined with a multi-objective Bayesian shape optimization strategy to reduce the far field overall sound pressure level. The acoustic near field is predicted by coupling the solution of the acoustic perturbation equations to the flow field simulation. This approach is extended using the FWH equation based on the acoustic near-field data to predict the acoustic far-field. First, highly resolved aeroacoustic simulations of the jet noise from a benchmark chevron nozzle at a Reynolds number of Re = 1 · 10^6 are discussed. The noise spectra show good agreement with the reference data across a wide frequency range, where the predicted sound pressure levels are insensitive to the chosen location of the FWH surface. Computational performance results of the hybrid CFD-CAA method on up to 4096 compute nodes of a CPU based HPC system demonstrate that the numerical approach is highly scalable, enabling the execution of large-scale simulations in short time frames. A simulation setup with medium mesh resolution and a parameterized chevron nozzle shape is then used for a simplified shape optimization. Since the simulations are based on hierarchical Cartesian grids with automatic mesh generation, a prediction of the sound field can be fully automated. This enables the integration of the aeroacoustic workflow in an optimization loop to perform a chevron nozzle shape optimization for noise and thrust loss minimization. Two studies for the optimization of a chevron nozzle first described by two and then four shape parameters are presented. In total about 100 simulation runs are performed. The results demonstrate the viability of the approach to perform a multi-objective shape optimization, allowing to balance the competing targets of noise reduction and minimal thrust loss.

Authors 5

1. Ansgar Niemöller Aachen

   RWTH Aachen University

   Affiliation as printed

   Rheinisch-Westfalische Technische Hochschule Aachen

2. Fabian Hübenthal Aachen

   RWTH Aachen University

   Affiliation as printed

   Rheinisch-Westfalische Technische Hochschule Aachen

3. Vaishali Ravishankar

   Fraunhofer Institute for Algorithms and Scientific Computing

   Affiliation as printed

   Fraunhofer-Institut fur Algorithmen und Wissenschaftliches Rechnen SCAI

4. Matthias Meinke Aachen

   RWTH Aachen University

   Affiliation as printed

   Rheinisch-Westfalische Technische Hochschule Aachen

5. Dominik J. Krug Aachen

   RWTH Aachen University

   Affiliation as printed

   Rheinisch-Westfalische Technische Hochschule Aachen