Comparison of Aerodynamic Noise Propagation Techniques Markus P. Rumpfkeil * University of Dayton, Ohio, 45469, USA Darrel K. Robertson † University of Dayton Research Institute, Dayton, Ohio, 45469, USA Miguel R. Visbal ‡ Air Force Research Laboratory, Wright-Patterson AFB, Ohio, 45433, USA In this paper the computation of aerodynamic noise via different hybrid noise prediction methods is presented. An unsteady Reynolds-averaged Navier-Stokes (URANS) and a large eddy simulation (LES) solver are coupled to several wave propagation codes such as an in-house two-dimensional Ffowcs Williams and Hawkings (FW-H) acoustic analogy method and several solvers from the commercially available LMS Virtual Lab package. Application examples include the turbulent flows over a cylinder, a NACA 0012 as well as a SD 7003 airfoil. Both the aerodynamic and acoustic results show reasonable agreement with experimental data. I. Introduction and Motivation Airframe-generated noise is an important component of the total noise radiated from aircraft, especially during aircraft approach and landing, when engines operate at reduced thrust, and airframe components (such as high-lift devices) are in the deployed state. 1–4 Future Federal Aviation Administration noise regulations, the projected growth in air travel, and the increase in population density near airports will require future aircraft to be substantially quieter than the current ones. Consequently, the attempt to understand and reduce airframe noise has become an important research topic. 5 A computational obstacle that immediately arises in problems concerning noise is that accurate prop- agation of the pressure (noise) signatures over a large number of wavelengths can only be obtained with very small computational mesh spacings. This makes all external aerodynamic noise problems, where the computational domain has to cover tens or hundreds of chord lengths, infeasible for even today’s largest computers. A typical approach to tackle noise problems nonetheless is to represent the computational fluid dynamics (CFD) solution on a reasonable computational mesh that does not extend too far from the aircraft. The location of a fixed near-field plane within the computational mesh can then be specified as shown in Figure 1. This near-field plane or surface serves as an interface between the CFD solution and a wave propagation program based on principles of geometrical acoustics and/or nonlinear wave propagation. 6–8 Such a program is able to model the wave propagation and to calculate the pressure fluctuations at a user specified ground plane which can then be used as a measure of the generated noise. Several prediction methodologies for far-field signals based on near-field flow inputs are currently available and this approach is known as hybrid noise prediction method. The most popular prediction methodologies are the Kirchhoff approach 9, 10 and the Ffowcs Williams and Hawkings (FW-H) approach 11 which is based on the Lighthill acoustic analogy. 12 Examples of hybrid predictions using unsteady Reynolds-averaged Navier-Stokes (URANS) solutions as near-field inputs can be found in the literature for a supersonic cavity flow, 13 radiated sound for a circular * Assistant Professor, Dept. of Mechanical and Aerospace Eng., Markus.Rumpfkeil@udayton.edu, Senior Member AIAA † Senior Research Engineer, Senior Member AIAA ‡ Technical Area Leader, AIAA Fellow 1 of 13 American Institute of Aeronautics and Astronautics Downloaded by Markus Rumpfkeil on February 4, 2014 | http://arc.aiaa.org | DOI: 10.2514/6.2014-0021 52nd Aerospace Sciences Meeting 13-17 January 2014, National Harbor, Maryland AIAA 2014-0021 Copyright © 2014 by Markus P. Rumpfkeil, Darrel K. Robertson and Miguel R. Visbal. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission. AIAA SciTech