A diffraction integral based turbomachinery noise shielding method

Abstract

A current research focus in subsonic aeronautics is the reduction of noise, emissions and fuel burn. The Silent Aircraft Initiative, NASA N+2 and N+3 projects are examples of recent efforts investigating innovative aircraft configurations to meet the future goals of air transportation. This requires novel methodologies to assess unconventional aircraft designs. This thesis is part of the N+2 program and focuses on the development of a method for the assessment of turbomachinery noise shielding in hybrid wing body aircraft. The preliminary design and assessment of novel aircraft configurations require both low computational cost and versatility of the shielding method. High fidelity methods, such as for example boundary element methods, are computationally expensive and not amenable for optimization framework integration. On the other hand, low fidelity methods, such as the barrier shielding method, are limited in their source and geometry definitions. The diffraction integral method is a simplified ray tracing method capturing edge diffracted rays. Creeping rays and reflected rays are not accounted for making the method suitable for flat geometries with sharp edges. It is based on the Maggi-Rubinowicz formulation of the Kirchoff diffraction theory for monopole source descriptions and is inherently a high frequency method. The diffraction line integral requires numerical integration and does not account for flight effects. A new method described in this thesis was developed to address these shortcomings. It is based on the Miyamoto and Wolf formulation of the boundary diffraction theory to allow the definition of source directivity inherent to turbomachinery noise. It is amenable to multipole and directional point source descriptions. Bulk flight effects are modelled with a generalized Prandtl-Glauert approach. Computational cost is dramatically decreased using uniform asymptotic theory to reduce the diffraction integral into a simple Fresnel integral. The Fresnel integral is solved via an analytical approximation such that the resulting shielding method does not require numerical integration. The method is applicable to three-dimensional aircraft configurations and comparison with an equivalent source method for sphere and disk shielding test cases show good agreement at high frequencies. Its analytical formulation offers compatibility with optimization frameworks facilitating new perspectives in aircraft design for noise reduction.