Flux Reconstruction as a Direct Method for Near-Field Computational Aeroacoustics

The noise produced from commercial aviation is detrimental and strictly regulated at an international level. To satisfy stringent forthcoming noise reduction requirements, current industry standard, lower order aeroacoustic methods used to approximate acoustic fields must be replaced with high-order methods that can more accurately compute the acoustic field, providing invaluable insight into noise generation and potential design optimization processes. In this thesis, the accuracy and performance of high-order numerical methods, as applied in the scope of computational aeroacoustics, are evaluated. Specifically, the high-order flux reconstruction method’s ability to directly compute acoustic fields is assessed. The field of computational aeroacoustics is intrinsically dissimilar to the field of computational fluid dynamics and thus contains highly distinctive numerical challenges. Several verification studies are performed, for a range of polynomial orders, each addressing an individual numerical challenge. It is shown that the high order flux reconstruction method sufficiently resolves each of these numerical challenges, with higher order polynomials providing more accurate and efficient results on a per degree of freedom basis. The high order flux reconstruction method’s proficiency for direct computation of near-field acoustics is validated by performing simulations of flow over a cylinder and a deep cavity and comparing the results against experimental data. Finally, the performance of the high-order flux reconstruction method in industrial applications is assessed by directly computing the acoustic field produced by a NACA0012 airfoil at varying angles of attack.