An immersed boundary method (IBM) suitable for aeroacoustic analysis is developed using the block-structured Cartesian mesh method called the building-cube method. The IBM using a ghost cell approach in conjunction with an image point is employed. A hybrid interpolation scheme combining the wavenumber-based method and inverse distance weighting method is introduced to the IBM. This method is evaluated using four test cases. The results of interpolation at an image point indicate that the wavenumber-based method provides more accurate results than standard interpolation methods for aeroacoustic problems. The results of acoustic scattering around a sphere show that the proposed IBM achieves better than second-order accuracy. In the case of noise propagation from a generic inlet, the proposed IBM is shown to yield more accurate predictions for the far-field sound pressure level than the inverse distance weighting-based IBM. In a different application involving the JT15D nacelle, the far-field sound pressure level based on the proposed IBM is in good agreement with the experiment and computational results reported in a previous study. These results validate the conclusion that the proposed IBM can compute noise propagation with a mean flowfield and accurately estimate the far-field sound pressure level.
ASJC Scopus subject areas