Investigation of Interpolation Strategies for Hybrid Schemes in Computational Aeroacoustics

One key challenge in the numerical simulation of aeroacoustic fields is the huge disparity of scales between flow structures and audible acoustic wavelengths. A volume discretization, resolving both scales from the generating vortex to the desired microphone position, not only leads to a very high number of cells but also to a very small time step size in order to minimize dissipation of the acoustic waves. Hybrid schemes in computational aeroacoustics (CAA) separate the flow from the acoustic computation by using aeroacoustic analogies. It seems adequate to use different computational grids for computing each field in an optimal manner. As a result, a factor of 100 between the cell edge lengths of flow and acoustic grids is not uncommon in low Mach number flows. The fundamental requirement is an accurate data transfer from one grid to the other with minimal interpolation errors. To cope with this task, different strategies can be applied, starting from low complexity nearest neighbour interpolation to complex volume intersections between flow and acoustic grid. Within our contribution we will investigate two approaches. The first one is a conservative interpolation strategy which features a low computational complexity but can lead to errors if certain conditions for CFD and acoustic grids do not meet certain conditions. The second scheme can be seen as an extension an copes with situations in which the flow grid has a cell size in the same range or bigger than the acoustic grid. Both schemes are investigated for two different aeroacoustic analogies.