Computational aeroacoustics of human phonation

Abstract The current paper presents a CFD model of flow past vibrating vocal folds coupled to an acoustic solver, which calculates the sound sources from the flow field in a hybrid approach. The CFD model is based on the numerical solution of 3D Navier-Stokes equations on a time-dependent domain, solved by cell-centered finite volume method. To capture the fine turbulent scales important for the acoustic source calculations, the equations are discretized and solved on large computational meshes up to 3.2M elements. The CFD simulations were run in parallel using domain decomposition method and OpenMPI implementation of the MPI standard. Aeroacoustic simulations are calculated in a separate step by Lighthill’s acoustic analogy, which determines the acoustic sources based on the fluid field. This is done with the research code CFS++ which employs the finite element method (FEM).