Numerical Simulation of the Self-oscillating Vocal Folds in Interaction with Vocal Tract Shaped for Particular Czech Vowels

摘要

The study presents a two-dimensional (2D) finite element (FE) model which consists of the vocal folds (VF), the trachea and idealized vocal tract (VT) shaped for Czech vowels [a:], [i:] and [u:] created from magnetic resonance images (MRI). Such configuration enables solving fluid-structure-acoustic interaction, flow-induced self-oscillations of the VF and acoustic wave propagation in the VT by explicit coupling scheme with two separate solvers for structure and fluid domain. The FE model of the VF includes the VF pretension and setting to phonatory position, large deformation of tissues and VF contact. Unsteady viscous compressible airflow through the FE model of the trachea, glottis and the VT is modelled by using Navier-Stokes (NS) equations. Moving boundary of the fluid domain (according to the VF motion) is solved by the Arbitrary Lagrangian-Eulerian approach. The solution obtained for the FE models is analyzed and the effect of the VT shape on the spectra of the generated acoustic pressure at the lips is discussed and the results are compared with measured data published in literature.