Modeling Vocal Fold Motion with a Continuum Fluid Dynamic Model, I. Derivation and Analysis

Abstract

Vocal fold (VF) motion is fundamental to voice production and diagnosis in speech and health sciences. The motion is a consequence of air flow interacting with elastic vocal fold structures. Motivated by existing lumped mass models and known flow properties, we propose to model the continuous shape of vocal fold in motion by the two dimensional compressible Navier-Stokes equations coupled with an elastic damped driven wave equation on the fold cover. In this paper, instead of pursuing a direct two dimensional numerical simulation, we derive reduced quasi-one-dimensional model equations by averaging two dimensional solutions along the flow cross sections. We then analyze the oscillation modes of the linearized system about a flat fold, and found that the fold motion goes through a Hopf bifurcation into temporal oscillation if the flow energy is sufficient to overcome the damping in the fold consistent with the early models. We also analyze the further reduced system under the quasi-steady approximation and compare the resulting vocal fold equation in the small vibration regime with that of the Titze model. Our model shares several qualitative features with the Titze model yet differs in the specific form of energy input from the air flow to the fold. Numerical issues and results of the quasi-one-dimensional model system will be presented in part II (view resulting web VF animation at http://www.ma.utexas.edu/users/jxin).