Modeling of aerodynamic interaction between vocal folds and vocal tract during production of a vowel-voiceless plosive-vowel sequence.

Abstract

The context of this study is the physical modeling of speech production. The objective is, by using a mechanical replica of the vocal tract, to test quantitatively an aerodynamic model of the interaction between the vocal folds and the vocal tract during the production of a vowel-voiceless plosive-vowel sequence. The first step is to realize acoustic and aerodynamic measurements on a speaker during the production of an /apa/ sequence. The aperture and width of the lips are also derived from a high-speed video recording of the subject's face. Theoretical models to describe the flow through the lips and the effect of an expansion of the supraglottal cavity are proposed and validated by comparison with measurements made using a self-oscillating replica of the phonatory system. Finally, using these models, numerical simulations of an /apa/ sequence are performed using the measured lip parameters as the only time-varying input parameters. The results of these simulations suggest that the realization of an occlusion of the vocal tract produces a passive increase in glottal area associated with a voice offset and that the expansion of the supraglottal cavity is responsible for the extension of the phonation up to 40 ms after closure of the lips.