Theoretical and experimental study of glottal geometry in phonation.

Most existing theoretical models of phonation assume that the vocal folds are parallel and that the glottis forms a two-dimensional channel for the flow. However, during phonation the vocal folds can be very accurately abducted or adducted using intrinsic muscles acting on the arytenoid cartilages. The resulting shape of the glottis can then vary between an almost uniform slit (when the vocal folds are parallel) to a V shape (with an angle between the vocal folds up to 20 °). Further, the vocal folds surface can present some irregularity which can be severe, in some pathological cases such as cysts or nodules. In this paper, we present a theoretical and experimental study performed in order to evaluate and to predict these geometrical effects. A theoretical model to predict the pressure losses in such a non-uniform glottis will be presented and tested against an in-vitro experimental set-up using a self-oscillating latex replica of the vocal folds. The angle between the artificial folds could be controlled using micrometers while surface irregularities could be simulated by inserting small spheres with various masses and diameters.