A review of lumped-element models of voiced speech
Speech Communication
| Hi-index | 0.00 |
Human phonation does not merely depend upon the vibration of the vocal folds. The false vocal fold (FVF), as an important laryngeal constriction, has also been found by more and more research both in clinically and computer simulations that it plays an important role during phonation and contributes significantly to the aerodynamics and sound generation processes of human voice production. Among many parameters which are used to determine and describe the geometry of the false vocal folds, the false vocal fold gap (GFVF), which means the minimal distance between the two false vocal folds, is regarded as an important and dominant parameter. Therefore, this study explores the effects of the FVF gaps on the intralaryngeal pressure distributions, laryngeal resistance and flows using both three-dimensional Plexiglas model and commercially available computational fluid dynamics code. Three glottal angles, divergent 40°, uniform 0°, and convergent -40°were used for this study to explore the effects of FVF gaps, as they represent the basic glottal shapes typically expected in phonation, the angle values also were typically expected for most phonation in modal Register. A wide variety of FVF gaps (GFVF) were parameterized with 12 different values: 0.02, 0.04, 0.06, 0.08, 0.09, 0.1, 0.11, 0.12, 0.16, 0.2, 0.4, 0.6 cm to represent important geometries often appearing within phonatory vibratory cycles. These gaps were used for each glottal angle. The specific design of the FVFs followed prior literature. The minimal glottal diameter (Dg) was constantly at 0.06 cm in this study for each FVF gaps, and the translaryngeal pressure were held constant at 8 cm H2O. A nonvibrating laryngeal airway Plexiglas model, which had linear dimensions 1.732 times of a normal male larynx, was used in this study. In order to measure pressures inside the Plexiglas model, twelve cylindrical ducts were made on the midline of the laryngeal wall of the model. The diameter of each duct was 0.07 cm human size (0.12 cm in the model), so that the connector of an Entran EPE-551 pressure transducer could fit snugly into the holes. The distance between the centers of each hole was 0.14 cm human size. FLUENT (Fluent, Inc., Lebanon, NH), a commercially available computational fluid dynamics code was also used to obtain estimates of the normal wall pressures along the laryngeal surfaces (including the surfaces of vocal folds, ventricles, and false vocal folds) as a function of the FVF gaps and the glottal angles. The code is based on the control-volume technique and was used to solve the Navier-Stokes equations for constant shapes (not for vibrating vocal folds), laminar, incompressible airflow physics occurring inside the symmetric laryngeal geometries. The flow field was assumed to be symmetric across the midline of the glottis in this study, and therefore only the half flow field was modeled. The results suggest that (1) the intralaryngeal pressure was lowest and the flow was highest (least flow resistance) when the FVF gap was 1.5-2 times of Dg, the intralaryngeal pressures decreased and flows increased as smaller FVF gaps increased, and the intralaryngeal pressures increased and flows decreased as larger FVF gaps increased, indicating that the least pressure drop for any given flow (that is, the least flow resistance) was found to correspond to the 1.5-2 times of Dg for different glottal angle. Suggesting that the 1.5-2 times of Dg might be the optimal gap for pressure, and efficient phonation may involve laryngeal shaping of this condition. Therefore, the positioning and existing structure of the FVFs can aid in phonation by reducing energy losses and increasing airflow in the larynx when positioned appropriately; (2) both the pressure and flow were unaffected when the FVF gaps larger than 0.4 cm; (3) the divergent glottal angle gave lower pressure and greater flow than the convergent and uniform glottal angle as no FVF conditions; (4) the present of the FVF decreased the effects of the glottal angle on both the intralaryngeal pressure and flow to some extent, and the smaller the FVF gaps, the smaller this effect. Perhaps more important, (5) the present of the FVF also moving the separation points downstream, straitening the glottal jet for a longer distance, decreasing overall laryngeal resistance, and reducing the energy dissipation, suggesting that the FVF would be of importance to efficient voice production; (6) the empirical pressure distributions were supported by computational results. The results suggest that the intralaryngeal pressure distributions and the laryngeal flow resistance are highly affected by the presence of the FVFs, and the FVFs can aid in phonation when by reducing energy losses positioned appropriately. Therefore, the results might be helpful not only in maintaining healthy vocal habits, but also in exploring surgical and rehabilitative intervention of related voice problem. The results also suggest that they may be incorporated in the phonatory models (physical or computational) for better understanding of vocal mechanics.