Phonation threshold pressure using a 3-mass model of phonation with empirical pressure values.

Understanding the control parameters that influence phonation threshold pressure can have important implications for ease of phonation. Using a computer model of phonation can aid in studying parameters not easily controllable through human experimental work and may provide a means of explaining variations seen across human participants. A vertical 3-mass computer model of phonation with empirical driving pressures was used to obtain phonation threshold pressures for a variety of prephonatory conditions that may be realistically produced by humans. The resulting phonation threshold pressures are reasonable compared to results from human studies and may extend beyond the range of phonatory control parameters studied in human experiments. In addition, the present work adds a formula for calculating phonation threshold pressure based on the prephonatory glottal angle, the tension of the vocal folds, and the prephonatory diameter. Of special interest is that, as the prephonatory angle of convergence increases from 0 degrees (the rectangular glottis condition), the phonation threshold pressure increases in a nearly linear fashion.

Modeling coupled aerodynamics and vocal fold dynamics using immersed boundary methods.

The penalty immersed boundary (PIB) method, originally introduced by Peskin (1972) to model the function of the mammalian heart, is tested as a fluid-structure interaction model of the closely coupled dynamics of the vocal folds and aerodynamics in phonation. Two-dimensional vocal folds are simulated with material properties chosen to result in self-oscillation and volume flows in physiological frequency ranges. Properties of the glottal flow field, including vorticity, are studied in conjunction with the dynamic vocal fold motion. The results of using the PIB method to model self-oscillating vocal folds for the case of 8 cm H20 as the transglottal pressure gradient are described. The volume flow at 8 cm H20, the transglottal pressure, and vortex dynamics associated with the self-oscillating model are shown. Volume flow is also given for 2, 4, and 12 cm H2O, illustrating the robustness of the model to a range of transglottal pressures. The results indicate that the PIB method applied to modeling phonation has good potential for the study of the interdependence of aerodynamics and vocal fold motion.

Analytic representation of volume flow as a function of geometry and pressure in a static physical model of the glottis.

A static physical model of the larynx (model M5) was used to obtain a large set of volume flows as a function of symmetric glottal geometry and transglottal pressure. The measurements cover ranges of these variables relevant to human phonation. A generalized equation was created to accurately estimate the glottal volume flow given specific glottal geometries and transglottal pressures. Both the data and the generalized formula give insights into the flow behavior for different glottal geometries, especially the contrast between convergent and divergent glottal angles at different glottal diameters. The generalized equation produced a fit to the entire M5 dataset (267 points) with an average accuracy of 3.4%. The accuracy was about seven times better than that of the Ishizaka-Flanagan approach to glottal flow and about four times better than that of a pressure coefficient approach. Thus, for synthesis purposes, the generalized equation presented here should provide more realistic glottal flows (based on steady flow conditions) as suitable inputs to the vocal tract, for given values of transglottal pressure and glottal geometry. Applications of the generalized formula to pulses generated by vocal fold motions typical of those produced by the Ishizaka-Flanagan coupled-oscillator model and the more recent body-cover model of Story and Titze are also included.