Turbulent Flow Dynamics in Sustained Vowel Production: A Computational Study

Abstract

Objective: The current study investigates the fluid flow dynamics in the vocal tract for the sustained phonation of vowel /ɑ/. Previous research utilized the true geometry of the vocal tract to study flow dynamics for several fricatives in the absence of vocal folds’ motion. The present study aims to incorporate the true geometry of the vocal tract into a computational model of vowel phonation while correlating the vocal folds vibration with the aero-acoustics of the fluid flow at different speech intensities and frequencies.

Methods: A two-dimensional true vocal tract geometry, extending from the vocal folds to the mouth cavity for the sustained phonation of vowel /ɑ/ and based on MRI data, is adopted in the present study. Laryngeal high-speed videoendoscopy (HSV) data during production of sustained vowel /ɑ/ was recorded from a normophonic speaker. The glottal area waveform was extracted from the HSV data using a deep neural network approach. This waveform was used for validating the designed computational fluid dynamics model. The flow field is simulated by solving the Navier-Stokes equations using the velocity-vorticity formation (V2F) turbulence model, which captures the flow turbulence.

Results and Conclusions: The results show for different types of speech intensities and vibrational frequencies, the glottal jet skews, highlighting the presence of the Coanda effect. Through our analysis, we can specify the vertical position, where the skewness is initiated. We also see how the location and degree of glottal jet skewness, length scale, and spatial distribution of the vortical structures vary with time and we can correlate them directly to the sustained phonation of the vowel /ɑ/. Further, we have observed variations of the boundary layer and pressure profile along the vocal tract wall with time, speech intensity, and the vibrational frequency. Through these flow field analysis techniques, the current study determines the effects of turbulence on flow dynamics for the sustained phonation of vowel /ɑ/.

Acknowledgments: We acknowledge the support from NIH NIDCD K01DC017751 and R21DC020003, the ARO Young Investigator Program award W911NF-19-1-0444, and NSF award DMS-1923201. The High-performance Computing resources were provided by the Institute for Cyber-Enabled Research at Michigan State University.