Elucidating Fluid Mechanisms of Droplet Formation during Human Vocalization

Vocalization during human speech is known to generate substantial quantities of expiratory droplets. While expiratory droplets are believed to transmit viruses, the fluid dynamics behind their formation during vocalization remain unclear. In this study, we simulate salivary filament stretching between two oscillating walls to examine their elongation and eventual breakup into droplets as observed in the human larynx. We model a slender, axisymmetric column of polymeric fluid held between two sinusoidally oscillating solid walls, representing the rhythmic opening and closing of vocal folds during vocalization. To capture its dynamics, we numerically solve a lubrication approximation-based equation incorporating the Oldroyd-B constitutive model. We apply a coordinate transformation to handle the moving boundaries and solve the resulting nonlinear system using an adaptive WENO (Weighted Essentially Non-Oscillatory) finite difference method. We use this framework to investigate salivary filament mechanics during physiologically relevant vocal fold operation. We quantify the number of droplets generated in each vocal cycle as a function of capillary, inertial and viscoelastic properties as well as operating conditions such as oscillatory frequency and fluid-structure pinning conditions. Put together, these results help clarify the mechanisms underlying salivary droplet formation during human vocalization, contributing to a better understanding of airborne disease transmission.