Linearly diverging vocal folds and the Coanda effect

Abstract

We would like to verify whether the Coanda effect has a significant impact when incorporated into theoretical vocal fold models that assume a piecewise linear shape of the vocal fold walls, as many do. We model the intraglottal flow with the equations of Thwaites. Thwaites boundary layer theory gives simple criteria for the glottal jet separation point if the vocal folds diverge linearly, and even validates well-known empirical observations for the jet width at flow separation. We test this criteria against flow experiments with rigid vocal fold replicas. The experiments involve symmetric and asymmetric vocal fold configurations, as well as steady and unsteady flow. We then validate the significance of the predicted Coanda effect on several numerical models of human vocal folds. We test the significance of the effect both on mechanically symmetric vocal fold models and on ones with mechanical asymmetries. We find limited effects on symmetric vocal folds and varying degrees of impact on asymmetric vocal folds.