Synthetic and computational vocal fold modeling: Advances and issues.
Scott L. Thomson · The Journal of the Acoustical Society of America · 2009
The use of computational and synthetic models to simulate vocal fold vibration is discussed. First, issues pertaining to computational modeling are addressed. The results of simulations using compressible and incompressible flow models coupled with identical finite element vocal fold models are compared. It is demonstrated that because incompressible flow models lack acoustic coupling, non-physical vocal fold motion may be predicted. Also discussed is the common use of contact segments to restrict medial vocal fold motion in order to prevent complete fluid mesh collapse and subsequent solver failure. The sensitivity of vocal fold vibration and intraglottal flow response to contact segment medial placement is quantified. Two advances in synthetic modeling are then presented. Using high-speed imaging, it is shown that models based on realistic vocal fold geometry (e.g., from MRI data) produce more life-like motion than models with similar sizes and material properties but more idealized geometries. Fabrication methods for modeling the thin vocal fold epithelial layer and the flexible superficial lamina propria layer are described, and corresponding improvements to vocal fold motion are demonstrated. [Work supported by NIH.]