Characterization of Contact Forces on Synthetic Vocal Folds Using Two-Dimensional Digital Image Correlation Open Access
Downloadable ContentDownload PDF
Background and Motivation: From the cry of an infant to the Gettysburg address, the human voice is a powerful instrument in society. However, abnormal functioning of the phonatory system can lead to vocal fold disorders that impair voiced speech, such as polyps, nodules, and tissue bowing. Approximately one-third of the population will experience a vocal impairment at some point in their lifetime. Because this oral communication is such a fundamental tool for humans, it is important to understand the development of voice disorders and how they affect speech. Objectives: This study specifically aims to measure the contact forces experienced by vocal folds during self-sustained oscillation. Excessively high contact forces are believed to lead to formation of nodules. To accomplish the objective, an open source, two-dimensional digital image correlation (2D DIC) software package called "Ncorr" ( http://www.ncorr.com/) will be applied to analyze synthetic vocal fold models tested using an in vitro vocal tract with interchangeable supraglottal tract inserts that replicate the effects of throat and mouth shape for different vowel sounds. Methods and Results: To validate the software, a computer simulation as well as an experimental assessment of the software were completed. The computer simulation used Photoshop to impose known displacements on a synthetically developed image created in MATLAB. To experimentally validate the software, a tensile coupon representative of human vocal folds (Material: EcoflexTM 00-30) was fabricated, and a random speckle pattern was airbrushed onto its surface using Krylon Leather Brown Gloss. Known strains were imposed on the coupon using a custom-made tensile testing setup. Imaging was performed using an IDT NX4 camera and Motion Studio software. Post-processing was completed using the Ncorr-software to generate displacement and strain fields for both validation tests, which resulted in simulation and experimental strain errors of 8.3% and 10.3%, respectively. Discussion: Synthetic vocal fold models are being prepared for additional experiments and 2D DIC strain analysis. These models are fabricated using a standardized "M5" geometry and materials that match the physiological properties of the different vocal fold layers. The vocal fold models will be tested using an in vitro vocal tract where compressed air is introduced using an adjustable volume, acoustically-treated, constant pressure chamber that simulates the pressure developed in the lungs during phonation. Acknowledgements: Supported by the GW Center for Biomimetics and Bioinspired Engineering and the SEAS SUPER program.