Tailoring biomechanical model meshes for aero-acoustic simulations

To simulate the airflow and acoustic wave propagation associated with voice production, a closed surface mesh representing the vocal tract is needed. Biomechanically, the vocal tract is composed of surfaces from several different anatomical structures. We present a method for assembling a dynamic vocal tract mesh by trimming and stitching surface meshes tracking biomechanical models of relevant structures. Two algorithms, one for trimming and one for stitching, are used to first isolate surface mesh patches that are in contact with the airway and then merge them into a closed surface. The algorithms rely on manually selected boundaries and are able to cover gaps between mesh patches. Test cases are used to illustrate how the algorithms behave in various situations. The algorithms are implemented in the toolkit ArtiSynth where many relevant biomechanical models are already available.