Numerical simulation of the sibilant /s/ sound articulation process

The numerical simulation of the articulation process of the sibilant /s/ sound was conducted with the moving realistic vocal tract geometry of the Japanese word /usui/ using an implicit compressible flow solver. The moving immersed boundary method with the hierarchical structure grid was adopted to approach the complex geometry of the human speech organs. The sound emitted from the vocal tract was analyzed by the spectrogram to clarify the effect of velopharyngeal closure. The sibilant /s/ sound was gradually generated by reducing the constriction size. However, while the velopharyngeal valve was not completely closed, the sound was affected by the nasal emission. After the closure of the valve, a typical /s/ sound was produced by the turbulent airflow in the vocal tract. These results provide the underlying insights necessary for clinical treatment of speech difficulty for sibilant sound.