Hysteresis of aeroacoustic sound generation in the articulation of [s]

Abstract

A fricative consonant (e.g., [s]) is known to be pronounced by a turbulent jet flow inside the oral cavity. In this study, the effects of tongue motion on the aeroacoustic sound generation during the articulation of [s] were investigated through the large eddy simulation of compressible flow using a simplified vocal tract model. The walls of the simplified model were expressed using a volume penalization approach as an immersed boundary method, and the tongue geometry was ascended and descended from the position of /u/ to /s/ with tongue speeds of 40 mm/s, 60 mm/s, and 80 mm/s. The simulated acoustic pressure at a far-field sampling point was compared with previous experimental measurements, and the acoustic characteristics of the simulated sound reasonably agreed with those of the experiment. The overall acoustic amplitudes increased and decreased in accordance with tongue ascent and descent, and these transitions in amplitudes were almost the same for the different tongue speeds. Meanwhile, we found a hysteresis effect on the overall acoustic amplitudes between tongue ascent and tongue descent. This hysteresis was caused by the larger velocity fluctuations and vortices near the upper and lower teeth during tongue descent, and the results indicated that these flow differences occurred owing to the inertia of the turbulent flow structures and the aerodynamic pressures over the constriction of the vocal tract. This study suggests that these phenomena cause a delay in the sound generation of [s].