Fluid flow and frequency-dependent friction in the human vocal system

Abstract

A numerical procedure is developed to include friction in the water hammer model of the human vocal system. The developed procedure is based on the method of characteristics and takes into consideration frictional effects in the subglottal and supraglottal tracts. Three friction models are considered in the present study; the inviscid model, the quasi-steady friction model and the frequency-dependent friction model. Numerical results from the three models show that the quasi-steady component of the shear stress has negligible effect on the flow in the human vocal system. The frequency-dependent component of the shear stress, on the other hand, is shown to attenuate flow and pressure peaks in the subglottal and supraglottal tracts by up to 12 %, compared to the idealized inviscid flow case. Present results demonstrate that the frequency-dependent component of the shear stress dominates the quasi-steady component and should be taken into consideration for more accurate modeling of the time-varying flow in the human vocal system.