A numerical investigation of sound radiated from subsonic jets with application to human phonation
Abstract
In this thesis, numerical studies of sound radiation from subsonic jets with application to human speech production are presented. A highly accurate numerical method which features high-order compact finite difference schemes and pseudo-spectral methods for spatial discretization and a fourth-order Runge-Kutta method for time advancement is implemented. The vortex ring generated by a starting jet is simulated and the results show very good agreement with experimental data. Direct Numerical Simulation (DNS) of the sound radiated by a subsonic axisymmetric jet is performed. A Kirchhoff method and Lighthill's acoustic analogy are used to predict the far-field sound using the DNS near-field results. Both predictions are in excellent agreement with the directly computed sound field. The effect of spatial filtering on the sound radiation is assessed by an a priori study. Large Eddy Simulation (LES) of a turbulent jet is performed using a dynamic Smagorinsky SGS model and a dynamic mixed model. (qualitative comparison of the near field results are made to the previous computational and experimental results with similar flow conditions. The far-field sound directivity is predicted using the Kirchhoff method, and the results are in good agreement with experimental data. Unsteady flow and sound radiation of axisymmetric pulsating jets inside simplified static and dynamic models of the human vocal tract are simulated. Ffowcs Williams-Hawkings equation is used to identify different sources of sound for human voice production. Finally, a quasi-one-dimensional model is developed for speech production. The model shows some promising prospects for future practical application.
Degree
Ph.D.
Advisors
Mongeau, Purdue University.
Subject Area
Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.