The Mechanics of Human Speech

Batten, Ross (2006). The Mechanics of Human Speech Honours Thesis, School of Engineering, The University of Queensland.

       
Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
Batten_Ross_Thesis.pdf Full Text application/pdf 1.20MB 0
Author Batten, Ross
Thesis Title The Mechanics of Human Speech
Formatted title
The mechanics of human speech involve specialist functioning of various parts of the human respiratory system. Most influential of these is the production of sound waves by the vocal cords at the level of the larynx. Vibration of the vocal cords results in an oscillating flow rate of air through the larynx, causing pressure waves to propagate up through the windpipe to be perceived as what is known as human speech.

Research on the mechanism of vocal cord vibration began with the Myoelastic- Aerodynamic Theory (Van Den Berg, 1958). More recent computer models (Flanagan and Landgraf, 1968, Ishizaka and Flanagan, 1972, Lucero, 2000, Gardner et. al., 2001), based on this theory, all lack an accurate representation of the mechanism that causes vocal cord vibrations. This deficiency is the motivation behind this thesis.

The thesis aims to analyze the properties of the Flanagan and Landgraf Model on which currently used models are based, and develop an alternative low-dimensional model of the vocal cords that better represents the vibration mechanism.

It was found that the Flanagan and Landgraf model is incapable of producing a stable vibration at realistic amplitude. When cord closure is achieved, excitation forces become infinite, preventing separation. The alternative Batten Model does not contain this flaw and can accurately simulate cord oscillation at realistic amplitude and frequency. Both of these models were observed to be incapable of negating a non-zero damping coefficient.

It is recommended that further development of the basic features of the Batten Model be done to address the issue of unbalanced damping effects before the dimensionality of the system is increased to allow further refinement of parameters. It is intended that this model be developed to the state at which realism rivals that of the Ishizaka and Flanagan Model.
School, Centre or Institute School of Engineering
Institution The University of Queensland
Publication date 2006
Thesis type Honours Thesis
Supervisor Paul Meehan
Total pages 61
Language eng
Subjects 0913 Mechanical Engineering
Formatted abstract
The mechanics of human speech involve specialist functioning of various parts of the human respiratory system. Most influential of these is the production of sound waves by the vocal cords at the level of the larynx. Vibration of the vocal cords results in an oscillating flow rate of air through the larynx, causing pressure waves to propagate up through the windpipe to be perceived as what is known as human speech.

Research on the mechanism of vocal cord vibration began with the Myoelastic- Aerodynamic Theory (Van Den Berg, 1958). More recent computer models (Flanagan and Landgraf, 1968, Ishizaka and Flanagan, 1972, Lucero, 2000, Gardner et. al., 2001), based on this theory, all lack an accurate representation of the mechanism that causes vocal cord vibrations. This deficiency is the motivation behind this thesis.

The thesis aims to analyze the properties of the Flanagan and Landgraf Model on which currently used models are based, and develop an alternative low-dimensional model of the vocal cords that better represents the vibration mechanism.

It was found that the Flanagan and Landgraf model is incapable of producing a stable vibration at realistic amplitude. When cord closure is achieved, excitation forces become infinite, preventing separation. The alternative Batten Model does not contain this flaw and can accurately simulate cord oscillation at realistic amplitude and frequency. Both of these models were observed to be incapable of negating a non-zero damping coefficient.

It is recommended that further development of the basic features of the Batten Model be done to address the issue of unbalanced damping effects before the dimensionality of the system is increased to allow further refinement of parameters. It is intended that this model be developed to the state at which realism rivals that of the Ishizaka and Flanagan Model.
Keyword Human Speech

Document type: Thesis
Collection: UQ Theses (non-RHD) - UQ staff and students only
 
Citation counts: Google Scholar Search Google Scholar
Created: Thu, 14 May 2015, 12:11:19 EST by Asma Asrar Qureshi on behalf of Scholarly Communication and Digitisation Service