Negative Coulomb damping, limit cycles, and self-oscillation of the vocal folds

An effective one-mass model of phonation is developed. It borrows the salient features of the classic two-mass model of human speech developed by Ishizaka, Matsudaira, and Flanagan. Their model is based on the idea that the oscillating vocal folds maintain their motion by deriving energy from the flow of air through the glottis. We argue that the essence of the action of the aerodynamic forces on the vocal folds is captured by negative Coulomb damping, which acts on the oscillator to energize it. A viscous force is added to include the effects of tissue damping. The solutions to this single oscillator model show that when it is excited by negative Coulomb damping, it will reach a limit cycle. Displacements, phase portraits, and energy histories are presented for two underdamped linear oscillators. A nonlinear force is added so that the variations of the fundamental frequency and the open quotient with lung pressure are comparable to the behavior of the two-mass model.