A.

Abstract
A review on mathematical and mechanical models of the vocal cords is given.
The basic model is a two-mass nonlinear oscillator system which is accepted to be the
basic one for mechanical description in voice production. The model is not only extended
into three, five, and more mass systems, systems with time variable parameters and three-
dimensional systems, but also simplified into one-mass system with coupled two-
direction deflection and damping functions. The corresponding mathematical models are
the systems of coupled second-order differential equations which describe the vibrations
of the symmetric and asymmetric vocal folds. The models give the conditions for the
regular and irregular motions like bifurcation and deterministic chaos in vocal folds. The
obtained results are of special interest for detecting the pathology of vocal cords, when
there are no visual effects of disease. Based on the results given in the paper, the
objectives for future investigation in this matter are given.

B. Introduction
Model the movement of the vocal chords can be useful to the development of better
synthetic speech sensors. In this study case, we try to approximate the displacement over
time of a vocal cord with both a constant force and impulse stimuli. We applied a
simplified model of an oscillatory mass and spring system with damping, to account for
the properties of the environment of the vocal cords. However, to create a proper model,
we must first understand the anatomy of the voice box.

C. The Vocal fold : Anatomy

There are two sets of vocal cords, the posterior or ventricular chords, and the
inferior set of “true” folds. The main functions of
the ventricular cords are to lubricate the true vocal
chords and to block food and liquid from entering
the airway. Ventricular cords are not actively
involved in phonation, except for yelling or
grunting. The inferior set of vocal cords are
primarily involved in phonation.They are protected
by a layer of a mucous membrane and are attached
to the adductor muscles. When the adductor
muscles close, the vocal cords join together, and
produce resistance to exhaled air from the lungs.
Then, the air rushes through the vocal cords
pushing them aside. Therefore the pressure between
the cords drops and sucks them together causing a
“Bernoulli Effect”. This vibration then produces
sound which is then shaped by muscular changes in

Model for Vocal Fold Vibration Analysis 1

 the throat, jaw, tongue, palate, and lips, creating speech.
Injuries to the Vocal Chords Proper understanding of the physiology of the voice
box improves medical practices on the system, and also personal use of the voice. The
most common source of injury to the vocal cords is related to improper voice use.
Whether it be voice overuse, misuse due to poor singing techniques, coughing, or even
acid reflux, any treatment of the vocal cords is difficult due to the fact that they are in
constant use. With the mathematical model, we can see how any injury in this area would
greatly impair the patient’s speech abilities and cause it very slow for the healing process
to occur, so any technological development leading to the expediting of this process is
worth investigating.

D. Mathematical Model of the System in Second Order Differential Equation

As discussed above, the vocal folds are mucous membranes that vibrate to produce
sound. This process is dependent on many variables, including glottal flow, muscle
interaction with vocal folds, feedback from the vagus nerve, and material properties of
the vocal folds. In order to simplify this simulation, we made several assumptions and
simplifications:
1) Symmetry: since the vocal folds are symmetrical, we assume that we can model one
vocal fold independently from the other vocal fold.
2) One Dimensional: for simplicity, only vibrations in the x-direction are being
modeled. It signify the assumption that vibrations in the y and z direction are
independent from vibrations in the x-direction.
3) Viscoelastic Material: We assume that vocal folds can be modeled as a viscoelastic
material with a stiffness k and viscosity b.
4) Isotropic Material: Consider that material properties are constant throughout,
allowing for material coefficients k, b, and m (mass), to be constant across space (x).
5) Time Independent Coefficients: Due to various types of feedback in the vocal system,
material coefficients can be time dependent. For simplicity, we assume that we are
modeling for a short enough period of time at coefficients can be approximated as
constant over time.
6) Constant Glottal Flow: for simplicity, we assume glottal flow is constant over a short
period of time

With the assumptions made above, the vocal fold model can be simplified as a
single mass-damper-spring model with the equation below:

Model for Vocal Fold Vibration Analysis 2

 𝒅𝟐 𝒙(𝒕) 𝒅𝒙(𝒕)
𝒎 + 𝒃 + 𝒌𝒙(𝒕) = 𝑭𝒙 (𝒕)
𝒅𝒕𝟐 𝒅𝒕

m = mass (kg)
b = Damping Coefficient
k = Spring Constant (N/m)
F = force (N)

Where 𝐹𝑥 = 𝑃𝐴 is the Coulomb

force, P is the lung pressure, and A is the
medial surface of the glottis. An elevated
lung pressure gives rise to a flow of air
through the glottis and produces a series of alternating converging and diverging shapes
of the vertical dimensions of the vocal folds. The pressure distributions in the glottis
resulting from the series of shapes are alternately higher and lower than the pressures in
the vocal tract. These pressure variations are in phase with the motion of the vocal folds
and add energy to the oscillator in the same way as negative Coulomb damping does.
Limit cycle of the oscillator with negative Coulomb damping provides a natural
explanation of the self-oscillation property of the model.

Table of Parameters
Mass (m) 10-4 kg
Spring Constant (k) 35-80 N/m
Damping Coefficient (b) 0.010955
Damping Ratio (ƺ) 0.1
Fx 0.1 N
The parameters of the single mass-damper-spring model system.

Apply the Laplace operator on both side of the equation (i), which is a linear
operator. Then we can get

𝐿(𝐹𝑥 ) = 𝐿 (𝑚 ∙ 𝑥 ′′ + 𝑏 ∙ 𝑥 ′ + 𝑘 ∙ 𝑥)(𝑠)

= 𝑚𝐿(𝑥 ′′ )(𝑠) + 𝑏𝐿(𝑥 ′ )(𝑠) + 𝑘𝐿(𝑥)(𝑠) (𝑖𝑖)

According to the table of Laplace transforms, we can further simplify the equation (ii),

10−4 ∙ [𝑠 2 𝐿𝑥 (𝑠) − 𝑠 ∙ 𝑥(0) − 𝑥 ′ (0)] + 0.010955 ∙ [𝑠𝐿𝑥 (𝑠) − 𝑥(0)] + 30𝐿𝑥 (𝑠) = 1⁄(10𝑠)
With x(0) = x’(0) = 0.

By separating 𝐿𝑥 (𝑠) the and decomposing the denominator, we get:

10−4 ∙ [𝑠 2 𝐿𝑥 (𝑠) ] + 0.010955 ∙ [𝑠𝐿𝑥 (𝑠)] + 30𝐿𝑥 (𝑠) = 1⁄(10𝑠),

Model for Vocal Fold Vibration Analysis 3

 𝑖 𝑖
103 103 3 (𝑠1 −𝑠2 ) (𝑠1 −𝑠2 )
i.e., 𝐿𝑥 (𝑠) = = = 10 ∙ [ 𝑠(𝑠−𝑠 ) − ]
𝑠∙(𝑠2 +109.555𝑠+3×105 ) 𝑠∙(𝑠−𝑠2 )(𝑠−𝑠1 ) 1 𝑠(𝑠−𝑠2 )

103 1 1 1 1 1 1
= ∙[ ∙( − )− ∙( − )]
(𝑠1 − 𝑠2 ) 𝑠1 𝑠 − 𝑠1 𝑠 𝑠2 𝑠 − 𝑠2 𝑠
103 1 103 1 103 1
= ∙ −𝑠 ∙ +𝑠 ∙ 𝑠 (𝑖𝑖𝑖),
𝑠1 (𝑠1 −𝑠2 ) 𝑠−𝑠 1 2 (𝑠1 −𝑠2 ) 𝑠−𝑠 2 1 ∙𝑠2

−109.55±√109.552 −4×3×105
with 𝑠1 , 𝑠2 = 2

Then we could apply the inverse Laplace transforms on the both sides of the equation
(iii), i.e. . 𝑥(𝑡) = 𝐿𝑠 −1 [𝐿𝑥 (𝑠)] We could get:
103 1 103 1 103 1
𝑥(𝑡) = 𝐿𝑠 −1 [ 𝑠 ∙ −𝑠 ∙ +𝑠 ∙ 𝑠]
1 (𝑠1 −𝑠2 ) 𝑠−𝑠 1 2 (𝑠1 −𝑠2 ) 𝑠−𝑠 2 1 ∙𝑠2

From the table of Laplace transform, we can get the x = x(t) by plugging in the value of
𝑠1 𝑎𝑛𝑑 𝑠2 :

𝑥(𝑡) = −8.80449 × 10−5 × 𝑒 −109.55𝑡 × sin(536.655𝑡) − 3.333 × 10−4 × 𝑒 −109.55𝑡

× cos(536.655𝑡) + 3.333 × 10−4 , 𝑡 ≥ 0
Here, we just show the analytical solution by using Laplace transform since we
have the I.C.
Model Limitations
The most significant discrepancy between the analytical model and real vocal
folds is that it is an under-damped as opposed to a system that sustains oscillations.
However, a system that sustains oscillations, assuming a damper is present in the model,
would need an appropriate driving force, Fx. Although setting Fx to a constant was
convenient for finding a solution to the model, it does not allow for the interesting
dynamics observed in real vocal folds.

E. Characteristics of the System

𝒅𝟐 𝒙(𝒕) 𝒅𝒙(𝒕)
𝒎 𝟐
+ 𝒃 + 𝒌𝒙(𝒕) = 𝑭𝒙 (𝒕)
𝒅𝒕 𝒅𝒕
 Causality
Causal and memoryless, because the system depends on present input.

 Stability
Stable, because there is limit for the system so the value will not overshoot.

 Linearity

Model for Vocal Fold Vibration Analysis 4

 𝑑2 𝑥1 (𝑡) 𝑑𝑥1 (𝑡)
𝑦1 (𝑡)= 𝑚 + 𝑏 + 𝑘𝑥1 (𝑡)
𝑑𝑡 2 𝑑𝑡
2
𝑑 𝑥2 (𝑡) 𝑑𝑥2 (𝑡)
𝑦2 (𝑡)= 𝑚 + 𝑏 + 𝑘𝑥2 (𝑡)
𝑑𝑡 2 𝑑𝑡

𝑦3 (𝑡) = A 𝑦1 (𝑡) + B 𝑦2 (𝑡)

So, the system is linear.

 Time Variance
Time varrying, because if the time is shifted, x(t-t0) ≠ y (t-t0).

F. Pulse Signal of the system

Our model input is variable “x”. Where x here is the air flowing through the the
pharynx so our vocal cord shakes so it sounds like human vocals or sound. For the pulse
signal, we consider it is similar to sine wave, so we can assume that the pulse signal is
odd signal.The signal is periodic. Because the average of amplitude of the signal is 0,
therefore 𝑎0 = 0.

G. Fourier Series Representation.

I. Input (pulse signal)

Model for Vocal Fold Vibration Analysis 5

 II. Fourier representation of the input signal

III. The comparasion of the input signal before and after fourier

Model for Vocal Fold Vibration Analysis 6

H. Model of the system after the pulse signal is applied

I. Model of the system before the pulse signal is applied

Model for Vocal Fold Vibration Analysis 7

II. Model of he system after the pulse is applied (the fourier one)

Conclusions

In order to simply the model of the vocal folds, many assumptions were made.
The geometry was lumped which makes the model inaccurate when the geometry in the
x-direction changes, such as in the case of the child and adult. Interactions between
different tissues and different direction were taken to be independent from the movement
of the vocal folds in the xdirection when it is well known that the vocal folds respond to
feedback from surrounding tissue and neural input. The glottal flow was also simplified.
However, despite all the simplifications, the model still is able to demonstrate how
changes in physiology can alter the change in displacement and frequency output.

Model for Vocal Fold Vibration Analysis 8

 SOURCE CODE ATTACHMENT
PULSE SIGNAL
N = 100
x = linspace(-1,1,N);
f = sign(x);
sum = 0;
plot(x,f)

PULSE SIGNAL AFTER FOURIER

N = 100
x = linspace(-1,1,N);
f = sign(x);
sum = 0
M = 10
for j = 1:2:M
sum = sum + 4/pi*sin(j*pi*x)/j;
end
plot(x, sum)

THE COMPARISON OF PULSE SIGNAL BEFORE AND AFTER FOURIER

N = 100
x = linspace(-1,1,N);
f = sign(x);

Model for Vocal Fold Vibration Analysis 9

sum = 0;
M = 10
for j = 1:2:M
sum = sum + 4/pi*sin(j*pi*x)/j;
end
plot(x, sum)
hold on
plot(x,f)

THE SYSTEM BEFORE THE INPUT IS APPLIED

t=0:10e-6:0.1;
x=-6.80449e-5*exp(-109.55*t).*sin(536.655*t)-3.333e-4*exp(
109.55*t).*cos(536.655*t)+3.333e-4;
plot(t,x*1e6)

THE SYSTEM AFTER THE INPUT IS APPLIED

N = 1000
x = linspace(-1,1,N);
f = sign(x);
sum = 0
M = 1000
for j = 1:2:M
sum = sum + 4/pi*sin(j*pi*x)/j;
end
y=sum;
t=0:10e-6:0.1;
y=-6.80449e-5*exp(-109.55*t).*sin(536.655*t)-3.333e-4*exp(-
109.55*t).*cos(536.655*t)+3.333e-4;
plot(t,y*1e6)

Model for Vocal Fold Vibration Analysis 10

References
[1] Cataldo, Edson, Christian Soize, Christophe Desceliers, and Rubens Sampaio. "Uncertainties
in mechanical models of larynx and vocal tract for voice production." In Proceedings of the XII
International Symposium on Dynamics Problems of Mechanics (DINAME 2007). 2007.
[2] Flanagan, J., and Lois Landgraf. "Self-oscillating source for vocal-tract synthesizers." Audio
and Electroacoustics, IEEE Transactions on 16, no. 1 (1968): 57-64.
[3] Lulich, Steven M. "ESTIMATION OF LUMPED VOCAL FOLD MECHANICAL
PROPERTIES FROM NON-INVASIVE MICROPHONE RECORDINGS.“
[4] Gómez, P., C. Lázaro, R. Fernández, A. Nieto, J. I. Godino, R. Martínez, F. Díaz et al.
"Using biomechanical parameter estimates in voice pathology detection." In Proc of 4th
International Workshop on Models and Analysis of Vocal Emissions for Biomedical
Applications (MABEVA05), pp. 29-31. 2005.
[5] Mehta, Daryush D., Matías Zañartu, Thomas F. Quatieri, Dimitar D. Deliyski, and Robert E.
Hillman. "Investigating acoustic correlates of human vocal fold vibratory phase asymmetry
through modeling and laryngeal high-speed videoendoscopy." the Journal of the Acoustical
Society of America 130, no. 6 (2011): 3999.

Model for Vocal Fold Vibration Analysis 11