Applied Acoustics 68 (2007) 603–615

www.elsevier.com/locate/apacoust

Analysis of disc brake squeal using

the complex eigenvalue method
a,*
P. Liu , H. Zheng a, C. Cai a, Y.Y. Wang a, C. Lu a,
K.H. Ang b, G.R. Liu c
a
Institute of High Performance Computing, 1 Science Park Road, #01-01 The Capricorn Singapore
Science Park II, Singapore 117528, Singapore
b
Sunstar Logistic Singapore Pte Ltd., 10 Science Park Road, #04-16/17 The Alpha Singapore Science Park
II, Singapore 117684, Singapore
c
National University of Singapore, 9 Engineering Drive 1, Singapore 117576, Singapore

Received 25 April 2005; received in revised form 30 March 2006; accepted 30 March 2006
Available online 5 June 2006

Abstract

A new functionality of ABAQUS/Standard, which allows for a nonlinear analysis prior to a com-
plex eigenvalue extraction in order to study the stability of brake systems, is used to analyse disc
brake squeal. An attempt is made to investigate the effects of system parameters, such as the hydrau-
lic pressure, the rotational velocity of the disc, the friction coefficient of the contact interactions
between the pads and the disc, the stiffness of the disc, and the stiffness of the back plates of the pads,
on the disc squeal. The simulation results show that significant pad bending vibration may be
responsible for the disc brake squeal. The squeal can be reduced by decreasing the friction coefficient,
increasing the stiffness of the disc, using damping material on the back plates of the pads, and
modifying the shape of the brake pads.
Ó 2006 Elsevier Ltd. All rights reserved.

Keywords: Disc brake squeal; Complex eigenvalue extraction; Friction coefficient; Stiffness; Damping ratio

*
Corresponding author. Tel.: +65 64191218; fax: +65 64191280.
E-mail address: [email protected] (P. Liu).

0003-682X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apacoust.2006.03.012
604 P. Liu et al. / Applied Acoustics 68 (2007) 603–615

1. Introduction

Brake squeal, which usually occurs in the frequency range between 1 and 16 kHz, has
been one of the most difficult concerns associated with vehicle brake systems. It causes cus-
tomer dissatisfaction and increases warranty costs. Although substantial research has been
conducted into predicting and eliminating brake squeal, it is still difficult to predict its
occurrence due to the complexity of the mechanisms that cause brake squeal [1].
Several theories have been formulated to explain the mechanisms of brake squeal, and
numerous studies have tried with varied success to apply them to the dynamics of disc brakes
[2]. There are many models for analysing disc brake squeal. For example, the effect of surface
topography of the pad/disc assembly on squeal generation was reported [3] and a distributed-
parameter model of a disc brake has been developed to simulate friction-induced vibrations
in the form of high-frequency squeal [4]. A two-degree-of-freedom model has been used to
investigate the basic mechanisms of instability of the disc brake system and demonstrates
the conditions necessary for preventing the instability [5]. Brake squeal has also been studied
from an energy perspective using feed-in energy analysis and results indicate a squeal ten-
dency of the brake system [6]. The use of viscoelastic material (damping material) on the back
of the back plates of the pads can be effective in reducing squeal when there is significant pad
bending vibration [7] and another reported effective method is to modify the shape of the
brake pads to change the coupling between the pads and the disc [8].
Brake noise is mainly caused by friction-induced dynamic instability. There are two main
categories of numerical methods that are used to study this problem: (1) transient dynamic
analysis and (2) complex eigenvalue analysis. Currently the complex eigenvalue method is
preferred and widely used in predicting the squeal propensity of the brake system including
damping and contact [9–12], since the transient dynamic analysis is computationally expen-
sive. The main idea of the complex eigenvalue method involves symmetry arguments of the
stiffness matrix and the formulation of a friction coupling. This method is more efficient and
provides more insight to the friction-induced dynamic instability of the disc brake system.
In the present study, an investigation of disc brake squeal is performed by using the new
complex eigenvalue capability of the finite element (FE) software ABAQUS version 6.4
[13]. This FE method uses nonlinear static analysis to calculate the friction coupling prior
to the complex eigenvalue extraction, as opposed to the direct matrix input approach that
requires the user to tailor the friction coupling to stiffness matrix, Thus, the effect of non-
uniform contact and other nonlinear effects are incorporated. A systematic analysis is done
to investigate the effects of system parameters, such as the hydraulic pressure, the rota-
tional velocity of the disc, the friction coefficient of the contact interactions between the
pads and the disc, the stiffness of the disc, and the stiffness of the back plates of the pads,
on the disc squeal. The simulations performed in this work present a guideline to reduce
the squeal noise of the disc brake system.

2. Methodology and numerical model

2.1. Complex eigenvalue extraction

For brake squeal analysis, the most important source of nonlinearity is the frictional
sliding contact between the disc and the pads. ABAQUS allows for a convenient, but gen-
eral definition of contact interfaces by specifying the contact surface and the properties of
 P. Liu et al. / Applied Acoustics 68 (2007) 603–615 605

the interfaces. ABAQUS version 6.4 has developed a new approach of complex eigenvalue
analysis to simulate the disc brake squeal. Starting from preloading the brake, rotating the
disc, and then extracting natural frequencies and complex eigenvalues, this new approach
combines all steps in one seamless run [13]. The complex eigenproblem is solved using the
subspace projection method, thus a natural frequency extraction must be performed first
in order to determine the projection subspace. The governing equation of the system is
Mx € þ C x_ þ Kx ¼ 0; ð1Þ
where M is the mass matrix, C is the damping matrix, which includes friction-induced con-
tributions, and K is the stiffness matrix, which is unsymmetric due to friction. The govern-
ing equation can be rewritten as
ðl2 M þ lC þ KÞU ¼ 0; ð2Þ
where l is the eigenvalue and U is the corresponding eigenvector. Both eigenvalues and
eigenvectors may be complex. In order to solve the complex eigenproblem, this system
is symmetrized by ignoring the damping matrix C and the unsymmetric contributions to
the stiffness matrix K. Then this symmetric eigenvalue problem is solved to find the pro-
jection subspace. The N eigenvectors obtained from the symmetric eigenvalue problem
are expressed in a matrix as [/1, . . . , /N]. Next, the original matrices are projected onto
the subspace of N eigenvectors
M  ¼ ½/1 ; . . . ; /N T M½/1 ; . . . ; /N ; ð3aÞ
 T
C ¼ ½/1 ; . . . ; /N  C½/1 ; . . . ; /N  ð3bÞ
and
T
K  ¼ ½/1 ; . . . ; /N  K½/1 ; . . . ; /N : ð3cÞ
Then the projected complex eigenproblem becomes
ðl2 M  þ lC  þ K  ÞU ¼ 0: ð4Þ
Finally, the complex eigenvectors of the original system can be obtained by
U ¼ ½/1 ; . . . ; /N T U : ð5Þ
A more detailed description of the algorithm may be found in [13]. The complex eigen-
value l, can be expressed as l = a ± ix where a is the real part of l, Re(l), indicating
the stability of the system, and x is the imaginary part of l, Im(l), indicating the mode
frequency. The generalized displacement of the disc system, x, can then be expressed as
x ¼ A elt ¼ eat ðA1 cos xt þ A2 sin xtÞ: ð6Þ
This analysis determines the stability of the system. When the system is unstable, a becomes
positive and squeal noise occurs. An extra term, damping ratio, is defined as a/(p|x|). If the
damping ratio is negative, the system becomes unstable, and vice versa. The main aim of this
analysis is to reduce the damping ratio of the dominant unstable modes.

2.2. Finite element model

A disc brake system consists of a disc that rotates about the axis of a wheel, a calliper–
piston assembly where the piston slides inside the calliper, that is mounted to the vehicle
606 P. Liu et al. / Applied Acoustics 68 (2007) 603–615

Fig. 1. Geometry and finite element mesh of the simplified disc brake system.

suspension system, and a pair of brake pads. When hydraulic pressure is applied, the pis-
ton is pushed forward to press the inner pad against the disc and simultaneously the outer
pad is pressed by the calliper against the disc. The brake model used in this study is a sim-
plified version of a disc brake system which consists of a disc and a pair of brake pads. The
disc has a diameter of 292 mm and a thickness with typical value of 5.08 mm and is made
of cast iron. The pair of brake pads, which consist of contact plates and back plates, are
pressed against the disc in order to generate a friction torque to slow the disc rotation.

Fig. 2. Constraints and loadings of the disc brake system.

 P. Liu et al. / Applied Acoustics 68 (2007) 603–615 607

The contact plates are made of an organic friction material and the back plates are made
of steel. The FE mesh is generated using three-dimensional continuum elements for the
disc and pads as shown in Fig. 1, where a fine mesh is used in the contact regions. The
friction contact interactions are defined between both sides of the disc and the contact
plates of the pads. A constant friction coefficient and a constant angular velocity of the
disc are used for simulation purposes. Figs. 2(a)–(c) present the constraints and loadings
for the pads and disc assembly. The disc is completely fixed at the five counter-bolt holes as
shown in Fig. 2(a) and the ears of the pads are constrained to allow only axial directional
movements as shown in Figs. 2(b) and (c). The calliper–piston assembly is not defined in
the simplified model of the disc brake system, hence the hydraulic pressure, which has a
typical value of 0.5 MPa, is directly applied to the back plates at the contact regions
between the inner pad and the piston and between the outer pad and the calliper as shown
in Figs. 2(b) and (c), and it is assumed that an equal magnitude of force acts on each pad.
The analysis procedure contains the following four steps: (1) nonlinear static analysis for
the application of brake pressure; (2) nonlinear static analysis to impose a rotational veloc-
ity on the disc; (3) normal mode analysis to extract the natural frequency to find the pro-
jection subspace; and (4) complex eigenvalue analysis to incorporate the effect of friction
coupling.

Fig. 3. (a) Variation of the damping ratio with frequency for different friction coefficients; (b) variation of the
damping ratio with friction coefficient at frequency 12 kHz.
608 P. Liu et al. / Applied Acoustics 68 (2007) 603–615

3. Results and discussion

The effects of the system parameters, such as the hydraulic pressure P, the rotational
velocity of the disc W, the friction coefficient of the contact interactions between the pads
and the disc u, the stiffness of the disc, and the stiffness of the back plates of the pads, on
the disc squeal are investigated by the simulation model. The effect of the stiffness of the
disc can be changed by varying Young’s modulus ED and the disc thickness TD of the disc
while the effect of the stiffness of the back plates of the pads can be changed by varying
Young’s modulus EP of the back plates of the pads. The complex eigenvalue analysis is
performed up to 13 kHz which is the range of squeal occurrence for the present disc model.
As mentioned previously, if the damping ratio is negative, the system becomes unstable,
and vice versa; when the disc system is unstable, the squeal propensity increases with an
increased value of the damping ratio (absolute values are used). For clarity, only negative
values of the damping ratio are plotted. The typical values for the system parameters used
in the simulation are: P = 0.5 MPa, W = 1.5 rad/s, u = 0.653, ED = 219.669 GPa,
TD = 5.08 mm, and EP = 210 GPa. Analysis is carried out by changing the values of each
parameter while retaining the respective typical values for the others.

Fig. 4. (a) Variation of the damping ratio with frequency for different hydraulic pressures; (b) variation of the
damping ratio with hydraulic pressure at frequency 12 kHz.
 P. Liu et al. / Applied Acoustics 68 (2007) 603–615 609

3.1. Effect of friction coefficient

Disc squeal is believed to be caused mainly by friction-induced dynamic instability. This

section presents the effect of the friction coefficient of the contact interactions between the
pads and the disc on the disc squeal, in which the friction coefficient u varies from 0.2 to
0.8. Fig. 3(a) shows results in the form of the damping ratio as a function of frequency for
different friction coefficients. It can be seen that the major squeal frequency is approxi-
mately 12 kHz. The value of the damping ratio is decreased significantly with a decrease
of the friction coefficient as shown in Fig. 3(b) at a frequency of 12 kHz. It is understand-
able that with an increase in the friction coefficient, there is an accompanying increase in
the instability of the system, thus an increase in the damping ratios. This means that the
most fundamental method of eliminating brake squeal is to reduce the friction between the
pads and the disc. However, this obviously reduces braking performance and is not a pref-
erable method to employ.

3.2. Effect of hydraulic pressure

The effect of the hydraulic pressure P on the squeal propensity is studied by varying P
from 0.5 MPa to 2.0 MPa. Fig. 4(a) shows the change of the damping ratio with frequency

Fig. 5. (a) Variation of the damping ratio with frequency for different rotational velocities of the disc; (b)
variation of the damping ratio with rotational velocity of the disc at frequency 12 kHz.
610 P. Liu et al. / Applied Acoustics 68 (2007) 603–615

for different hydraulic pressures. The major squeal frequency is approximately 12 kHz. It
can be seen from Fig. 4(b) that with an increase in P, the value of the damping ratio is
increased, so the squeal propensity is increased. This is due to a larger hydraulic pressure
inducing more friction between the pads and the disc. However, the simulation results also
show that the effect of the hydraulic pressure on the disc brake squeal is not significant
because the value of the damping ratio only changes from 0.17 to 0.193 when P increases
from 0.5 MPa to 2.0 MPa.

3.3. Effect of rotational velocity of the disc

Fig. 5(a) presents the variation of the damping ratio with the frequency for different disc
angular velocities W (0.7–8.0 rad/s). The dominant squeal frequency is approximately
12 kHz. As the angular velocity increases, the value of the damping ratio gradually
decreases. However, as with the previous case, when changing the hydraulic pressure,
the effect of changing the angular velocity on the squeal propensity is also not obvious:
this can be seen from Fig. 5(b) which shows the value of the damping ratio varies with
an increase in the rotational velocity of the disc.

Fig. 6. (a) Variation of the damping ratio with frequency for different Young’s moduli of the disc; (b) variation of
the damping ratio with Young’s moduli of the disc at frequency 12 kHz.
 P. Liu et al. / Applied Acoustics 68 (2007) 603–615 611

3.4. Effect of stiffness of the disc

The effect of the stiffness of the disc on the disc brake squeal is studied by changing
Young’s modulus ED and the thickness TD of the disc. Fig. 6(a) shows results of the damp-
ing ratio versus frequency for different Young’s modulus ED, i.e. ED = 0.8ED0, 0.9ED0,
1.0ED0, 1.1ED0 and 1.2ED0, where ED0 is the typical value of Young’s modulus of the disc,
which is 219.669 GPa. It can be seen that the major squeal frequency does not change for
different disc Young’s moduli. The value of the major squeal frequency is approximately
12 kHz. As Young’s modulus ED is increased and hence as the stiffness of the disc is
increased, the value of the damping ratio decreases greatly. Fig. 6(b) presents the damping
ratio versus Young’s modulus of the disc at a frequency of 12 kHz. It is found that a larger
disc stiffness can reduce the squeal propensity of the disc system. It is believed that a stiff-
ening of the disc can reduce the disc vibration magnitude, as a result, the squeal propensity
of the disc system can be reduced. The stiffness of the disc is also changed by varying its
thickness TD. Four cases were studied, i.e. TD = 0.9TD0, 1.0TD0, 1.1TD0 and 1.2TD0,
where TD0 = 5.08 mm is the typical value for disc thickness. Fig. 7(a) shows results of
the damping ratio plotted against frequency for different disc thicknesses and Fig. 7(b) pre-
sents the damping ratio versus disc thickness at a frequency of 12 kHz. The thicker the

Fig. 7. (a) Variation of the damping ratio with frequency for different disc thicknesses; (b) variation of the
damping ratio with thickness of the disc at frequency 12 kHz.
612 P. Liu et al. / Applied Acoustics 68 (2007) 603–615

Fig. 8. (a) Variation of the damping ratio with frequency for different Young’s moduli of the back plates of the
pads; (b) variation of the damping ratio with Young’s moduli of the back plates of the pads at frequency 12 kHz.

disc, the higher its stiffness, the smaller the damping ratio, and thus the lower the squeal
propensity.

3.5. Effect of stiffness of the back plates of the pad

Brake pads consist of contact plates which are made of a friction material and back
plates. In this study, the effect of Young’s modulus EP of the back plates of the pads on
the disc squeal is investigated, in which EP = 0.8EP0, 0.9EP0, 1.0EP0, 1.1EP0 and 1.2EP0,
where EP0 = 210 GPa, is the typical value of Young’s modulus for the back plates of pads.
Fig. 8(a) shows results of the damping ratio versus frequency for different Young’s moduli
EP. It can be seen that the dominant squeal occurs at a frequency of approximately
12 kHz. As Young’s modulus EP, is increased, corresponding to an increase in stiffness
of the back plates of the pads, the value of the damping ratio increases significantly as
shown in Fig. 8(b); here the variation of the damping ratio with Young’s modulus of
the back plates at a frequency of 12 kHz is shown. This important observation implies that
the stiffer back plates of pads cause a higher squeal propensity. This is so since the friction
material connected to the back plates is very soft compared with the back plate material.
Hence the higher the stiffness of the back plates, the greater the uneven deformation and
vibration magnitude of the pad, and hence the higher the damping ratio. So an effective
 P. Liu et al. / Applied Acoustics 68 (2007) 603–615 613

Fig. 9. Vibration mode of the disc brake system at frequency 12 kHz.

method to reduce squeal propensity of disc brake system is to use a damping material for
the back plates of the pads.

3.6. Unstable modes of disc brake system

The simulation results show that for all the cases owe large damping ratios, the unstable
frequencies are approximately 12 kHz. There is a significant pad bending vibration for
these cases. Fig. 9 gives an example of the vibration mode of the disc brake system at a
frequency of 12 kHz, where all the system parameters are the typical values. It can be seen
that the disc has only slight out-of-plane modes of vibration as shown in Fig. 9(a), but the
pads have serious out-of-plane modes of vibration which occur mainly at the bottom parts
of the pads as shown in Fig. 9(b). This suggests that the brake pads may be the source of
the disc brake squeal. So methods which can reduce the pad bending vibration should be
used to eliminate the disc squeal. One of the methods reported is to use viscoelastic mate-
rial (damping material) on the back of the back plates of the pads [7] and another effective
method is to modify the shape of the brake pads to change the coupling between the pads
and the disc [8]. Except the unstable vibration modes which occur at frequency 12 kHz and
are caused mainly by the pads vibration, the other unstable vibration modes are caused
mainly by the disc vibration. Figs. 10(a) and (b) give an example of the unstable vibration
614 P. Liu et al. / Applied Acoustics 68 (2007) 603–615

Fig. 10. Vibration mode of the disc brake system at frequency 9766 Hz.

mode of the disc brake system at the frequency of 9766 Hz, where all the system param-
eters are the typical values. It can be seen that the disc has significant out-of-plane vibra-
tion compared with the vibration of pads.

4. Conclusion

Friction-induced disc brake squeal is investigated using the new function of ABAQUS
version 6.4, which combines a nonlinear static analysis and a complex eigenvalue extrac-
tion method. The nonlinear effects can be taken into account in the preloading steps in
order to more accurately model a deformed configuration at which a complex eigenvalue
analysis is performed. The systematic analysis here shows that significant pad bending
vibration may be responsible for causing the disc brake squeal and the major squeal fre-
quency is approximately 12 kHz for the present disc brake system. The effects of the fric-
tion between the pads and the disc, the stiffness of the disc, and the stiffness of the back
plates of the pads, on disc squeal, are significant, but the effects of the hydraulic pressure
and the angular velocity of the disc on disc squeal are not obvious. The squeal can be
 P. Liu et al. / Applied Acoustics 68 (2007) 603–615 615

reduced by decreasing the friction coefficient, increasing the stiffness of the disc, using
damping material on the back of the pads, and modifying the shape of the brake pads.

Acknowledgements

The authors thank Mr. Y.C. Tse for providing some input data and Dr. Bud Fox for
proofreading.

References

[1] Papinniemi A, Lai JCS, Zhao J, Loader L. Brake squeal: a literature review. Appl Acoust 2002;63:391–400.
[2] Kinkaid NM, O’Reilly OM, Papadopoulos P. Automotive disc brake squeal. J Sound Vibrat
2003;267:105–66.
[3] Sherif HA. Investigation on effect of surface topography of pad/disc assembly on squeal generation. Wear
2004;257:687–95.
[4] Flint J. Lining-deformation-induced modal coupling as squeal generator in a distributed parameter disc
brake model. J Sound Vibrat 2002;254:1–21.
[5] Shin K, Brennan MJ, Oh JE, Harris CJ. Analysis of disc brake noise using a two-degree-of-freedom model. J
Sound Vibrat 2002;254:837–48.
[6] Guan DH, Huang JC. The method of feed-in energy on disc brake squeal. J Sound Vibrat 2003;261:297–307.
[7] Hoffman CT. Damper design and development for use on disc brake shoe and lining assemblies. SAE Paper,
No. 880254; 1988.
[8] Flint J. Instabilities in brake system. SAE Paper, No. 920432; 1992.
[9] Kung SW, Stelzer G, Belsky V, Bajer A. Brake squeal analysis incorporating contact conditions and other
nonlinear effects. SAE Paper, 2003-01-3343; 2003.
[10] Bajer A, Belsky V, Zeng LJ. Combining a nonlinear static analysis and complex eigenvalue extraction in
brake squeal simulation. SAE Paper, 2003-01-3349; 2003.
[11] Lee L, Xu K, Malott B, Matsuzaki M, Lou G. A systematic approach to brake squeal simulation using
MacNeal method. SAE Paper, 2002-01-2610; 2002.
[12] Blaschke P, Tan M, Wang A. On the analysis of brake squeal propensity using finite element method. SAE
Paper, 2000-01-2765; 2000.
[13] ABAQUS Analysis User’s Manual, Version6.4.