Navnath Kalel1, Bhaskaranand Bhatt1, Ashish Darpe2, Jayashree Bijwe1. 1. Centre for Automotive Research and Tribology (CART), Indian Institute of Technology Delhi, New Delhi 110016, India. 2. Department of Mechanical Engineering, Indian Institute of Technology Delhi, New Delhi 110016, India.
Abstract
Aramid pulp/fiber is the most vital ingredient of brake friction material (FM) formulation. It is perpetually added to achieve quality brake pads/shoes and improve the overall friction and wear performance. Additionally, novel Zylon fibers have a superior property to aramid fibers. However, no studies give insights on their influence on brake noise and vibration (NV) performance. In the current work, a series of six different types of eco-friendly brake pads was developed. The first five contain aramid pulp, aramid short fibers, and Zylon fibers of different sizes (1, 3, and 6 mm) as the theme ingredients (3 wt %) by keeping the parent composition identical. Additionally, one more pad was developed that contains no aramid/Zylon fibers (i.e., reference pad). The pads were characterized for physical and mechanical properties. The damping and natural frequencies of pads were measured experimentally and numerically. All brake pads were evaluated for detailed NV performance by following the SAE J 2521 test schedule. In addition, numerical simulation was performed to validate the experimental brake squeal results. Results revealed that aramid/Zylon fiber-based pads improved the porosity, damping, and compressibility. Overall, brake noise and vibrations were improved for aramid/Zylon fiber-based pads by 1.2-1.5 dBA and 20-25%, respectively, compared to the reference pad. The complex eigenvalue analysis (CEA) proved that squeal was mainly influenced by the damping and density of the pad materials. Thus, aramid/Zylon fiber-based pads can effectively suppress the instability of the brake system and reduce the brake squeal propensity.
Aramid pulp/fiber is the most vital ingredient of brake friction material (FM) formulation. It is perpetually added to achieve quality brake pads/shoes and improve the overall friction and wear performance. Additionally, novel Zylon fibers have a superior property to aramid fibers. However, no studies give insights on their influence on brake noise and vibration (NV) performance. In the current work, a series of six different types of eco-friendly brake pads was developed. The first five contain aramid pulp, aramid short fibers, and Zylon fibers of different sizes (1, 3, and 6 mm) as the theme ingredients (3 wt %) by keeping the parent composition identical. Additionally, one more pad was developed that contains no aramid/Zylon fibers (i.e., reference pad). The pads were characterized for physical and mechanical properties. The damping and natural frequencies of pads were measured experimentally and numerically. All brake pads were evaluated for detailed NV performance by following the SAE J 2521 test schedule. In addition, numerical simulation was performed to validate the experimental brake squeal results. Results revealed that aramid/Zylon fiber-based pads improved the porosity, damping, and compressibility. Overall, brake noise and vibrations were improved for aramid/Zylon fiber-based pads by 1.2-1.5 dBA and 20-25%, respectively, compared to the reference pad. The complex eigenvalue analysis (CEA) proved that squeal was mainly influenced by the damping and density of the pad materials. Thus, aramid/Zylon fiber-based pads can effectively suppress the instability of the brake system and reduce the brake squeal propensity.
The brake is a crucial member of an automobile that ensures operational
safety for a wide range of operating conditions. With brake noise
pollution and related customer dissatisfaction, several attempts have
been made to predict and eliminate the brake noise issues by the researchers.
In general, brake noise is classified into various categories depending
upon the frequency ranges, such as judder (<100 Hz), groan (<500
Hz), moan (<500 Hz), howl (500 Hz to 1 kHz), wire brush (>1
kHz),
squeal (>1 kHz), etc.[1−4] The friction-induced vibration brake noise problems
were widely
studied in the 20th century and summarized in review articles.[1,2] Several researchers have studied brake squeal problems using experimental[5−13] and numerical[14−22] techniques in the past decades. Generally, the design is first determined
by the brake pad/lining materials in the brake industry due to their
main role in the braking performance and the squeal propensity. However,
unfortunately, the composition and material properties of brake pad/lining
materials usually considered proprietarily are not known in specific
details. Generally, changing the pad/lining materials, using a damping
shim, and modifying the pad/lining/disc geometry are common ways to
minimize brake noise and vibration issues.[23,24]Recently, a few researchers have attempted to minimize the
brake
noise by changing the brake pad’s composition. Brake squeal
is strongly associated with the natural frequencies of the key components
of the brake system, i.e., the brake pad and disc.[25−27] For instance,
Masoomi et al.[28] studied the role of thermoplastic
elastomer (TPE) ingredients in friction materials (FMs), and the results
suggested that damping properties can be strongly affected by TPE
and the high content of TPE proved to be the best for noise reduction.
Bhatt et al.[29] explored the potential of
calcium silicate (commercially Promaxon-D) to ameliorate the brake
noise and vibration (NV) performance. Results revealed that 20 wt
% Promaxon-D in Cu-free brake pads proved to be the best for the overall
NV performance. Kalel et al.[30] studied
the role of different types of stainless-steel particles (SSPs) on
brake NV performance, and a comparative study was done with a Cu particle-based
pad. Results showed a marginal difference in NV performance of pads,
and one of the investigated SSPs can be a suitable choice in Cu-free
pads. Kamioka et al.[31] revealed cashew
dust and rubber particles in the brake pads, which proved to be beneficial
in suppressing the noise due to their superior viscoelastic properties.
Yamashita et al.[31−33] reported that mica and vermiculite are commonly used
as fillers in brake pads/shoes, reducing the low-frequency noise due
to the needle-like structure and porosity.Aramid in pulp form
is an almost inevitable ingredient of good-quality
non-asbestos organic (NAO) friction materials (FMs). It provides better
processability, sufficient strength to preform, stability in μ,
and increased resistance to fade and wear of FMs.[34−39] Although aramid fibers/pulp is explored extensively in the brake
pads for improving the tribological performance, no studies are available
on the role of aramid fibers/pulp on brake NV performance. In contrast,
Zylon fibers (i.e., PBO: p-phenylene-2,6-benzobisoxazole)
have superior properties (1.6 times higher tensile strength, Young’s
modulus, thermal stability, etc.) to aramid fibers.[40] But, it has not been explored yet in brake pads.Based on the important aspects of materials for brake NV performance,
in the current work, a series of six types of brake pads was developed.
The first five contain aramid pulp, short aramid fibers (1 mm), and
Zylon fibers with varying lengths (i.e., 1, 3, and 6 mm) as the theme
ingredients (3 wt %) by keeping the parent composition identical.
The selected theme ingredient content (3 wt %) was based on the previous
studies.[29,30,36,37,50,51] One more type of brake pad was developed without aramid/Zylon fibers
(reference pad) for comparison. The detailed comparative noise and
vibration studies were carried out, and the results are compiled and
discussed in the subsequent sections.
Materials
and Methodology
Materials and Fabrication
of Brake Pads
The developed brake pad composition consists
of a total of 13 ingredients,
including the theme ingredient. A series of six Cu-free (low metallic)
brake pads was developed. The first five types of pads were based
on theme ingredients (3 wt %), aramid/Zylon fibers, by keeping the
parent composition identical. As mentioned earlier, an additional
pad was developed without the theme ingredient (i.e., a reference
pad). The development process of the brake pads is reported in the
earlier work.[41,42] The composition and designation
of developed brake pads in the laboratory are summarized in Table .
Table 1
Composition and Designation of Developed
Brake Padsb
The
developed brake pads were characterized for physical (density and
porosity) and mechanical (hardness and compressibility) properties
as per the standards. The density was measured as per the Archimedes
principle, porosity (JIS D 4418), hardness (ASTM D 785), and compressibility
(ISO 6310) standards. Each test was done five times, and the average
values of results are reported.
Modal
Analysis of Brake Pads
The
experimental modal analysis helps evaluate the parameters of a mechanical
vibratory system and gives information of its natural frequencies,
mode shapes, and damping ratios. The experimental modal analysis and
numerical eigenvalue analysis were carried out for the brake disc
and brake pads.
Experimental Technique
This subsection
describes the experimental modal analysis of the developed brake pads
and the brake disc to obtain the natural frequencies and damping ratio. Figure shows the experimental
test setup and schematic diagram of the test procedure.
Figure 1
(a) Experimental
test setup. (b) Schematic diagram of the modal
analysis procedure.
(a) Experimental
test setup. (b) Schematic diagram of the modal
analysis procedure.The damping property
of a structure plays a vital role in controlling
its noise and vibration characteristics. Experimental modal analysis
is carried out by simultaneous measurement of the input excitation
force and the output vibration response to get the frequency response
function (FRF). A fast Fourier transform (FFT) analyzer (model CF
7200, ONO SOKKI, Japan) was used to acquire the impact hammer (model
8206, B&K, Denmark) and accelerometer (model 352C03, PCB Piezoelectronics,
USA) signals and compute the corresponding FRF. The acquired raw data
is then input to the ICAT modal analysis software. In ICAT software,
the circle fit method was used to find the natural modes and the corresponding
natural frequency and damping ratio for the disc and brake pads. The
several sets of measurement data of force excitation and vibration
were recorded using a modal hammer and accelerometer, and corresponding
FRFs were found for disc and each type of brake pad. In the present
work, the measurements were done in a free-free condition of the pad
samples (Figure a)
to avoid the influence of the boundary conditions (say, support clamping
force) on the modal properties of the sample.
Numerical Modal Analysis
The mode
shapes and natural frequencies for the disc and developed brake pads
were also studied in ANSYS19 software. To match the free-free boundary
conditions used in the experimental measurements, the same boundary
conditions were applied in the simulations. The finite element models
of the disc and pad were constructed using tetrahedral elements. Both
models were fine-meshed, and the requisite mesh quality was ensured
using quality metrics such as aspect ratio and skewness. In addition,
the node convergence study was also done for the disc and brake pad
models to check the accuracy of the numerical results, and a model
with an optimum element count was chosen. The numbers of finite elements
chosen for the disc and the brake pads were 21388 and 3228, respectively.Table shows the
results of the convergence analysis for the initial four modes of
the disc and brake pad sample. A mesh with 2–4 mm element sizes
was shown as a satisfactory result in relation to experimental and
numerical accuracy. The corresponding numbers of nodes and elements
for the different element sizes are also studied and reported in Table . In addition, the
percentage change in the natural frequencies for the initial four
modes is shown in Table . Finally, the 2 mm element size was chosen as the natural frequencies
were close to the experimental natural frequencies within an error
of 2%.
Table 2
Convergence Analysis of Disc and Brake
Pad Sample FE Models for Different Element Sizes, Numbers of Elements,
and Nodesa
disc
brake pad sample
element size
(mm)
number of
nodes (N) and elements (E)
natural frequencies
(Nf), (Hz)
% change
in NF with a reduction in element size
number of
nodes (N) and elements (E)
natural frequencies
(Nf), (Hz)
% change
in Nf with a reduction in element size
4 mm
N, 21118; E, 11257
Nf1, 1755; Nf2, 3347; Nf3, 4972; Nf4, 6971
N, 1312; E, 196
Nf1, 9940; Nf2, 12168; Nf3, 13229;
Nf4, 13956
3 mm
N, 36471; E, 19905
Nf1, 1729; Nf2, 3335; Nf3, 4957; Nf4, 6957
Nf1, 1.48; Nf2, 0.36; Nf3, 0.30; Nf4, 0.20
N, 2460; E, 405
Nf1, 9938; Nf2, 12166; Nf3, 13220; Nf4, 13956
Nf1, 0.02; Nf2, 0.02; Nf3, 0.07; Nf4, 0.03
2 mm
N, 81245; E, 45362
Nf1, 1726; Nf2, 3326; Nf3, 4947; Nf4, 6945
Nf1, 1.65; Nf2, 0.63; Nf3, 0.50; Nf4, 0.37
N, 6412; E, 1183
Nf1, 9754; Nf2, 12095; Nf3, 13072; Nf4, 13929
Nf1, 1.87; Nf2, 0.66; Nf3, 1.19; Nf4, 0.22
N, nodes;
E, elements; Nf, natural
frequency (Hz).
N, nodes;
E, elements; Nf, natural
frequency (Hz).
Brake Noise and Vibration Evaluation
The developed
brake pads were characterized for noise and vibration
(NV) performance on the in-house-developed NV test rig (the test rig
specifications are detailed in ref (30) as shown in Figure ).
Figure 2
(a) 3D solid work model and (b) top view of
the brake noise–vibration
test rig.
(a) 3D solid work model and (b) top view of
the brake noise–vibration
test rig.The brake pad samples (size :
25 mm × 25 mm) were cut from
the developed pads and fixed in the pad holder slots as shown. In
addition, the brake disc (material, gray cast iron) was fixed opposite
to the pads. The hydraulic pressure system was used to perform the
braking operation. The sound pressure and vibration signals were recorded
using a microphone (model: PCB make model 377C13) and accelerometer
(model: PCB make model 352C03) during braking operations. The microphone
position was set as per the SAE J 2521 test standard, 50 cm above
and 10 cm offset from the center of the disc. The acquisition of vibration
and sound pressure data and subsequent processing were done using
an NI DAQ express system using an NI-9234 data card. The sampling
rate of 25600/s was chosen for the NV tests. Table shows the design of experiments (DOE) for
brake NV performance evaluation. Before starting the actual test,
a bedding operation was performed to get the conformal contact for
all pads. All the tests were performed under the same test conditions
(i.e., background noise level of <45 dBA, relative humidity of
50 ± 10% RH, and temperature of 30 ± 2 °C) in the laboratory.
The fresh brake pads and disc were used for each test. Brake NV signals
were recorded for 10 braking applications in each cycle and used for
the NV analysis. The octave band analysis of the acquired acoustic
pressure time-domain vibration of acceleration data was carried out
in an in-house-developed MATLAB code.
Table 3
Design
of Experiment (DOE) for NV
Measurements
braking cycle
pressure
(bar)
speed (kmph)
number of
brake application
number of
NV measurement
bedding
30
80
100
performance
10, 20, 30
50 and 80
10 each combination
10 each combination
forward/backward
10
20 and −20
10 each combination
10 each combination
drag
10, 20, 30
10 and
20
10 each combination
10 each combination
Brake
Squeal Prophecy Model
Modeling of the Disc–Brake
Pad System
For the brake pad–disc assembly, it was
decided to investigate
the eigenvalue analysis in ABAQUS. A solid model of disc and pads
was prepared as per the dimensions of the discs and pad samples as
shown in Figure .
Later, this solid model was imported into ABAQUS/CAE and was fine-meshed.
The mesh quality was confirmed using elemental failure criteria with
right values of shape factor and aspect ratio used, and analysis errors
and warning were checked and found to be 0%.
Figure 3
Disc brake system with
(a) solid work model, (b) disc–pad
boundary conditions, and (c) fine mesh model.
Disc brake system with
(a) solid work model, (b) disc–pad
boundary conditions, and (c) fine mesh model.The equation of motion of the disc brake unit can be written as
follows:[3]where {x}
is a nodal displacement vector, and [M], [C], and [K] are the mass, damping, and stiffness matrices of the system, respectively.The above matrices are asymmetric due to friction. The greater
the asymmetry, the more easy the possibility of friction-induced vibration
of the disc brake system.[48]In the
current work, the complex eigenvalue analysis (CEA) method
was used to solve the equation of motion of the disc–pad system.
The CEA helps in finding all unstable eigenvalues and corresponding
vibration modes under the condition of linearized nonlinearities,
which are close to steady sliding states. In the CEA, the eigenvalue
equation can be represented as follows:where λ is the eigenvalue
and ϕ is the corresponding eigenvector. Additionally, the general
solution can be found using the subspace iteration technique, which
is as follows:where “t” is the time and α is
the real part and ω is the imaginary
part for the ith eigenvalue. If the real part of
the eigenvalue (α) is a positive value (i.e., α > 0),
then it indicates an unstable vibration of the brake system. The brake
system thus vibrates with a large amplitude of vibration. Generally,
the effective damping ratio (ζE) is used to quantify
the intensity of unstable vibration, which is defined as,[10]The unstable vibrations can occur if the effective damping ratio
is negative (i.e., ζE < 0). The smaller the ζE value, the more the corresponding squealing vibrations are.
The effective damping ratio (ζE) can be obtained
from the CEA.
Model Constraints and
Parameters
Figure shows the
disc brake system consisting of brake pads and a disc. For carrying
out the eigenvalue analysis, the FE model of pads and the disc is
created. In the fine mesh model, there are a total of 49382 nodes
and 46062 elements (types: disc, C3D10; pad and backplate, C3D8R).
All the nodes of the central holes of the disc were fixed in all directions
(i.e., x, y, and z). The brake disc and pads are the friction surface pairs. The pads
were connected to backplates using tie-type constraints of ABAQUS.
The tangential behavior with the penalty method was used as a contact
property for the disc–pad pair. The friction coefficient was
considered as a constant value during the sliding for the selected
operating pressure and speed condition. The material properties of
brake discs, pads, and backplates are enlisted in Table . The value of the positive
real part has a strong correlation with the friction-induced vibrations.
Thus, the eigenvalue with the positive real part indicates an unstable
frequency of vibration that in turn reflects into a squeal frequency.
Table 4
Material Properties Used for FEA Analysis
brake pad (low metallic)
property
disc (gray
cast iron)
Ref
Zylon fiber
backplate
(mild steel)
Young’s modulus (GPa)
125
11
11
210
Poisson’s ratio
0.25
0.34
0.34
0.3
density (g/cc)
7.16
2.3
2.1
7.85
damping ratio
0.01
0.027
0.040
0.007
Results and Discussion
Characterizations of Brake
Pads
Table enlists the physical
and mechanical properties of the developed pads.[43] The density was the highest for the reference pad due to
the high amount of higher density space filler (i.e., barite). However,
aramid/Zylon fiber-based pads showed more or less similar density
values except the Z6 pad. The reference pad showed the
lowest porosity and compressibility, while higher compressibility
was found for aramid fiber- and Zylon fiber-based pads. Slightly higher
compressibility was observed for Z3 and Z1 pads
compared to aramid fiber-based pads. The marginal difference in hardness
was observed for all pads. Porosity, compressibility, and hardness
can affect brake NV performance.[44]
Table 5
Properties of the Developed Brake
Pads[43]
property
stds.
units
Ref
AP
AF1
Z1
Z3
Z6
density
g/cc
2.30
2.16
2.11
2.12
2.14
2.19
water porosity
JIS D 4418
%
8.45
10.28
9.10
9.92
9.94
9.62
oil porosity
JIS
D 4418
%
9.81
11.8
10.23
10.64
11.58
10.43
hardness
ASTM D 785
HRR
90.9
93.7
91.8
95.3
92.1
96.6
compressibility
ISO 6310
μm
57
76
72
81
87
74
Friction and Wear Studies
Figure a–e shows
the friction and wear results for the tested pads during the NV cycles.
Figure 4
(a–e)
Average coefficient of friction in braking cycles
and (f) weight loss of brake pads.
(a–e)
Average coefficient of friction in braking cycles
and (f) weight loss of brake pads.The average coefficient of friction (μ) values of 10 brakes
with standard deviations are reported for all braking cycles. It was
observed that the average μ was more or less similar for all
pads except a few cases. The aramid short fibers showed slightly higher
μ compared to aramid pulp-based pads. However, with the increasing
size of Zylon fibers, a marginal difference was observed in all cycles.
Overall, the reference pad showed slightly more fluctuation in μ
compared to aramid fiber- and Zylon fiber-based pads. Figure f shows the weight loss of
the brake pads after the competition of test. The wear resistance
was compared to all other types of pads. Aramid and Zylon short fibers
proved to be the best, while the reference pad proved to be the worst
in wear resistance. Wear was increased slightly with the increase
in the size of fibers.
Modeling of the Brake Pad
and Brake Disc
Using the FE model for pads and discs, eigenvalue
analysis has
been carried out, and Table details the natural frequencies and mode shapes for the brake
disc and brake pad samples.
Table 6
Natural Frequencies
and Mode Shapes
for the Disc and Brake Pad Samples
Corresponding to the disc natural frequencies, the
reference pad
frequencies such as mode 2 (12095 Hz) and mode 5 (15153 Hz) are very
close to the natural modes of the disc (modes 5 and 6). The same is
observed in mode 4 of the aramid/Zylon fiber-based pad. The closeness
leads to a ready mode coupling and stronger vibration levels. However,
the overall modal damping can play an important role in the unstable
vibration.
Experimental Modal Analysis
Figure shows the experimental
FRF based on measurement data for the developed brake pads. The FRF
shows the variation of the ratio of the amplitude (acceleration, m/s2) and applied force (N) with frequency (Hz). The blue curve
shows the FRF for the pads. The circle fit method was used to find
and locate the modal frequencies (for the natural modes). The blue
curve displays the FRF plots based on the measured data, while the
red curve represents the curve-fitted FRF based on the circle fit
method using the resonances picked up and chosen from the measured
FRF plot.[45] The initial four modes of vibrations
for each brake pad were chosen, and the corresponding identified natural
frequencies and damping ratios are reported in Table .
Figure 5
FRF plots for the developed brake pads (blue:
measured; red: curve-fitted).
Table 7
Experimental Natural Frequencies and
Damping Ratios of Padsa
brake
pad
Ref
AP
AF1
Z1
Z3
Z6
mode
ω (Hz)
ζ
ω (Hz)
ζ
ω (Hz)
ζ
ω (Hz)
ζ
ω (Hz)
ζ
ω (Hz)
ζ
1
4175
0.045
4175
0.048
4275
0.065
4125
0.051
4150
0.035
4225
0.047
2
6375
0.023
6250
0.035
6600
0.024
6125
0.048
6525
0.055
6100
0.038
3
9350
0.010
7850
0.019
9375
0.055
8975
0.036
9400
0.024
8850
0.029
4
12575
0.029
12300
0.031
10625
0.014
11350
0.023
11900
0.021
12,625
0.036
avg. damping
0.027
0.033
0.040
0.040
0.034
0.038
Here, ω is the natural frequency
(Hz), and ζ is the damping ratio.
FRF plots for the developed brake pads (blue:
measured; red: curve-fitted).Here, ω is the natural frequency
(Hz), and ζ is the damping ratio.Overall, it may be noted from the amplitudes of the
peaks in the
FRF that due to the inclusion of aramid and Zylon fibers, damping
improved significantly, and natural frequencies were reduced for most
cases. However, the short fibers (1 mm) of aramid and Zylon proved
to be the best for the average damping ratio, while the reference
proved to be the worst.Figure shows the
comparison of the natural frequencies obtained from experimental and
numerical (FEA) methods with corresponding mode shapes (B, bending;
T, twisting) for reference brake pads. The natural frequencies of
pads using experimental and numerical methods were observed to be
more or less similar with a marginal difference (<5%) except the
second mode. Additionally, a slight difference (4–8%) in the
natural frequency of the disc rotor was observed for experimental
and numerical methods. Overall, it is noted that the numerical FE
model of the brake components is reasonably accurate.
Figure 6
Experimental and numerical
natural frequency results for (a) brake
pads and (b) disc rotor with corresponding mode shapes.
Experimental and numerical
natural frequency results for (a) brake
pads and (b) disc rotor with corresponding mode shapes.
Brake Noise and Vibration Performance
The brake pads were assessed for experimental NV performance for
four braking cycles (viz., performance, forward, backward, and drag)
under various pressure and speed conditions. The pad–disc touch
times of 1 s for the performance cycle, half-second for forward/backward
cycles, and 4 s for the drag cycle were kept fixed during the braking
operations. Figures and 8 highlight the time-domain and fast
Fourier transform (FFT) vibration (g) spectra for braking and no-braking
conditions. The acquisition of braking NV signals was started before
the engagement of brake pads with the disc and stopped after disengagement
for all the brake cycles. The analysis of braking signals was done
for the actual braking time, as shown in the time domain in Figures c and 8c. Figure a,b shows the time domain and FFT plot (entire frequency range: 0–12,800
Hz) for the no-braking condition (80 kmph). As expected, the amplitude
of vibration increased due to the braking action compared to the no-braking
condition. The frequency spectra (Figure b) show the presence of some small amplitude
(<0.01 g) vibrational frequencies (49 and 61 Hz), which could be
related to the rotational motion of the disc or structural vibration. Figure c,d shows the time
domain and FFT plot for the braking action (reference pad). The FFT
plot (Figure d) clearly
indicates that the vibration amplitude drastically increased due to
friction-induced vibration between tribo-couples (pad–disc).
Due to the braking action, new vibrational frequencies (47, 193, 383,
and 580 Hz and some high-frequency vibrations) appeared, which are
highlighted in Figure d. The higher frequency vibrations (10–12 kHz) were observed
due to structural resonances. FFT plots of low-frequency regions for
all pads are shown in Figure d,f–j.
Figure 7
(a, c, e) Time-domain and (b, d, f–j) FFT spectra
of vibration
(g) without (background) and with braking conditions (performance
cycle, 80 kmph, 10 bar).
Figure 8
(a, c, e) Time-domain
vibration response and (b, d, f–j)
corresponding frequency spectra with and without braking conditions
(drag cycle, 10 kmph, 10 bar).
(a, c, e) Time-domain and (b, d, f–j) FFT spectra
of vibration
(g) without (background) and with braking conditions (performance
cycle, 80 kmph, 10 bar).(a, c, e) Time-domain
vibration response and (b, d, f–j)
corresponding frequency spectra with and without braking conditions
(drag cycle, 10 kmph, 10 bar).As discussed later, the brake noise spectra were dominated by mainly
the low-frequency components (<500 Hz).The time- and frequency-domain
vibration data for the drag cycle
(speed, 10 kmph; pressure, 10 bar condition) are shown in Figure . The braking action
led to an increase in vibration level significantly compared to the
no-braking condition, as shown in Figure a,c,e. The reference pad showed more unsteady
and higher vibrations compared to aramid/Zylon fiber-based pads due
to their lower damping, porosity, and compressibility values. Brake
vibrations are directly proportional to the damping property of materials.[29,30,44,45]Tables and 6 show that the porosity, compressibility, and damping
values were lower for reference pads than aramid/Zylon fiber-based
pads. The damping performance order of brake pads was as follows:
AF1 (0.40) = Z1 (0.40) > Z6 (0.38)
> Z3 (0.34) > AP (0.33) > Ref (0.27).In
addition, there was a more or less similar trend observed in
the vibration amplitude for other pads, as shown in Figure f–j.Overall,
the higher values of compressibility, porosity, and damping
resulted in attenuation of the vibration amplitude during the braking
action for all the cycles, which finally influenced better noise performance,
which will be discussed in the later section.
Root
Mean Square (RMS) Vibration
The noise and vibration measurements
were repeated 10 times for all
the brake pads in each braking cycle, and the measured data were averaged
out. The noise and vibration readings are therefore an average value
of 10 readings. Figure a–e shows the average RMS vibration during braking under different
brake conditions for all the pads. The average RMS vibration (in g)
values were calculated for the entire frequency range (0–12800
Hz).
Figure 9
(a–e) Average RMS vibration (g) during braking conditions
for different brake pads. (f) Overall improvement in RMS vibration
(%) relative to the reference pad considering all braking cycles.
(a–e) Average RMS vibration (g) during braking conditions
for different brake pads. (f) Overall improvement in RMS vibration
(%) relative to the reference pad considering all braking cycles.It was observed that the RMS vibration increased
for the same pressure
conditions with an increase in speed. However, there was a slight
drop in the RMS vibration with an increase in pressure for the same
speed condition. Overall, the aramid/Zylon fiber-based brake pads
showed lower average RMS vibrations in all the brake cycles compared
to the reference pad. Figure f shows the overall improvement in the RMS vibration (%) for
aramid/Zylon fiber-based pads compared to the reference pad considering
all the braking cycles. The Z6 pad proved to be the best
with a maximum improvement (26.3%) in average RMS vibration compared
to the reference pad. Additionally, the other two pads, short aramid
(AF1) and Zylon (Z1), showed more or less similar
improvement (21–22%) in average RMS vibration.
Brake Noise Performance
Figure a–e shows
brake noise level (dBA) results during the braking action for various
test conditions. According to the SAE J 2521 test schedule, the background
noise level was maintained below 60 dBA inside the acoustic chamber
for all the speeds in the no-braking conditions. It was observed that
the noise level increased with an increase in speed for the same pressure
condition. The brake noise level was raised as the pressure increased
for all pads except a few cases in the performance cycles. The reference
pad showed the highest noise level compared to all pads. The brake
noise level dropped (drag: 1–1.5 dBA ± 0.3–0.4
dBA; performance: 1–1.5 dBA ± 0.5–0.7 dBA; forward/backward:
2–2.5 dBA ± 1–1.5 dBA) for aramid/Zylon fiber-based
pads compared to the reference pad. Overall, the reduction in brake
noise (dBA) due to aramid/Zylon fibers in the pad is shown in Figure f. The AF1 and Z3 pads showed a maximum reduction in the overall
noise level (1.5 dBA). Additionally, AP, Z1, and Z6 pads showed an almost similar reduction in the overall noise
level (1.2 dBA). Overall, the short aramid fiber (AF1)-
and Zylon fiber (Z3)-based pads proved to be the best in
noise performance among all pads.
Figure 10
(a–e) Overall sound pressure level
(dBA) during braking
conditions for different brake pads. (f) Overall reduction in SPL
(dBA) relative to the reference pad considering all braking cycles.
(a–e) Overall sound pressure level
(dBA) during braking
conditions for different brake pads. (f) Overall reduction in SPL
(dBA) relative to the reference pad considering all braking cycles.
Brake Squeal Quantification
Experimental Analysis
Figure shows the time-domain
and vibration (g) spectra for reference and Zylon fiber-based pads
in the forward brake cycle. Brake squeal was observed under low-speed,
low-pressure test conditions for the reference pad only in the forward
cycle. No squeal was observed for high-speed and high-pressure (>10
bar) conditions. However, in the forward cycle, friction was observed
in the range of 0.42–0.44.
Figure 11
Time-domain and FFT spectra of brake
vibration (g) for (a, b) reference
and (c, d) Zylon fiber-based pads in a forward cycle.
Time-domain and FFT spectra of brake
vibration (g) for (a, b) reference
and (c, d) Zylon fiber-based pads in a forward cycle.The high-frequency brake squeal was observed at a frequency
of
9381 Hz for the reference pad, and the corresponding time-domain and
vibration (g) spectra are highlighted in Figure a,b. It may be noted that the third mode
of the brake pad–disc assembly is found to be 9350 Hz (Table ). The main reasons
of squeal in the case of the reference pad (i.e., without aramid/Zylon
fibers) are higher density and lower damping of pad material in addition
to friction. All the aramid fiber- and Zylon fiber-based pads showed
no squeal tendency for all selected test conditions due to their lower
density and higher damping of pad material, although friction was
almost in a similar range as that of the reference pad. Figure c,d shows the time
domain and FFT plot of vibration for the Zylon fiber-based pads. A
similar trend in vibration spectra was observed for other Zylon fiber-
and aramid fiber-based pads. The brake vibration amplitude was lower
and without any squeal frequency in aramid fiber- and Zylon fiber-based
pads compared to the reference pad (without aramid/Zylon fibers). Figure d shows no squeal
frequency in the brake vibration spectra.
Numerical
Analysis
Complex eigenvalue
analysis (CEA) on a mechanical system helps one to understand the
nature of the eigenvalues of the system. The real part of the complex
eigenvalue thus obtained gives information on the stability of the
vibratory system. It is used to find the unstable modes of vibration,
and the corresponding mode shapes can be extracted. It is common to
model the vibratory system using the finite element approach by discretizing
the continuum into several discrete finite elements. This discretized
model and subsequent numerical approach to find the set of complex
eigenvalues with their real and imaginary parts help identify the
rate of growth and unstable frequencies of the system.[10,46] In the current study, CEA is performed to predict the possible unstable
eigenfrequencies of the brake pad–disc assembly between 0 and
16 kHz for the forward cycle (speed, 20 kmph; pressure, 10 bar condition)
for the brake squeal prophecy of the developed pads.In the
CEA, the actual density, damping, and friction properties were considered,
while the elastic modulus (E) and Poisson’s
ratio were assumed to be constant for all pads. Table shows the material properties of brake system
components. Figure shows the mode shapes of the brake pad–disc assembly obtained
through the FE simulation. The results in Table show three dominant unstable frequencies
(i.e., 9226, 11,809, and 14,593 Hz) for the reference pad and corresponding
absolute values of the real part of the eigenvalue (i.e., 14.2, 353.7,
and 1352) and effective damping ratio (i.e., −0.003, −0.061,
and −0.185). The simulated natural frequency of the reference
brake pad (9226 Hz) matches close to the experimentally observed natural
frequency of 9350 Hz (Table ). The additional one unstable vibration frequency close to
11800 Hz (Figure b,e) is observed in numerical analysis (CEA) in contrast to experimental
analysis for the selected frequency range (0–12800 Hz). The
brake pad and disc may be coupled at higher frequencies due to friction
when the pad–disc natural frequencies come closer. If modal
frequencies of the pad–disc pair are close, then modal coupling
occurs and generates the brake squeal. In the current study, the modal
frequencies of the disc (9299 Hz) and reference pad sample (9754 Hz)
are closer to the observed squeal frequency. However, the first natural
frequency of aramid fiber- and Zylon fiber-based pads occurred around
10356 Hz, which was far from the disc modal frequency (9299 Hz). In
addition to this, the next modal frequencies for the disc and pads
are closer to a higher frequency range beyond 12000 Hz and may not
be readily excited. The probability of squeal noise generation can
be reduced by reinforcing high damping aramid/Zylon fibers in the
pad and ensuring the adequate difference in the modal frequencies
of the brake pad and disc.
Table 8
Material Properties of Brake System
Components
brake
pad
property
disc
backplate
Ref
Zylon fiber
Young’s modulus (GPa)
125
210
11
11
Poisson’s ratio
0.23
0.3
0.34
0.34
density (g/cc)
7.16
7.85
2.3
2.1
damping ratio
0.01
0.007
0.027
0.04
Figure 12
Mode of vibration for (a–c) reference
pad-unstable modes,
(d) Zylon fiber pad-stable mode (no squeal event), and (e, f) Zylon
fiber pad-unstable modes (fs, squeal frequency).
Table 9
Real Parts
of Complex Eigenvalue and
Effective Damping Ratio for Squeal Frequencies
reference
pad
Zylon fiber pad
squeal frequency
(Hz)
eigenvalue
(real part)
effective
damping ratio (ζE)
squeal frequency
(Hz)
eigenvalue
(real part)
effective
damping ratio (ζE)
mode 29: 9226
14.2
–0.003
mode 29: 9232
–16.2
0.004
mode 39: 11809
353.7
–0.060
mode
39: 11821
295.1
–0.050
mode 50: 14593
1352
–0.185
mode 50: 14628
1165.1
–0.159
Mode of vibration for (a–c) reference
pad-unstable modes,
(d) Zylon fiber pad-stable mode (no squeal event), and (e, f) Zylon
fiber pad-unstable modes (fs, squeal frequency).In addition to the above studies, the effect of damping
was studied
for reference and aramid/Zylon fiber-based pads, as shown in Figure . The results revealed
that with an increase in the damping of the pad, the effective damping
ratio (i.e., brake squeal intensity) decreased slightly for both pads.
Overall, it was found that the density of the brake pad plays a vital
role in addition to the damping in reducing the brake squeal propensity.[49]
Figure 13
Effect of pad damping on squeal for (a) the reference
pad and (b)
aramid/Zylon pad.
Effect of pad damping on squeal for (a) the reference
pad and (b)
aramid/Zylon pad.Generally, the magnitude
of the real part of the eigenvalue shows
the brake squeal propensity to a certain degree. However, there are
various unstable frequencies that show the squeal propensity. Thus,
to quantify the squeal propensity for these three types (Ref, aramid,
and Zylon fibers) of pads, the tendency of instability (TOI) is calculated
using the given formula.[47]where α and ω are the
real and imaginary parts of the jth eigenvalue, and n is the total number of positive eigenvalues.Figure shows
the TOI of three types of brake pads. The squeal propensity is larger
for higher TOI. While the reference pad shows the TOI of 127.3, the
aramid fiber-based pad showed the lowest (100.3) TOI, followed by
Zylon fiber-based pads with more or less similar values (102.9). The
TOI was mainly controlled by enhancing the damping of brake pads as
shown in Table . The
reference brake pad comprises barite (BaSO4) having a higher
density ( = 4.4
g/cc) instead of lighter ingredients ( = 1.44–1.54 g/cc) such
as aramid and Zylon fibers, thus decreasing the density of brake pads.
The inclusion of fibers leads to higher porosity and, hence, compressibility
compared to particulate ingredients, which further leads to enhanced
damping.[43] Overall, the improvement in
the squeal propensity is 21.2 and 19.2% for aramid fiber- and Zylon
fiber-based pads relative to the reference pad, respectively. Hence,
the inclusion of aramid or Zylon fibers is an effective method to
minimize the squeal propensity through damping enhancement.
Figure 14
Tendency
of instability (TOI) of three types of pads.
Tendency
of instability (TOI) of three types of pads.In addition, the wear mechanism is already explained in our earlier
tribological article,[43] and this article
was a part of the noise–vibration study. The reference pad
showed more unstable secondary plateaus and more wear debris on the
surface. More secondary plateaus support higher wear of the pad.[51] The friction materials with large contact plateaus
produced squeal noise due to the high impact energy transferred to
the brake assembly. In contrast, the specimens with small plateaus
produced chatter without squeal noise. The control of contact plateaus
could be the key to reduced friction-induced vibrations.[11] In the earlier tribological study, similar mechanisms
(more contact plateaus) were observed without aramid/Zylon fiber-based
pads.[43]
Conclusions
In the present work, various types of aramid/Zylon fiber-based
brake pads were evaluated for noise and vibration characteristics
on the developed brake noise–vibration test rig. In addition,
the analysis of brake squeal frequencies was done using both methods
(i.e., experimental and numerical). The following are the salient
conclusions that were drawn.The inclusion of aramid/Zylon fibers in the pads aided
in decreasing the density and increasing the porosity, compressibility,
and damping properties of the brake pads compared to the reference
pad.Experimental results of the noise–vibration
(NV)
test rig showed that damping, compressibility, and density control
the intensity of noise and vibration. It was found that both low-
and high-frequency noise and vibration can be suppressed using high
damping aramid and novel Zylon fibers in the pads without affecting
the friction performance.Overall reduction
in the brake noise (1.2–1.5
dBA) and improvement in vibration (20–25%) were observed for
aramid fiber- and Zylon fiber-based pads relative to the reference
pad.Complex eigenvalue analysis of the
brake system showed
that brake squeal was mainly influenced by the damping and density
of the pad materials.Aramid/Zylon fiber-based
brake pads can effectively
suppress the instability of the brake system and reduce the brake
squeal propensity.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14