Tribology International 119 (2018) 585–592

Contents lists available at ScienceDirect

Tribology International
journal homepage: www.elsevier.com/locate/triboint

Mechanical and tribological behaviors of copper metal matrix composites

for brake pads used in high-speed trains
Yelong Xiao a, 1, Zhongyi Zhang a, 1, Pingping Yao a, *, Kunyang Fan b, Haibin Zhou a,
Taimin Gong a, Lin Zhao a, Minwen Deng a
a
State Key Laboratory of Powder Metallurgy, Central South University, Changsha 410083, PR China
b
College of Materials Science and Engineering, Xihua University, Chengdu 610039, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: As one of the most important components in high-speed trains, demands are improved on the mechanical and
Metal matrix composites tribological properties of materials for brake pads. In this study, a newly developed copper metal matrix com-
Mechanical properties posite (Cu-MMC) for the aforementioned brake pads was fabricated by powder metallurgy route. The micro-
Friction materials structure and mechanical properties of Cu-MMC were investigated. Cu-MMC was tribo-evaluated by a full-scale
Brake pad dynamometer, and special attention was paid on the braking performances during emergency stop-braking at
initial speeds from 300 to 380 km/h. Examination and analysis of the worn surface and subsurface corroborated
the wear mechanism. The results indicate that Cu-MMC exhibits excellent properties and can meet the technical
requirements, so it holds great promise for applications in high-speed trains.

1. Introduction
Compared with other forms of transportation, high-speed trains have
the characteristics of large capacity, fast running speed, high transport
efficiency, low operational cost and high safety coefficient [1,2].
Consequently, a serious interest in this rapidly expanding transport mode
has been increasing all over the world. China has successfully developed
and commercially operated CRH380 high-speed trains with the
maximum speed of 350–380 km/h [3–5].
High speed means more dynamic actions and potential danger. In
order to ensure safe operations, braking systems of high-speed trains are
required to keep enormous braking forces under control. In braking
systems, in addition to the regenerative braking system, the friction
braking system is used to guarantee the performances and safety, espe-
cially at low speed and in case of emergency or failure [6]. In case of
emergency, the friction braking system must be capable of stopping a
CRH380 train running at 300 km/h at a maximum acceptable braking
distance of 3800 m. Huge kinetic energy would be generated during the
emergent braking process, and it is required to be dissipated in brake
discs and pads in a short time (up to 19 MJ per disc in less than 2 min).
The flash temperature at the disc/pad interface may reach up to
900
C [7].
In order for high-speed trains to operate safely and comfortably, de-
mands are proposed on the properties of brake pads, such as the pre-
dictable friction behavior, satisfactory wear resistance, low aggressivity
against the brake disc (e.g., burn marks and thermally initiated cracks),
high comfort qualities (i.e. low squeal noise and anti-shudder), and suf-
ficient mechanical characteristics [6,8,9]. Various materials have been
developed as candidate for brake pads. Organic brake pads, defined as
polymer-based friction materials containing multiple ingredients, have
been broadly applied to railroad passenger coaches and locomotives
[10,11]. However, poor thermal stability of organic materials, especially
the thermal fade, extremely limits applications of organic brake pads,
which can only be used at temperatures lower than 300
C generally
[12]. Compared to organic friction materials, metal matrix composites
can typically withstand higher temperatures and can be used to tem-
peratures over 1000
C. Especially, Cu-MMCs are the top choices as brake
materials for high-speed trains due to their attractive properties such as
high thermal conductivity, excellent tribological properties, and adapt-
ability to working conditions [8,9,13–16].
Considerable attention has been paid to the improvement in prop-
erties of Cu-MMCs. Gyimah et al. [14] investigated the wear behavior of
a novel Cu-based friction material for train brake pads with a high
pressure pad-on-disc braking tester. Su et al. [15] evaluated the

* Corresponding author.
E-mail address: yaopingpingxx@sohu.com (P. Yao).
1
Yelong Xiao and Zhongyi Zhang contributed equally to this work and should be considered co-first authors.

https://doi.org/10.1016/j.triboint.2017.11.038
Received 25 July 2017; Received in revised form 23 November 2017; Accepted 26 November 2017
Available online 2 December 2017
0301-679X/© 2017 Elsevier Ltd. All rights reserved.
Y. Xiao et al. Tribology International 119 (2018) 585–592

important concern on brake pads.

Nomenclature The current work aims to present a newly developed powder metal-
lurgy processed Cu-MMC for high-speed trains, and introduce the prep-
P load during the shear test (N) aration, mechanical properties and full-scale dynamometer test results of
L length of the specimens during the shear test (mm) this material.
B width of the specimens during the shear test (mm)
FB clamping force (namely total nominal contact force) 2. Experimental procedures
during braking (N)
Ft total braking force during braking (N) 2.1. Materials preparation
Fb total contact force during braking (N)
S2 braking distance (m) In this study, the chemical composition of Cu-MMC is shown in
m0 weight of the brake pad before the friction tests (g) Table 1. The commercial-grade powders of electrolytic Cu (99.7% purity,
mi weight of the brake pad after the friction tests (g) <74 μm), reduced Fe (98% purity, <74 μm), graphite (97% purity,
E braking energy during braking (MJ) 150–600 μm), MoS2 (96% purity, <6.5 μm), ZrO2 (92% purity,
100–250 μm), and FeCr intermetallic compound (92% purity, <74 μm,
Greek letters
hereafter referred to simply as “FeCr”) were used as the main
τ average shear strength during the shear test (MPa)
raw materials.
μa instantaneous coefficient of friction during braking
As illustrated in Fig. 1, the powder metallurgy route for
μm average coefficient of friction during braking
manufacturing Cu-MMC involves the following sequential steps. The
ω wear rate of the friction material (cm3/MJ)
powders listed in Table 1 were accurately weighed out in amounts cor-
ρ density of the friction material (g/cm3)
responding to the desired proportions, and thoroughly mixed in a V-cone
blender for 8 h. After the blending process, the mixed powders were
consolidated into compacts via cold die compaction, under a pressure of
Table 1 650 MPa at room temperature. Subsequently, the green compacts were
Chemical composition of Cu-MMC. sintered directly to the steel backing plates under a pressure of 2.5 MPa,
Element Cu Fe Graphite MoS2 ZrO2 FeCr Others
in a protection atmosphere of H2 or cracked ammonia, at a temperature
of 970
C for 180 min. After sintering, the samples were cooled down to
Content (wt. 40–55 10–12 17–19 2–3 6–8 6–8 4–10
room temperature in the furnace.
%)

2.2. Experimental apparatus

tribological properties of copper-based friction materials with exoge-
nous copper powder third body by a pin-on-disc tribometer. Goo et al.
A part of Cu-MMC was cold mounted for the metallographic exami-
[9] developed three Cu-MMCs for brake pads, and found that the fric-
nation. The micro-images were taken using an optical microscope (Leica
tion coefficients of the developed materials were higher than those of
DM4500P, Germany). The bulk density of Cu-MMC was measured by the
materials in service during pin-on-disc experiments. Chung et al. [16]
Archmedes' method at room temperature. The surface hardness of Cu-
examined the brake performances of sintered friction materials devel-
MMC was evaluated via a Brinell hardness tester (310 HBS-3000,
oped for high-speed trains on a 1/5 scale dynamometer, and claimed
China) using a standard 10 mm diameter tungsten carbide ball
that the improvement in tribological properties of the newly developed
indenter, at a normal load of 250 kgf and a dwelling time of 30 s.
friction materials was attributed to the high graphite content. Kim [17]
The schematic illustrations of the shear strength testing apparatus are
reported a full-scale dynamometer that was designed to simulate the
given in Fig. 2. The shear tests were conducted in a universal testing
brake characteristics of high-speed trains, and gave a brief description
machine (Instron 3369, USA) at a cross-head rate of 1 mm/min, using
of the braking performance evaluation on brake pads. For practical
specimens of approximate dimensions 15  15 mm2 cut from Cu-MMC.
applications, it is essential to assess the safety of brake pads by verifying
The shear planes of the specimens were parallel to the friction surfaces
their braking performances on full-scale dynamometers. Nevertheless,
of Cu-MMC. The load and displacements were monitored continuously
reports concerning the braking performances of Cu-MMCs for brake
until failures of the specimens, the load (P) registered at debonding was
pads tested on full-scale dynamometers are limited, especially at speeds
converted into an average shear strength (τ) by
over 300 km/h.
In addition, for brake pads in service, there are still unsolved prob- P
lems concerning the low mechanical strength of the friction materials, as τ¼ (1)
Lb
well as the insufficient bonding strength between the friction materials
and the metal backing plate, frequently resulting in spallation of the where L is the length, and b is the width.
friction materials and detachment of the friction material from its According to TJ/CL 307–2014C.4 test programme issued by China
backing plate [18]. Thereby, improving mechanical properties is also an Railway Corp. [19], the friction and wear tests of Cu-MMC, simulating the

Fig. 1. Processing steps for Cu-MMC.

586
Y. Xiao et al. Tribology International 119 (2018) 585–592

Fig. 2. Schematic illustrations of the shear tests performed: (a) shear strength of the friction material, and (b) bonding strength between the friction material and its backing plate.

real operating conditions of CRH380 train brake pads, were conducted on a
full-scale dynamometer (Fig. 3) used to simulate the brake performance of
the railway vehicle. The dynamometer consists mainly of motor, inter-
changeable flywheels, emergency braking unit (for affording the protection
in case of an accident), clutch, static-friction actuator (for measuring the
static coefficient of friction) and disc brake system. The dynamic energy was
supplied by the flywheels, which was driven by the motor. As shown in
Fig. 4(a), the brake pad rubbed on the brake disc around the horizontal axle.
The schematic view of Cu-MMC brake pad is shown in Fig. 4(b).
TJ/CL 307–2014 specifies the frictional, mechanical, physical and
chemical features of brake pads for high-speed trains in China [19]. The
approval procedure for brake pads including the test programmes to be
used for each type is described. The instantaneous and average coefficients
of friction to be achieved are shown in diagrammatic form. To bed-in the
brake pad up to at least 85% contact area, the adequate pre-tests (bed-
ding-in brake applications) should be conducted. The brake applications
under wet conditions are carried out with a water quantity of 25 L/h. So
the brake applications of Cu-MMC brake pad were carried out in the given
sequence. The main test parameters employed in the full-scale dyna-
mometer tests are given in Table 2. CRH380 trains that run at speeds over
250 km/h have a two stage pressure system, so the clamping forces (FB,
namely total nominal contact forces per disc) consisted of two stages in the
stop-braking tests performed at 250, 300, 320, 350, and 380 km/h. FB was
45 N during parking braking. The surface temperature was measured by a
thermocouple embedded in the brake disc.
The instantaneous coefficient of friction (μa), determined at any
instant of braking by the ratio of the total braking force (Ft) to the total
contact force (Fb), was calculated as [20]

Fig. 3. Schematic diagram of the full-scale dynamometer. (1) Concrete base, (2) motor, Ft
(3) emergency braking unit, (4) clutch, (5) interchangeable flywheels, (6) static-friction
μa ¼ (2)
Fb
actuator, (7) disc brake system.

and the average coefficient of friction (μm), determined by integrating μa

from reaching 95% of FB μa over the braking distance (S2), was calculated
as [20].

1 S2
μm ¼ ∫ μ ds (3)
S2 0 a
During full-scale dynamometer tests, the wear rate (ω) was defined as
m0  mi
ω¼ (4)
ρE

Table 2
Main test parameters employed in the full-scale dynamometer tests.

Parameters Values

Brake pad Surface area, 400 cm2

Brake disc Φ 640  80 mm, with reduced fan capacity, made of
cast steel
Mass per brake disc (t) 5.7
Wheel diameter (mm) 890
Initial stop-braking speeds 50, 80, 120, 160, 200, 250, 300, 320, 350,
(km/h) 220 380
Clamping forces (N) 12, 23, 31.5 FB1/FB2 (18/23, 24/
31.5)a
a
FB in two stages: FB1 from initial stop-braking speed to 215 km/h and FB2 from 215 km/
Fig. 4. View of (a) the disc brake system and schematic drawing of (b) the brake pad. h to 0 km/h.

587
Y. Xiao et al.
Fig. 5. Optical microscope images of Cu-MMC: (a) parallel to the friction surface, and (b) perpendicular to the friction surface.
where m0 is the weight of the brake pad before the friction tests, mi is the
weight of the brake pad after the friction tests, ρ is the density of the
friction material, and E is the braking energy.
The amount of wear was determined by weighting the brake pad
before and after tests using a precision electronic balance with an accu-
racy of 0.2 g. The morphologies of the worn surface and subsurface were
observed using scanning electron microscopy (SEM) equipped with en-
ergy dispersion spectrum (EDS) (Quanta FEG 250, USA).
3. Results and discussion
3.1. Microstructure and mechanical properties
The matrix of Cu-MMC was formed by sintered copper with the
addition of iron, and solid lubricants (graphite, MoS2) and abrasive
particles (ZrO2, FeCr) were introduced to improve the wear-resistance
and stabilize the coefficient of friction around a given value. Character-
istic microstructure parallel to the friction surface of Cu-MMC, as seen in
a light microscope, is given in Fig. 5(a). The constituents dispersed ho-
mogeneously in the light orange copper matrix. Due to the particle
strengthening contribution and small solid solubility of iron in copper,
the presence of gray-white iron in the matrix made the composite
stronger [21]. The metallic constituents employed in Cu-MMC provided
strength, high temperature stability, oxidation resistance, and high
thermal conductivity [22]. The dark coarse ZrO2 particles and white fine
FeCr particles with high hardness, excellent thermal and chemical sta-
bilities were used to provide the desired level of coefficient of friction,
and protect against the negative effects of lubricants [23,24].
Graphite and MoS2 applied in Cu-MMC imparted the smoother and
vibration less engagement during braking process, and also provided the
anti-seizure characteristics [25]. In our previous work [26], it was found
that solid-solid phase reactions occurred between MoS2 and other in-
gredients during the sintering process. Some new phases acting as
Fig. 6. Cross sectional microstructure of Cu-MMC.
588
Tribology International 119 (2018) 585–592
lubricants were responsible for the desired effect of friction force stabi-
lization. However other new compounds acting as abrasives and playing
a function of dispersion strengthening, could enhance the friction resis-
tance and improve the wear-resistance to some extent. The grey graphite
in the matrix was exploited to form a self-sustaining lubricant layer on
the friction surface, and increase the thermal conductivity of the com-
posite. Graphite in the flake form can improve the lubrication properties,
while graphite in the granular form is able to dissipate the heat generated
during braking more effectively [27]. So the isotropic synthetic graphite
in the coarse granular form with the addition of the natural graphite flake
was added to Cu-MMC for their synergistic effects on properties. During
the compression process, the anisotropy natural graphite flake generated
displacement and rotation simultaneously [21], hence the grey strip in
Fig. 5(b) is the graphite flake flattened perpendicular to the pressing
direction, namely parallel to the friction surface as shown in Fig. 5(a).
The cross sectional view of Cu-MMC shown in Fig. 6 presents three
distinct layers: (i) the friction material, (ii) an intermediate transition
layer with a thickness of approximately 500 μm, and (iii) the steel
backing plate. AISI 1020 steel was chosen as the material for the steel
carrier of the friction material. The purpose of the steel backing plate was
to protect Cu-MMC from sudden catastrophic failures caused by cyclic
dynamic loading during braking. In addition, it also enhanced the
dimensional stability of the brake pad [28]. Since materials towards
either side of the interface were entirely different, debonding occurred
frequently during actual applications. In order to overcome the poor
joint, the bond between the steel backing plate and the friction material
was achieved by the intermediate transition layer. This intermediate
layer, comprising copper, iron, stannum and nickel, avoided the direct
contacts of the nonmetallic constituents (e.g., graphite) in the friction
material with the steel backing plate, leading to an adequate joint be-
tween the friction material and the backing plate.
The mechanical properties of Cu-MMC are listed in Table 3. The
density of Cu-MMC, containing high volume fraction of particulates in
lower densities, was slightly higher than half that of copper (8.96 g/cm3)
[29]. As specified in TJ/CL 307–2014, the composites are noted within a
surface hardness range of 10–30 HBW. Graphite is a well-known soft
material, and hence Cu-MMC comprising high volume content of
graphite exhibited lower surface hardness value compared with copper
[29]. Nevertheless, this negative effect on the hardness and strength
could be partly compensated by the strengthening contributions of the
Table 3
Mechanical properties of Cu-MMC.
Properties Technical Values
requirements
Density (g/cm3) / 4.8
Surface hardness (HBW 10/250/30) 10–30 15
Shear strength (MPa) 6 16
Bonding strength between the friction material and its 7 37
backing plate (MPa)
Y. Xiao et al. Tribology International 119 (2018) 585–592

powder metallurgy processed Cu-MMC was desirable to eliminate the

common problems (e.g., spallation and detachment) of brake pads
in service.

3.2. Dynamometer tests

Fig. 7 shows the relation between the average coefficient of friction

and the initial stop-braking speed for Cu-MMC tested under dry and wet
conditions. As specified in TJ/CL 307–2014, the tolerance range for the
average coefficient of friction of brake pads for CRH380 high-speed
trains under dry conditions is illustrated in Fig. 7. It is evident that the
average coefficients of friction of Cu-MMC under dry conditions were in
the scope of the frictional requirements. All other things being equal, the
friction coefficient was slightly sensitive to the clamping force and
decreased with the latter. This sliding movement occurred in very small
areas at the asperities over time. An increase in the clamping force
accelerated the rupture of the asperities and the deformation of the
matrix, and then enhanced the compaction of wear debris, eventually
caused the increase in the actual contact area and the decrease of the
Fig. 7. Results of the average coefficient of friction of Cu-MMC. friction coefficient [26]. The friction coefficient was not greatly influ-
enced by the initial stop-braking speed below 220 km/h, whereas it
Table 4 progressively decreased with increasing the initial stop-braking speed
Tribological performances of Cu-MMC tested at a clamping force of 24/31.5 kN and several
over 250 km/h. As the brake speed increased, the brake energy was
initial stop-braking speeds under emergency stop-braking conditions.
sufficient to make the matrix softer and bring about a smooth friction
Initial stop- Braking Brake Average Maximum layer (the accumulation of the oxides and wear debris), resulting in the
braking speeds distance time coefficient of temperature of the
decrease of the friction coefficient.
(km/h) (m) (s) friction disc surface (
C)
Though the permissible average coefficient of friction under wet
300 3464 79.5 0.378 472
conditions is not described in TJ/CL 307–2014, it is desirable that the
320 4127 88.2 0.364 505
350 5475 105.6 0.339 543 friction coefficient of brake pads shall be independent of the weather
380 6946 123.1 0.321 566 effects (e.g., rain) in applications. When the water spray was introduced
to the sliding interface, water not only removed most of the heat that
generated during the tests, but also might develop the water film and
above-mentioned constituents. As shown in Table 3, the shear strength of prevent solid-solid contact, resulting in low friction coefficient (as shown
Cu-MMC (namely the friction material) was more than double the pre- in Fig. 7). Compared with those obtained under dry conditions, the
scriptive value illustrated in TJ/CL 307–2014 (the minimum acceptable average coefficients of friction of Cu-MMC did not vary by more than
shear strength of 6 MPa). The steel backing plate was coated with a ±15% with the highest braking pressure (31.5 kN) and in the presence of
bonding aid (copper plating, 10–20 μm thickness). The alloying elements wetness, and they could immediately return to those values under dry
(stannum and nickel) diffused from the transition layer towards the conditions once this disruptive influence has ceased. Furthermore, with
copper plating layer and the friction material during pressure sintering, increasing the initial stop-baking speed or the clamping force, it was
leading to improved joint strength [30]. The bonding strength between increasingly difficult to build up a thick water film to support the
the friction material and its backing plate was more than five times that clamping force and prevent solid-solid contact, so a significant increase
specified in TJ/CL 307–2014 (the minimum acceptable bonding strength was observed in the friction coefficient. Still, Cu-MMC brake pad under
of 7 MPa). With remarkably improved mechanical properties, this novel wet conditions could meet the friction requirements specified in UIC

Fig. 8. Changes on the instantaneous coefficients of friction of Cu-MMC tested at a clamping force of 24/31.5 kN and several initial stop-braking speeds under emergency stop-
braking conditions.

589
Y. Xiao et al. Tribology International 119 (2018) 585–592

Fig. 9. Variation of the instantaneous coefficients of friction with time during continuous brakings at 20 and 30 kW for 20 min.

CODE 541-3 [20]. (namely downhill braking). During the continuous braking test per-
This study concerns a series of emergency stop-braking applications formed at 20 kW for 20 min (the total amount of braking energy was
of Cu-MMC brake pad tested within the initial stop-baking speed range of 23.14 MJ), the friction coefficient increased and tended to stabilize
300–380 km/h under dry conditions. Correspondingly, the amount of gradually along with the brake application time. In the long downhill
energy dissipated by the brake pads and disc was 19.79–31.75 MJ, and braking process executed at 30 kW for 20 min (the total amount of
the braking power exceeding 250 kW was generated during each stop- braking energy was 34.69 MJ), Cu-MMC brake pad exhibited the steady-
braking. The tribological performances of Cu-MMC in emergency situa- state friction coefficient at the beginning of braking, and then the friction
tions are listed in Table 4. As mentioned above, the average coefficient of coefficient decreased slowly with the increase of braking time. These
friction declined step by step with the increase in the initial stop-baking phenomena might be relevant to the thermal equilibrium between the
speed from 300 to 380 km/h. The brake pads converted primarily the frictional heat generation (accumulation) and dissipation (namely high
kinetic energy into thermal energy by friction, so a huge amount of temperature might cause the thermal fade). With the static coefficient of
frictional heat was generated at the disc/pads interface, resulting in the friction equal to about 0.400, Cu-MMC brake pad could provide a parking
increase in temperature of the disc surface with increasing the initial brake force to prevent movement of a stopped or parked CRH380 train.
stop-baking speed. But the highest temperature of the brake disc surface Fig. 10 shows the maximum temperatures on the disc surface under
was only 566
C in the full-scale dynamometer tests. As set by China various conditions. As indicated above, most of the braking energy was
Railway Corp., brake pads are obliged to stop a CRH380 train running at dissipated as heat through the brake disc and pads to the environment.
300, 350, and 380 km/h at maximum acceptable braking distances of All other things being equal, with increasing the initial stop-braking
3800, 6500, and 8500 m, respectively. Table 4 indicates that Cu-MMC speed, the braking energy gradually increased, leading to the increase
brake pad can meet the braking distance requirements. According to in the temperature of the disc surface. Meanwhile, the braking time
the specification of TJ/CL 307–2014, the permissible wear of the com- decreased with the increase of the clamping force, resulting in the in-
posite pads for the brake applications is 0.35 cm3/MJ. After a sequence of crease in the braking power and the surface temperature. Under wet
repeated stop braking, the wear rate of Cu-MMC brake pad was only conditions, water can absorb and take away part of the frictional heat
0.20 cm3/MJ. during stop-braking, so the surface temperatures were less than those
Fig. 8 shows the evolutions of the instantaneous coefficient of friction
of Cu-MMC brake pad tested within the initial stop-baking speed range of
300–380 km/h under dry conditions. The upper and lower curves of
instantaneous coefficient of friction in Fig. 8 indicate the scope of the
technical requirement according to the specification of TJ/CL 307–2014.
On these curves, stop braking started from the right and speed decreased
down to zero at the left. Thus, curves had to be followed backwards
during a braking cycle. It is clear that the brake curves were relatively
smooth and in the scope of the technical requirements. And another
noteworthy aspect of the results is that, the instantaneous coefficients of
friction experienced slight fluctuations at the beginning of stop-barking,
which became progressively smaller. These phenomena might be caused
by changes in the contact pressure and dynamics as a result of thermo-
elastic friction instabilities [7,31].
Brake pads are used in high-speed trains to enable deceleration,
control acceleration (downhill) or to keep them standing when parked.
When verifying the braking performances of brake pads, one should
consider not only the service braking and emergency braking operations
(both called stop braking), but also the downhill braking and parking
braking operations. Fig. 9 presents the instantaneous coefficients of Fig. 10. Results of the maximum temperature on the disc surface under various
friction of Cu-MMC as functions of time under dry continuous braking environments.

590
Y. Xiao et al. Tribology International 119 (2018) 585–592

Fig. 11. SEM observations of (a) the worn surface and (b, c) the subsurface regions.

under dry conditions. Meanwhile, due to the weak mechanical bond between the subsurface
layer and the friction material, as well as graphite with low shear strength
along the layer planes, under the combined effects of the normal pres-
3.3. Wear mechanism sure, tangential stress and frictional heat, the microcracks originated
from the subsurface region and graphite particles during braking process.
After braking operations, the representative morphology of Cu-MMC Under these conditions, these microcracks propagated from the subsur-
is shown in Fig. 11(a). It is can be seen that the relatively smooth worn face region and graphite to the worn surface. So one can see pits after the
surface interspersed with some spalling pits. There is no evidence of subsurface layer spalling off. This process is a manifestation of delami-
ploughed furrows on the worn surface. Generally, the presence of pits nation wear.
means that microcrack may originate and propagate at a certain depth
beneath the contact surface. The subsurface layer perpendicular to the
worn surface can indicate changes in the near-surface materials [32]. As
the friction process proceeded, the subsurface microcracks were gener-
ated as seen in Fig. 11(b), and then the microcracks propagated to the
worn surface where pits are found, as shown in Fig. 11(c).
EDS analysis of the entire region in Fig. 11(a) was performed to identify
the nature of the worn surface. Fig. 12 shows the primary peak of the O
element as proof of the formation of oxide. A large of quantity of frictional
heat was generated during braking. Frictional heat could lead to suffi-
ciently high surface temperature that caused oxidation loss. The Zr and Cr
peaks revealed that the hard particles (ZrO2 and FeCr) penetrating in the
friction layer could plow the mating surface and enhance the friction co-
efficient, so an iron peak with relatively high intensity was found in the
spectra. Simultaneously, the hard particles, incorporating in the soft and
ductile metal matrix, could relieve the extent of deformation in the sub-
surface region and improve the wear resistance. The C peak indicated that
the worn surface was covered by a certain amount of graphite.
The accumulation and compaction of the oxides as well as absorption
layer and wear debris (e.g., graphite) give rise to generation of the sub-
surface layer. This layer was contributed to prevention of the direct
contact of the sliding surfaces, reduced ploughing and eliminated adhe-
sive wear, resulting in low aggressivity against the brake disc. Fig. 12. EDS analysis of the worn surface.

591
Y. Xiao et al.
4. Conclusions
In the present study, the microstructure, mechanical properties and
tribological behaviors of a powder metallurgy processed Cu-MMC were
evaluated. The main conclusions resulting from the above study are
as follows:
The metal matrix, lubricant components and friction components
were adequately combined in view of their respective advantages, and an
intermediate transition layer and the steel backing plate coated with a
bonding aid were employed in Cu-MMC, resulting in a significant
improvement in the mechanical properties. The shear strength of the
friction material was 16 MPa, and the bonding strength between the
friction material and its backing plate was 37 MPa, so Cu-MMC was
desirable to eliminate common spallation and detachment problems
associated with brake pads in service.
Cu-MMC exhibited excellent tribological properties during stop
braking, downhill braking, and parking braking. The average and
instantaneous coefficients of friction were in the scope of the frictional
requirements. The static coefficient of friction was about 0.400. The
temperature of the disc surface increased with the increase of the initial
stop-braking speed and the clamping force. But the highest temperature
of the brake disc surface was only 566
C. The delamination and oxida-
tion were considered as the dominant wear mechanism of Cu-MMC, but
the wear rate was only 0.20 cm3/MJ.
According to the technical requirements specified in TJ/CL 307–2014
and UIC CODE 541-3, Cu-MMC is a promising candidate material for
brake pads in high-speed trains.
Acknowledgments
This work was supported by the National Natural Science Foundation
of China (No. 51175516, No. 51475476); and the Fundamental Research
Funds for the Central Universities of Central South University (No.
2014 zzts023).
References
[1] Raturi V, Verma A. Analyzing competition between high speed rail and bus mode
using market entry game analysis. Transp Res Procedia 2017;25:2373–84.
[2] Fu XW, Zhang AM, Lei Z. Will China's airline industry survive the entry of high-
speed rail? Res Transp Econ 2012;35:13–25.
[3] Utsunomiya M, Hodota K. Financial lessons from Asian experience in constructing
and operating high speed train networks. Transportation 2011;38:753–64.
[4] Deng CY, Yin J, Zhang HB, Xiong X, Wang P, Sun M. The tribological properties of
Cf/Cu/C composites under applied electric current. Tribol Int 2017;116:84–94.
[5] Zhou L, Shen ZY. Progress in high-speed train technology around the world. J Mod
Transp 2011;19:1–6.
[6] Desplanques Y, Roussette O, Degallaix G, Copin R, Berthier Y. Analysis of
tribological behaviour of pad-disc contact in railway braking: Part 1. Laboratory test
development, compromises between actual and simulated tribological triplets.
Wear 2007;262:582–91.
[7] Kasem H, Brunel JF, Defr
enoy P, Siroux M, Desmet B. Thermal levels and subsurface
damage induced by the occurrence of hot spots during high-energy braking. Wear
2011;270:355–64.
592
Tribology International 119 (2018) 585–592
[8] Gyimah GK, Guo ZN, Huang P, Chen D. Application of powder metallurgy methods
for production of a novel Cu-based composite frictional train brake material. In:
Dobrzanski LA, editor. Powder metallurgy - fundamentals and case studies. Rijeka:
InTech; 2017. p. 349–83.
[9] Goo BC, Ryu SS, Kim SW. Development and characterization of sintered friction
materials for brake pads. In: 2013 International conference on quality, reliability,
risk, maintenance, and safety engineering, Emeishan, China; July, 2013. p. 803–7.
[10] Lyu YZ, Bergseth E, Tu MH, Olofsson U. Effect of humidity on the tribological
behaviour and airborne particle emissions of railway brake block materials. Tribol
Int 2018;118:360–7.
[11] Cai P, Wang YM, Wang TM, Wang QH. Effect of resins on thermal, mechanical and
tribological properties of friction materials. Tribol Int 2015;87:1–10.
[12] Filip P. Friction brakes for automotive and aircraft. In: Wang QJ, Chung YW,
editors. Encyclopedia of tribology. New York: Springer US; 2013. p. 1296–304.
[13] Zhang XJ, Dong PY, Chen Y, Yang WC, Zhan YZ, Wu KF, et al. Fabrication and
tribological properties of copper matrix composite with short carbon fiber/reduced
graphene oxide filler. Tribol Int 2016;103:406–11.
[14] Gyimah GK, Chen D, Huang P. Dry sliding wear studies of Cu-based powder
metallurgy brake materials. In: Proceedings of the ASME 2012 international
mechanical engineering congress & exposition, houston, USA; November, 2012.
p. 1–8.
[15] Su LL, Gao F, Han XM, Chen JG. Effect of copper powder third body on tribological
property of copper-based friction materials. Tribol Int 2015;90:420–5.
[16] Chung SL, Hong US, Jang H. The brake performance of sintered friction materials
developed for high speed trains. J KSTLE 2007;23:266–71.
[17] Kim MS. Development of the braking performance evaluation technology for high-
speed brake dynamometer. Int J Syst Appl Eng Dev 2012;6:122–9.
[18] Ru HQ, Li JY, Wang W. 3D network SiC-metals composites for heavy duty brake
applications. In: Cuppoletti J, editor. Metal, ceramic and polymeric composites for
various uses. Rijeka: InTech; 2011. p. 461–74.
[19] TJ/CL 307–2014. Brakes – disc brakes and their application – provisional
conditions for the approval of brake pads intended for high-speed EMUs. issued by
China Railway Corp.. May, 2014 (in Chinese)
[20] UIC CODE 541–543. Brakes – disc brakes and their application – general conditions
for the approval of brake pads. seventh ed. October, 2010.
[21] Xiong X, Chen J, Yao PP, Li SP, Huang BY. Friction and wear behaviors and
mechanisms of Fe and SiO2 in Cu-based P/M friction materials. Wear 2007;262:
1182–6.
[22] Asif M, Chandra K, Misra PS. Characterization of iron based hot powder brake pads
for heavy duty applications. IJMME 2013;8:94–104.
[23] Akinci A, Sen S, Sen U. Friction and wear behavior of zirconium oxide reinforced
PMMA composites. Compos Part B 2014;56:42–7.
[24] Yılmaz SO, Aksoy M, Ozel C, Pıhtılı H, Gür M. Effect of FeCr intermetallic on wear
resistance of Fe-based composites. In: Pihtili H, editor. Tribology in engineering.
Rijeka: InTech; 2013. p. 3–14.
€
[25] Osterle W, Dmitriev AI. The role of solid lubricants for brake friction materials.
Lubricants 2016;4:1–22.
[26] Yao PP, Xiao YL, Deng JW. Study on space copper-based powder metallurgy friction
material and its tribological properties. Adv Mater Res 2011;284–286:479–87.
[27] Gujrathi TV, Damale AV. A review on friction materials of automobile disc brake
pad. ARDIJEET 2015;3:1–4.
[28] Prabhu TR, Varma VK, Vedantam S. Effect of reinforcement type, size, and volume
fraction on the tribological behavior of Fe matrix composites at high sliding speed
conditions. Wear 2014;309:247–55.
[29] Hammond CR. The elements, in handbook of chemistry and physics. Boca Raton:
CRC Press; 2004.
[30] Singaravelu DL, Chandra K, Misra PS. New processing technology for manufacture
of heavy duty metalloceramic brake pads. Powder Metall 2011;54:118–26.
[31] Kasem H, Defr
enoy P, Desplanques Y. Relationships between surface thermal
gradients and disc distortion during stop-braking with high energy dissipation.
Tribol Lett 2012;48:169–81.
[32] Greiner C, Liu ZL, Strassberger L, Gumbsch P. Sequence of stages in the
microstructure evolution in copper under mild reciprocating tribological loading.
ACS Appl Mater Interfaces 2016;8:15809–19.