Engineering Failure Analysis 118 (2020) 104848

Contents lists available at ScienceDirect

Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/engfailanal

Analysis of failure mechanism of W1-type fastening clip in high

T
speed railway and structure study of damping composite
⁎
Xiaogang Gao , Anbin Wang, Lang Liu, Yu He, Longhua Ju
School of Urban Rail Transportation, Shanghai University of Engineering Science, Shanghai 201620, PR China

A R T IC LE I N F O ABS TRA CT

Keywords: In this paper, the fastener clip in high-speed railway with the speed of 300–350 km/h is taken as
Rail clip failure the research object, it studied the relationship among parameters such as high-speed rail wear,
Wheel and rail wear dynamic frequency response of clip and dynamic stress of clip, revealed the dynamic failure
High-frequency excitation mechanism of clip damage in high-speed rail caused by high-frequency excitation (580–680 Hz).
Resonance
In order to improve the service life of existing W1-type fastener in high-speed rail on the basis
Clip damping optimization
of guaranteeing design requirements and installation interchangeability of clips, taking “keeping
away from excitation frequency band and reducing vibration amplitude” as the research objec-
tive, the clip geometric parameter optimization based on the correlation theory and the com-
posite damping structure design of clip based on the modal damping theory are mainly promoted.
Under the conditions of installation and use of rail clips, high damping clips with different
combination of inner and outer diameters are obtained and strength simulation analysis,
damping clip samples and dynamic frequency response test of optimized damping clips are
carried out. The results show that, (1) with the increase of the equal ratio of outer diameter and
damping diameter of high damping clip, the attentive mode frequency of clip increases up to
18.5%;(2) As the ratio of outer diameter and damping diameter of high damping clip increases,
the response amplitude of high damping clip significantly decreases, up to 31.6%, (3) As the ratio
of outer diameter and damping diameter of high damping clip increases, the overall mass of
fastener clip also decreases, up to 9.7%.To a certain extent, the design structure of high damping
composite clip is not only far away from the excitation frequency of wheel and rail, but also
improves the dynamic fatigue life of clip, saves the cost of fastener system and has more ap-
plicability for high-speed railway lines.

1. Introduction

With the continuous development of high-speed railway, rail corrugation and wheel polygon have gradually become an indis-
pensable problem in wheel-rail wear of high-speed railway. Wheel- rail contact wear in high-speed railways will cause additional
high-frequency excitation between wheel and rail, which will then cause severe impact on the track structure. It will not only affect
the stability and safety of vehicle movement, but also damage and destroy the components of acceleration track such as fastener clips.
The ω-type fastener clip in high speed railway is installed on the baseplate and insulation pad block under the action of the bolt
pressure, which can fix the rail, maintain the gauge. During the operation of the train, the response state of the fastener system itself
will change greatly under the influence of external conditions such as running speed, train axle load and wheel-rail wear. For

⁎
Corresponding author.
E-mail address: [email protected] (X. Gao).

https://doi.org/10.1016/j.engfailanal.2020.104848
Received 28 June 2020; Received in revised form 31 July 2020; Accepted 10 August 2020
Available online 13 August 2020
1350-6307/ © 2020 Elsevier Ltd. All rights reserved.
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

example, aging of insulating pad, stiffness change of elastic pad plate, plastic deformation of clip, etc., it will further change stress
distribution of clip of fastener and increase vibration level of clip, which will result in damage effect of elastic fastener clip.
Continuous damage to the fastener clip will destroy the continuity of the track structure support and change the track geometry. In
turn, it will lead to increased wheel-rail interaction and track dynamic response in the fastener failure area, thus accelerating the
destruction of track components. Therefore, how to change the wheel-rail matching performance and the frequency response
characteristics of the rail system at the fixed speed of the train to avoid the intense interaction between the wheel and rail is an
important issue to be solved in the process of high-speed railway line design and operation and maintenance.
In recent years, with the improvement of fastener processing technology and maintenance technology in high-speed railway, as
well as the deepening of the research on wheel polygon wear and rail corrugation, a lot of research achievements have been made in
the aspects of dynamics analysis of railway fasteners, characteristics analysis of clip fatigue and material processing technology. In
terms of track dynamics analysis, Ling et al. [1] believes that the fatigue failure of the clip is caused not only by the static preload of
the clip, but also by the cyclic dynamic excitation load of the wheel under the wheel-rail coupling. Under normal operation of the
train, the rail clip will be subjected to abnormal loads such as welded parts of the rail, rail wear, wheel polygon, etc., which are 60%
of the wheel-rail loads. Under the action of clamp force, cyclic wheel-rail force and excitation load of abnormal wheel-rail, the clip toe
displacement will be greatly increased and the fatigue failure of clip will be accelerated. In this paper, to some extent, the phe-
nomenon of wheel-rail wear increases the acting force between the wheel and rail, and increases the additional dynamic acting force
on the rail clip under operating conditions, which accelerates the dynamic fatigue of clip. Hasap et al. [2] performed a fatigue
experiment and finite element analysis of the failure mechanism of an e-clip. They reported that under normal wheel load, e-clips
with high, normal and low toe loads were run-out at 5 × 106 loading cycles. While under the contribution of impact on wheel load,
the fatigue life of e-clips was reduced to 5468 and 16,839 loading cycles. In addition, the fatigue crack was found to nucleate at the
location of highest stress. Hasap has experimentally obtained that the addition of additional impact load on the rail clip has a
significant impact on the life of the clip, which is an extension of the research results in literature [1]. In addition to consider the
speed, axle load and capacity, based on the state of abnormal vibration of the vehicle and track irregularity and the wear between the
wheel/rail article of the additional stress is also a important focus.D.J. Thompson et al.[3] measured the vertical and lateral dynamic
stiffness of rail fastener systems under preload in the range 100–1000 Hz and clarified the influence of vibration at higher frequencies
on the stiffness of fastener systems. Thompson first proposed the stiffness variation of rail fastener system under different frequencies
under preload, that is, the frequency variation characteristics of fastener under preload. With the increase of wheel-rail excitation
frequency, the stiffness of the fastener will increase slowly, which corrects our understanding of fastener frequency-variable stiffness,
and will have a certain influence on the damage of the rail clip. Xiao et al. [4] systematically investigated fractures of a fastening with
the aim of revealing the fracture mechanism and proposing effective remedy. Results show that the stress concentration in fastening
clips caused by the unreasonable installation in the term of excessive inserting depths and the vibration fastener resonance resulting
from the rail corrugation as well as its second-order natural frequency led to the fracture. Xiao also revealed for the first time that the
failure of e-type rail clip is caused by resonance caused by rail corrugation excitation. Although e-type rail clip is used in urban rail
transit with low speed and low wheel wear, the wheel-rail wear excitation has a certain influence on the injury of e-type clip, which
inspires the research on the significance of the damage of high-speed rail fastener. Zhu [5] found that the vibration response of rail
clips on tracks with rail corrugation was about 10 times larger than that of rail clips on tracks without rail corrugation, resulting in
the acceleration of clip fatigue fracture. Xiao [6] considers that periodic wear of corrugated rails and periodic wear of polygons of
wheels of units will occur in the long-term operation of high-speed railways and the dynamic action between wheels and rails will
increase. It reveals the important mechanism that the high-frequency excitation of wheel and rail causes resonance of clip when it
approaches the natural frequency of fastening clip, thus causing the damage of clip. On the basis of literature [4], this paper puts
forward the mechanism of resonance fracture caused by high-frequency excitation of wheel and rail of high-speed railway and the
coincidence of modal frequency of rail clip. Gao [7] took SKL15-type fastener clip commonly used in high-speed railway as research
object, it is revealed that the natural frequency range of SKL15-type clip in high-speed rail is close to HF excitation frequency of wheel
rail under installation, which results in resonance and damage of fastener clip. In order to verify the mechanism of damage caused by
high-frequency excitation of clip, a high-frequency fatigue testing machine for clip has been independently designed and developed.
During high-frequency excitation test of ω clip, the loading excitation frequency is 590 Hz, the loading amplitude is 0.04 mm, and the
clip break and fail after 2.05 h of operation. This paper is a further extension of the research results in literature [6], as SKL15 clip
high-frequency excitation resonance fracture for high-speed railway has evolved through the development of test equipment. Yu et al.
[8] calculated and analyzed the effect of fatigue life under horizontal force on the performance of X2-type clip by considering the
contact between clip and other parts of fastener system. In order to study the fracture causes of clip components in fastener system of
ballastless track of high-speed railway, Xiang et al. [9] took WJ-7 fastener as the research object to analyze the mechanical char-
acteristics of fastener clip under three conditions: fastener installation, wheel polygon wear and curve line type, and it is believed that
the fastener will be in place when the clip preload reaches 24 kN; An increase in the wear order of the wheel polygon will increase the
stress value of the fastener clip. The stress increment of 3rd order wear time is 36 MPa, which is 5 times higher than that of non-wear.
When the radius of line curve is 4000 m, the life of clip is 20,000 times, which is more than 98% less than that of 8000 m. Mo-
hammadzadeh [10] proposed a reliability analysis method for fatigue failure of clip under operating conditions, which is obtaining
the displacement–time domain diagram of clip by finite element analysis. The clip crack zone is numerically simulated by rain flow
counting method and Palmgren-Miner linear failure rule. However, fatigue life during experimentation and use is not validated by
comparing their results. In terms of clip material, Guo et al. [11] has carried out macro-micro observation, metallographic structure
examination, hardness examination and material composition examination on the fracture surface of 60Si2MnA material, and
analyzed the cause of clip fatigue, it is found that the network ferrite formed by decarburization of the surface of the clip becomes the

2
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

source of fatigue crack. Xu et al. [12] took 60Si2Mn as the research object, and the surface of the clip is galvanized and the fracture
reason of the clip is analyzed by means of macro-examination, chemical composition analysis, hardness test and micro-examination.
The results show that the clip is fatigue fracture, the fatigue source is located at the junction of the surface of the clip and the zinc
coating. The fracture location is basically in the area where the clip is subjected to the greatest fatigue alternating load. During the
fatigue test, the clip is subjected to combined bending and torsion stresses, and fatigue cracks are initiated and propagated at the
junction between the zinc-coated layer and the surface of the clip, resulting in fatigue fracture. Wang et al. [13] analyzed the cause of
failure from the point of view of crack and fracture morphology of SKL15 clip used for high-speed railway, it is believed that there
was a decarburization layer on the inner surface of clip followed by small rounded corners, and there were tension ditches and small
notches. Under the alternating stress, stress concentration occurred near the clip, which then initiated cracks and failed.
The above literature studies respectively from the dynamics of wheel rail, wear excitation of wheel rail, clip fatigue in high-speed
railway and materials, etc. to explain the dynamic effect of wheel rail on the performance of fastener clip, and the relevant con-
clusions have certain reference value, However, there is no relevant report on how to prevent the damage of fastener clips and
improve service life of high-speed railway. This paper takes clip of domestic high-speed rail operation line as research object, analyses
and summarizes the wheel-rail wear excitation and dynamic stress by the experiment and simulation, and reveals the fatigue failure
mechanism of clip caused by high-frequency action of wheel-rail and dynamic high stress under the installation condition of fastener
clip. In order to improve the service life of existing W1-type clip in high speed railway, under the premise of guaranteeing design
requirements and installation interchangeability of fastener clip in high-speed railway, the research objective of the clip is that
“keeping away from the excitation frequency band and reducing the amplitude of vibration”. Based on structural modal theory and
damping vibration reduction technique, the structural parameters of clip in high-speed rail and high damping composite materials are
designed and analyzed. Under the conditions of installation and use of clip, high damping clip structures with different inner and
outer diameters are obtained. Combined the fatigue fracture accumulation theory, the strength and fatigue simulation, dynamic
frequency response test of the optimized damped clip are carried out, the results show that the design structure of high damping
composite clip is not only far away from the excitation frequency of wheel rail, but also improves the dynamic fatigue life of clip,
saves the cost of fastener system and has more applicability for high-speed railway lines.

2. Failure analysis of clip in high-speed railway

When the high-speed train runs on the track, the dynamic force exerted by the wheels on the rail head causes deformation and
vibration of the rail, which in turn causes vibration of fastener system and other components. With the difference of axle load and
design speed of high-speed rail trains, the magnitude of impact force applied by wheels on rails usually varies from 50 kN to 300 kN
[6].In order to ensure the safe operation of high-speed train, the design of railway fastener system requires that clip have proper
clamping force and strength, and at the same time, the dynamic displacement of rails can be guaranteed within the design range
under the impact force of wheels. The material of fastener clip in high-speed rail is high-strength clip steel basically. It is hot-
processed into the most common ω-type clip. This clip is widely used due to its high strength, adjustable buckle pressure and
convenient installation. However, fatigue fracture of ω-type clip in railway occurs frequently during its use.
In this part, the structure of clip of commonly used fasteners for High-speed railway in China is firstly described, and the failure
causes of clip are investigated by microscopic analysis of clip materials and fracture morphology. At the same time, the manuscript
analyzes whether the strength of the clip satisfies the material design characteristics from the two aspects of static preload stress and
simulated abnormal abrasion excitation dynamic stress through the experimental.

2.1. Damage type

The research object of this paper is a type W1 fastener clip in high-speed rail with a diameter of 14 mm, which is characterized by
two relatively symmetrical arched arms. This clip is characterized by two relatively symmetrical arched arms with one end of the arch
being a toe which is buckled to the surface of the rail foot during installation. The other end is connected with a semi-circular clip
which is connected to the central half-circle ring by a straight line segment. In practical application, the fastener system consisting of

Fig. 1. Current fastener system in high-speed rail.

3
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Fig. 2. Failure mode of clip.

clips and other parts is bolted to the sleeper or ballast bed as shown in Fig. 1. Fig. 2 shows a fractured clip filmed on site. One of the
two elastic arches is completely disconnected at the clip heel, resulting in the failure of the clip and loss of fastener pressure. The
consequence of pressure loss on the fastener clip reduces the stability of the track and increases the possibility of the train derailing.
As can be seen from Fig. 3, the fracture occurs at the support point of the clip heal. The expression form of efficiency is the initial
surface yield, and there are basically three zones in the fracture failure of clip: initiation crack stage, fatigue crack growth zone and
the final instantaneous fracture zone.
By checking the material at the fracture area of the clip, Fig. 3(a)–(c), hardness 430–440HV (index requirement 400–460 HV),
decarburization layer 62.58 μm (index requirement < 200um), metallographic examination shows that the amount of ferrite is large,
but the material of clip requires uniform tempered yield and tempered sorbite, while a small amount of ferrite and a small amount of
bainite are allowed. In general, the performance requirements of the clip are basically met by material analysis.

2.2. Dynamic fatigue stress of fastener clip

Failure of fastener clips in railway is caused by many factors. In addition to vibration excitation caused by dynamic frequency
response, fatigue strength of fastener spring itself is also an important consideration in design, operation and maintenance of high-
speed rail fasteners.
The clip is subject to tension and compression, bending, torsion and shear composite effect in the working status. Any point on the
clip is in complex stress state, as shown in Fig. 4 and Eq. (1), σx is the normal stress led by axial tension and compression, τxy 、τzx is the
shear stress led by bending and torsion, respectively. The material used in high-speed rail is spring steel, which belongs to the plastic
material. The strength theory commonly used to judge the yield of plastic materials mainly includes the maximum shear stress theory.
The maximum shear stress theory holds that the material is expected to yield when the maximum shear stress under three stress states
equals or exceeds the maximum shear stress under axial tension (compression) yield of the material.
The equivalent stress of the clip calculated according to the third strength check theory is:

σx 3 = σx2 + 4(τxy
2 2
+ τzx ) (1)

σ = εE (2)

In order to study the cause of fastener clip from the strength point of view, an acceleration sensor is arranged at the support failure
position of clip heel with strain gauge and clip arch as shown in Fig. 4 and Eq. (2). The pre-loading bucking pressure process was
simulated during the tightening process of the fastener clip, and the static stress under the pre-loading of the clip was obtained.
Meanwhile, the dynamic strain of the high-speed rail clip was tested by the modal frequency (loading frequency 685 Hz) measured by

a) b) c)

Failure port of clip Material decarburization layer Fracture metallographic decarburization layer
Fig. 3. Failure form of clip and material inspection.

4
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

z x

x
y
y
Fig. 4. Clip stress state at any point and testing apparatus.

the acceleration sensor. The stress level of the clip can be obtained by using the strength checking theory.
Through the above experimental study, as shown in Fig. 5, it is obtained that the static strain of the clip is 6098με and the
corresponding stress is 1250 MPa during the application of clamping force of fastener clips; Under the dynamic excitation of assembly
frequency of the fastener clips , the dynamic strain of spring strip is 630με , and the corresponding dynamic stress is 136 MPa, as
shown in Fig. 6;The stress of fastener clip under static installation is 1250 MPa, which is close to the yield strength of spring steel
material. At the same time, under the dynamic excitation of wheel-rail, clip bears the dynamic alternating stress under static in-
stallation stress, and its amplitude is much higher than that of spring steel material. Micro-yield and cracking occur on the surface of
the local area of the clip heel where the clip stress is very concentrated.

3. Failure mechanism analysis of clip in high-speed railway

According to the failure form and static strength analysis results of rail clip in Section 2, in order to further analyze the failure
mechanism of clip, first of all, dynamic vibration parameters of clip during train operation are collected on site in this section, and the
concerned modal parameters and vibration amplitude of clip at different speeds are analyzed. Meanwhile, the vibration amplitude of
the clip is measured on line with or without wheel-rail wear. Finally, based on the theory of modal resonance, the failure mechanism
of the rail clip is revealed by combining the “butterfly effect” modal parameters of the clip and the characteristics of wheel-rail wear
in the high speed railway.

3.1. Modal parameters of fastener clip

When a high-speed train runs on ballastless track, the wheels exert impact excitation on the tread of the rail head, which results in
deformation and vibration of the rail and transmits it to the rail fastener system and other components of rail system. The purpose of
the test modal analysis and the test under the running of the train is to find out the dynamic characteristic value of the ω-type clip in
high-speed rail, including the natural frequency and mode vibration under the normal installation state, and to find out the modal
deformation characteristics which cause the stress concentration point on the surface of the clip and the inner arc. In order to
simulate the actual installation conditions of W1-type clip, the clip is installed with clip component on a 60 kg/m rails as shown in
Fig. 7.The clip modal test boundary is set to Table 1. 29 test points are arranged even along the clip, which are excited in turn by
incentive hammer. Unlike on-line testing, the clip vibration response replaces the accelerometer with a microphone fixed above the
clip to eliminate the additional mass introduced by the acceleration on the clip. The experimental results show that the clip has two
main modal frequencies below 1000 Hz, among which the modal frequencies are 692 Hz, which well agrees with the butterfly type

Fig. 5. Dynamic strain of clip during application of clamping force.

5
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Fig. 6. Simulating dynamic strain of clip with Wheel-rail excitation.

Fig. 7. Modal analysis of W1-type assembly clip.

Table 1
Rail clip modal analysis boundary condition setting.
Type Clip structure Clip heel constraint Clip toe constraint Clip central load

Laboratory test China high-Speed WJ-8 insulated WJ-8 insulation block 27kN (Bolt)
rail W1-type clip shoulder block
FEA W1-type clip model Set spring constraint, longitudinal 80kN/ Set spring constraint, longitudinal 80kN/ Clip central impose
mm vertical 28kN/mmtransverse 16kN/ mm vertical 28kN/mm transverse 16kN/ concentrated load of 27kN
mm mm

model vibration [6].

According to the test in the laboratory, assembly mode simulation analysis of the clip is introduced. The finite element analysis
model is shown in Fig. 8. The clip modal test boundary is set to Table 1 The simulation model of butterfly type model vibration is the
second mode, and the error rate with the test is 1.1%. The finite element simulation analysis further confirms that the stress response

Fig. 8. Finite element analysis of W1-type assembly clip.

6
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Table 2
Comparison of modal test and simulation analysis of clip assembly.
Modal Experimental Simulation Error Modal shape
order modal/Hz modal/Hz

1 584 579 0.8% Clip two arms move along the longitudinal direction, one high and one low.
2 685 692 1.1% Clip two arms turn inward, similar to the “butterfly effect”.
3 1455 1465 0.6% Clip two heels and front toes move alone longitudinally, but in the opposite direction.
4 1512 1535 1.5% Clip two heels are expanded backward, and the two toes are squeezed in the same direction
5 1726 1738 0.7% The middle of the clip moves longitudinally, and the heel and toe move in reverse.

of the internal arc surface of the elastic toe of the clip mainly comes from the natural frequency of the modal frequency of
580–690 Hz, as shown in Table 2.

3.2. Dynamic parameters of fastener clip

In an operation section of domestic high-speed railway, special tooling and vertical acceleration sensor are respectively set on the
elastic arch of the W1-type clip for the research target. When the high-speed train passes through the section at 300–350 km/h, the
dynamic response parameters of the clip is measured and the dynamic frequency response result is obtained as shown in Fig. 9. It can
be seen from the diagram that the main response frequency of the clip on running trains is 685 Hz and the vibration amplitude is
about 23 g.
In conclusion, the static structure parameters of W1-type clip have a great influence on the self-strength of the clip. In subsequent
optimization, priority should be given to increasing the parameters with positive effects and avoiding negative ones. According to the
results of the modal test under the train running condition and the laboratory test, the main peak frequency of the clip is about
690 Hz, and the corresponding modal of the elastic arm is butterfly type vibration, while such alternating vibration of the elastic arm
introduces alternating stress at the elastic heel. If the excitation frequency of the train is also in this range, the clip will resonate and
generate high frequency and high dynamic stress, which will lead to fatigue fracture.

3.3. High-frequency excitation under wheel-rail wear

In recent years, almost all the breaking points of the problematic clip were found in fasteners of high-speed railway similar to
those shown in Fig. 2. During the inspection and maintenance of the track system, it is found that under the condition of domestic
high-speed railway running at 250–350 km/h, most of the areas where the breakage and damage of spring of fasteners occur are
accompanied by obvious wheel polygon wear (order 18–22) and rail corrugation (obvious wavelength 150 mm) [6].
When the train is in operation, the wheels are subjected to periodic excitation which comes from inherent characteristics of the
vehicle-track system. The uneven wear frequency could be presented per Eq. (3):
v
f= ×n
πD (3)

where, f is the excitation frequency of the wheel polygon, v is the train speed, D is the wheel diameter, and n is the numbers of wheel
wear orders.
The excitation frequency of rail corrugation, at a given train speed, can be described as per Eq. (4):

Fig. 9. On-site test and output of frequency contents of rail clip vibration.

7
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Fig. 10. Vibration energy of clip at different speeds.

v
fc =
λ (4)

in which fc is the excitation frequency of corrugation, v is train speed, λ is the wavelength of corrugation.
Calculated by the above excitation frequency formulas (3), (4), the dynamic excitation frequency range introduced by wheel-rail
wear is basically 580–690 Hz, which belongs to the irregular high-frequency excitation characteristics of wheel-rail coupling. Under
the periodic operation of high-speed rail, the rail vibration acceleration under different driving speeds and wheel-rail wear is tested,
and the vibration energy level of the clip under uneven high frequency is obtained by means of formula (5).

E= ∫ |s (t )|2 dt = ∫ |S (ω)|2 dω (5)

In the formula, s(t) is the time-domain acceleration value under rail vibration and S (ω ) is the frequency acceleration value under
rail vibration.
From Fig. 10, it can be seen that with the increase of running speed of high-speed train, the excitation frequency of wheel and rail
increases, and the vibration acceleration and energy of clip increase. Under the condition of 250 km/h speed level, the excitation
frequency of wheel and rail is not consistent with the natural frequency of the clip, and the vibration energy of the clip is small. Under
the speed levels of 280 km/h, 300 km/h and 345 km/h, the wheel-rail force excitation is consistent with the natural frequency of the
clip, and the clip vibration energy is large. From Fig. 11, at 300 km/h speed, the wheel-rail excitation generated by wheel polygon
(18-order) and rail corrugation (140–160 mm wavelength) is 580–690 Hz, which results in an increase in the energy of clip vibration.
At the same time, when both wheel polygon and rail corrugation exist, the amplitude of wheel-rail force and spring acceleration is the
largest, and the superposition effect is obvious. The vibration energy is 10 times higher than that without wear. Therefore, fatigue
fracture will occur to the clip under the action of periodic operation of the train and alternating stress of high frequency excitation.
Under the installation condition, the second order modal of W1-type clip is “butterfly effect”, which conforms to the actual
working state of clip. However, the frequency (685 Hz) corresponding of the butterfly modal coincides with the high frequency range
(580–700 Hz) of wheel-rail wear excitation. In addition, the high alternating stress concentration emerges at the internal arc surface
of the clip back toe, which finally leads to high frequency fatigue fracture. In conclusion, based on the high-frequency excitation of
wear in high-speed rail and the dynamic stress under the assembly of the clips, the dynamic failure mechanism of W1-type clip in
high-speed rail is revealed, which is a representative research achievement at present.

Fig. 11. Vibration energy of clip with or without wear.

8
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

4. Research on geometric parameters of clips

In view of the large area damage and failure characteristics of fastener clips in high-speed rail, many factors, such as clamping
force of fastener, maximum stress and dynamic frequency response, need to be considered in the design of clip in railway. The
existing W1-type clip in high-speed rail has become a difficulty in design and optimization of fastener clip system due to its com-
plicated linearity, multiple parameters and interrelated relationship. In this section, after studying the clip structure of W1-type in
high speed railway, the geometric parameters of the clip are modeled, and the influence of its geometric size, modal parameters and
dynamic vibration on the clip performance is analyzed, which lays a theoretical foundation for the structural optimization and
performance research of the clip.
Due to various uncertainties of the “vehicle-track” system on high-speed railway, although the fracture of clip on high-speed rail is
single, how to enhance the strength of clip as much as possible and reduce the vibration energy on the body of clip is the direction to
be considered in the design and optimization in this section. On the one hand, Based on the parameter correlation theory, the model
parametric design of the existing W1-type clip is carried out in the manuscript, and 11 main parameters of the clip center curve are
obtained. the strength, clamping force and stress of the model generated by the main parameters are simulated and analyzed to
evaluate the advantages and disadvantages of each parameter. On the other hand, combining with the unit mass energy storage
theory in literature [14], the optimal center curve structure of the clip is obtained. It can improve the service life of clips in high-
speed rail to some extent.

4.1. Geometric parameters of clip

The Structural form of W1-type clip in high speed rail is complex and several parameters are interrelated. According to the
analysis, the central axis of the clip is on the cylinder with radius R01, R02, R03 and R04, and the cylinder with radius R01 and R03 are
tangent with each other, while the cylinder with radius R02 and R04 are tangent with each other as shown in Fig. 12.
There are up to 11 parameters affecting the geometrical size of the clip model, but not all of them have a significant impact on the
performance of the clip. According to the design requirements of fastener in high-speed rail, three parameters of the fastener clips are
mainly studied: assembly clamping force P, maximum assembly stress σ and optimization index Q. Among them, the optimization
index adopts the energy viewpoint put forward by Xiao [14] et al.: after the assembly of the fastener clip, the energy Q stored per unit
mass at the same stress level is used as the index to evaluate the rationality and economy of the design of the clip.
Q= PT/(mσmax ) (6)

In the formula (6), P is the clamping force, T is the projectile range, m is the mass of the clip, and
σmax is the maximum stress after the assembly of the clip.
By studying the relationship among various parameters of the clip, as shown in Fig. 13, it is found that as the length of the front
toe of the clip B02 increases, the clamping force and the clip stress decrease. With the increase of heel radius R06, the clip deflection P

Fig. 12. Plane expansion diagram of clip in high-speed railway.

9
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

(a) Influence of toe length on clip performance (b) Influence of heel radius on clip performance

(c) Influence of arch width on clip performance (d) Influence of arch height on clip performance
Fig. 13. Centerline Parameter Optimization of clip Structure.

force also increased. With the increase of arch width B03, the clamping force and clip stress also increase. With the increase of arch
height H0, the clamping force and the stress of the clip did not increase obviously. The length of the front toe of the clip has a negative
effect for the clip performance, while the increase of the heel radius, arch width and arch height of the clip has a positive effect for the
clip performance. According to the analysis in Table 3, the performance of rail clip (single clamping force greater than 9kN, material
strength of 60Si2Mn, σb = 1350MPa ) and unit energy storage Q are the highest. The optimal structure of clip is shown in Table 4. The

Table 3
Optimization analysis of parameters of fastener clip.
Item H0/mm B03/mm R06/mm B02/mm D/mm F/kN T/mm P/kN Stress/MPa Mass /kg Q

1 44 20 25 34 14 27 13.65 7.15 1286 0.71 0.107

2 44 20 25 32 14 27 13.89 7.45 1267 0.70 0.117
3 44 20 25 30 14 27 13.95 7.72 1252 0.69 0.125
4 44 20 25 28 14 27 14.04 7.98 1239 0.68 0.133
5 44 20 25 26 14 27 14.13 8.19 1223 0.67 0.141
6 44 20 24 34 14 27 13.48 8.26 1306 0.66 0.129
7 44 20 22 34 14 27 13.21 8.51 1318 0.69 0.124
8 44 20 26 34 14 27 13.74 9.67 1264 0.70 0.150
9 44 20 27 34 14 27 14.05 9.86 1175 0.73 0.162
10 44 19 25 34 14 27 13.49 9.55 1316 0.71 0.138
11 44 18 25 34 14 27 13.23 9.38 1343 0.69 0.134
12 44 17 25 34 14 27 13.13 9.31 1368 0.68 0.131
13 44 16 25 34 14 27 13.05 9.25 1389 0.67 0.130
14 46 20 25 34 14 27 13.45 7.89 1185 0.73 0.123
15 45 20 25 34 14 27 13.38 7.47 1243 0.72 0.112
16 43 20 25 34 14 27 13.28 6.99 1320 0.79 0.089
17 42 20 25 34 14 27 13.25 6.87 1337 0.68 0.100

10
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Table 4
Optimal dimension parameters of rail clip.
H0/mm B03/mm R06/mm B02/mm D/mm F/kN T/mm P/kN Stress/MPa Mass /kg

44 20 27 34 14 27 14.05 9.86 1175 0.73

center curve of rail clip can be generated through the optimal parameters.

4.2. Geometric parameter optimization of clip

In combination with the correlation study of Section 4.1, a new center curve of clip is obtained after analyzing the main para-
meters of clip, as shown in Fig. 14 and Table 4. It can be seen that the front toe of the clip is shorter, radius of the clip heel is larger
and the arch height is reduced.

5. Research on optimal composite damping clip

It has been proved by experimental research and finite element simulation analysis that butterfly-type alternating stress is pro-
duced due to natural frequency and wheel-rail excitation when the clip in high-speed rail is damaged, resulting in the yield of the heel
strength of the clips.
In addition to improving the clip geometric parameter of Section 4 to solve the failure of rail clip, this section will optimize and
compare the structure of clip from two perspectives of “enhance modal frequency” and “reducing vibration amplitude”. Based on
modal theory and damping vibration reduction technology, the research and optimization direction will be considered compre-
hensively from three performance parameters of clip: stiffness, mass and damping. First of all, with the help of the research con-
clusion in Section 4, the optimized clip center curve is used to generate hollow clip structure. It is necessary to improve the stiffness of
the clip so that its natural frequency can avoid the range of excitation frequency, but the increase in stiffness should be larger than the
increase in mass; The second is to set the composite damping clip by the hollow clip structure, reduce the vibration amplitude of the
clip, which can improve the vibration attenuation and energy of the clip. Thirdly, with the help of the impact load test, the amplitude
test was carried out on the damped composite clip with different structure, and the amplitude-frequency characteristics of the
damped composite clip with different structure were compared. Lastly, the clip optimization not only considers from the curve of
structure optimization, but also ensures the practicability and economic feasibility of the clip in order to facilitate the installation and
replace ability of the existing fasteners.

5.1. Geometric parameter optimization and strength analysis

In combination with the study of Section 4.1, the optimal center of clip curve is used to generate hollow modal and analyze the
characteristics. The analysis boundary conditions are shown in Table 5.
The normal contact is defined using HARD mode in Abaqus analysis. The Coulomb friction model is used for tangential contact, in
which the elastic slip is considered between the states of bond and slip. The clip, shoulder block and insulation block are all defined

Fig. 14. Comparison before and after optimization of the clip centerline.

11
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Table 5
Setting of boundary conditions for strength analysis of hollow clip.
Component Elasticity modulus/MPa Poisson's ratio Density Boundary and loading force Settings
kg/m3

Hollow clip 210,000 0.3 7800 Contact restraint is provided with the shoulder block, insulation block and washer.
Shoulder block 20 0.4 1600 Shoulder block all direction constraints;
Insulation block 20 0.4 1600 Insulation block all directional constraints;
Metal washer 210,000 0.3 7800 The middle of clip applied concentrated force of 27kN.

with the non-coordinated mode unit C3D8I.The assembling strength and mode changes of clip under different outer diameter and
different inner diameter of excavation are obtained. As shown in Fig. 15, it can be obtained from 14.0 mm-H-0 mm solid clip
(referring to the clip of outer diameter 14 mm, internal aperture 0 mm),hollow 14.4 mm-H-2 mm, hollow 14.6 mm-H-4 mm and
hollow 14.8 mm-H-6 mm, the assembly clamping force of the clip is 9.5kN, 7.6kN, 8.3kN and 9.6kN respectively. The clamping force
first decreased and then increased, while the stress change first increased and then decreased, and the maximum value was
1420 MPa, corresponding to the hollow 14.4 mm-H-2 mm structure.
As can be seen from Fig. 16, from the solid clip 14.0 mm-H-0 mm clip, hollow 14.4 mm-H-2 mm, hollow 14.6 mm-H-4 mm and
hollow 14.8 mm-H-6 mm, each mode of clip of high-speed railway is improved with the increase of outer diameter and inner diameter
of the clip, and as shown in Fig. 17, the butterfly type vibration frequency of the second order model increases from solid 692 Hz to
821 Hz, which is far away from the excitation frequency of the wheel and rail interaction corresponding to train speed of
300–350 km/h.

5.2. Optimization of clip fatigue life analysis

Rail fasteners are subjected to impact loads of the train while the train is running. Under the effect of repeated impact force load,
the stress strain of the bolt is generated under the pre-tightening force of the bolt. Once this cyclic stress exceeds the fatigue strength
of the spring, fatigue failure occurs. In order to ensure the safe operation environment of the railway line, it is necessary to estimate
the fatigue life of the clip under the impact load of the train. On the basis of static analysis, dynamic vertical displacement D is applied
to the optimum clip. As shown in Fig. 18, the fatigue life of the clip under train operation is simulated, and the position where fatigue
failure first occurs and the influence of rail vertical force on the fatigue performance of the clip are obtained.

14.0mm-H-0mm 14.4mm-H-2mm

14.6mm-H-4mm 14.8mm-H-6mm
Fig. 15. Results of strength and clamping force of different hollow clip structures.

12
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Fig. 16. Characteristics of clip with different diameter.

Fig. 17. Modal frequencies of clip with different diameter.

Fig. 18. Dynamic fatigue loading displacement of clip.

From Fig. 19 and Fig. 20, it can be seen that the minimum life of solid clip with 14 mm diameter is 397,900 times, while that of
structures with 14 mm-H-2 mm is the lowest and the life is 1,892,000 times, which indicates that the fatigue life of clips with hollow
structure is only half of that of solid structure. At the same time, the fatigue life of 14.6 mm-H-4 mm and 14.8 mm-H-6 mm is higher
than that of solid clip when the diameter of clip is increased, which indicates that the hollow clip meets the design requirements from
the perspective of fatigue.

5.3. Frequency response characteristics of highly damped clip

Most of the existing rail fastener clips are made of circular spring steel with equal section. The design of clip is generally three-

13
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Fig. 19. Comparison of the fatigure life of hollow and damping structure of rai clip.

a Fatigue life of 14.0mm-K-0mm b Fatigue life of 14.4mm-H-2mm

c Fatigue life of 14.6mm-H-4mm d Fatigue life of 14.8mm-H-6mm

Fig. 20. Fatigue life of different clip structures.

point stress design, i.e. buckle point, support point and restraint point. The clip design is generally three-point stress layout, that is,
the compression point, the support point and the constraint point. Low damping spring steel is normal used as the clip material. The
modal frequency of the clip is in the vehicle-track system range. The clip would generates high-amplitude vibration under the high-
frequency excitation of the vehicle-track system [15–17]. Especially in the high stress area of clip, the resonant amplitude value is
much higher than the high frequency alternating stress of the wheel passing frequency, which results in that the clip fatigue life
would be shortened greatly.
In order to improve fatigue life and reduce resonance damage of W1-type clip, the hollow elastic strips with different diameters
designed in 4.1 above are filled with non-metallic damping material by grouting. That is, damping material is filled in the middle
hollow area of the optimized clip, in which the bright part is hollow metal clip and the black part is filled with damping material. The
solid structure is shown in Fig. 21. Damping characteristics of the filled damping material are much larger than that of the clip body,

14
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Fig. 21. W1-type composite damping clip.

i.e. the amplitude of the clip vibration is reduced, the vibration attenuation of the clip is increased, and the fatigue life of the clip is
further improved.
The process of filling damping material for hollow clip is as follows: by selecting hollow bars with different internal and external
diameters, heating and forging them into hollow clip, it is necessary to ensure that the hollow diameter of each surface section of the
clip is basically set; Clean the hollow structure inside the clip and spray with adhesive. Material composition and properties of
damping clip were shown in Table 6. When the damping material is heated to 150℃, the inner hole of the clip is filled under pressure.
Natural drying and solidification of damping material filled with the same material is an optimized high damping clip. In order to
compare the characteristics of the clips before and after optimization, drop hammer impact attenuation tests were carried out on the
grouting damping clips as shown in Fig. 22.
In order to compare the vibration amplitude-frequency characteristics of the damped composite clip before and after optimiza-
tion, ensure the same installation boundary conditions, the left and right sides of the same sleeper and the same fastener are re-
spectively placed for the clip before and after. Ensure that the installation constraints on both sides of sleeper are consistent, and the
intermediate bolts are applied with a 27kN preload. The height of each drop hammer was set to 25 cm, and the average value was
taken for a total of three drop hammer impacts.
As can be seen from Fig. 23 and Table 7, after the damping material is filled, the diameter of the clip increases from solid
(14.0 mm-D-0 mm, referring to the clip of outer diameter 14 mm, internal damping aperture 0 mm) to 14.8 mm-D-6 mm for the high-
damping clip, The mass of the clip decreased by 9.7% from 0.71 kg solid to 0.64 kg highly damped, the frequency of the clip increases
by 18.4% from 680 Hz to 805 Hz. The frequency response amplitude corresponding to the high damping clip butterfly mode de-
creased from 256 dB to 175 dB, decreasing by 31.6%.
In conclusion, in addition to ensuring the characteristics of fastener clips, the response amplitude of high damping decreases
significantly as the ratio of outer diameter to damping diameter increases, and the overall quality of clips decreases. To a certain
extent, not only the excitation frequency of wheel rails is kept away, but also the dynamic fatigue life of clips is improved and the cost
of clips is saved.

6. Conclusion

Fastener clip in High-speed rail is widely used due to its high strength, adjustable fastening pressure and convenient installation.
This paper analyzes the failure characteristics and structural parameters of the W1-type clip, reveals the mechanism of the fatigue
fracture of the fastener clip, and puts forward a kind of high-speed clip with high composite damping material, and draws the
following conclusions:
(1) Wheel polygon and rail corrugation on 300–350 km/h high-speed rail line may be the excitation source of vibration in rail and
fastener, which is characterized by high frequency and high amplitude with main frequency range of 550–700 Hz. The dynamic
response frequency of W1-type rail clip is 580–690 Hz, which falls within the range of wheel and rail excitation frequency. Therefore,
under the action of the excitation force caused by wheel and rail wear, the clip resonates and damage.
(2) According to the high-speed clip assembly stress and dynamic motivation research, W1-type clip in the normal assembly

Table 6
Material composition and properties of damping clip.
Material Elasticity modulus /MPa Poisson's ratio Density/kg/m3

Clip parent metal 210,000 0.3 7.8 × 103

Hollow bonding material 5.5 0.48 1300
Fill in damping material 8.5 0.45 1300

15
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Original clip Damping clip

Fig. 22. Impact frequency response test of different damped clip.

Fig. 23. Frequency response of drop Hammer for different damping clips.

Table 7
Impact test results of different damped clip.
Damped bars with different radius 14.0-D-0 mm 14.4-D-2 mm 14.6-D-4 mm 14.8-D-6 mm

Frequency /Hz 680 715 752 805

Amplitude /dB 256 207 193 175
Mass /kg 0.72 0.69 0.67 0.65

situation is close to the yield stress of the material, the clip stress increased by 10.9% on the base of the installation dynamic stress
with the excitation of the high-speed wheel/rail, which is also the potential reason of clip failure.
(3) Based on the parameter correlation theory, the research on the structural parameters of high-speed clip shows that the front
toe length of W1-type clip has a negative influence on the main performance of the clip, while increasing the heel radius, arch width
and arch height of the clip has a positive influence on the performance of the clip. At the same time, the equal ratio hollow clip also
has an impact on the performance of the clip. Under the different optimized structures of the solid clip 14.0 mm-H-0 mm, hollow
14.4 mm-H-2 mm, hollow 14.6 mm-H-4 mm and hollow 14.8 mm-H-6 mm, the clamping force of the clip first decreases and then
increases, while the change of stress first increases and then decreases. The maximum value is 1420 MPa, corresponding to the hollow
14.4 mm-H-2 mm structure.
(4) Based on modal theory and damping vibration reduction technology, research on high damping composite W1-type clip
obtained by filling optimized hollow clip with damping material is carried out, the diameter of the clip increases from 14.0 mm-D-
0 mm to 14.8 mm-D-6 mm for the high-damping clip. As the ratio of outer diameter of high damping clip and damping diameter
increases, the attention to natural mode frequencies of hollow clips was increased by 19% respectively after optimization. The
response amplitude of high-damping composite clip decreases significantly, with a drop rate of up to 31.6%, and the overall quality of
the clip decreases. To a certain extent, the structural design of high-damping composite clip is not only far away from the excitation
frequency of wheel and rail, but also improves the dynamic fatigue life of the clip and saves the cost of fastener system.

16
X. Gao, et al. Engineering Failure Analysis 118 (2020) 104848

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgements

This work is funded by the National Natural Science Foundation of China (U1834201).

References

[1] L. Ling, W. Li, H. Shang, X. Xiao, Z. Wen, X. Jin, Experimental and numerical investigation of the effect of rail corrugation on the behaviour of rail fastenings,
Veh. Syst. Dyn. 52 (9) (2014) 1211–1231.
[2] A. Hasap, P. Paitekul, N. Noraphaiphipaksa, C. Kanchanomai, Analysis of the fatigue performance of elastic rail clip, Eng. Fail. Anal. 92 (2018) 195–204.
[3] D.J. Thompson, J.W. Verheij, The dynamic behavior of rail fasteners at high frequencies, Appl. Acoust. 52 (1) (1997) 1–17.
[4] H. Xiao, J.-B. Wang, Y.-R. Zhang, The fractures of e-type fastening clips used in the subway: Theory and experiment, Eng. Fail. Anal. 81 (2017) 57–68.
[5] S.Y. Zhu, C.B. Cai, Q. Yin, et al., Dynamic analysis of rail fastening clip in high-speed railway, J. Eng. Mech. 6 (2013) 254–258 (in Chinese).
[6] J.H. Xiao, Z.Q. Yan, Y.H. Tu, On Impact of wheel-rail Vibration on HSR Fastener Damage, Chin. Railways (2017) 11 (in Chinese).
[7] X.G. Gao, A.B. Wang, Y. He, et al., Structural improvement of the omega-type High-Speed Rail clip based on a study of its failure mechanism, Shock Vib. 7
(2019).
[8] Z.R. Yu, Y. Yuan, Y.Q. Zhang, et al., Fatigue properties of elastic bars of fastening systems installed with high-speed railways, Journal of the China Railway
Society 7 (2014) 90–95.
[9] J. Xiang, J. Yang, C. Yu, et al., Analysis of elastic bar fracture causes of fasteners in ballastless track of high-speed railway, Journal of the China Railway Society
16 (7) (2019).
[10] S. Mohanmmadzadeh, S. Ahadi, M. Nouri, Stress-based fatigue reliability analysis of the rail fastening spring clip under traffic loads, Latin Amer. J. Solids
Structure 11 (6) (2014) 993–1011.
[11] H.P. Guo, B. Li, Y.S. Li, Fracture analysis of 60Si2MnA spring rod, Fail. Anal. Prev. 6 (2) (2011) 90–93 (in Chinese).
[12] X. Xu, D.Q. Zou, D.Q. Yang, et al., Analysis of fatigue fracture reason of railway W1-clip, Chinese Railways 11 (2016) 24–27.
[13] W.X. Wang, Z. Wang, L. Zhao, et al., Failure analysis of SKL15 spring fasteners for high-speed railway[J], Heat Treat. Met. 12 (2013) 108–112.
[14] J.H. Xiao, R.K. Zhao, Study on Optimum Design of Bar-Spring Clip, China Railway Sci. 016 (004) (1995) 42–47.
[15] J. Ryue, D.J. Thompson, P.R. White, D.R. Thompson, Investigations of propagating wave types in railway tracks at high frequencies, J. Sound Vib. 315 (1-2)
(2008) 157–175.
[16] D.R. Ahlbeck, L.E. Daniels, A review of rail corrugation processes under different operating modes, Proceedings of the 1990 ASME/IEEE Joint Railroad
California. New York, 1990.
[17] S.L. Grassie, Rail corrugation: Characteristics, causes, and treatments, Proc. Inst. Mech. Eng. Part F: J. Rail Rapid Transit. 223 (6) (2009) 581–596.

17