Criteria for repairing damages of CA mortar for prefabricated framework-type slab track

Construction and Building Materials 110 (2016) 300–311

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Criteria for repairing damages of CA mortar for prefabricated framework-type slab track Ren Juanjuan ⇑, Li Xiao, Yang Rongshan, Wang Ping, Xie Peng MOE Key Laboratory of High-speed Railway Engineering, Southwest Jiaotong University, Chengdu 610031, PR China School of Civil Engineering, Southwest Jiaotong University, Chengdu 610031, PR China

h i g h l i g h t s
Influence of CA mortar damages on the structure’s carrying capacity was studied.
A 3-D FEM of framework-type slab track reflecting CA mortar damages is established.
A dynamic test of the slab track with damage to CA mortar has been conducted on site.
Recommendation of the classification and criteria of mortar damages was put forward.

a r t i c l e

i n f o

Article history: Received 5 August 2015 Received in revised form 15 December 2015 Accepted 15 February 2016 Available online 19 February 2016 Keywords: CA mortar damage Finite element method Carrying capacity Criteria for repairing damages

a b s t r a c t Damage of bagged cement–emulsified asphalt (CA) mortar tends to create voids underneath the track slab and will reduce the carrying capacity of the track structure, which is contrary to its design purpose of providing a long-term, safe and reliable service. In this paper, a study on the criteria for repairing damages of the CA mortar was conducted by establishing a 3-D FEM of the prefabricated framework-type slab track on elastic foundation. In order to verify the calculation results, a field test was carried out by comparison of the slab with or without the repair of damaged CA mortar. Based on the limited results in this study, the recommended values for criteria of repairing damages to CA mortar are preliminarily put forward. The results show that the mortar damages have the most adverse impacts on the carrying capacity of the track slab and CA mortar, but less impact on the concrete base. The numerical simulation indicates the critical length of 0.20 m and critical width of 0.40 m for the transverse damage type, and the critical length of 0.60 m and critical width of 0.20 m for the longitudinal damage type. For the damage type of the interfacial loss of bond, a critical gap size of about 2 mm was also deduced. If the actual dimensions of the damage to CA mortar are greater than the critical values, the tensile stress of track slab or the compressive stress of CA mortar will increase significantly and may exceed their characteristic value; and the dynamic deformation and vibration performance of track structure becomes apparent; the fatigue damage of CA mortar under the area adjacent to the damage would also be accelerated. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Slab track is characterized by its high stability, availability, reliability, and low maintenance associated costs, which reveals that it is a developing trend for high speed railways. By the end of 2015, China is expected to complete the construction of 18,000 km highspeed railways, among which the slab track technology accounts for about 78%. The Japanese prefabricated concrete slab track (J-slab track for short) has been widely used on high-speed railway or passenger dedicated lines in China with the total length about ⇑ Corresponding author at: MOE Key Laboratory of High-speed Railway Engineering, Southwest Jiaotong University, Chengdu 610031, PR China. E-mail address: [email protected] (J. Ren). http://dx.doi.org/10.1016/j.conbuildmat.2016.02.036 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.

2203 km (double track) for its convenient construction, easy maintenance and low structure height [1,2]. Although high-speed railway places strict requirements on the regularity and reliability of railway track, the frequent train loading and its operation in harsh natural conditions subject the parts of track structure to a longterm degradation [3], which will inevitably give rise to a variety of damages. Framework-type prefabricated slab track is one main kind of Jslab track applied to China’s high-speed railways. This slab track structure is mainly composed of the steel rail, the WJ-7 fastening system, the framework-type track slab, the bagged cement–emulsified asphalt (CA) mortar, the cylindrical stopper/bollard and the concrete base. It has the following advantages: reducing the warping of track slab caused by temperature change, which can cut

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

down the repair work of CA mortar; reducing the volume and dead weight of the track slab and the dosage of CA mortar, which can accordingly reduce the production cost and freight and get better engineering and economic benefits; and improving the workability, which makes the distribution of the CA mortar filled underneath the slab more uniform. Fig. 1(a) and (b) are the schematic diagram and the after-laying site picture of framework-type slab track structure. With high fluidity, the 40–100 mm thick cement–emulsified asphalt mortar (CA mortar) layer can be poured into the gap between the prefabricated track slab and the concrete base purely by means of its own gravity, playing the roles of leveling, supporting and damping [4]. It is a key component of the prefabricated slab track and its mechanical properties (such as strength, stiffness, and damping) play an important role for a smooth and safe ride. CA mortar is a metastable suspension consisting of cement, emulsified asphalt, sand, water and admixtures (such as aluminum powder and superplasticizer), which has unique properties that differ from concrete and asphalt alone as this hybrid material combines strength of cement mortar and flexibility of asphalt binder [4,5]. This slab track type adopts an assembled structure and its CA mortar with low elastic modulus (E = 100–300 MPa) has relatively weak constraints on the track slab. Therefore the unstable vertical layered structure tends to cause damages to the weakest CA mortar and then possibly to the track slab. The previous investigations mainly focused on the material performance and engineering applications of CA mortar. By conducting dynamic compression tests of CA mortar within its strain rate range of 105–102 s1, Wang Ping et al. [6] investigated the constitutive relation of CA mortar, and the impact of strain rate on the CA mortar’s compressive strength, elastic modulus and peak strain, and presented relevant recommended values for the CA mortar’s stress–strain curve equation under different strain rates. Xie Youjun et al. [7] investigated the dynamic mechanical properties of CA mortar through the Split Hopkinson Pressure Bar (SHPB) test. The results provided the theoretical basis for the structural design of slab track. Jin Shouhua et al. [8] proposed the mortar formulation design and established the quality inspection system according to its application performance. They worked out a reasonable formulation of new materials based on the inversion of the aggregate properties of CA mortar and on the analysis of the aggregate properties. At present, the investigations on the influence of CA mortar damage on the track system have been carried out mainly on the flat-type prefabricated slab tracks, involving less framework-type prefabricated slab track. In addition, the main factor to be considered during the analysis process is the contingency of the influence of CA mortar damage on the track system, not taking into account of the cumulative effect of CA mortar damage and its influence on the structural durability of the track system. Zeng Zhen [9] studied the carrying capacity of the track structure with different contact

areas between the CA mortar and the track slab using the ‘‘Element Birth and Death” and the embeddable random function methods. Based on the dynamics of the wheel-track system, Li Peigang et al. [10] established a 2-D vertical vibration model of the traintrack-bridge by which the influence of CA mortar void on the dynamic characteristics of the slab track element on bridge is investigated. Zhu Shengyang [11] analyzed the influence of the damage to the interface between the flat-type prefabricated slab track and high elastic modulus CA mortar on the vibration of slab track system under the effect of temperature loads and train loads. Xiang Jun et al. [12] studied the influence of suspended track slab on track vibration response which was caused by the degradation of CA mortar based on the dynamics analysis of the train-slab track coupling system. It should be also drawn attention that the similar problem on conventional railway track – the effect of unsupported sleepers has been investigated. The main difference between models for FEM calculations (regardless of the variety of parameters related to the vehicle itself and track system) consists in different modeling of track construction beneath the fastening system. The ballast at some sections of railway tracks subsides seriously and unevenly, which is also mainly caused by the contact vibration of the wheel and the rail due to the existence of various irregularities of the track and the wheelset. Nielsen and Igeland [13] investigated the vertical dynamic behavior for a railway bogie moving on a rail which is discretely supported, via rail pads, by sleepers resting on an elastic foundation, and analyzed the influences of three types of practically important imperfections in the compound vehicletrack system. Shuguang Zhang et al. [14] built a coupling dynamic model of vehicle-track which can consider the effect of the discrete support by sleepers on the coupling dynamic behavior of the vehicle and track. A nonlinear spring and a nonlinear damper are used to simulate a gap between the unsupported sleeper and the ballast mass. Zhu et al. [15] studied the dynamic wheel-rail interaction force due to one or multiple unsupported sleepers, and put forward that the vehicle speed, the gap size and the number of unsupported sleepers primarily dictate the magnitude of impact load which can be significant. Shi et al. [16] also investigated the behavior of a section of existing track on the ballasted heavy haul railway line, and showed that the maximum displacement of the rails and sleepers increases significantly with the number of consecutive unsupported sleepers.

2. Criteria for repairing CA mortar damages The structure of slab track is composed of multiple structures or components which have different materials and characteristics. Any change of any component in the performance, strength and structure will impose an influence on the working conditions of

Framework-type prefabricated slab

Concrete base

Fastening system CA mortar Cylindrical stopper

(a) schematic diagram of framework-type slab track

301

(b) site picture of framework-type slab track

Fig. 1. Framework-type prefabricated slab track.

302

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

other components, thus having direct effects on the quality of train operation. The high-speed railways and passenger-dedicated railways require high durability and reliability of the track structure. However, the mortar damage or defects can lead to mutation of track stiffness, which will affect the whole life cycle of the track system [12,17]. Therefore, the slab track systems assessment should be enforced, reasonable maintenance cycle and standard should be formulated, and scientific systems of maintenance should be established, aimed at ensuring a stable and credible state of the slab track system during the safety service life. The slab track system has high reliability but poor maintainability and it is difficult to be repaired if damaged. Therefore, the concept of comprehensive repair should apply across the whole process of design, construction and operation. With the passage of time, the components of track will result in degradation or damage. Proper repair within a reasonable time can effectively extend the service life of track components and keep the track system stable and reliable [12,18]. As shown in Fig. 2, the development process of the wear of track components, which can be divided into the three stages of initial wear, normal wear and severe wear, poses certain regularity. Apparently, the normal wear stage is the best time for repair. The repair level of the track component damage is an important consideration for the temporary repair. The principle of setting the repair level is to reduce the workload of repair and extend the safety service life of the track structure as far as possible on the premise of ensuring the safety of train operation. Learning from the domestic and foreign operation practices, the levels of mortar damage can be divided into three levels, Level I, Level II and Level III. The respective coping approaches to all levels of mortar damage are as follows: for the damage of Level I, we should pay attention, make records and keep it under observation; for the damage of Level II, we should work out a repair plan and remedy the damage as the case may be; for the damage of Level III, we should remedy it as soon as practicable. Because repair of the mortar damage is closely related to the environment, time, material and process, considering the particularly severe environment conditions, repair of Level III damage should guarantee the running safety and riding comfort, as well as reserve enough time for repair. 3. Objective and scope In this paper, the finite element method incorporating the contact analysis and the ‘‘Element Birth and Death” technology [19,20] is used to accurately simulate the contact relation between the CA mortar and the track slab, as well as the failure of local CA mortar. A 3-D Finite Element Model of framework-type slab track structure reflecting CA mortar damages is then established, to investigate the influence of CA mortar damages in different levels on the carrying capacity the track structure. In addition, a dynamic test has been conducted on site on the damage to CA mortar for the framework-type slab track laid in the slab track test section.

Fig. 2. Schematic diagram of track components wear.

Finally, the levels and criteria for repairing mortar damage are presented from the perspectives of the mortar fatigue life, the durability and the principle of preventative maintenance of track structure. These criteria for repairing damages can be used as the theoretical basis for maintenance management of the slab tracks in China. 4. Characteristics and damages of the material CA mortar damages include cracking, stripping, rupture, excessive plastic deformation and water accumulation caused by insufficient drainage, as shown in Fig. 3. CA mortar is the adjustment layer and the load-bearing layer of a slab track structure. The causes and the development mechanism of CA mortar damages or defects are very complicated, generally related to the coupling effect of multiple adverse factors, such as inadequate structural design, deteriorated material performance, unfavorable natural environment, repetitive train loading and poor construction quality. Since the slab track adopts the assembled structure with vertical stratification, the interfaces between its structural layers are often the weak links. Because the low elastic modulus CA mortar has a high water–cement ratio and contains rich asphalt, with the micro structure of mortar materials, it is susceptible to material performance degradation [21–24]. The work environment of CA mortar is very complicated and the coupling effect of the environmental factors like water and temperature is one of the major factors that cause the degradation of mortar material performance. The train load’s frequent impacting on the mortar surface seriously degrades the mortar stress conditions and causes degradation of the mortar material performance. The construction defects will aggravate the degradation of the stress conditions of CA mortar during its service life, further affecting the durability of the structure. 5. 3-D model of framework-type slab track structure 5.1. Computation model Bagged CA mortar with low elastic modulus imposes smaller constraint to the track slab in the longitudinal and transverse directions and larger constraint in the vertical direction. In order to investigate the influence of CA mortar damages on the carrying capacity of the track structure, a 3-D model of framework-type slab track structure on elastic foundation is built by using ANSYS software (see Fig. 4). Three track slabs are established in the computation model (corresponding to a total length of 15 m) to eliminate the boundary effect by computation, and the middle track slab is taken as the research object. Contact element is used to simulate the sliding and friction between the CA mortar and the track slab, assuming there is no bonding force between the track slab and the CA mortar with a friction factor of 0.3 and there is bonding force between concrete base and CA mortar and ignoring the constraint which cylindrical stopper/bollard imposes on the warping deformation of the slab [25]. In the numerical simulation, the sliding and friction between the CA mortar and the concrete base is neglected. The main element types of the computation model are set as follows: the rail with the linear density 60 kg/m is applied and simulated by the discretely supported Euler beam. The [beam 188] in the finite element model is adopted in ANSYS software. Both of the fastener and subgrade are considered as the spring-damping element and are simulated by [COMBIN14] in ANSYS. The track slab, CA mortar and the concrete base are all simulated by the solid element [SOLID45]. In the model, ‘‘Element Birth and Death” technology is used to simulate the contact relation between the

303

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

(a) excessive plastic deformation of CA mortar

(b) cracking and stripping of CA mortar

(c) water accumulation in framework-type slab

Fig. 3. CA mortar damages or defects.

(a) side view

(b) 3-D model Fig. 4. 3-D model of framework-type slab track structure.

CA mortar and the track slab. The elements of damaged CA mortar can be ‘‘killed” in order to simulate the damaged part, and the elements are set to be ‘‘active” to simulate the undamaged part. The main parameters for the framework-type slab track and subgrade proceed in Table 1.

5.2. Values of temperature gradient and train loading Under the actions of natural environmental factors such as the warming of sunlight and the cooling of strong cold air, temperature difference or temperature gradient tends to appear in the thickness direction of the track slab, which will cause warping deformation and warping stress inside the track slab. When analyzing the influence of the mortar damage on the carrying capacity of track structure, the commonly-used temperature gradient and wheel load are adopted. The values of commonly-used positive and negative temperature gradient [19] are respectively set to be 45 °C/m and 22.5 °C/m. The thickness of track slab is 190 mm and the correction coefficient of the track slab thickness is 1.08 [25]. The term of correction coefficient is used for getting the maximum temperature gradient of concrete pavement [26], if the thickness of slab is not equal to a standard one (22 cm). Using the linearly interpolating method, we can get the correction coefficient of 1.08 for the J-slab with thickness of 19 cm [27].

Table 1 Basic parameters of the slab track and subgrade. Parameters

Symbol

Unit

Magnitude

Vertical stiffness of fastener Distance between fasteners or rail supports Length of slab Width of slab Thickness of slab Length of concrete base Width of concrete base Thickness of concrete base Thickness of CA mortar Length of frame hole Width of frame hole Young’s modulus of CA mortar Young’s modulus of track slab Young’s modulus of the concrete base Poisson’s ratio of CA mortar Poisson’s ratio of track slab Poisson’s ratio of concrete base Vertical surface stiffness of subgrade

kf df

kN/mm m

50 0.625

LS Ws Ts Lcb Wcb Tcb TCA LFH WFH ECA ES ECB

m m m m m m mm m m MPa MPa MPa

4.95 2.4 0.19 15 3.2 0.3 50 2.80 0.70 200 3.60  104 3.25  104

mCA mS mCB

– – – MPa/m

0.15 0.20 0.20 76

k76

When considering the temperature gradient, positive temperature gradient is based on the slab bottom for the reference temperature (0 °C), while negative temperature gradient is based on the

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

w

w

304

l (a) mortar damage under the corner of slab

(b) mortar damage under the edge of slab

l

l

l (c) mortar damage under the end of slab

w

(d) mortar damage under the middle of slab

Fig. 5. Schematic diagram of the assumption of four typical forms of mortar damage.

slab surface for the reference temperature (0 °C), and the temperature gradient load is applied layer by layer along the vertical direction of the track slab. The wheel load is applied by the single-axle and double-wheel method, with the Chinese commonly-used wheel load of 150 kN. 5.3. The simplified forms and representation of the mortar damage For excessive loading under service condition, a loss of bond at the interface between the slab and CA mortar or spalling of CA mortar might occur for the slab track. Both damage can easily tend to create voids underneath the track slab. According to the actual situation, mortar damage forms can be simplified in the finite element model and it is assumed that mortar exhibits a complete damage along the vertical direction, as shown in Fig. 5. The two parameters, length (l) and width (w), are only used to describe the mortar damage because of the assumption that the mortar exhibits a complete damage along the vertical/thickness direction. Specifically, l is the length of mortar damage along the longitudinal direction of the track, and w is the width of the mortar damage along the transverse direction of the track. As shown in Fig. 5, the length of mortar damage under the edge of slab (Fig. 5 (b)) is set as the length of the track slab, and the width of mortar damage under the end of slab (Fig. 5(c)) is set as the width of the track slab. 6. Influence of mortar damage on the carrying capacity of track structure Theoretical researches [19] show that local damage of CA mortar is very bad for the carrying capacity of the track slab and the CA mortar itself. After CA mortar damage leads to voids underneath the track slab, vertical deformation of rail and the track slab will significantly increase under the effect of train loading. Besides, it will also have an influence on the carrying capacity of fastening system and reduce the service life of the track structure. 6.1. The influence of CA mortar damage on the carrying capacity of track slab Because the low elastic modulus mortar is characterized by the deformation compatibility, mortar damage has limited impact on

longitudinal and transverse warping stresses of the track slab under the effect of temperature gradient. Fig. 6 shows the deformation of the track slab without mortar damage and with voids under the corner of slab, respectively, when the positive temperature gradient takes effect on it. Under positive temperature gradient, the track slab have an upwarp deformation, part of the track slab breaks away from the CA mortar and the supporting force is obvious in the corner of slab. As shown in Fig. 6(a) and (b), compared with the mortar without damage, upwarp deformation in the middle of slab and the vertical displacement in the corner of slab increase significantly with voids under the corner of the track slab. When the CA mortar underneath the corner of concrete slab exhibits a damage, with the length and width of both 0.6 m, the corners of slab track will reach a downward warping displacement of 0.484 mm. Compared with the displacement under normal situation (undamaged CA mortar), the downward displacement with damaged CA mortar increased by 72%. In addition, the supporting in the bottom of the track slab becomes uneven because of local mortar failure. Under the effect of train loading, the track slab produces uneven warping deformation with additional bending moment or even reverse bending moment. The overall carrying capacity becomes complicated and adverse. As revealed by researches [10], the influence that the mortar damage under the corner of slab (Fig. 5(a)) and under the middle of slab (Fig. 5(d)) exercises on the longitudinal and transverse tensile stresses of the track slab is relatively smaller, and the mortar damage under the end of slab (Fig. 5(c)) is the most adverse. So in this section, the mortar damage under the end of slab is mainly considered in the following analysis. Fig. 7 shows the influence of the length of mortar damage on longitudinal tensile stresses along the track slab when the train loading and temperature gradient are applied onto the end of slab. As shown in Fig. 7, the longitudinal tensile stress of the track slab will generally grow with the increase of mortar damage length. When the length of mortar damage along the longitudinal direction, l, exceeds 0.3 m, the growth of longitudinal tensile stress of the track slab obviously accelerates. When l exceeds about 0.6 m, the longitudinal tensile stress of the track slab increases to several times than that of the normal and is very likely to be over the characteristic value of tensile strength of slab concrete (about 3 MPa). At this time, the surface of track slab is prone to penetrate crack, which is very bad for the carrying capacity and durability of the track slab. It is worth noticing that the longitudinal tensile

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

(a) slab warping deformation with undamaged mortar

305

(b) slab warping deformation with voids under the corner of slab

Longitudinal tensile stress of track slab, MPa

Fig. 6. Influence of voids area under the corner of slab on the warping displacement of track slab.

7.0

6.2. Influence of mortar damage on the carrying capacity of CA mortar

Train load Train load + positive temperature gradient Train load + negative temperature gradient

6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Length of mortar damage under the end of slab, m Fig. 7. Influence of mortar damage length under the end of slab on the longitudinal tensile stress of track slab.

stress of track slab under the co-action of train loading and positive temperature gradient possessed an initial drop. The reason should include two sides. For one thing, under the single train loading with the position of the end of slab (distance to the end of slab is 0.3125 m), the slab would warp upwards, for another, the slab would warp downwards under the action of positive temperature gradient. The two results will co-work with the increase of mortar damage length. Within a distance of about 0.3125 m, the positive temperature gradient becomes the dominant factor; and over the distance, the train loading becomes the dominant factor. Fig. 8 shows the influence of the width of mortar damage under the edge of slab on the longitudinal and transverse tensile stresses of the track slab when the train loading and temperature gradient are applied under the end or middle of slab. Both the longitudinal and transverse tensile stresses of the track slab basically tend to grow along with the increase of the width of mortar damage under the edge of slab. When the transverse width of mortar damage, w, exceeds 0.3 m, the growth rate of longitudinal and transverse tensile stresses of the track slab basically accelerates. Especially, when w exceeds 0.40 m, i.e. extending to the underneath of the rail, the longitudinal and transverse tensile stresses of the track slab will increase abruptly and obviously. The mortar damage under the edge of slab has relatively larger influence on the transverse tensile stress of the track slab. It is also worth noticing that, only one case of longitudinal tensile stress didn’t follow the variation, which is loaded under the co-action of train loading and positive temperature gradient. This abnormal trend could be also caused by the co-action of the two different loadings in the direction of longitudinal tensile stress.

Theoretical analysis [10] shows that mortar damage will not cause significant changes of the carrying capacity of the CA mortar under the effect of temperature gradient. But under the effect of train loading, mortar damage will directly lead to the reducing of supporting area of the track slab, which is inevitably adverse for the vertical carrying capacity of the CA mortar in the adjacent area of the damage. Fig. 9 shows the influence of mortar damage length or width under the corner and under the end of slab on the vertical compressive stress of CA mortar when the train’s load and temperature gradient are applied to the end of slab. As shown in Fig. 9(a), the vertical compressive stress of CA mortar will increase significantly along with the increase of the length and width of mortar damage under the corner of slab. When the length (l) or width (w) of mortar damage exceeds 0.20 m, the vertical compressive stress of the mortar grows sharply. As a result of a combination of the train loading and the positive temperature gradient, the vertical compressive stress of CA mortar is the largest. When l = 0.6 m and w = 0.6 m, the vertical compressive stress of CA mortar reaches 0.72 MPa. Compared with that of the normal situation, it increases by about 56%. The similar regularity can be obtained from Fig. 9(b) that the vertical compressive stress of CA mortar will increase significantly along with the increase of the length of mortar damage under the end of slab and the trend is relatively linear. As a result of a combination of the train loading and the positive temperature gradient, the vertical compressive stress of CA mortar is the largest. When the length of mortar damage along the longitudinal direction, l, exceeds 0.30 m, the vertical compressive stress of CA mortar exceeds 0.3 MPa, which is equal to the compressive strength of CA mortar at 3 days. When l = 0.9 m, the vertical compressive stress of CA mortar reaches 0.82 MPa, which corresponds to the compressive strength of CA mortar at 7 days. Compared with that of the normal situation, it increased by about 78%. Fig. 10(a) shows the variation of the vertical compressive stress of CA mortar along with the width of mortar damage under the edge of slab when the train loading is applied to the end or middle of slab. Fig. 10(b) shows the variation of the vertical compressive stress of CA mortar along with the width of mortar damage under the middle of slab along the transverse direction when the train loading is applied to the middle of slab. The length of mortar damage, l, under the middle of slab is 1.25 m, namely double of the fastening spacing. As shown in Fig. 10, the vertical compressive stress of CA mortar will increase significantly along with the increase of the width of mortar damage under the edge of slab. When the width of mortar damage, w, exceeds 0.20 m, the vertical compressive stress of the mortar grows significantly. As a result of a combination of the train loading under the end of slab and the positive temperature gradient, the vertical compressive stress of CA mortar is the largest.

6.0

Transverse tensile stress of track slab, MPa

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

Longitudinal tensile stress of track slab, MPa

306

Train load under the end of slab Train load under the end of slab + positive temperature gradient Train load under the end of slab + negative temperature gradient Train load under the middle of slab Train load under the middle of slab + positive temperature gradient Train load under the middle of slab + negative temperaturn gradient

5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

0.1

0.2

0.3

0.4

0.5

5.0

Train load under the end of slab Train load under the end of slab + positive temperature gradient Train load under the end of slab + negative temperature gradient Train load under the middle of slab Train load under the middle of slab + positive temperature gradient Train load under the middle of slab + negative temperaturn gradient

4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.0

Width of mortar damage under the edge of slab, m

0.1

0.2

0.3

0.4

0.5

Width of mortar damage under the edge of slab, m

(a) longitudinal tensile stress

(b) transverse tensile stress

0.8

Vertical compressive stress of mortar layer, MPa

Vertical compressive stress of mortar layer, MPa

Fig. 8. Influence of mortar damage width under the edge of slab on the tensile stress of track slab.

Train load Train load + positive temperature gradient Train load + negative temperature gradient

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0*0

0.05*0.05 0.13*0.13 0.21*0.21 0.29*0.30 0.37*0.38 0.48*0.450.60*0.60

0.9

Train load Train load+positive temperature gradient Train load+negative temperature gradient

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

2

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Length of mortar damage under the end of slab, m

Length and width of mortar damage under the corner of slab, m

(a) mortar damage under the corner of slab

(b) mortar damage under the end of slab

Train load under the end of slab Train load under the end of slab + positive temperature gradient Train load under the end of slab + negative temperature gradient Train load under the middle of slab Train load under the middle of slab + positive temperature gradient Train load under the middle of slab + negative temperature gradient

1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

0.1

0.2

0.3

0.4

0.5

Vertical compressive stress of mortar layer, MPa

Longitudinal tensile stress of track slab, /MPa

Fig. 9. Influence of the length or width of mortar damage on the vertical compressive stress of CA mortar.

0.5

Train load Train load+positive temperature gradient Train load+negative temperature gradient

0.4

0.3

0.2

0.1

0.0 0.0

0.1

0.2

0.3

0.4

0.5

Width of mortar damage under the edge of slab, m

Width of mortar damager under the middle of slab, m

(a) mortar damage under the edge of slab

(b) mortar damage under the middle of slab

Fig. 10. Influence of the width of mortar damage on the vertical compressive stress of CA mortar.

When w = 0.525 m, i.e. the damage extends to the underneath of the rail, the vertical compressive stress of CA mortar reaches 0.85 MPa. Compared with that of the normal situation, it increases by about 84%. The vertical compressive stress of CA mortar tends to

grow along with the increase of the length of mortar damage under the middle of slab. With the combination of the train loading and the positive temperature gradient, the vertical compressive stress of the CA mortar is the largest.

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

307

Fig. 11. Side view of dynamic test on mortar damage of framework-type slab track. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

7. Site test of mortar damages 7.1. Test overview A dynamic test has been conducted on site on the framework-type slab track with damages to CA mortar, which was laid in the slab track test section of the Suining-Chongqing Railway. Fig. 11 shows a site test and the positions of mortar damage are indicted by red circles. An obvious contact loss between the slab and the CA mortar and between the CA mortar and the concrete base can be observed in the left circle, with the length of one point five times of fastener spacing, and the gap size of about 2 mm. A repair work using the repair glue and gap-filling glue was conducted at the right side of the slab, shown in the right circle. The mechanical displacement meter is used to measure the displacements, while the built-in IC piezoelectric acceleration transducer is used to measure the accelerations. The test involves the following cases according to the time sequence of repair: Case I: The CA mortar damage at End A (with the length and width equal to about 0.20 m) is not repaired and the CA mortar damage at End B (with the length and width equal to about 0.20 m) is repaired with SKD803 repair glue. Case II: Both the CA mortar damages at End A and End B are jointed with SKD803 repair glue, and then sealed with SKD801 gap-filling glue. The vertical accelerations of rail and slab, and the vertical relative displacement between the rail and slab as well as the vertical relative displacement between – the slab and the concrete base under the effect of train loading are tested in the two cases at site, and the layout of test points are as shown in Fig. 12. In the test, data acquisition and processing are performed mainly based on the running of CRH1 train, which is operated for passenger trains, with an axle load of 16 t and a test speed of 170 km/h. 7.2. Test results and analysis The sampling rates for displacement meter is 5000 Hz, and the sampling rates for piezoelectric acceleration transducer is

20,000 Hz. The following items are measured during the test: the vertical accelerations (at End A and End B) of the rail, the vertical relative displacements (at End A, End C and End B) between the rail and slab at one side of the track with CA mortar damage, the vertical acceleration responses (at End A, End C and End B) of the track slab and the vertical relative displacements between the slab and the concrete base at one side of the track with CA mortar damage. As the rail directly bears the train loading and the track slab is directly placed on the CA mortar and bearing the dynamic loads from rails and fastenings, it is able to effectively evaluate the influence of CA mortar damage on the train running safety and riding comfort by testing and analyzing the vertical dynamic responses of the rail and track slab. At End A, the track slab has no contact with the mortar or the concrete base when the train passing by. At End B, the damaged CA mortar is repaired by using the repair glue or gap-filling glue, thus the slab has contact with the mortar. In the two cases, the maximum vertical accelerations and the maximum vertical relative displacements of the track slab with CA mortar damage are as shown in Table 2. The relative displacements can be calculated from the absolute displacement test results using the mechanical displacement meter at the speed of 170 km/h. And the maximum accelerations can be directly tested by the piezoelectric acceleration transducer. It can be seen from Table 2 that under the effect of train loading, the vertical acceleration of rail at End A is basically consistent with that at End B in Case I, and the acceleration of rail at End A is only 0.23% more than that at End B. After the CA mortar damages are repaired in Case II, the accelerations of rail at both ends decrease. Compared with the accelerations of rail at the same position in the two cases, the acceleration of rail at End A decreases by 5.47% and that at End B decreases by 4%. Before and after the repair of CA mortar damage, the dynamic response changes of the acceleration of the track slab are consistent with that of the rail. After the CA mortar damage is repaired, the accelerations of track slab are evenly distributed along the longitudinal direction, indicating that the overall carrying capacity of the track slab structure is improved after the repair of CA mortar damage. From Table 2, it is also worth noticing that the slab middle (End C) has a greater relative displacement between rail and slab than the

Fig. 12. Schematic diagram of test point layout for the test on damage to CA mortar.

308

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

Table 2 Dynamic responses of rail and slab under the effect of train loading. Test item

Case

End A

End C

End B

Acceleration of rail (g)

Case I Case II

87.7 82.9

– –

87.5 84.0

Relative displacement between rail and slab (mm)

Case I Case II

0.10 0.14

0.12 0.15

0.10 0.12

Acceleration of track slab (g)

Case I Case II

6.5 5.4

1.7 2.7

5.8 3.9

Relative displacement between slab and concrete base (mm)

Case I Case II

0.16 0.04

0.04 0.04

0.52 0.06

edges of slab (End A and B). It is an indication that a deflection pattern of ‘‘lift up” of the fastening in the slab middle occurs, which is a reflection of the degradation of compression performance of fastening at the edges of slab after the damage of CA mortar. Comparing Case I and Case II, the relative displacement between rail and slab increased, which is further an indication of the improvement of compression performance of fastening spring and the decrease of the fastening ‘‘lift up”. Besides, the reduction of the relative displacement between slab and concrete base implies a remarkable function of the repair material and a better contact status between slab and CA mortar, which accordingly leads to a more stable supporting of CA mortar. Based on the test and analysis of the damage to CA mortar of framework-type slab track, the following conclusions are drawn out: (1) Under the effect of train loading, the vertical accelerations and the relative displacements between rail and slab are not sensitive to CA mortar damage, by comparing the test results at End A and End B in Table 2. (2) After the CA mortar damage is repaired, the vertical accelerations of the track slab tend to be evenly distributed along the longitudinal direction; however, the vertical accelerations of the track slab decrease after the damage repair, and the structural vibration of the track slab tends to be mitigated after repair. (3) From the comparison of the slab with or without the repair of damaged CA mortar (with the length and width of about 0.20 m), it is clear that the accelerations of rail and slab, the relative displacement between the slab and concrete base for the structure with CA mortar damage possessed greater values, and those test results decreased after the repair of CA mortar. This trend explained if the length or width of mortar damage reaches 0.20 m, a clear deterioration of the dynamic performance of track structure can be tested, which reflects a severe repair level of 0.20 m. (4) The positions with CA mortar damage of this railway have been repaired for more than 3 years and the repair effect is currently good based on site detection, identification and repair. 8. Criteria for repairing CA mortar damages According to the forms and characteristics of mortar damage, we can use three parameters, i.e. length (l), width (w) and gap size (h) for description. Among them, l is the longitudinal length of mortar damage along the track; w is the transverse width of mortar damage along the track; h is the gap size between the mortar and the track slab or between the mortar and the concrete base, and generally it refers in particular to the gap size between mortar and the track slab. 8.1. Length and width On-site inspections show that mortar damage occurs mainly under the end and edge of slab. The further the position of the

damage is away from the rail supporting points, the smaller the influence of the damage on the carrying capacity of track structure and train operation. So only the characteristic damage length becomes the dominant factor. Therefore, the mortar damage could be divided into two kinds, the damage along the transverse direction and the damage along the longitudinal direction, namely w P l for transverse damage and w < l for longitudinal damage, as shown in Fig. 13. The transverse damage mainly appears under the end of slab and is often caused by construction defects or excessive loading. The influence of w is relatively smaller, while the influence of the length of l is larger. The researches in reference [10,19] show that even if w is set as the maximum width of CA mortar, when l meets the condition that l 6 0.20–0.30 m, the influence of damage on the carrying capacity of track structure is not obvious. Fig. 14 shows the variation of the vertical static stresses and the fatigue life of mortar along with the change of mortar damage length. Fig. 14(a) illustrates the vertical compressive stress of mortar considering the effect of an axle load of 300 kN. By experimental study, Wang Tao and Wang Fazhou [28,29] got the fatigue equations of low elastic modulus mortar at the temperatures of 20 °C and 20 °C, namely

S ¼ 0:10207lgN þ 1:3388

ð1Þ

and

S ¼ 0:09938lgN þ 1:1343

ð2Þ

in which, S is the stress ratio, S = r/fc;
r is the compressive stress of CA mortar (MPa);
fc is the compressive strength of CA mortar (MPa), fc = 2.26 MPa;
N is the fatigue cycle number. Experiment results [28] show that the lower the temperature, the shorter the fatigue life of mortar. Therefore, a theoretical fatigue life analysis of mortar is made by considering the low temperature condition of 20 °C as the most adverse operating condition, taking ECA = 300 MPa and using the mortar stress obtained in numerical calculation, as shown in Fig. 14(b). In addition, it is assumed that the track carries 20 trains of CRH2 EMU per hour and each train has 8 car units. It is uncertain if the slab track can reach a life span of 60 years [1,2]. However, a life cycle of at least 30 years is desired for the track structures, even for the structures working with some acceptable small damages. As shown in Fig. 14, with the increase of mortar damage length, the vertical compressive stress of mortar goes stronger and the mortar fatigue life decreases sharply. It indicates that the fatigue life of mortar is very sensitive to the change of its vertical compressive stress. It can be also seen that a fatigue life of 30 years corresponds to the damage length of about 0.28 m. Besides, it is clear from Fig. 9(a) that the vertical compressive stress of the mortar grows sharply if the damage length (l) exceeds 0.20 m. When the damage length exceeds 0.20 m, the mortar damage is likely to get fatigue growth and further deterioration.

309

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

(b) w < l

(a) w ≥ l

1.6

ECA=100MPa ECA=200MPa ECA=300MPa

1.4 1.2

Fatigue life, year

Vertical compressive stress of mortar layer, MPa

Fig. 13. Division of mortar damage.

1.0 0.8 0.6 0.4 0.2

720 660 600 540 480 420 360 300 240 180 120 60 0

-20

0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Mortar damage length, m

(a) vertical compressive stress of mortar

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Mortar damage length, m

(b) fatigue life of mortar

Fig. 14. Influence of damage length on the vertical static stresses and fatigue life of mortar.

Therefore, the length of mortar damage, l, should be theoretically kept below 0.20 m. In conclusion, for the transverse damage, the width of mortar damage, w, should not be more than 0.40 m, i.e. extending to the underneath of the rail, while the length of mortar damage, l, should not exceed 0.20 m. When it comes to the longitudinal damage, the influence of the width, w, is relatively larger, while the influence of the length, l, is smaller, which mainly occurs under the edge of slab. Based on the analysis above and the field operation experience [10,23], it can be summarized that the length of mortar damage, l, should not be more than 0.6 m, and the width of mortar damage, w, not more than 0.2 m. 8.2. Gap height Due to excessive loading, the mortar damage is not always completely along the thickness direction, but can also cause the loss of bond or a certain degree of gap between the mortar and the track slab. When the gap size exceeds a certain degree, even with the effect of train loading, it is hard for the track slab in the damage area to have a contact interaction with the surface of mortar, which will give rise to voids underneath the track slab. Take the mortar damage under the end of slab for an example to make a static analysis and assume that the mortar exhibits a complete damage along the thickness direction. Fig. 15 shows the variation of the vertical relative displacement between the track slab and concrete base under the end of slab along with the change of mortar damage length, under the effect of a train loading of 300 kN and with different elastic modulus of mortar. Fig. 15(a) shows the vertical displacement of the track slab and concrete base and the transverse distribution of the relative displacement between them along the track, when ECA = 200 MPa and the mortar

damage length just reaches the position of the first fastener at the slab end. Fig. 15(b) illustrates the variation of vertical relative displacement between the track slab and concrete base with the length of mortar damage changes, when different elastic modulus of mortar is applied. The gap size can be obtained from the change of the relative displacements between the track slab and the concrete base. If we define the relative displacement between the track slab and the concrete base as h0 when the CA mortar is undamaged, the relative displacement between them as h1 when the CA mortar is damaged with a certain length, then the gap size h can be calculated by h = h1  h0. As shown in Fig. 15, the vertical relative displacement between the track slab and concrete base, getting larger and larger along with the increase of mortar damage length, reaches the maximum in the position underneath the rail. When the mortar is undamaged, the vertical relative displacement between the track slab and concrete base h0 indicates the amount of mortar’s vertical compression. However, when the mortar exhibits a damage, it can be inferred that if the relative displacement between the mortar and the track slab in the damaged position h1 exceeds the vertical relative displacement between the track slab and concrete base h0, it would be difficult for the track slab to have contact with mortar under the effect of train loading. Because of the strong bending stiffness of the track slab, when the gap size between the mortar and the track slab exceeds about 1 mm and 2 mm, the mortar damage length should be more than about 0.40 m and 0.60 m respectively, as illustrated in Fig. 14(b), and only in this way can the voids underneath the track slab appear. Therefore, it is suggested that the gap between the mortar and the track slab should not exceed 2 mm; otherwise, voids are very likely to appear underneath the track slab.

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

0.0

Track slab Concrete base Track slab relative to concrete base

-0.2

Vertical displacement, mm

Vertical relative displacement between track slab and concrete base, mm

310

-0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

4.5

ECA=100MPa ECA=200MPa ECA=300MPa

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Transverse position of track slab, m

Mortar damage length, m

(a) transverse distribution of vertical

(b) influence of mortar damage length

displacement Fig. 15. Distribution of vertical relative displacement between track slab and concrete base.

Table 3 Recommended values for the criteria for repairing damages to low elastic modulus cement–emulsified asphalt mortar. Type

Criteria

Spalling

Transverse damage ðl < w 6 0:40 mÞ

Level I Level II Level III

l < 0:10 m 0:10 m 6 l < 0:20 m l P 0:20 m

Longitudinal damage (w < l 6 0:6 m)

Level I Level II Level III

w < 0:10 m 0:10 m 6 w < 0:20 m w P 0:20 m

Gap size h

Level I Level II Level III

h < 1:0 mm 1:0 mm 6 h < 2:0 mm h P 2:0 mm

Loss of bond

Level

8.3. Repair level In general, two types of defect can occur, including spalling of CA mortar and loss of bond at the interface. As shown in Table 3, the theoretical maintenance levels of the two types of damage to CA mortar for the framework-type slab track are preliminarily proposed according to the analysis above, based on the structural stress and fatigue life and considering the domestic and foreign criteria for repairing damages of slab track. From Fig. 8, it can be get that the longitudinal and transverse tensile stresses of the track slab will increase abruptly and obviously, if w exceeds about 0.40 m. From Fig. 9 and the fatigue life calculation, it can be obtained that the vertical compressive stress of the mortar grows sharply if the l exceeds 0.20 m. So for the transverse damage, the width of mortar damage, w, should not be more than 0.40 m, while the length of mortar damage, l, should not exceed 0.20 m. Thus the damage width should not exceed 0.40 m for the transverse damage (l < w), then we have l < w 6 0.40 m. Also, for the transverse damage, the damage length l should not exceed 0.20 m, and consequently we define l P 0:20 m as Level III. In order to expediently divide Level II and Level I, we choose 0.10 m as the critical value, so we have 0:10 m 6 l < 0:20 m for Level II and l < 0:10 m for Level I. Similarly, from Fig. 7, it can be obtained that for the longitudinal damage, the longitudinal tensile stress of the track slab increases to the characteristic value of tensile strength of slab concrete (about

Sketch

h

3 MPa), if l exceeds 0.6 m. From Fig. 9 and Fig. 10, if damage width (w) exceeds 0.20 m, the vertical compressive stress of the mortar grows sharply. Thus, for the longitudinal damage we have w < l < 0:8 m, and we define the worst condition w P 0:20 m as Level III. In order to expediently divide Level II and Level I, we also choose 0.10 m as the critical value, so we have 0:10 m 6 w < 0:20 m for Level II and w < 0:10 m for Level I. For the damage type of the interfacial loss of bond, it is clear from Fig. 14(b) that a gap size of about 1 mm corresponds to a damage length of 0.4 m, and a gap size of 2 mm corresponds to a damage length of 0.6 m. Thus, we define the worst condition h P 2:0 mm as Level III. For convenience in the measurement and repair work, we define 1:0 mm 6 h < 2:0 mm as Level II, and h < 1:0 mm as Level I.

9. Conclusions As a leveling, supporting and damping layer, the CA mortar is a necessity for the prefabricated framework-type slab rack. In this paper, a study on the criteria for repairing damages of the CA mortar was conducted by establishing a 3-D FEM of the slab track on elastic foundation. A field test was carried out in order to verify the numerical analysis results. Based on the limited results in this study, the recommended values for criteria of repairing damages to CA mortar are preliminarily put forward.

J. Ren et al. / Construction and Building Materials 110 (2016) 300–311

(1). As one of the main hazards of slab track, the damage of CA mortar greatly affects the carrying capacity and durability of the track slab and CA mortar, but has relatively less impacts on that of the concrete base. The numerical simulation indicates for the transverse damage, the critical width of damage is about 0.40 m, and the critical length of mortar damage is about 0.20 m; and for the longitudinal damage, the critical length of damage is about 0.6 m, the critical width of damage is about 0.20 m. For the damage type of the interfacial loss of bond, a critical gap size of about 2 mm was deduced. (2). If the actual dimensions of the damage to CA mortar are greater than the critical values, the tensile stress of track slab or the compressive stress of CA mortar will increase significantly and may exceed their characteristic value. And the fatigue damage of CA mortar under the area adjacent to the damage would also be accelerated. (3). The comparison of the slab with or without the repair of damaged CA mortar in the field test also explained the trend that if the length or width of mortar damage reaches a critical value of about 0.20 m, a clear deterioration of the dynamic performance of track structure can be tested. Thus the field test could be correlated with the calculation results. In summary, slab track is a kind of track structure needing ‘‘less repair”, but not ‘‘free of repair”. With advanced repair theory as a guide, the lack of repair and eliminate excessive repair can be avoided, so as to realize an optimal repair. Therefore, we need to enhance the state assessment of slab track system, define the reasonable repair cycle and standards and establish a scientific repair system, which is aimed at maintaining the stable and reliable state of the slab track system during its safe service life. Acknowledgments The authors thank China Railway Corporation (Grant No. 2015G001-F), and National Natural Science Foundation of China (Grant Nos. 51278431, 51578472) for funding part of this work. The useful contribution and discussions from project partners are also acknowledged. References [1] Juanjuan Ren, Rui Xiang, Xueyi Liu, Force characteristics of longitudinal coupled slab track turnout on bridges under temperature action, Transp. Res. Rec.: J. Transp. Res. Board 2159 (2010) 85–90. [2] Xiao-lin Song, Wan-ming Zhai, Shao-lin Wang, Analysis of vertical displacement distributions of ballastless track infrastructure of high-speed railways, J. Civ. Eng. 45 (5) (2012) 163–168 (in Chinese). [3] X.Y. Liu, P. Wang, Vehicle-Track-Subgrade Coupled Dynamics, Southwest Jiaotong University Press, Chengdu, 2011 (in Chinese). [4] X.H. Zeng, Y.J. Xie, D.H. Deng, Conductivity behavior of the fresh CA mortar and its relationship with the fluidity properties, Constr. Build. Mater. 36 (2012) 890–894.

311

[5] T. Rutherford, Z.J. Wang, X. Shu, B.S. Huang, D. Clarke, Laboratory investigation into mechanical properties of cement emulsified asphalt mortar, Constr. Build. Mater. 65 (2014) 76–83. [6] Ping Wang, Hao Xu, Rong Chen, Effect of cement asphalt mortar debonding on dynamic properties of CRTS II slab ballastless track, Adv. Mater. Sci. Eng. 2014 (2014) 1–8. [7] Y.J. Xie, Q. Fu, K.R. Zheng, Y. Qiang, S. Hao, Dynamic mechanical properties of cement and asphalt mortar based on SHPB test, Constr. Build. Mater. 70 (2014) 217–225. [8] S.H. Jin, X.F. Chen, J. Yang, Key technologies of CA mortar for slab track, China Railway Sci. 27 (2) (2006) 20–25 (in Chinese). [9] Z. Zeng, Study on The Damage of Slab Ballastless Track Structure and The Railway Maintenance System for High-speed Railway, Southwest Jiaotong University, Chengdu, 2009 (in Chinese). [10] P.G. Li, X.Y. Liu, G.Q. Li, Influence of CA mortar void on dynamic characteristics of unit slab track on bridge, China Railway Sci. 35 (3) (2014) 20–27 (in Chinese). [11] S.Y. Zhu, C.B. Cai, Interface damage and its effect on vibrations of slab track under temperature and vehicle dynamic loads, Int. J. Non-Linear Mech. 58 (2014) 222–232. [12] J. Xiang, D. He, Q.Y. Zeng, Effect of cement asphalt mortar disease on dynamic performance of slab track, J. Cent. S. Univ. (Sci. Technol.) 40 (3) (2009) 791–796 (in Chinese). [13] J.C.O. Nielsen, A. Igeland, Vertical dynamic interaction between train and track-influence of wheel and track imperfections, J. Sound Vib. 187 (5) (1995) 825–839. [14] Shuguang Zhang, Xinbiao Xiao, Zefeng Wen, et al., Effect of unsupported sleepers on wheel/rail normal load, Soil Dyn. Earthquake Eng. 28 (2008) 662– 673. [15] J.J. Zhu, A.K.W. Ahmed, Subhash Rakheja, Amir Khajepour, Development of a vehicle-track model assembly and numerical method for simulation of wheelrail dynamic interaction due to unsupported sleepers, Veh. Syst. Dyn. 48 (12) (2010) 1535–1552. [16] Jin Shi, Andrew H. Chan, Michael P.N. Burrow, Influence of unsupported sleepers on the dynamic response of a heavy haul railway embankment, Proc. Inst. Mech. Eng. Part [F] 227 (6) (2013) 657–667. [17] Z.W. Lu, Overall comments on track technology of high-speed railway, J. Railway Eng. Soc. 1 (1) (2007) 41–53 (in Chinese). [18] X.C. Chen, Modern Maintenance Theory, National Defense Industry Press, Beijing, 2010 (in Chinese). [19] K.F. Liu, Research on Cement Asphalt Mortar Damage and Maintenance Standards for frame Type of Slab Track, Southwest Jiaotong University, Chengdu, 2013 (in Chinese). [20] W. Zhao, Study on Damages and The Longitudinal Forces of Element Slab Ballastless Track, Southwest Jiaotong University, Chengdu, 2008 (in Chinese). [21] F.Z. Wang, Z.C. Liu, S.G. Hu, Early age volume change of cement asphalt mortar in the presence of aluminum powder, Mater. Struct. 43 (2010) 493–498. [22] Y.J. Mei, Z.Y. Li, N.X. Liang, Fatigue life of asphalt concrete under moisture– temperature–loading coupling action, J. Hunan Univ. (Nat. Sci.) 38 (2) (2011) 13–17 (in Chinese). [23] O. Kazuyosi, I. Yosio, Compressive strength of the CA mortar and its temperature-susceptibility, Memoirs Fac. Technol. 10 (2) (1976) 1–13. [24] Y. Harada, S. Tottori, N. Itai, Development of cement–asphalt mortar for slab tracks in cold climate, Q. Rep. RTTI (Railway Technical Research Institute) 15 (1) (1983) 62–67. [25] X.Y. Liu, P.R. Zhao, R.S. Yang, P. Wang, Research on the Design Theory and Method for Ballastless Track on Passenger Dedicated Line, Southwest Jiaotong University Press, Chengdu, 2010 (in Chinese). [26] E.S. Barber, Calculation of maximum pavement temperature from weather reports, Highw. Res. Board Bull. 168 (1957). [27] G.Z. Xie, H. Yuan, Z.K. Yao, The maximum temperature gradient of cement concrete pavement, East China Highway 6 (1982) 9–20 (in Chinese). [28] T. Wang, Research and Application on CA Mortar in Ballastless Slab Track of High Speed Railway, Wuhan University of Technology, Wuhan, 2008 (in Chinese). [29] F.Z. Wang, Z.C. Liu, Research on the fatigue behavior of CA mortar used in ballastless slab track of high speed railway, J. Wuhan Univ. Technol. 30 (11) (2008) 79–81 (in Chinese).