The use of elastic elements in railway tracks: A state of the art review

Construction and Building Materials 75 (2015) 293–305

Contents lists available at ScienceDirect

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

Review

The use of elastic elements in railway tracks: A state of the art review Miguel Sol-Sánchez ⇑, Fernando Moreno-Navarro, Mª Carmen Rubio-Gámez Laboratorio de Ingeniería de la Construcción de la Universidad de Granada, C/Severo Ochoa s/n, 18071 Granada, Spain

h i g h l i g h t s  An analysis of elastic elements to improve railway track behavior and durability.  A review of studies about rail pads, under-sleeper pads and under-ballast mats.  The effect of diverse parameters on elastic elements behavior is described.  Recommendations about stiffness of elastic elements have been drawn.

a r t i c l e

i n f o

Article history: Received 21 July 2014 Received in revised form 3 October 2014 Accepted 12 November 2014

Keywords: Railway Elastic elements Rail pads Under-sleeper pads Under-ballast mats Review

a b s t r a c t Railway is envisaged as the transportation mode of the future, but in spite of its advantages, its development is not exempt from technical difficulties that lead to track deterioration. To overcome these drawbacks, research in this field needs to be developed. Geometry degradation, as well as noise and vibration, have been identified as problems that need to be reduced, which could be possible by modifying track vertical stiffness and obtaining a more homogeneous value along the track. One measure to minimize these problems involves the installation of elastic elements (e.g. rail pads, under-sleeper pads, and under-ballast mats) in the railway track. In fact, this has now become the most effective means to vary track vertical stiffness as well as to abate noise emission and vibrations caused by the passage of trains. This paper discusses the problems associated with track stiffness, geometry degradation, and vibrations, and at the same time, studies the characteristics of elastic elements as well as the research carried out to test and evaluate their effectiveness. After reviewing and analyzing a wide range of research initiatives, this paper proposes a set of recommendations and guidelines for the use of elastic elements in railway infrastructure as well as highlighting a series of possible further investigations. Ó 2014 Elsevier Ltd. All rights reserved.

Contents 1. 2.

3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main problems of railway tracks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Track deterioration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Vibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1. Ground vibrations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2. Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Relevance of vertical stiffness of the track . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Elastic elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Rail pads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1. Main characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.2. Cases and studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Under sleeper pads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1. Main characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2. Cases and studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Under ballast mats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

⇑ Corresponding author. Tel.: +34 958249445; fax: +34 958246138. E-mail addresses: [email protected] (M. Sol-Sánchez), [email protected] (F. Moreno-Navarro), [email protected] (M.C. Rubio-Gámez). http://dx.doi.org/10.1016/j.conbuildmat.2014.11.027 0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.

294 294 294 295 295 296 296 296 297 297 297 299 299 299 301

294

4.

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

3.3.1. Main characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2. Cases and studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Since their invention, railway trains have become one of the most popular transportation modes in the world. Their freight capacity, high efficiency and functionality, combined with their minimal environmental impact, have made trains one of the most frequently chosen options for transporting people and merchandise from one place to another. Moreover, high-speed train travel has transformed the railway into one of the most attractive transportation modes with features that significantly enhance rapid communication between cities. Such properties make high-speed trains preferable to other modes, especially on busy routes with a heavy transportation demand [1]. Although railway transport has advantages, the increase in rail freight being transported and train velocity make it not exempt from technical difficulties, and the study and research for finding new solutions is essential. The increase in speed and the load transported has incurred higher forces on the track as well as the increase of the noise and vibrations caused by the trains [1,2]. In addition, the higher speed leads to an increase in the dynamic overload that may accelerate the track deterioration, a problem which is particularly marked on tracks with inappropriate values of vertical stiffness [3,4], making it necessary to obtain an optimal global stiffness of the infrastructure. Consequently, if railway systems are to continue to grow in socioeconomic importance, finding solutions for the negative effects of those technical difficulties is imperative since this will undoubtedly facilitate the future development and evolution of railway transportation method. In this regard, the most frequent measure taken to reduce stresses on railway tracks and to abate noise emissions and vibrations is the incorporation of elastic elements into railway tracks [5]. The purpose of this is to improve track performance and overcome problems stemming from highspeed train traffic. Broadly speaking, elastic elements used in ballasted tracks fall into three categories: (i) rail pads (installed between rails and sleepers); (ii) under-sleeper pads (embedded beneath sleepers); and (iii) under-ballast mats (installed on the granular layer in the case of ballast tracks, and underneath the slab in the case of slab tracks). This paper provides a review of the technical properties of elastic components within the context of recent research and experience initiatives. The first section briefly describes the main problems on railway tracks which could be mitigated or reduced by using elastic elements. The following section studies the main types of elastic component, such as rail pads, under-sleeper pads, and under-ballast mats, and analyzes their most salient features. It also explains the way in which the installation of these components modifies the properties and parameters of railway infrastructure. 2. Main problems of railway tracks There are different types of damage in the railway system (deterioration of the geometrical quality of the track, defects on the track surface, settlement of the granular layers, fatigue of materials, etc.), which need to be identified and analyzed in order to develop specific solutions to reduce track deterioration and maintenance costs, while achieving higher durability of the

301 302 303 303

infrastructure [6]. In addition, the emission of air-born noise and the propagation of waves through the ground constitute serious social and environmental problems that should be studied and reduced. In this regard, the importance of track stiffness in terms of both its long-term performance and the reduction of other modes of track deterioration should be taken into account. 2.1. Track deterioration According to Nielsen et al. [7] and Teixeira [8], there are various types of deterioration of the track quality in reference to the damage source, causing different modes of failure. The deterioration modes could be associated with the track component degradation: the service life of rails, sleepers and fastener systems plays an essential role in the railway infrastructure since their failure could cause train derailment as well as important maintenance costs. Nonetheless, these components, due to the materials which they are made of, have a high fatigue strength and durability. Thus, when they reach the end of their service life, the best solution is to replace them [9]. Granular layers also present a further progressive deformation with the passage of trains, causing an accumulative deterioration of the track geometry, which is a fundamental parameter, especially on high-speed lines. This failure mode is due to the settlement of granular layers as a consequence of the loss of contact between particles or the breakage of them caused by the repeated dynamic loads. The granular layer settlement is equal to the sum of the deformation of the diverse layers used in order to distribute the loads transmitted to the sub-layers. Selig and Waters [10] showed that the ballast is the layer with highest contribution to the track settlement (up to 50–70% of the total vertical deformation). Fig. 1 shows an example of ballast settlement in a railway track. Various authors [11,12], using measurements from French and Japanese railway tracks, have shown that the ballast settlement accords with an exponential law, with the greater deformations occurring at the beginning of the track service (due to low material compaction), followed by a progressive vertical deterioration. On the other hand, some authors [10] have found that there is a lineal relationship between the mean settlement and the differential vertical deformation of the track, which indicates that an increase in ballast deterioration can lead to important degradation of the track geometry (longitudinal and transversal levelling, alignment and buckles). In order to reduce these problems, there are various possible solutions, such as the use of ballast particles with high performance; an increase in the number of maintenance tasks; or finally the replacement of the granular material. However, these solutions can lead to important increases in social, economic and environmental costs due to the necessity of using more specific material (longer hauling distance, high consumption of raw materials, etc.) as well as a higher quantity of ballast during maintenance and rehabilitation tasks, whose frequency is also increased. Given these problems, a more efficient solution is the reduction of load transmitted to the ballast layer in order to obtain lower settlement, and therefore, higher durability of the railway tracks. For this purpose, elastic elements are used since their resilient behavior allows for a reduction of the stress on the ballast layer at the same time that the contact area between components is increased.

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

295

Fig. 1. Visual apparence of the effect of ballast settlement on railway track [6].

Fig. 2. Rail corrugation [21].

Fig. 3. Visual apparence of auxiliar rails used to reduce the effect of track transition [22].

2.2. Vibrations The vibrations generated by the trains cause two distinct effects: noise, and track vibration [13,14]. Given the importance that these effects can have on the deterioration of the tracks and on the wellbeing of residents living near to railways, the propagation of waves has been the focus of attention since the beginning of the railway system, with studies from South [15] and Hyde and Lintern [16]. 2.2.1. Ground vibrations The vibration waves that are transmitted through the railway track and the underlying ground are generated by the passing of trains, there being 3 distinguishable sources:

 The effect of vehicle speed since it generates dynamic loads which are applied through an elastic half space, causing a periodical movement of the track components. The higher the train speed, the higher the frequency of vibrations and the greater the effect of the propagation of waves [17]. In lines where the train speed is higher than 250–300 km/h, studies in track dynamics are required.  Irregularities at the wheel–rail contact, which can be divided into 2 groups (long and short waves) depending on their vibration frequency [18]. Long waves (k > 300 mm) are mainly caused by rail waves and deformations of several meters due to rail manufacturing or the effect of ‘‘out-of-roundness (o-o-r)’’ [19]. Short waves caused by impact loads due to the presence of multiple singularities in wheel–rail contact such as wheelflats, rail joints or turnout crossings [7,20]. In addition, rail corrugation (Fig. 2) is a source of vibrations whose frequency depends on the characteristics of the rail deformation; the main solutions for this problem consist of the rail and wheel maintenance like rail grinding, although the stiffness of the rail pads used under the rail have an important effect also [21].  Irregularities in the track which can be caused by several phenomenon such as differential settlements of the ballast layer, sleeper spacing, stiffness variations (as track transitions), heterogeneities of subgrade soil, and others. All of them provoke an effect that increases vibrations amplitude, and therefore, track deterioration. This causes an increase in damage to the infrastructure, leading to a relentless degradation process. Due to the effect of this type of irregularities, diverse authors have proposed different solutions, focusing on reducing track transitions by the incorporation of auxiliary rails (Fig. 3) [22] or the most common technique consisting of elastic elements with different properties [23]. These different vibration sources cause acceleration (on the vertical, transversal, and longitudinal plane) of the different components of the railway infrastructure [24]. The vertical vibration leads to the fatigue of elements such as the fastener system, as well as the ballast expansion that could take place, which is accentuated when critical levels of particle acceleration are reached (1.4 g and 1.6 g), leading to ballast liquefaction [25,26]. In addition, when vertical vibration waves are transmitted through soft soil, energy can accumulate underneath the weight of the train and cause the known phenomenon of resonance when vibration waves oscillate at train speed, leading to the rapid deterioration of the infrastructure [14,27]. Thus, track stiffness can have a very important role to play in the generation of vertical ground vibrations. As a consequence, diverse solutions, like elastic elements, have been developed to vary the track stiffness [28,29] with the aim of reducing vibrations. With respect to lateral vibrations, they mainly produce an effect known as corrugation, which manifests itself as a periodic irregularity or waviness that develops on the rail. This corrugation effect

296

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

causes a periodic application of impacts that produce a cyclical process of structural damage, loud noise, uneven movement, and deterioration of the track and train cars [30,31]. These effects can be reduced by using elastic elements with appropriate stiffness, according to a study developed by Egana et al. [21]. The third possible vibration type (longitudinal movements) is best interpreted as compression waves within the rails, which can lead to fatigue in rail axes, and can also loosen the fastening system [24]. 2.2.2. Noise Trains are regarded as one of the most environmentally-friendly modes of transportation. Nevertheless, the noise generated by train traffic is a serious environmental problem, particularly at locations with nearby residential areas that are exposed to this noise [32]. There are basically three types of railway noise: (i) traction and auxiliary noise; (ii) aerodynamic noise and (iii) rolling noise, the latter two being the most predominant types. Aerodynamic noise is known to increase faster with the train speed than rolling noise, so this type of noise could be the most predominant on high-speed lines. Aerodynamic sound is caused by unstable air currents that take place due to the separation between railcars, bogies, traffic lights, etc. On the other hand, structural noise caused by train rolling can be predominant in conventional railways, being emitted by the mechanical vibrations of the wheels over a rough, irregular surface (Fig. 4) [33] and has a high frequency range (over 1500 Hz). In contrast, the noise emitted by sleepers is generally in a medium frequency range (500– 1500 Hz), whereas the noise associated with the ballast layer is in a lower frequency range [32]. Thus, it would be necessary to have different solutions, depending on the rank of frequencies in which the acoustic transmissions need to be reduced. It is also necessary to consider that the stiffness of the structure in general, and its components in particular, has a great impact on the noise transmitted by the trains. Thus, Lichtberger [35] compared the noise produced on a slab track with the noise produced on a ballast track and concluded that the track over concrete produces an increase of 5 dB (particularly between frequencies of 250 Hz and 1000 Hz) over the noise produced on a ballast track, given the absence of the granular layer (shock-absorbing element) in the slab track and the difference in the stiffness between the different railway infrastructures. Following the study of the effects and sources of vibrations, it is understood that the use of elastic components could be an effective

solution to mitigate vibrations, since these elements, due to their resilient characteristics, allow damping movements, and therefore, vibrations at the same time that the can modify the track stiffness (an important parameter in noise and waves propagation). 2.3. Relevance of vertical stiffness of the track The vertical stiffness of the track structure has an impact on the dynamic overloads, given that Prud´homme [36] demonstrated that the increase of this parameter also increases the demands on the railway track. This could speed up the deterioration of the track geometry and components, as well as increase the level of vibrations and noise transmitted by the passage of trains. Thus, some authors [37–40] confirmed the advantages of reducing the stiffness of the track given that it will decrease the vertical effort transmitted to the infrastructure. However, another study by Fortin [41], concerned with the high-speed line between Paris and Lyon, showed that a big decrease in the track stiffness could produce an increase in the resistance to forward movement of the vehicles and therefore, of the energy consumed by them. In addition to this, the decrease in the track stiffness could increase the settlement of the track and the fatigue of its elements thus increasing the deterioration of the track [42]. Consequently, various authors [10,43] have focused their studies on analyzing how much the track stiffness could be reduced without affecting other parameters of the mechanical performance of the infrastructure. Lopez Pita [44] established an optimal value of vertical stiffness around 50–78 kN/mm (taking into account the dynamic loads and the dissipated energy), and Teixeira [8] fixed this value at 70–80 kN considering the energy costs associated with exploiting the track and its deterioration (maintenance costs). This latter rank for optimal stiffness of the system was established after the maintenance studies for the Paris–Lyon and Madrid–Sevilla railways [23]. To obtain an optimal value of stiffness, the most common solution is to incorporate elastic elements with different properties into the railway infrastructure, given that the modification of the stiffness of other components such as rail, sleepers or ballast can cause the deterioration of the stability and resistance of the track [8]. In addition, the elastic elements allow for a reduction in the variations of the stiffness along the track [23,45] as well as obtaining gradual changes of the stiffness between sections with different type of structures (e.g. between embankment and concrete structures, or between ballasted track and slab track). This could not only reduce the effect of dynamic overloads, but also the extent of track degradation and intensity of the vibrations. 3. Elastic elements

Fig. 4. Scheme of rolling noise generated by rough contact between rail and wheels [34].

With the increase in rail freight traffic and train speed, many countries have incorporated elastic elements into their railway systems as standard practice [24]. Such elements modify the stiffness of the track and mitigate phenomena such as ballast liquefaction, noise emission, and wave propagation. This is made possible by the fact that elastic components can be manufactured with different stiffness levels as well as a high damping capacity. The polymeric nature of these elements means that they are lightweight, highly resistant, corrosion-proof, and easy to mould. Nevertheless, one of the main problems of elastic elements used in railroads is the deterioration produced by environmental agents such as temperature, oxidation, or hydrolysis. For this reason, they have a useful life of approximately 20 years [45,46]. There are many types of elastic components (embedded rail, elastic pad under rail fastenings, elastic under-sleeper pads,

297

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

geogrids, etc.) which are not only intended to reduce vibrations and stress on components, but are also used as track reinforcement in the same sense that elements like geogrid are employed in other infrastructures [47]. However, the most commonly used devices are rail pads, under-sleeper pads, and under-ballast mats, which can distribute loads and reduce the noise emissions and vibrations stemming from the movement of rails, sleepers, and ballast, whilst at the same time mitigate the impact of these elements on each other [45,48]. 3.1. Rail pads 3.1.1. Main characteristics Resilient elastic pads are generally installed beneath train track rails since it became standard practice to make sleepers out of concrete. Rail pads are normally made out of rubber, high-density polyethylene (HDPE), thermoplastic polyester elastomer (TPE), and ethylene vinyl acetate (EVA) [49,50]. Nonetheless, over the last few years, new elastic elements have been developed from alternative materials, with emphasis upon those made from used tires [51,52]. As a general rule, these pads (see Fig. 5) come in various designs in order to better adapt to the railway system, and can thus range in thickness from 4.5 to 15.0 mm. With regard the horizontal geometry, rail pads are usually 180 mm long and 140 mm wide under rail type UIC 54, and 180 mm long and 148 mm wide under rail UIC 60. The use of this component improves load distribution, which means a smoother ride, and a better conservation of the superstructure. Furthermore, rail pads provide electrical insulation (between track rails) and damp the vibrations that the rail transmits to the sleepers. This prevents the concrete from cracking and reduces ballast wear. The characteristic parameter of the rail pads is their vertical static stiffness. Given that these materials have a non-lineal performance, in order to study its influence the value of the tangent stiffness is usually taken into account for the design load, or the

Fig. 5. Types of rail pad [49,53].

value of the secant between the minimum and maximum load design [42]. In addition, when modeling of the pad behavior is required, nonlinear models that follow the hysteretic curve are recommended due to the non-lineal performance of these materials [54]. As previously mentioned, rail pads come in a wide range of stiffness (k, kN/mm), which means that they can be used in trams and light railway systems as well as in infrastructures with heavy axle loads. Thus, their secant vertical stiffness is the main parameter to characterize the behavior of rail pads. However, the materials classification according to this parameter can be different depending on the literature reviewed, since their properties vary in relation to the track characteristics and the main aim of the application of the rail pads. In this sense, Table 1 shows some examples of rail pad classifications, proving the differences depending on the consulted bibliography [9,55]. In the present paper, the classification proposed by López Pita [9] will be used in order to study the effect of rail pads in the railway system. 3.1.2. Cases and studies In certain European countries, such as Germany and Spain, the first high-speed lines had stiff rail pads (approx. 400–500 kN/ mm) which, in combination with an increased number of foundation layers, made the stiffness of the system excessive. This factor was responsible for the increased deterioration of the ballast particles. This experience led to the use of rail pads that were more resilient with stiffness values of less than 60–100 kN/mm [56]. This tendency to decrease the stiffness of the rail pads also occurred in France (Paris-Lyon railway), where the elastic elements with 4.5 mm of thickness and 150 kN/mm of stiffness were replaced by 9.0 mm thick rail pads and 90 kN/mm of stiffness in order to reduce the global vertical stiffness. Therefore, this last example demonstrates the influence of the thickness of the rail pads. This trend toward reducing the stiffness of rail pads was confirmed by the results of research carried out after cracks appeared in the concrete sleepers of a Greek railway line after fewer than 15 years of service [57]. The evaluations showed that the use of stiff rail pads increased stresses on the sleepers, causing them to crack. Results showed that replacing stiff rail pads (250 kN/mm) with more flexible ones (40 kN/mm) reduced stresses transmitted to the sleeper by up to 20% (in a ballasted track). Further, to determine the capacity of the rail pads to attenuate impacts between rail and sleepers, a study from the University of Wollongong [58] evaluated the energy dissipated by the HDPE pads against the impact of a 6 m high mace. From these results, they concluded that these elastic elements can attenuate up to 50% of the energy transmitted by an impact load, and the remainder is absorbed by the vehicles suspension, the fastening system, and the ballast particles. Similarly, Carrascal et al. [59] showed that the use of TPE pads with 7.0 mm of thickness (stiffness equal to 100 kN/mm) lead to an attenuation of 50% of the impact loads, reducing sleeper deformation when the rail pad stiffness is decreased. Fig. 6 displays the influence of pad stiffness on the capacity of impact attenuation, showing that softer pads lead to lower sleeper deformation on its top and bottom surface. Regarding the corrugation phenomenon, studies focused on lateral movements have shown that the rail pad stiffness is one of the

Table 1 Rail pad classification. Differences depending on the consulted bibliography [9,54]. Author

Author 1 [9] Author 2 [54]

Type of rail pads depending on their vertical secant stiffness, k (kN/mm) Soft

Medium

Stiff

Very stiff

Extremely stiff

<80 130

80 < k < 150 410

>150 1300

– 4100

– 13,000

298

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

Fig. 6. Effect of rail pad stiffness on the attenuation capacity for load impacts [58].

most influential parameters, given that they limit the movement of the track as well as make the contact between the wheel-track less aggressive [60]. Egana et al. [21] measured rail deformations before and after replacing pads with a stiffness equal to 90 kN/mm by 60 kN/mm pads. They found that the last type of pads allowed for a reduction in the length and amplitude of rail waves of up to 55% in comparison with the use of stiff pads. Moreover, the pad stiffness also has an important influence on variations of track stiffness. For example, the transition between sections with 40 kN/mm of stiffness to others with 80 kN/mm could lead to an increase of up to 40% in the stress over the ballast layer. However, the decrease of rail pad stiffness below 100 kN/mm could markedly reduce the changes in track behavior [8,9]. In the evaluation of the effect of rail pads on noise emission, these elements are employed as discrete absorber of noise, and therefore specific formulations should be used to predict the effect of rail pads, according to diverse authors [29,32]. Based on these models, Wu and Thompson [60] affirm that even though soft rail pads (68.8 kN/mm, while the rest of pads studied had stiffness values equal to 270 kN/mm and 1,190 kN/mm) increase the movements and vibrations in rail as well as increasing noise from wheel–rail contact up to 3 dB (A), this is compensated by the fact that the sleepers emit less noise since their movement is reduced. However, when stiff rail pads are used, precisely the opposite occurs. More specifically, the noise made by the sleepers increases, whilst the noise made by the rails decreases [55]. The effect of pad stiffness on noise is shown in Fig. 7. The influence of rail pads on sound level variations particularly affects frequencies of up to 250 Hz, 450 Hz, and 800 Hz in the case of pads with low stiffness, medium stiffness, and high stiffness, respectively [61]. Similarly, Leykauf and Stahl [62] showed that the reduction of rail pad stiffness resulted in a lower speed of the vibration of ballast particles. It was demonstrated by measurements recorded in a conventional German line where the most favorable behavior was presented by a section with rail pads with stiffness equal to 27 kN/ mm, followed by 60 kN/mm pads, and finally 500 kN/mm pads. In addition, it was seen that the rail pad stiffness had a greater effect at the frequency between 16 and 250 Hz, since differences near 90% were recorded in that frequency range. With respect to general track behavior, it is known that the use of softer rail pads produced larger rail deflection, which could lead to the fatigue of this component or others such as the fastener system. Nonetheless, they lead to a more even distribution of stiffness throughout the railway system. In contrast, it was found that stiff

Fig. 7. Influence of pad stiffness on noise generation [34].

pads cause greater dynamic actions on the infrastructure and ballast material. However, on the more positive side, they have a longer service life and reduce rail vibrations. Thus, from the experiences and studies recorded in this article, Table 2 shows a scheme of the optimal field of application of the rail pads according to their vertical static stiffness. On the other hand, due to the importance of the rail pad stiffness in the behavior of the railway track, some authors have focused their studies on analyzing the effect of various parameters on pad stiffness. Thus, Carrascal et al. [63] evaluated the influence of temperature in the mechanical performance of TPE (thermoplastic elastomer) pads. They found that an increase of 30 °C could lead to a reduction in the material stiffness of over 30%, this effect being more notable in the static response of the material. This trend is also seen when the load frequency is increased [59], showing that dynamic loads cause stiffening (up to 25%) of the material in comparison with static loads. In stiffness tests for elastic elements, the fixed preload value is more important than the frequency value, since the pad becomes stiffer mostly when the preload is increased and only slightly when the load application frequency is raised. Some studies [64] have proved that an increase in preload of 20 kN can produce an increase in dynamic stiffness up to 50%, according to Fig. 8.

Table 2 Field of application of rail pads according to its stiffness. Field of application Reduction of damage in sleepers Decrease in the stress transmitted to sublayers Impact attenuation Reduction in corrugation Decrease in rail deflection. Lower energy consumption Reduction in stiffness changes Lower rail movements. Longer life of fastener system Reduction in rail vibrations Lower level of noise from wheel–rail contact Reduction in sleeper and ballast vibration

Stiff pads

Soft pads U U U U

U U U U U U

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

299

3.2. Under sleeper pads

Fig. 8. Effect of preload values on the dynamic stiffness of rail pads [63].

However, the loss factor is practically independent of the preload applied and only increases very slightly with the load frequency [65,66]. Furthermore, Carrascal Vaquero [49], who studied the dynamic response of TPE pads, points out that when evaluating the performance of these materials, it is necessary to take into account that with the repetition of loads, pad stiffness could increase up to 18% and at the same time the dissipated energy falls to approximately 40% (a change that is further accentuated at temperatures of 40–60 °C). The increase in stiffness and reduction in capacity to dissipate energy during the fatigue process has also been demonstrated by other authors [67] who have analyzed the behavior of rail pads manufactured from end-of-life tires. In addition, they found that the stiffness (static and dynamic) of rubber pads fits with a power law in reference to the pad thickness, which showed that the increase in thickness can lead to an important reduction in material stiffness. The evaluation of the material stiffness during its service life is also important. Thus, some authors [68] have analyzed the influence of different deterioration processes (physical–chemical and mechanical actions that can take place on railway tracks) on rail pads made of TPE (ThermoPlastic Elastomer). It was confirmed that such environmental factors cause the pads to progressively increase in stiffness, a value that can reach 33–41% for a service life of 1–3 years, respectively. Stiffness variation in rail pads has also been evaluated by testing equipment developed by the University of Wollongong in Australia (a non-destructive testing methodology based on the vibration response to frequencies of 0–1.000 Hz) [69]. In relation to HDPE rail pads (5.5 mm thick), the annual rate of deterioration of the dynamic stiffness and the damping constant was found to be approximately 12 kN/mm and 108 Ns/m, respectively [70,71]. Table 3 shows the ways in which various parameters can vary the stiffness of rail pads.

3.2.1. Main characteristics Elastic pads between the ballast and sleepers are a popular solution since they can reduce the load and vibrations from concrete sleepers, whose damping power is not sufficiently potent from a structural viewpoint. Under-sleeper pads (USP) have been used for over twenty years, though a wider utilization of these pads has been developed with the construction of high speed railway tracks. They are installed to reduce the track stiffness as an alternative to increase the thickness of the ballast or sub-ballast layers, which makes the compaction of this layer more difficult [45]. These elastic pads (see Fig. 9), which are installed under the sleepers, are normally made of polyurethane elastomers, rubber, and EVA [72,73] as well as elastic waste material could be used [74]. USP have a thickness of 10–20 mm and are usually 1 m long (with the exception of pads for diblock sleepers where the length is similar to that of the sleeper block) while their width depends on the sleeper geometry (typically close to 20–30 cm) [2,17,75]. Under-sleeper pads usually have two layers: (i) a cellular layer inserted inside the sleeper, thus providing it with the necessary damping characteristics, and (ii) an elastic layer on the lower part of the sleeper that protects the hard cellular layer from the damage caused by repeated impacts with the ballast. The use of these pads helps to distribute the load transmitted by the rail between a greater number of sleepers. At the same time, the design and installation of various types of sleeper pad makes it possible to even out differences in stiffness along turnouts, as well as transition zones between building and construction works and natural soil [48]. Unlike rail pads, there is currently no standard classification of under-sleeper pads. However, certain authors, such as Witt [2], categorize this component in terms of its vertical stiffness as follows: soft (50 kN/mm); medium (400 kN/mm); and stiff (3.000 kN/mm). However, the parameter used to form the undersleeper pads is the static bedding modulus C (N/mm3), which is obtained as the material stiffness per unit area relating to the area of the USP [76]. According to this parameter they can be classified as stiff (0.25–0.35 N/mm3), medium (0.15–0.25 N/mm3), soft (0.10–0.15 N/mm3) and very soft (less than 0.10 N/mm3). From the data collected by the UIC (International Union of Railways), the medium and hard under-sleeper pads can be appropriated to improve the track quality, to reduce the stiffness between consecutive sections, to reduce the thickness of the ballast layer, and to avoid the contraction of the track; whereas the soft pads are more suitable for reducing the vibrations caused by the trains. 3.2.2. Cases and studies One of the first experiences with elastic pads under sleepers was developed on the line between Tokaido and Shinkasen in the 70s. The USP used had a vertical stiffness close to 68 kN/mm. It was seen that these elements allowed for a reduction in the ballast

Fig. 9. Under-sleeper pads [48].

300

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

vibration of 22% as well as decreasing the stress on the granular layers, thereby demonstrating the effectiveness of USP in reducing ballast deterioration [9]. From this experience, the use of these pads has spread throughout Europe due to the construction of high-speed railways [72]. In 2003, sleepers with USP were used in a track section in Copenhagen in order to reduce ballast thickness. The railway line is still in good condition and there has been no unforeseen maintenance. The improvement of the track quality was also shown by the use of USP (2–3 mm of thickness, and made of polyurethane) in Austrian railway lines such as that between Langenlebarn and Tulln [75], which is still in service, and various tests have shown that the track presents good levelling values after more than 2 decades. On the other hand, a study developed between 1997 and 2000 in a bridge of the German Hannover-Gottingen line showed that USP (with a stiffness between 30 and 70 kN/mm) led to the reduction of geometry deterioration (the rate of railway defects was reduced by 25–30% in comparison to a railway without USP), which allowed for an increase in the interval time of maintenance tasks, and therefore reduced costs [9]. It was also seen in studies developed between 2001 and 2006 in Austrian lines since tracks with USP (with a bedding modulus close to 0.2 N/mm3) presented low rates of defects [77] in comparison with sections without USP (differences higher than 1.0 mm, obtaining the greater differences in the transition between sleeper with and without USP), increasing the time between ballast tamping [48]. Other studies carried out in Germany have analyzed the effect of under-sleeper pads on the vibration velocity of ballast in reference to a sleeper with only a pad under the rail. Fig. 10 shows that USP reduce up to 45% (in the case of a frequency equal to 63 Hz) the vibrations transmitted by the sleepers to underlying layers [78], which decreases the damage to the ballast and avoids the development of the liquefaction phenomenon. It was observed that the use of USP in tracks with stiff rail pads leads to a comparable mechanical performance with that recorded in tracks with soft rail pads and without USP. In addition, Austrian studies have shown that USP allows for a reduction in rail corrugation in small radius curves by using soft pads. Thus, a study developed in a tight curve (radius equal to 265 m) in the Austrian line of Markerforf showed that the use of USP with a bedding modulus near 0.2–0.3 N/mm3 reduced the

length of rail waves as well as decreasing the amplitude of the deformations by up to 50% after 25 GMT [77]. The influence of under-sleeper pads on rail corrugation development has also been evaluated in a narrow curve (R = 288 m) in a test track in Czech Republic [75], obtaining good behavior of the track with USP since it was built in 2008. In addition, different studies focused on the use of USP in small radius curves have shown that these elastic components could modify the lateral resistance of the track. However, there is no general consensus about whether USP increases or decreases the lateral track stability [75]. With the use of USP (0.13 N/mm3) in a 17 km Swiss track (a new line between Mattstetten–Rothrist) it was possible to reduce structure borne vibrations. With this purpose, USP were also used in Zurich´s main station. In addition, other European studies have evaluated the changes in the noise emission of train passages after the installation of under-sleeper pads. In these studies, a resonance frequency of 16.0–31.5 Hz was observed for all the under-sleeper pads studied. This signified greater noise levels for that frequency range. However, there was an effective noise reduction from

Fig. 11. Effect of USP stiffness on rail movements, depending on rail pad and ballast characteristic [71].

Fig. 10. Comparison of ballast vibration between a section with USP and another without this type of elastic element [77].

301

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

frequencies above 40 Hz, which reached a level of between 8 and 15 dB as insertion loss [73]. These levels of noise reduction were also recorded in an experimental study carried out in Croatia, in which there was a more marked effect of USP among 5–250 Hz [45]. In relation to the effect of the USP stiffness, the DB (Deutsche Bahn) developed a study in an experimental section where USP with stiffness equal to 70 kN/mm and 80 kN/mm were used, along with an analysis of sleepers without elastic elements as a control. It was observed that the decrease in stiffness led to the increase (30– 50%) in rail and sleeper movements. In addition, it was shown that USP allowed for a reduction of up to 35% of the stress on the ballast [17]. In addition to these various experiences, a number of authors have analyzed the response of railway structures that have incorporated this element, studying USP with diverse stiffness values. The results show that under-sleeper pads (3 cm thick) reduce the intensity of vibrations transmitted by the railway system by up to 30% [39]. However, the use of soft USP (35 kN/mm) can produce higher movements (up to 35% more, depending on the characteristics of the rail pads and ballast modulus) and accelerations (increase near 50% for frequencies around 150 Hz) of rail and sleepers, according to a theoretical study developed by Johansson et al. [72] in which the influence of USP on the dynamic behavior of the railway track was analyzed. Fig. 11 shows an example of the effect of USP stiffness in rail movements depending on rail pad and ballast stiffness. It is evident that under-sleeper pad stiffness has an important influence on the dynamic performance of the railway infrastructure. Accordingly, various authors have studied the role of this parameter in the behavior of the railway system. It has been found that under-sleeper pads with a bedding modulus equal to 0.3 N/mm3 reduce the overall stiffness of the track as well as stresses on the sleepers and ballast layer [2,79]. However, since this type of pad causes more movement and vibrations in the rails and sleepers, it was necessary to use stiffer pads to avoid such problems. Stiffer pads also reduce the bending moment of the rail, but have the disadvantage of increasing the load transmitted to the under layers and can lead to the potential sagging of the sleepers. A number of authors [2,23] have also investigated the influence of under-sleeper pad stiffness on the dynamic forces exerted on the railway track. Results showed that the use of stiff pads (in this study with a bedding modulus higher than 0.5 N/mm3) has little effect on the variation of contact forces between wheel and rail. However, soft under-sleeper pads (0.05 N/mm3 in this case) allow for a reduction in the variation of forces and stress in track sections with changing vertical stiffness. In addition, these types of pads reduce the loads transmitted to the ballast and subgrade layers, obtaining a decrease higher than 140% [80]. The use of USP is responsible for the increase in the area of interaction between the sleeper and ballast, since it is possible to reach values higher than 30% (depending on the stiffness) of the total surface of the sleeper, whereas such values are normally 3–4% on ballasted tracks without USP [2,81]. This fact has been confirmed in laboratory tests where USP from tire pads were used [74], although in this case the increase in area was lower due to the tire slots. With the aim of summarizing the influence of USP stiffness on railway tracks, Table 4 shows some field of application for these elements, depending on its vertical stiffness. 3.3. Under ballast mats 3.3.1. Main characteristics Another measure adopted for the damping of vibrations and energy absorption from train passage is the use of an elastic

Table 3 Influence of various parameters on rail pad stiffness. Parameters Temperature (increase) Dynamic loads Frequency (increase) Pre-load (increase) Fatigue process Pad thickness (increase) Mechanical deterioration Thermal ageing

Stiffness increase

Stiffness reduction U

U U Low influence U High influence U High influence U U High influence U

Table 4 Field of application of USP according to its vertical stiffness. Field of application

Stiff USP

Reduction in rail movements and vibrations Decrease in rail deflection Reduction in corrugation (curve sections) Reduction in corrugation (straight sections) Reduction in sleeper movements and vibrations Decrease in ballast settlement due to stress reduction Reduction in ballast vibrations Decrease in ballast layer thickness Reduction in track stiffness variations

U U U U U

Soft USP

U U U U (less effect than soft)

U

Table 5 Field of application for UBM with different vertical stiffnesses. Field of application

Stiff mats

Soft mats

Reduction in ballast pressure Increase in track flexibility Decrease of the ballast layer thickness Reduction in ballast degradation in stiff sections Reduction in vibrations transmitted through the ground

U

U U U U U

U U

anti-vibration mat between the ballast layer and the substructure (ballasted track) or beneath the concrete slab (slab track). This mat is essential in areas (tunnels, bridges, elevated stations, cuts, switches, etc.) where elastic components are required in the railway structure to reduce stress on the substructure as well as vibrations and noise levels [82,83], due to the deformation capacity and mechanical energy dissipation of this material. These properties depend on the thickness and density of the mat as well as on the size and type of compound that the material is composed of (see Fig. 12). Elastic mats usually have a thickness of 15–30 mm whereas their horizontal dimensions depend on the technique developed during the construction. UBM are often composed of a single layer of polymeric material or of two layers: (i) a distribution layer to uniformly distribute loads, and (ii) an elastic layer to dampen loads. In addition to the conventional elastic materials used in mat manufacturing, alternative composites are being developed from waste tires in order to reduce costs [84]. Despite the variation in design and composition, all of these mats are able to absorb the impact of loads and the vibrations produced by the railway system. They can also maintain performance even after a repeated number of loads and in diverse climate conditions. Whilst the rail pads and under sleeper pads (elastic elements near the surface) are more suitable for increasing the elasticity of

302

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

Fig. 12. Under-ballast mats [52].

the rail, the under-ballast mats (UBM) are more efficient for reducing the fast deterioration of the ballast in contact with the planks and with rigid substructures such as tunnels or bridges. The mats are defined by their dynamic bedding modulus Cdyn (N/mm3), and they can be classified as hard (>0.22 N/mm3), medium (0.09– 0.22 N/mm3), soft (0.05–0.09 N/mm3) and very soft (0.03–0.05 N/ mm3) [83]. 3.3.2. Cases and studies The use of elastic under-ballast mats in railway infrastructure began in the first high-speed railway line between Tokyo and Osaka in 1964. The first under-ballast mats were made from used automobile tires, and were installed to increase vertical flexibility all along the track system. At the same time, the objective was to reduce stresses on the ballast, and thus avoid the rapid breakage and crushing of ballast particles, which was occurring in the areas near the deck of the concrete bridges of the high-speed railway [9]. Since this first experience in Japan, other studies have been developed. In particular, it is worth highlighting the work developed in European countries where the UBMs are used as specific solutions. For example, the instruction I-AM 05/02 established by the SBB (Swiss National Railways) indicate the employment of UBM to be appropriate when the ballast layer thickness is lower than 30 cm, since the elastic mats allow for an increase in the track flexibility. Similarly, the DB (Deutsche Bahn) recommends UBM in

lines with ballast thickness lower than 30 cm and traffic higher than 10,000 tonnes per day [83]. In the United Kingdom, of the different cases, the use of under ballast mats in a tunnel section of the Gospel Oak line (London) is particularly worthy of note, as in this case the ballast layer thickness was reduced by up to 13 cm. Thus, the main aim of the UBM was to increase the track flexibility and reduce ballast degradation [85]. In addition, UBM with a bedding modulus equal to 0.83 N/mm3 were placed in a bridge (1160 m length) from the high-speed line between Hannover and Würzburg (Germany) in 1987. After 21 years (traffic roughly 384 MGT), the elastic mats were taken off during maintenance and rehabilitation tasks in order to analyze their mechanical response. Laboratory studies confirmed that the mats stiffness had increased by only 11.2% whereas the visual aspect of the underlying granular layers was good, and in addition a lower number of maintenance tasks had been developed during the service life of the track in comparison with a section without UBM [86]. This work indicates the potential of UBM to be used long-term whilst at the same time reducing track deterioration. This decrease in track degradation is related to the fact that the use of UBM allows for an increase in the area of contact between the ballast layer and the underlying ones [87,88], which reduces the stresses transmitted by the ballast to the under layers (thus producing fewer settlements). At the same time, this in turn

Fig. 13. Comparison of ballast vibration between a section without UBM and another where UBM were installed [88].

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

decreases the vibrations transmitted to the soil, and also reduces the noise stemming from the contact between ballast particles. In addition, a study [89] carried out in a tunnel section in Germany was focused on measuring the vibrations before and after (18 years) the use of UBM. Results (shown in Fig. 13) indicated that the speed of the waves transmitted through the ground decreased when the elastic mats were used in comparison with the results recorded before the maintenance task, obtaining a reduction of 15 dB (almost 30%) for frequencies higher than 31.5 Hz in comparison with a section without UBM. A further study [83] also indicates that mats reduce up to 19 dBV the vibrations transmitted to the structure when 0.3 N/ mm3 mats are used, these being more effective in reducing its dynamic bedding modulus, and also when used over rock or in tunnels instead of banks. It was also shown that the UBMs protect the ballast, given that they avoid contact between the ballast and the substructure, whist giving greater flexibility to the track so that the load of the trains is divided between more planks, therefore reducing the transmitted tension (see Table 5). In relation to the properties of these elements, Kimura [87] developed a simplified prediction procedure based on an original Wettschureck–Kurze model [90]. This model uses a finite termination impedance to represent the case of an anti-vibration mat installed beneath a concrete slab or asphalt layer. This prediction procedure was tested against measurements on at-grade installations on light rail transit and commuter railway installations in Baltimore and Boston. This comparative analysis showed that this method can be useful for the study of the properties of under-ballast mats because it can predict insertion losses for the range of low frequencies in which the transmission of vibrations to the soil is particularly important [91].

4. Conclusions In spite of the many advantages of railways (efficiency, transportation capacity, low environmental impact, etc.), there are some drawbacks that need to be addressed in order to improve their efficiency. Thereby, this paper presents a state of the art review focused on elastic elements as a common solutions applied in railway infrastructure to reduce the effect of the main problems associated with train traffic. From the study, the following conclusions can be drawn: Elastic elements are an adequate solution to modify the vertical stiffness of tracks and to dampen loads, vibrations and noises. The most commonly used elastic elements in railway infrastructure are rail pads, under-sleeper-pads, and under-ballast mats. The stiffness (k, kN/mm) of the rail pads is the main characteristic parameter of these elements. Nonetheless, this property is influenced by diverse factors like temperature, load frequency, preload and material degradation.  Different studies have shown that soft rail pads (close to 80 kN/ mm) could increase rail movements (and thus its vibrations and noise) and deflection, which could cause the fatigue of other railway components. Stiff pads are more appropriate to reduce noise and vibrations from wheel–rail contact.  Nonetheless, the decrease of rail pad stiffness can reduce the vibrations and noise from sleepers and ballast particles, whilst these types of pads can simultaneously allow for a more homogeneous distribution of stiffness throughout the track and a lower effect of the loads transmitted to underlayers. With regard to under-sleeper pads and under-ballast mats, their characteristic parameter is the bedding modulus (C, N/mm3) that is to say the element stiffness per unit area.

303

 When dealing with under-sleeper pads, it is advisable to use stiff pads to reduce vibrations in sleepers and rails (more effectives for frequencies higher than 40–50 Hz). In contrast, soft pads (approximately 0.10–0.15 N/mm3) should be installed to reduce stresses on the ballast, though medium stiffness pads (0.20 N/mm3) are the most suitable to achieve a progressive variation of stiffness throughout the track.  Under-ballast mats are mainly used to mitigate the lowfrequency vibrations transmitted to the soil and to increase the overall flexibility of the track, the use of soft mats being more appropriate (bedding modulus lower than 0.10 N/mm3). At the same time, these mats reduce stresses on the ballast layer as well as on the underlayers of the track bed. Thus, it is possible to reduce the ballast layer thickness when UBM are used. Based on the study of the state of the art of elastic elements for railway infrastructure, some lines of investigation could be proposed:  Only a few authors have quantified the changes in the lifespan under fatigue of materials such as the fastener when the stiffness of rail pads is modified. Thus, more in-depth research is needed in order to analyze the effect of this parameter on the deterioration of the track components.  The use of elastic pads under the sleeper could modify the lateral resistance of the track. Thus, some experimental studies have focused on this issue, but they conclude that a more indepth study is necessary to quantify the effect of USP on the lateral stability of tracks.  When under ballast mats are used it could be necessary to determine the maximum possible reduction of ballast thickness that would not modify track behavior.

References [1] López Pita A. Alta Velocidad en el Ferrocarril. Ediciones UPC, Universidad Politécnica de Cataluña, Barcelona; 2010. [2] Witt S. The influence of under sleeper pads on railway track dynamics. In: Proceedings of Linköping University, Department of Management and Engineering, Division of Solid Mechanics, Sweden; 2008. [3] Eisenmann J. Stress distribution in the permanent way due to heavy axle loads and high speeds. Proc AREA 1970;71:24–59. [4] Prud’Homme A. Forces and behaviour of railroad tracks at very high train speeds. Standards adopted by SNCF for its future high speed lines (250– 300 km/h). In: Proc Symp Rail Track Mech Tech. New Yersey: Princeton University; 1978. p. 79–108. [5] Hunt GA. Eurobalt optimises ballasted track. Railway gazette international, December 2000. p. 813–6. [6] Dinis Ferreira PA. Modeling and prediction of the dynamic behaviour of railway infrastructures at very high speeds. PhD dissertation Universidade técnica de Lisboa; 2010. [7] Nielsen J, Johansson A, Vernersson T. Train–track interaction and mechanisms of irregular wear on wheel and rail surfaces. Vehicle Syst Dynam 2003; 40(1–3):3–54. [8] Teixeira PF. Contribución a la reducción de los costes de mantenimiento de vías de alta velocidad mediante la optimización de su rigidez vertical. PhD Dissertion. E.T.S. Ingenieros de Caminos, Canales y Puertos, de Barcelona; 2003. [9] López Pita A. Infraestructuras ferroviarias. Edicions UPC, Universidad Politécnica de Cataluña, Barcelona; 2006. [10] Selig ET, y Waters JM. Track geotechnology and substructure management. London: Thomas Telford; 1994. [11] Prud’Homme A. Ligne à Grande Vitesse Paris-Sud-Est. Annnales de l’Institut Technique du Batiment et des Travaux Publics 1978;366:70–96. [12] Shenton MJ. Deformation of railway ballast under repeated loading conditions. In: Press P, editor, Railroad track mechanics and technology. Oxford; 1978. [13] Knothe K, Grassie SL. Modelling of railway track and vehicle/track interaction at high frequencies. Vehicle Syst Dynam 1960;22(1993):209–62. http:// dx.doi.org/10.1080/0042318969199. [14] Krylov VV, Dawson AR, Heelis ME, Collop AC. Rail movement and ground waves caused by high-speed trains approaching track-soil critical velocities. In: Proceedings of the Institution of Mechanical Engineers, Part F, vol. 214, 2000.

304

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305

[15] South J. Experiments, made at Watford, on the vibrations occasioned by railway trains passing through a tunnel. Phil Trans Roy Soc London, Ser A, vol. 13, 1863, p. 65–83. [16] Hyde JH, Lintern HR. The vibrations of roads and structures. Minutes Proc Inst Civ Eng 1929;227(1929):187–243. [17] Müller-Boruttau F, Kleinert U. Betonschwellen mit elastischer Sohle. ETR 2001;50(H3):90–8. [18] Sato Y, Matsumoto A, Knothe K. Review on rail corrugation studies. Wear 2002;253(1–2):130–9. [19] Dahlberg T. Railway track stiffness variations – a literature review for the REVA project. Sweden: Mechanical Engineering, IEI, Linköping University; 2007. [20] Dahlberg T. Railway track dynamics – a survey. Technical report LiTH-IKPR. Sweden: Linköping University; 2003. [21] Egana JI, Vinolas J, Seco M. Investigation of the influence of rail pad stiffness on rail corrugation on a transit system. Wear 2006;261(2):216–24. [22] Kang YS, Yang SC, Lee HS, Kim YB, Jang SY, Kim E. A study of track and train dynamic behavior of transition zone between concrete slab track and ballasted track. In: Proceedings of World Congress Railway Research, Seul, Korea; 2008. [23] Dahlberg T. Railway track stiffness variations-consequences and countermeasures. Int J Civ Eng 2010;8(1):1–12. [24] Kaewunruen S, Remennikov AM. Dynamic properties of railway track and its components: a state-of-the-art review. In: Weiss BN, editor. New research on acoustics. Hauppauge, New York: Nova Science Publishers; 2008. p. 197–220. [25] Morgan JGD, Markalnd E. The effect of vibration on ballast beds. Géotechnique 1981;31(3):367–86. [26] UIC (International Union of Railways) High Speed Department. Design of New Lines for Speeds of 300–350 km/h, S, First Report; 2001. [27] Álvarez R. Estudio de Vibraciones en Alta Velocidad. In: Proceedings of Jornadas de Vibraciones Causadas por el Ferrocarril, Análisis y Soluciones Correctoras, Barcelona; 2009. [28] Melis M. Terraplenes y Balasto en Alta Velocidad Ferroviaria [Primera parte]. Revista de Obras Públicas; 2006. [29] Thompson D. Railway noise and vibration: mechanisms. Oxford: Modelling and Means of Control Elsevier; 2009. [30] Kaewunruen S. Experimental and numerical studies for evaluating dynamic behaviour of prestressed concrete sleepers subject to severe impact loading. PhD Thesis, School of Civil, Mining and Environmental Engineering, University of Wollongong, NSW, Australia; 2007. [31] Remennikov AM, Kaewunruen S. A review on loading conditions for railway track structures due to train and track vertical interaction. Struct Control Health Monitor 2007;15(2):207–34. [32] Wu TX. On the railway track dynamics with rail vibration absorber for noise reduction. J Sound Vib 2008;309:739–55. http://dx.doi.org/10.1016/ j.jsv.2007.07.049. [33] Remington PJ. Wheel/rail rolling noise. What do we know? What don’t we know? Where do we go from here? J Sound Vib 1998;120:203–26. http:// dx.doi.org/10.1016/0022-460X(88)90430-0. [34] Thompson DJ, Jones CJC. Noise and vibration from railway vehicles. In: Iwnicki S, editor. Handbook of railway vehicle dynamics. CRC Press, Taylor & Francis; 2006. p. 279–325. [35] Lichtberger B. Track compendium. 1st ed. Eurail Press; 2005. [36] Prud’Homme A. La voie. Revue Générale des Chemins de Fer, vol. 1, 1970. p. 56–72. [37] Hirano M. Theoretical analysis of variation of wheel load. Railway Technical Research Institute. Quarterly Reports 1972;13(1):42–4. [38] Kuroda S. Dynamics variation of wheel load. Attributed to vertical deformation of rail end. Railway Technical Research Institute, Quarterly Reports, vol. 14(3), 1973. p. 143–4. [39] Jenkins J, Stephenson S, Clayton G, Morland G, Lyon DL. Incidences des paramètres caracteristiques de la voie et des vehicules sur les efforts dynamics verticaux qui se développent entre rail et roue. Rail International, vol. 10, 1974. [40] Amielin, SV. La vía y el servicio de vía. Translation from the Russian original by Fundación de los Ferrocarriles Españoles; 1974. [41] Fortin JP. La deformée dynamique de la voie. Revue Generale des Chemins de Fer.; 1982. p. 93–102. [42] AEA Technology Rail. Review of the effect of track stiffness on track performance. Rail Safety & Standards Board. Research project T372; 2005. [43] Raymond G, Bathurst R. Estimate of static track modulus using elastic foundation models. Transp Res Rec 1994;1470:65–72. [44] López Pita A. Posibilidades en la reducción de los costes de mantenimiento de la calidad geométrica de una vía, mediante la introducción de nuevos criterios en su diseño. XVI Pan American Railway Congress, Washington DC; 1984. p. 416–63. [45] Lakuši S, Ahac M, Haladin I. Experimental investigation of railway track with under sleeper pad. 10th Slovenian road and transportation congress. Portorozˇ, Slovenija; 2010. p. 20–2. [46] Ito M, Nagai K. Degradation issues of polymer materials used in railway field. Polym Degrad Stab 2008;93:1723–35. http://dx.doi.org/10.1016/ j.polymdegradstab.2008.07.011. [47] Zamora-Barraza D, Calzada-Pérez MA, Castro-Fresno D, Vega-Zamanillo A. Evaluation of anti-reflective cracking systems using geosynthetics in the interlayer zone. Geotext Geomembr 2010;29:130–6. http://dx.doi.org/ 10.1016/j.geotexmem.2010.10.005.

[48] Loy H. Under sleeper pads in turnouts. Proc Euro Rail Technol Rev 2009;49:35–8. [49] Carrascal Vaquero I. Optimización y análisis de comportamiento de sistemas de sujeción para vías de ferrocarril de alta velocidad española. PhD Thesis Dissertion, Universidad Cantabria; 2006. [50] Kaewunruen S, Remennikov AM. Sensitivity analysis of free vibration characteristics of an in situ railway concrete sleeper to variations of rail pad parameters. J Sound Vib 2006;298:453–61. http://dx.doi.org/10.1016/ j.jsv.2006.05.034. [51] Sol-Sánchez M, Moreno-Navarro F, Rubio-Gámez MC. Viability analysis of deconstructed tire as material for rail pads in high-speed railways. Mater Des 2014;64:407–14. [52] CDM. Manufacturer of elastic elements. [accessed on 15.09.14]. [53] Carrascal I, Casado JA, Polanco JA, Gutiérrez-Solana F. Comportamiento dinámico de placas de asiento de sujeción de vía de ferrocarril. Anal Mecán Fract 2005;22:372–7. [54] Gomes Correia A, Cunha J. Analysis of nonlinear soil modeling in the subgrade and rail track response under HST. Transport Geotech 2014 [in press]. [55] Vincent N, Bouvet P, Thompson DJ, Gautier PE. Theoretical optimization of track components to reduce rolling noise. J Sound Vib 1996;193(1):161–71. [56] López Pita A. La rigidez vertical de la vía y el deterioro de las líneas de alta velocidad. Revista de Obras Públicas 3415; 2001. p. 7–26. [57] Giannakos K. Influence of rail pad stiffness on track stressing, life-cycle and noise emission. In: Proceeding of the second international conference on sustainable construction materials and technologies, Ancona, Italy; 2010. [58] Kaewunruen S, Remennikov AM. An experimental evaluation of the attenuation effect of rail pad on flexural behaviour of railway concrete sleeper under severe impact loads. In: Proceeding of the 2008 Australian structural engineering conference, Melbourne, Australia; 2008. [59] Carrascal I, Casado JA, Diego S, Polanco JA. Atenuación frente a impacto en sistemas de sujeción ferroviaria de alta velocidad. J Anales Mecán Fract 2011;28(2):713–8. [60] Ilias H. The influence of rail pad stiffness on wheelset/track interaction and corrugation growth. J Sound Vib 1999;227:935–48. [61] Wu TX, Thompson DJ. The effects on railway rolling noise of wave reflections in the rail and support stiffening due to the presence of multiple wheels. Appl Acoust 2001;62:1249–66. [62] Leykauf G, Stahl W. Untersuchungen und Erfahrungen mit besohlten Schwellen. EI Der Eisenbahningenieur (55) 6, 2004, pp. 8–16. [63] Carrascal I, Casado JA, Polanco JA, Gutiérrez-Solana F. Dynamic behaviour of railway fastening setting pads. Eng Fail Anal 2007;14:364–73. [64] Thompson DJ, Verheij JW. The dynamic behaviour of rail fasteners at high frequencies. Appl Acoust 1997;52(1):1–17. [65] Fenander A. Modelling stiffness and damping by use of fractional calculus, with application to rail pads. PhD Dissertation, Division of Solid Mechanics, Chalmers University of Technology, Gothenburg, Sweden; 1997. [66] Thompson DJ, Van Vliet WJ, Verheij JW. Development of the indirect method for measuring the high frequency dynamic stiffness of resilient elements. J Sound Vib 1998;213:169–88. http://dx.doi.org/10.1006/jsvi.1998.1492. [67] Sol-Sánchez M, Moreno-Navarro F, Rubio-Gámez MC. The use of deconstructed tire rail pads in railroad tracks: Impact of thickness. Mater Des 2014;58:198–208. [68] Carrascal I, Casado JA, Diego S, Polanco JA, Gutiérrez-Solana F. Efecto del envejecimiento de placas de asiento de carril inyectadas con TPE en la elasticidad de la vía para Alta Velocidad. 31th Meeting of the Spanish Group of Crack ‘‘Grupo Español de Fractura’’, in San Lorenzo del Escorial, España; 2010. [69] Kaewunruen S, Remennikov AM. Monitoring structural degradation of rail pads in laboratory using impact excitation technique. In: Proceedings of 1st international conference on structural condition assessment, monitoring and improvement, Australia; 2005. [70] Remennikov AM, Kaewunruen S, Ikaunieks K. Deterioration of dynamic rail pad characteristics. Australia: Proceedings of University of Wollongong Research Online; 2006. [71] Kaewunruen S, Remennikov AM. Response and prediction of dynamic characteristics of worn rail pads under static preloads. In: Proceedings of the 14th international congress on sound and vibration, Australia; 2007. [72] Johansson A, Nielsen JCO, Bolmsvik R, Karlström A. Under sleeper pads – influence on dynamic train-track interaction. Wear 2008;265:1479–687. [73] UIC (International Union of Railways) USP Report. Under sleeper pads, Project no. I/05/U/440, Summarising report; 2009. [74] Sol-Sánchez M, Moreno-Navarro F, Rubio-Gámez MC. Viability of using end-oflife tire pads as under sleeper pads in railway. Constr Build Mater 2014;64:150–6. [75] UIC (International Union of Railways) USP recommendations. Under Sleeper Pads, Recommendations for Use. UIC-Leaflet, Draft Document; 2013. [76] UIC (International Union of Railways). Railway applications/track – concrete sleepers and bearers – concrete sleepers and bearers with under sleeper pads. CEN-document TC 256 WI 00256597; 2009. [77] Schilder R. UIC project no. I/05/U/440 USP-under sleeper pads. Applications and benefits of elastic elements in ballasted tracks UIC, Paris 17; 2006. [78] Stahl W. Improvement of ballasted tracks using sleeper pads – investigations and experiences in Germany Munich University of Technology; 2004. [79] Schneider P, Bolmsvik R, Nielsen JCO. In situ performance of a ballasted railway track with under sleeper pads. J Rail Rapid Transit 2011;225(3). ISSN.0954-4097.

M. Sol-Sánchez et al. / Construction and Building Materials 75 (2015) 293–305 [80] Krüger M. Theoretical investigations/calculations. Work package number 2a of UIC Project under sleeper pads; 2007. [81] Riessberger A. Ballast track for high speeds. In: Proceedings of tracks for highspeed railways, Porto, Portugal; 2006. p. 23–44. [82] Teixeira P. State-of-the-Art on the use of bituminous subballast on european high-speed rail lines. Congress of bearing capacity of roads, railways and airfields, Barcelona, Spain; 2009. [83] UIC Code 719-1 Recommendations for the use of under ballast mats. 1st ed. VII Way and Works; 2011. [84] Carrascal I, Casado JA, Diego S, Polanco JA. Comportamiento mecánico frente a esfuerzos dinámicos de manta elastomérica para la atenuación de vibraciones en ferrocarriles. [accessed on 12.06.14]. [85] Potocan S. Ballast mats within areas of reduced ballast depth. European railway review, Issue 5; 2010.

305

[86] Dold M, Potocan S. Long-term behaviour of Sylomer ballast mats. Rail technology review, vol. 13, 2013, ISSN 1869-7801. [87] Kimura Y. Control of ground-borne vibration from at-grade track using ballast mats. In: Proceedings of American public transit association rapid transit conference; 1995. [88] Brown SF, Brodrick BV, Thom NH, McDowell. The Nottingham railway test facility. In: Proceedings of the Institution of Civil Engineers – Transport; 2007. [89] Werkstoffe Getzner. Ballast mats. Characteristics of solutions in service. Paris: UIC; 2006. [90] Wettschureck R, Kurze UJ. Ground vibrations in the vicinity of shallow railroad tunnels and open tracks. Acustica 1985;58:170–6. [91] Hanson CE, Singleton HL. Performance of ballast mats on passenger railroads: Measurement vs. projections. J Sound Vib 2006;293:873–6.