Elsevier

Computers & Structures

Volume 228, February 2020, 106180
Computers & Structures

Dynamic calibration of slab track models for railway applications using full-scale testing

https://doi.org/10.1016/j.compstruc.2019.106180Get rights and content

Highlights

  • A detailed 3D slab track model that includes all components of the infrastructure.

  • Full-scale tests of the slab track in high-speed and conventional speed operations.

  • Full-scale tests of the fastening system in realistic service conditions.

  • Identification of representative rail track models in realistic operation scenarios.

Abstract

Research and development of technology for railways has found new impetus as society continues to search for cost effective and sustainable means of transport. This tasks engineers with using the state-of-the-art science and engineering for rolling stock development and advanced technologies for building high performance, reliable and cost-effective rail infrastructures. The main goal of this work is to develop detailed and validated three-dimensional slab track models using a finite element formulation, which include all components of the infrastructure. For this purpose, the parameters of the computational models are identified by performing full-scale tests of the fastening system and of the slab track, including all its material layers. The computational model proposed here is calibrated using this approach and a good agreement is obtained between experimental and numerical results. This work opens good perspectives to use this reliable track model to study the interaction with railway vehicles in realistic operation scenarios in order to assess the dynamic behaviour of the trains and to predict the long-term performance of the infrastructure and of its components.

Introduction

The health and long-time performance of the infrastructure is critical in any rail system, not only due to safety aspects but also owing to the high maintenance costs involved. Besides, it is extremely important to minimize any disturbance in the railway service due to the social and economic repercussions. Despite its importance, the performance and maintenance management of the track is, scientifically, one of the least understood and least predictable elements of railway systems. This fundamental understanding is central to reduce the Life-Cycle Costs (LCC) and to increase the safety, capacity and reliability of the rail networks.

During train operation, tracks are subjected to high impact and fatigue loads that can originate rapid degradation and unexpected, unpredictable failure. The conventional approach used by the infrastructure managers consists of performing regular preventive maintenance, fixed lifetime replacement or applying corrective measures when an incident occurs. This approach is neither effective nor economical so health/performance assessment and predictive maintenance has a great potential for innovation due to the high costs associated. Any efficiency improvement on this would represent a great advantage to the rail industry.

Currently, both ballasted and concrete slab tracks are being used for railways worldwide and it is recognised that both forms have advantages and disadvantages. When compared with ballast tracks, slab systems have advantages from a structural point of view, such as higher lateral and longitudinal stability, no rail buckling and lower sensitivity to differential settlements [1], [2], [3]. They have also operational advantages such as lower maintenance needs, lower structure height and not presenting the problems that are associated to ballast shaking and flight [4], [5], [6], [7]. On the other hand, slab tracks have some disadvantages associated to the higher installation costs and to the lower noise and vibration absorption provided [4], [8]. Due to the overall poor performance of ballast for increased train speeds, the use of concrete slab became more popular and various slab track forms have been produced and tested in recent years [2], [5], [9].

The analysis of the track structure conditions and its dynamic response has devoted the attention of many researchers aiming to support the rail industry in their developments [10], [11], [12], [13], [14], [15], [16]. For this purpose, numerical models have been developed mainly using Multibody (MB) and Finite Element (FE) methodologies. The initial models were 2D based on an elastic foundation formulation [17]. Later, Zhai et al. [18] proposed a 2D numerical model which reproduces the dynamic interaction between a lumped mass vehicle and a discretely supported continuous rail track. This type of model has also been used by other researchers [1], [19]. Cai et al [20] proposed a 3D model where the track is considered as a periodic elastically coupled beam system resting on a Winkler foundation and uses spring and dashpot elements to simulate the rail pads and the sleepers. A more recent 3D model was proposed by Poveda et al. [21] to analyse the fatigue life of concrete slabs, which was calibrated with lab tests [22]. Other authors have also proposed co-simulation methodologies between MB and FE formulations in order to study the track structure under realistic trainset loads [23], [24]. These developments open the possibility of integrating more detailed wheel-rail contact models [25], [26], [27], [28], [29], [30], [31], to consider track irregularities [32], [33] and other track singularities [7], [13], [34], [35], [36] in the studies aiming to assess the track performance and degradation evolution [11], [37], [38], [39], [40] in realistic operation conditions.

Due to the multidisciplinary areas of knowledge involved, all issues involving the complete characterization of the railway infrastructure are complex. Therefore, the use of reliable computational tools that are able to reproduce the dynamic response of the track when subjected to the loads induced by the railway traffic is essential. The numerical models have to represent all the structural layers of the track and the elements that are used to fix and support the rails, namely the fastening system. The main difficulty of building such models is the uncertainty associated to the properties of the layers and components that compose them. In order to overcome this uncertainty, field measurements and laboratory tests can be used to validate the numerical models. Nevertheless, many of the track measurements that are performed by the infrastructure managers are unavailable for scientific use due to industry restrictions/confidentiality. The work proposed here is a contribution in this field by proposing a detailed track model with properties that are calibrated with experimental results obtained in a full-scale test facility.

The material layers have non-linear properties that vary with load conditions, frequency, etc. The fastening system is the component that presents higher non-linearities and this is the reason why dedicated lab experiments on it are performed in this work. It should also be noted that the slab technology has a much more predictable behaviour than the ballast tracks, which have a non-linear performance as they are composed of granular material. In the literature there are authors that use linear elastic material models to describe the constitutive relations of the slab track components [41]. Poveda et al [21] use FE solid elements modelled as linear elastic materials to perform studies on the fatigue life design of concrete slab tracks. Zhu et al [42] show that the bilinear cohesive zone model can be employed to capture the mechanical behaviour of the concrete interface of slab tracks. Ren et al [43] show that when debonding occurs the slab track stops exhibiting linear elastic mechanical response. Zhang et al [44] use viscoelastic parameters in the FE models to better predict the initiation of interlayer debonding of track structure. El-Ghandour et al [45] use the modal frequencies extracted from the FE model, instead of the nodal degrees of freedom, and uses the floating frame of reference formulation to obtain the elastic response of the track system. Using a non-linear FE formulation in this work would represent a heavy burden to the code and would make it almost impossible to use the validated models for vehicle-track interaction studies.

Several authors have devoted their studies to highlight the importance of the fastenings to the rails’ dynamic response. Wei et al. [46] concluded that the properties of the rail pads have a non-linear behaviour and are dependent on the frequency and amplitude of the applied loads. Fenander [47] concluded that the rail pad stiffness increases with the preload and with the frequency, although the effect of the frequency is less relevant. Kaewunruen et al. [48] concluded that the level of preloading have a substantial influence on the natural frequencies and dynamic properties of rubber pads. Wei et al. [49] also showed the influence of the frequency in the pads stiffness especially for the vibrations produced by subways in tunnels. Zhu et al. [50] appreciated an increase in the dynamic stiffness and damping of the rail pads with the increment of the excitation frequency. Carrascal et al. [51] studied the deterioration in the rail pads produced by its normal working conditions and defined a test methodology to determine the dynamic behaviour of railway fastening setting pads in working conditions [52], [53].

Other authors have dedicated their research activities to investigate the performance of various parts of the railway track structure and have used experimental tests for the purpose. For example, full-scale model tests with simulated train moving loads hace been developed to explore the dynamic performance and long-term behaviour of concrete slab tracks [4], [22], [54]. In the case of ballasted track, a two-layer railway track model was developed and tested [55]. Pita et al. [56] and Colaço et al. [57] observed that a low track stiffness value can result in a flexible track with poor load distribution, whereas a high track stiffness can cause greater dynamic overloads on the rail, which induces increased vehicle-track interaction forces that lead to rail defects such as corrugation.

The main goal of this work is to develop and calibrate computational models of slab tracks that can be used for the consistent assessment of the performance and dynamic response of the railway infrastructure when subjected to realistic loads imposed by the rolling stock. The properties of the fastening elements have a noticeable uncertainty as they depend on the material, load, preload produced by the fastening system (toe load), frequency and on the degradation degree of the components. Furthermore, there are also uncertainties associated to the properties of the material layers that compose the slab track structure, such as the subgrade and the different concrete types. The main contribution of this work is that the uncertainties associated to the physical parameters of the material layers and of the components that compose the track system are overcome by performing full-scale tests of the slab track structure and of the fastening system. Such experimental data is used to calibrate the numerical models such that they present a frequency response similar to the real physical models.

It is foreseen that the calibrated slab track model proposed here can be used together with detailed vehicle models, in a co-simulation environment, to study the long-term behaviour of the rail infrastructure. This approach can then be used together with suitable track degradation models to develop decision support tools to promote the implementation of science-based maintenance strategies.

Section snippets

GRAFT II: Description of the test apparatus

A slab track system is generally composed by a track superstructure and a substructure. The superstructure includes the rail, fastening system, slab track and a concrete supporting layer. The track substructure comprises the roadbed, subgrade and subsoil [4]. The dynamic response of the slab track depends on the properties of all the different layers [58], [59]. In this work full-scale laboratory tests of a slab track system are performed comprising the layers represented in Fig. 1, namely

Description of the test apparatus

The rail fastening is the system used to fix the rails to the sleepers that, not only prevents the rails from rotating, but also provide elasticity to the track and damp the transmission of noise and vibrations to the infrastructure resulting from the train operation. The fastening system used in this work is the Vossloh system 300, represented in Fig. 5, which is a highly elastic solution for slab track with applications for both conventional and high-speed rail.

As shown in Fig. 5(b), the

General description

The purpose of this work is to develop a detailed and reliable FE slab track model that includes all components of the rail infrastructure. The characteristics of the track layers and of the rail supporting elements are identified by performing full-scale tests of the slab track and of the fastening system as previously described. The FE model is built and studied using the commercial software ANSYS and considering the dimensions shown in Table 1. The dimensions of the fastening system model

Computational model calibration

The calibration of the slab track numerical model is performed by comparing the computational results with the ones obtained in the laboratory experiments. To this end, the reference values for the properties of the track material layers are adjusted to get the better possible correspondence between the numerical and experimental results. The calibrated model considered here is the one with the material properties detailed in Tables 8 and 9 and the mesh size defined in Table 10.

The comparison

Conclusions and future developments

The aim of this work is to boost the overall performance of rail transport infrastructure by developing reliable numerical models that enable to predict the track behaviour, contributing to reduce the LCC of the infrastructure and to minimize the traffic disruptions, which are mainly caused by unpredicted failures or events. For this purpose, fastening characterization tests and full-scale slab track experiments are performed in realistic operation conditions. Then, a detailed 3D slab track

Declaration of Competing Interest

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

Acknowledgements

The authors are grateful to the Engineering and Physical Sciences Research Council (EPSRC) for funding this work under Grant Number EP/NOO9215/1. Tarmac, Tensar and Max-Bögl are also acknowledged for their support with regards to the experimental tests. This work was supported by FCT, through IDMEC, under LAETA, project UID/EMS/50022/2019.

References (68)

  • S. Liu et al.

    Modelling and in-situ measurement of dynamic behavior of asphalt supporting layer in slab track system

    Constr Build Mater

    (2019)
  • S. Zhu et al.

    Mechanical property and damage evolution of concrete interface of ballastless track in high-speed railway: Experiment and simulation

    Constr Build Mater

    (2018)
  • J. Ren et al.

    Influence of cement asphalt mortar debonding on the damage distribution and mechanical responses of CRTS I prefabricated slab

    Constr Build Mater

    (2020)
  • Y. Zhang et al.

    Study on the interlayer debonding and its effects on the mechanical properties of CRTS II slab track based on viscoelastic theory

    Constr Build Mater

    (2019)
  • I.A. Carrascal et al.

    Dynamic behaviour of railway fastening setting pads

    Eng Fail Anal

    (2007)
  • D. Ferreño et al.

    Experimental and finite element fatigue assessment of the spring clip of the SKL-1 railway fastening system

    Eng Struct

    (2019)
  • I.A. Carrascal et al.

    Experimental study of metal cushion pads for high speed railways

    Constr Build Mater

    (2018)
  • R. Chen et al.

    Experimental study on dynamic load magnification factor for ballastless track-subgrade of high-speed railway

    J Rock Mech Geotech Eng

    (2013)
  • T. Marolt Čebašek et al.

    Full scale laboratory testing of ballast and concrete slab tracks under phased cyclic loading

    Transp Geotech

    (2018)
  • C. Esveld

    Modern railway track

    (2001)
  • C. Esveld

    Developments in high-speed track design

    IABSE symp

    (2003)
  • O. Peltokangas et al.

    Evolution of railway track settlement after ballast tamping

    Int J Railw Technol

    (2015)
  • S. Costa D’Aguiar et al.

    Railway transitional zones: a case history from ballasted to ballastless track

    Int J Railw Technol

    (2014)
  • R. Sañudo et al.

    Study on different solutions to reduce the dynamic impacts in transition zones for high-speed rail

    J Theor Appl Vib Acoust

    (2017)
  • K. Nguyen et al.

    Dynamic effect of high speed railway traffic loads on the ballast track settlement

    Congr Métodos Numéricos Em Eng

    (2011)
  • N. Bosso et al.

    A comprehensive strategy to estimate track condition and its evolution

    Int J Railw Technol

    (2012)
  • B. Indraratna et al.

    Modernisation of rail tracks for higher speeds and greater freight

    Int J Railw Technol

    (2013)
  • E. Fortunato et al.

    Railway track transition zones: design, construction, monitoring and numerical modelling

    Int J Railw Technol

    (2013)
  • Y. Momoya et al.

    Improvement of degraded ballasted track to reduce maintenance work

    Int J Railw Technol

    (2016)
  • P.K. Woodward et al.

    Application of coupled train-track modelling of critical speeds for high-speed trains using three-dimensional non-linear finite elements

    Int J Railw Technol

    (2015)
  • Pita AL. Estudio de las deformabilidad del sistema balasto-plataforma en una vía férrea bajo la acción de cargas...
  • K.L. Knothe et al.

    Modelling of railway track and vehicle/track interaction at high frequencies

    Veh Syst Dyn

    (1993)
  • Z. Cai et al.

    Modelling the dynamic response of railway track to wheel/rail impact loading

    Struct Eng Mech

    (1994)
  • P. Antunes et al.

    A co-simulation approach to the wheel–rail contact with flexible railway track

    Multibody Syst Dyn

    (2019)
  • Cited by (34)

    • Parametric analysis of railway infrastructure for improved performance and lower life-cycle costs using machine learning techniques

      2023, Advances in Engineering Software
      Citation Excerpt :

      In order to transform these forces into load amplitudes supported by the fastening system, it is considered that: The fastening system underneath the wheelset supports 50% of the load, it being assumed that the remaining loads are supported by the adjacent sleepers [11,80]. The load considered is the maximum load, while the minimum is zero, meaning that the amplitude corresponds to 50% of the maximum load.

    • Thermal evolution of CRTS Ⅱ slab track under various environmental temperatures: Experimental study

      2022, Construction and Building Materials
      Citation Excerpt :

      In addition, Kang et al [17] used a full-scale model to test the mechanical characteristics of the CRTS-II slab track structure under vertical static train load. Sainz-Aja et al [18] calibrated material parameters of slab tracks and fastening systems by full-scale tests. Zhang et al [19] conducted a full-scale model test platform of double-block ballastless track to analyse the distribution law of rail support pressure and the load of track slab and support layer bottom.

    View all citing articles on Scopus
    View full text