Elsevier

Computers & Structures

Volume 254, 1 October 2021, 106519
Computers & Structures

A finite element methodology to model flexible tracks with arbitrary geometry for railway dynamics applications

https://doi.org/10.1016/j.compstruc.2021.106519Get rights and content

Highlights

  • A Timoshenko beam element to represent curved rails in realistic tracks models.

  • A methodology for flexible railway tracks and wheel-rail contact mechanics.

  • The automated construction of finite element models of 3D railway tracks.

  • A co-simulation between multibody dynamics and the finite element formulation.

  • A method that improves contact forces calculation and eliminates spurious vibrations.

Abstract

The dynamic analysis of railway vehicles requires an accurate representation of the vehicle, the track geometry and structure, and their interaction. Generally, flexible track models with curved geometries represent the rails with straight beam elements, which results in a piecewise linear representation of the rails. Consequently, the wheel-rail contact mechanics are not properly captured, and the wheel-rail contact forces present spurious high-frequency oscillations. This work proposes a novel approach to model flexible railway tracks with arbitrary geometries, in which the correct geometry in the wheel-rail contact mechanics is assured by modeling the rails as Timoshenko curved beam elements. This approach improves both the geometric representation of the rails and the accuracy of the wheel-rail contact forces calculation. A realistic operation scenario in which a multibody model of a railway vehicle runs on a flexible track with a curved geometry is used here to demonstrate the novel aspects of this work and then discuss the improvements over the conventional approaches. The results show that the proposed methodology greatly improves the computation of the wheel-rail interaction forces and prevents spurious oscillations from propagating to the vehicle.

Introduction

Vehicle-track studies involve a contact model that establishes the interaction between two subsystems: the vehicle and the track [1], [2], [3], [4], [5], [6]. Although several methodologies are available to represent these subsystems in vehicle-track interaction problems, the multibody dynamic analysis and the finite element method became the most used approaches. Studies concerning the vehicle subsystem favor a multibody dynamics approach and focus on describing the mechanical components of the vehicle and their interactions [7], [8], [9], [10], [11], [12], [13]. On the other hand, studies concerning the track subsystem tend to favor the finite element method and focus on describing the structural and material behavior of the track and its components [14], [15], [16], [17], [18], [19], [20], [21], [22].

Most commercial and in-house codes for railway dynamics applications use multibody dynamic analysis to study vehicle curving behavior, passenger comfort, and stability. The track is usually considered as a semi-rigid or rigid structure, and the representation of the rails is purely geometrical [1], [2], [23]. This modeling approach is reasonable in the low-frequency domain as the suspension of the vehicles filters most of the high-frequency excitations, and the track behaves as a stiff spring [24], [25], [26]. However, recent studies show that wheelset and track flexibility influence the dominant frequency of hunting motion [26], the critical speed of the vehicle [26], [27], and the wheel-rail contact forces [26], [27], [28], and thus should not be overlooked in stability studies. Furthermore, the flexibility of other track components becomes important in the mid and high-frequency range [29], [30], [31], [32], [33], [34], [35]. In the mid-frequency range to study the effects of unsupported sleepers [36], [37], [38], rail corrugation and wheel wear [10], [25], [28], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], and ground-borne vibration [49], [50], and in the high-frequency range to study noise and squeal [24], [51]. The finite element method is well established in the literature as a suitable approach to model track flexibility [14], [15], [52], [53].

The multibody dynamic analysis and the finite element method have different integration requirements and satisfying both within a single integration scheme can be challenging. Moreover, binding the two methodologies together increases the complexity of the problem and decreases its numerical efficiency. Therefore, it is common to find in the literature studies where a multibody model of a railway vehicle runs on a rigid track or, conversely, studies where the behavior of the vehicle is represented in a simplified manner on a finite element model of a railway track. The flexible multibody dynamic analysis provides a reasonable approach that represents track flexibility without compromising the complexity of the vehicle model, and unifies the multibody dynamic analysis and the finite element method under one integration scheme [54]. However, this approach involves modal synthesis [55], [56] and additional considerations [57] to keep the computational cost of the procedure under control. Alternatively, co-simulation procedures provide a general approach to simulate coupled systems [37], [58], [59], [60], [61], [62]. Instead of using a single formulation and unified integration, co-simulation allows the subsystems to have independent formulations and integration schemes.

In vehicle-track co-simulation, the coupling between the two subsystems is associated with the wheel-rail contact forces, which requires the assessment of contact conditions between the wheel and rail surfaces. The contact conditions depend on the relative motion and velocity between the wheel and the rail [2], [63], [64]. Therefore, when track flexibility is considered, the evaluation of the contact forces must also include the deformation of the rail. When modeling track flexibility, the rails are usually modeled as finite elements, and their position is evaluated using the corresponding shape functions. In the literature, the details regarding this implementation often refer to classic finite element references, i.e., unidimensional straight beam elements. This modeling approach is suitable if the track comprises only straight sections or when the wheel-rail interaction is replaced by simplified loading conditions. In curved sections, describing the rails with unidimensional straight elements results in high-frequency oscillations without physical meaning in the contact forces. This behavior is due to the wheel changing direction by rolling over straight elements with different orientations, instead of rolling over a smooth curve. These numerical difficulties are mitigated by using different interpolation schemes for the geometry and deformation of the rail [55], [56], or by improving the resolution of the rail geometry, i.e., increasing the number of straight beam elements used to represent the curve. However, increasing rail mesh density also increases the computational cost and only mitigates the undesired effects without solving them. Hence, making this approach less appealing for engineering applications.

This work proposes a new approach to model the rails as unidimensional curved Timoshenko beam elements. This approach improves the representation of rail geometry, with a direct impact on the prediction of the wheel-rail forces, and reduces the computational cost of dynamic studies involving vehicle-track interaction with flexible tracks. The impact of the finite element chosen to represent the rails is demonstrated with a case study of a vehicle running on a flexible track with curved geometry.

Section snippets

Track modeling methodology

The structure of the ballasted railway track is commonly divided into two subsystems: the superstructure and the substructure. The superstructure comprises the rails, rail pads, fastening systems, sleepers, ballast, and sub-ballast, while the substructure comprises the formation layer and the ground. The track model used in this work considers only the superstructure flexibility between the rail and the ballast, as shown in Fig. 1. The remaining components are disregarded because they do not

Railway vehicle modeling

A multibody model of a railway vehicle is characterized by a set of rigid and flexible bodies interconnected by joints and force elements. The equations of motion representing a multibody model of a railway vehicle are written as [73]MΦqTΦq0q¨λ=gγ,where M is a matrix comprising the mass and inertia properties of the rigid bodies, Φq is the Jacobian matrix comprising the partial derivatives of the constraints equations with respect to the coordinates, q¨ is the vector with the accelerations of

Vehicle-track interaction

In the vehicle-track co-simulation procedure, the wheel-rail contact establishes the constitutive relation between the vehicle and track subsystems [1], [2]. Both subsystems exchange information to assess contact conditions between the wheel and the rail surfaces, after which the wheel-rail forces are evaluated with a suitable contact model. Finally, the contact forces are transferred from the interacting surfaces to specific points in the models, i.e., to the centers of mass of the bodies on

Case study

A case study of a vehicle running on a flexible track demonstrates the implications of different rail modeling options, such as element choice and mesh refinement. Two different elements based on the Timoshenko beam theory are used to represent the rails: a quasi-exact two-node straight element [67], [68], and a three-node curved element [67]. The coupled dynamics of the railway vehicle and the flexible track are captured with a co-simulation procedure that ensures the synchronization,

Conclusions

In applications involving vehicle-track interaction and curved flexible tracks, modeling the rails using straight beam elements might compromise its geometric representation, which is fundamental in the evaluation of the wheel-rail contact forces. This work proposes the use of curved beam elements to describe the rails when dealing with vehicle-track interaction and flexible track models with arbitrary geometry. A case study demonstrates the benefits of the new approach over the typical

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.

Acknowledgments

The first author expresses his gratitude to the Portuguese Foundation for Science and Technology (Fundação para a Ciência e a Tecnologia) and the Luso-American Development Foundation (Fundação Luso-Americana para o Desenvolvimento) through the grants PD/BD/128138/2016 and Project – 140/2019, respectively. The third author expresses his gratitude to the Portuguese Foundation for Science and Technology through the PhD grant SFRH/BD/96695/2013. This work was supported by FCT, through IDMEC, under

References (96)

  • D. Ferreño et al.

    Prediction of mechanical properties of rail pads under in-service conditions through machine learning algorithms

    Adv Eng Softw

    (2021)
  • J.A. Sainz-Aja et al.

    Influence of the operational conditions on static and dynamic stiffness of rail pads

    Mech Mater

    (2020)
  • J. Pombo et al.

    Influence of track conditions and wheel wear state on the loads imposed on the infrastructure by railway vehicles

    Comput Struct

    (2011)
  • Y. Sato et al.

    Review on rail corrugation studies

    Wear

    (2002)
  • J. Pombo et al.

    Development of a wear prediction tool for steel railway wheels using three alternative wear functions

    Wear

    (2011)
  • D.J. Thompson et al.

    A review of the modelling of wheel/rail noise generation

    J Sound Vib

    (2000)
  • F. Marques et al.

    On the generation of enhanced lookup tables for wheel-rail contact models

    Wear

    (2019)
  • F. Marques et al.

    A three-dimensional approach for contact detection between realistic wheel and rail surfaces for improved railway dynamic analysis

    Mech Mach Theory

    (2020)
  • J. Baumgarte

    Stabilization of constraints and integrals of motion in dynamical systems

    Comput Methods Appl Mech Eng

    (1972)
  • J.N. Varandas et al.

    Dynamic behaviour of railway tracks on transitions zones

    Comput Struct

    (2011)
  • W.M. Zhai et al.

    Modelling and experiment of railway ballast vibrations

    J Sound Vib

    (2004)
  • J. Pombo et al.

    A new wheel–rail contact model for railway dynamics

    Veh Syst Dyn

    (2007)
  • H. Magalhães et al.

    Implementation of a non-Hertzian contact model for railway dynamic application

    Multibody Syst Dyn

    (2019)
  • H. Magalhães et al.

    Rail vehicle design optimization for operation in a mountainous railway track

    Innov Infrastruct Solut

    (2017)
  • A. Alonso et al.

    Wheel-rail contact: theoretical and experimental analysis

    Int J Railw Technol

    (2013)
  • M.S. Sichani et al.

    Non-elliptic wheel-rail contact modelling in vehicle dynamics simulation

    Int J Railw Technol

    (2014)
  • J. Pagaimo et al.

    Derailment study of railway cargo vehicles using a response surface methodology

    Veh Syst Dyn

    (2020)
  • E. Andersson et al.

    Towards Higher Speeds with Track-Friendly Bogies

    Int J Railw Technol

    (2012)
  • N. Kuka et al.

    Modelling and simulation of tilting vehicles

    Int J Railw Technol

    (2013)
  • O. Polach et al.

    Simulations of running dynamics for vehicle acceptance: application and validation

    Int J Railw Technol

    (2013)
  • S. Stichel et al.

    Modelling and simulation of freight wagon with special attention to the prediction of track damage

    Int J Railw Technol

    (2014)
  • N. Kuka et al.

    Impact of maintenance conditions of vehicle components on the vehicle–track interaction loads

    Proc Inst Mech Eng Part C J Mech Eng Sci

    (2018)
  • H. Magalhaes et al.

    Railway vehicle performance optimization using virtual homologation

    Veh Syst Dyn

    (2016)
  • N. Kuka et al.

    Impact of rail infrastructure maintenance conditions on the vehicle-track interaction loads

    Proc Inst Mech Eng Part C J Mech Eng Sci

    (2020)
  • A.E. Blanco-Saura et al.

    Study of the dynamic vehicle-track interaction in a railway turnout

    Multibody Syst Dyn

    (2018)
  • J. Sainz-Aja et al.

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

    Comput Struct

    (2020)
  • 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)
  • J. Pombo et al.

    An alternative method to include track irregularities in railway vehicle dynamic analyses

    Nonlinear Dyn

    (2012)
  • K. Knothe et al.

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

    Veh Syst Dyn

    (1993)
  • K. Popp et al.

    Vehicle-track dynamics in the mid-frequency range

    Veh Syst Dyn

    (1999)
  • I. Kaiser et al.

    The impact of structural flexibilities of wheelsets and rails on the hunting behaviour of a railway vehicle

    Veh Syst Dyn

    (2019)
  • J.L. Escalona et al.

    Modelling of structural flexiblity in multibody railroad vehicle systems

    Veh Syst Dyn

    (2013)
  • Å. Fenander

    Frequency dependent stiffness and damping of railpads

    Proc Inst Mech Eng Part F J Rail Rapid Transit

    (1997)
  • K. Wei et al.

    Influence of frequency-dependent dynamic parameters of rail pads on environmental vibration induced by subways in a tunnel

    Transp Res Rec J Transp Res Board

    (2015)
  • A.M. Recuero et al.

    Finite-element analysis of unsupported sleepers using three-dimensional wheel-rail contact formulation

    Proc Inst Mech Eng Part K J Multi-Body Dyn

    (2011)
  • Q. Wu et al.

    Parallel co-simulation method for railway vehicle-track dynamics

    J Comput Nonlinear Dyn

    (2018)
  • A. Lundqvist et al.

    Load impact on railway track due to unsupported sleepers

    Proc Inst Mech Eng Part F J Rail Rapid Transit

    (2005)
  • X.S. Jin et al.

    An investigation into rail corrugation, its mechanisms and effects on the dynamic behavior of metro trains and tracks in China

    Int J Railw Technol

    (2016)
  • Cited by (0)

    View full text