Railway infrastructure managers face the continuous challenge of maintaining the rail network within strict safety and operational parameters. A critical aspect of this maintenance is ensuring that rail wear remains below the thresholds imposed by standard regulations. Traditionally, rail wear inspection relies on dedicated track recording vehicles operating on a periodic basis. However, a key challenge in modern railway engineering is the development of on-board condition monitoring technologies capable of being installed on in-service vehicles for continuous monitoring aiming to optimize maintenance costs and avoid components failure. To address this gap, this study proposes a methodology to evaluate rail wear based on the dynamic response of operational vehicles. Recognizing that vehicle dynamics are influenced by rail profile degradation, numerical simulations are employed to analyse vehicle behaviour and generate comprehensive datasets correlating dynamic responses with specific rail wear conditions. Track models incorporating varying degrees of worn rails are developed using a dedicated computational tool originally designed to support the complex geometry of switches and crossings (S&C). Furthermore, an advanced wheel-rail contact model, previously benchmarked for S&C applications, is implemented to handle the complex contact mechanics associated with worn rail profiles. Ultimately, this approach provides a robust simulated framework to support the development of continuous, data-driven track inspection strategies.

The wheel-rail contact modeling is of paramount importance for the dynamics of railway vehicles since it represents the interaction between the vehicle and the track. Although, in most cases, the contact generated occurs between convex surfaces which results in planar contact areas, the contact might take place in concave surfaces when negotiation sharp curves or due to the wear of profiles. In that cases, the resulting contact area is not planar. This work proposes a methodology to determine the shape of the contact patch in a curved surface, where the normal direction varies along its lateral direction. This method is based on a semi-Hertzian approach and discretizes the contact into longitudinal strips. The normal pressure distribution is computed in each strip separately using a non-Hertzian contact model and it is summed in a vector form to obtain the total normal force magnitude. Regarding the tangential forces, a look up table approach is considered. Finally, a trailer vehicle negotiating a curve is used to demonstrate the effectiveness of this methodology.

The vehicle-track interaction plays a role in railway dynamics, since the developed wheel-rail forces are responsible for supporting and guiding the vehicle. However, the existence of high wheel-rail contact forces or abnormal contact conditions typically leads to several surface degradation mechanisms like wheel flats. Hence, wheels and rails with defects on their contacting geometries tend to amplify the vehicle-track interaction forces deteriorating other railway components from both the vehicle and infrastructure. Multibody system dynamics methodologies allow simulating vehicles negotiating a given railway track at specified operating conditions, which can be of interest to determine the causes and consequences of certain defects. In these studies, the computation of wheel-rail contact forces is not only fundamental to accurately and efficiently analyze the vehicle’s dynamic response, but also for assessing the development of degradation phenomena, such wear, rolling contact fatigue or plastic deformation. This work proposes a methodology to deal with the contact detection between rails with constant profile and wheels with variable profile (i.e., with flats and out-of-roundness) in order to take into account the wheel defects on the vehicle’s dynamic response. Both wheel and rail potential contacting surfaces are parametrized. The wheel is divided into strips, and the contact is evaluated between each strip and a locally straight rail close to the wheel. With this methodology, it is obtained the virtual penetration for each strip along the contact patch which allows a better characterization of the effective contact area and apply a non-Hertzian contact force model. The proposed approach is applied to a running scenario in which one wheel of the vehicle is defective. The defect is artificially manufactured by cutting the wheel with a rail profile, which leads to material removal in the tread and transition zone. Further studies must be performed to completely assess and validate the methodology.

O uso de metodologias computacionais para estudar o comportamento dinâmico de veículos ferroviários é cada vez mais comum, pois permite analisar o desempenho do veículo numa dada via, ajudar a projetar sistemas de suspensão, travagem ou tração, estudar o dano dos componentes, entre outros. Neste tipo de ferramentas, a modelação da interação do veículo com a via, traduzindo-se no contacto entre as rodas e os carris, tem particular importância. Posto isto, o presente trabalho visa desenvolver métodos computacionais que permitam uma análise mais detalhada do contacto propondo avanços relativos aos métodos existentes na literatura. Para a deteção de contacto entre a roda e o carril, foi desenvolvida uma abordagem tridimensional para identificação dos pontos de maior interferência, na qual as geometrias da roda e do carril são definidas como superfícies paramétricas. Este método tira partido do facto de a roda ser um corpo de revolução e o carril uma superfície de arrastamento, permitindo identificar o contacto entre perfis gerais, sem a necessidade recorrer a simplificações que evitem a existência de uma zona côncava na roda, isto é, permite a abordagem de contacto entre superfícies conformes. Nas abordagens atuais ao problema do contacto roda-carril, assume-se que a área de contacto é plana, contudo, essa hipótese deixa de ser válida para contactos onde exista alguma conformidade. Nesse sentido, foi desenvolvida uma metodologia para a determinação de uma zona de contacto curva em que a direção normal à superfície de contacto varia ao longo da mesma. Este método permite evitar descontinuidades no cálculo das forças de contacto normais como é aqui demonstrado em diversos casos de teste para o contacto estático entre rodas e carris. A avaliação das forças de contacto tangencial é determinante na interação roda-carril, sendo aqui proposta a utilização de tabelas de pesquisa para o cálculo dessas forças para permitir a diminuição do tempo de cálculo. Neste caso, uma tabela de procura com cinco varáveis de entrada, que considera uma região de contacto definida por uma dupla elipse, é utilizada para estudar o erro de interpolação devido à sua discretização. Com a metodologia proposta é possível estudar a exatidão da tabela de pesquisa sendo, neste processo, apresentadas duas versões melhoradas desta tabela, com minimização de erro ou com minimização da sua dimensão. O modelo multicorpo de uma carruagem do Metro de Lisboa é utilizado para demonstrar as metodologias desenvolvidas neste trabalho através da realização de análises dinâmicas em condições de operação. A cinemática do veículo, a localização dos pontos de contacto, as forças de contacto e as variáveis de entrada da tabela de procura são analisados em detalhe. Os resultados mostram que a utilização de perfis de rodas realistas em detrimento de perfis simplificados tem um impacto significativo na dinâmica do veículo e no tipo e número de contactos desenvolvidos. Da mesma forma, o uso de modelos não-Hertzianos para o cálculo das forças de contacto também tem uma influência considerável no comportamento dos veículos ferroviários relativamente às abordagens correntes.

The utilization of multibody systems formulation allows to study the railway vehicle dynamics, as well study local damaging phenomena that occur in the wheel-rail interaction. However, the analysis of these phenomena requires long simulations to perform a significant prediction of their evolution. Having that in mind, the utilization of simplified approaches for the different steps of calculation of wheel-rail contact interaction are addressed. Two parametric surfaces are used to describe both wheel and rail geometries, namely a revolution and an extruded body, respectively. Regarding the normal and tangential contact forces, an elastic approach is considered, therefore, these forces are treated as external forces acting on the multibody system and an Hertzian model with damping is employed for their evaluation. Finally, a trailer vehicle is considered as example of application to study different modeling strategies.

The use of reliable computational tools and of validated vehicle and track models allow studying the railway vehicle performance in realistic operation conditions. The use of such advanced tools permits performing the so-called virtual homologation, which means that most of the criteria defined in the standards and regulations for vehicle acceptance can be verified numerically. This approach reduces the need of the expensive and long on-track tests, and also permits performing design optimization of several vehicle components, namely, the suspension elements, to improve its operational performance in terms of running safety, ride quality and track loading. The realism of the numerical simulations depends strongly on the model assumptions. In this work, all the mechanical elements that compose the rail vehicle are modeled properly using advanced methodologies. Then, a realistic and fully three-dimensional track, containing the measured track irregularities, is used. Finally, for a realistic running representation of the vehicle in the track, a prescribed motion of the motor wheel sets is adopted to adjust the vehicle speed as function of the track characteristics, namely, its curvature, cant angle and grade. The aim of this research is to develop a methodology to optimize the design of a rail vehicle in a mountainous track based on virtual homologation procedure. For this purpose, an optimization method is used to run the numerical simulations in batch mode and the dynamic performance of the rail vehicle is quantified based on the safety and ride quality indices defined in the standards. In addition, the optimization procedure uses a penalty term that penalizes cases where the vehicle presents an unacceptable dynamic behavior. The design variables considered are the suspension characteristics. This work provides an optimal design tool for the rail vehicle performance that leads to optimal dynamic performance in terms of running safety and ride quality.

The dynamic analysis of railway systems using multibody systems formulation is nowadays a reliable tool for several studies, in which the wheelrail contact interaction has significant importance. The definition of contacting geometries affects directly the dynamic response of the system. For sake of simplicity and computational efficiency, some authors simplify the wheel profile to avoid its concave part. This work intends to study the impact of utilizing a simplified non-conformal profile instead of a real wheel profile. For this purpose, an Hertzian-based contact force model is used to compute either the normal and creep forces. A wheelset rolling over a straight track is used as demonstration case. The results show that although the location of the contact points is similar, the use of simplified wheel profile may have an extensive impact on the dynamic response of the system.

The top velocity of high-speed trains is generally limited by the ability to supply the proper amount of energy through the pantograph-catenary interface. The deterioration of this interaction can lead to the loss of contact, which interrupts the energy supply and originates arcing between the pantograph and the catenary, or to excessive contact forces that promote wear between the contacting elements. Another important issue is assessing on how the front pantograph influences the dynamic performance of the rear one in trainsets with two pantographs. In this work, the track and environmental conditions influence on the pantograph-catenary is addressed, with particular emphasis in the multiple pantograph operations. These studies are performed for high speed trains running at 300 km/h with relation to the separation between pantographs. Such studies contribute to identify the service conditions and the external factors influencing the contact quality on the overhead system. (C) 2013 Elsevier Ltd. All rights reserved.

The wheel-rail contact modeling problem assumes a preponderant role on the dynamic analysis of railway systems using multibody systems formulations. The accurate and efficient evaluation of both location and magnitude of the wheel-rail contact forces is fundamental for the development of reliable computational tools. The wheel concave zone might be a source of numerical difficulties when searching the contact points, which has been neglected in several works. Here, it is demonstrated that the minimum distance method does not always converge when the wheel surface is not fully convex, being an alternative methodology proposed to perform the contact detection. This approach examines independently the contact between each wheel strip and the rail, where the maximum virtual penetration is determined and associated with the location of the contact point. Then, an Hertzian-based force model is considered for both normal and tangential forces. The results obtained from dynamic simulations show that the minimum distance method and the proposed methodology provide a similar response for simplified wheel profiles. However, the new approach described here is reliable in the identification of the contact point when realistic wheel profiles are considered, which is not the case with the minimum distance method.