Steel Railway Bridges Research Articles

Fatigue Damage Identification by a Global-Local Integrated Procedure for Truss-Like Steel Bridges

Civil steel structures and infrastructures, such as truss railway bridges, are often subject to potential damage, mainly due to fatigue phenomena and corrosion. Therefore, damage detection algorithms should be designed and appropriately implemented to increase their structural health. Today, the vast amount of information provided by data processing techniques and measurements coming from a monitoring system constitutes a possible tool for damage identification in terms of both detection and description. For this reason, the research activity aims to develop a methodology for a preliminary description of the damage in steel railway bridges induced by fatigue phenomena. The proposed approach is developed through an integration of global and local procedures. At the global scale, vibration-based procedures will be applied to improve a forecast numerical model and, subsequently, to identify the zones most involved in fatigue problems. At the local scale, careful and refined local identification will be pursued via image processing techniques whose evidence will be analyzed and described through nonlinear numerical models. A case study of a historical railway bridge in Spain will illustrate the methodology’s performance, potentiality, and critical issue.

Probabilistic fatigue life assessment of steel rail-bridge in dual response surface framework

A new probabilistic fatigue life assessment (PFLA) procedure is proposed considering dynamic effects of train loading, possible enhancement of axle load, and uncertainty there-in. In conventional PFLA, Monte Carlo simulation (MCS) is used to simulate stochastic load. Thereafter, rainflow counting method is used to obtain mean stress ranges. However, it is well-established now that only mean stress range may not depict the detailed stochastic nature of train axle load time-histories. Moreover, such MCS-based PFLA is computationally demanding for real-life complex structures, and routine implementation of it is quite troublesome due to dearth of skilled engineers in advanced fatigue analysis and interface software. To circumvent this, the proposed PFLA procedure resorts to dual response surface method (DRSM), which yields explicit equations for mean as well as standard deviation of stress range. Using these equations, the proposed approach directly relates fatigue life with the input factors (train speed, axle load, etc.) through an excel spreadsheet without requiring MCS. The effectiveness of the proposed procedure is elucidated through an existing steel railway bridge in India. The accuracy of the approach is also validated with four other steel bridges practically failed in fatigue. The results show that the proposed PFLA is in excellent agreement with the most accurate direct MCS-based PFLA, and the practical data obtained from failed steel bridges; whereas, the conventional moving load analysis-based deterministic fatigue life assessment approach yields significantly deviated results. Also, the proposed DRSM-based PFLA approach is observed to be computationally more efficient than the conventional direct MCS-based PFLA.

A local response function approach for the stress investigation of a centenarian steel railway bridge

Local stress concentrations in some specific geometric details often play an essential role in time-dependent damage propagation. However, measured strain or stress data are generally available in locations that are not the most critical ones. Proper numerical models are thus required to aid the damage severity assessment. In this paper, a surrogate model procedure based on the local response function method is proposed to accurately predict stress time histories in particular regions of bridges through a multiscale numerical analysis. The methodology is calibrated on the basis of direct strain measurements on a railway bridge in Vänersborg, Sweden. The bridge is currently monitored with a set of strain gauges, accelerometers, and an inclinometer allowing the investigation of its in-service conditions during the train passages and the lifting of the bascule truss. The results show the reliability of the method for stress prediction in specific regions of the bridge without the availability of directly measured data together with more detailed knowledge about the stress histories that could lead to fatigue or corrosion damage.

Remaining fatigue endurance of gusset plate connections from a dismantled railway bridge

The replacement of an existing steel railway bridge in Sweden opened up for fatigue testing connections with a history of about 70 years of in-service loading. One decisive detail in the bridge was the gusset plates welded to the flange edge of I beams. This has low fatigue resistance in the governing standards and was deemed exhausted in theoretical assessments. To estimate the remaining fatigue endurance, 20 similar samples of these were cut out from the bridge and tested. During service, they were exposed to alternating compression and tension stresses. The testing was designed to determine their category for constant amplitude fatigue loading and to evaluate the influence of compressive stresses. It is shown that the compressive stresses have an influence on the fatigue endurance and a correction function has been fitted to the results. Despite many years of loading, they had a fatigue endurances significantly greater than that estimated theoretically. A detail category of 63 was derived, which, together with the correction function, provides a significant increase in fatigue life, which should be possible to use for similar details in other existing steel structures.

Intensity and location of corrosion on the reliability of a steel bridge

Visual inspections generally represent the initial step of the safety assessment of bridges. In particular, steel railway bridges are particularly prone to corrosion phenomena. Therefore, visual inspections provide the geometric characterization of the conservation state of each structural member, in particular, the size reduction originating from the progress of corrosion. Nonetheless, the geometric definition of corrosion is not directly related to structural safety, depending on the damage location. The paper quantifies and discusses the uncertainties in the safety assessment of trussed bridges associated with the unknown effect of damage location on the stress distribution, given initial knowledge of the number and intensity of damaged members. Is it possible to preliminary assess the structural safety for typological structures without knowing the effect of damage location on the stress re-distribution? This paper discusses these aspects by randomly simulating damage scenarios in a finite element model of a historic steel bridge. The model, validated against the experimental modal parameters and measured deflection under train transit, is used for estimating bidimensional fragility curves as a function of the damage intensity and the number of damaged members. Finally, a safety domain can support the preliminary safety assessment of typological steel railway bridges.

Multistrategy Cooperation Particle Swarm Optimization for FEM Model Update of the Continuous Warren Truss Steel Railway Bridge

It is very difficult to obtain an accurate finite element method (FEM) model to further analyze structural mechanical properties. Therefore, as the main means of establishing accurate models, the model update has become a research hotspot in the dominion of bridge engineering. Particle swarm optimization (PSO) has the characteristics of being easy to implement, but it is easy to fall into the local optimum. Therefore, multistrategy cooperation particle swarm optimization (MCPSO) that balances exploration and exploitation of particle swarm is proposed. This algorithm achieves the goal of balancing exploration and exploitation by adopting different combinations of particle swarm velocity update strategies in different iteration stages. The application effects of MCPSO in the FEM model update of the continuous Warren truss steel railway bridge are compared and analyzed, and the results show that the algorithm proposed in this paper outperforms the standard PSO (SPSO) algorithm. This paper provides a more effective algorithm for bridge model updates.

Validation of virtual sensing for the reconstruction of stresses in a railway bridge using field data of the KW51 bridge

Virtual sensing enables a model-based extrapolation of the response measured in a limited number of locations (accelerations, strains, etc.) to other points of interest on the structure. One possible application of virtual sensing consists of estimating the stresses in critical but unmeasured details for assessing fatigue accumulation of structures exposed to cyclic or dynamic loading. Since virtual sensing involves the combined use of response data and a model of the structure, the outcome is inherently prone to modeling errors. This paper investigates the use of virtual sensing for stress/strain monitoring on steel railway bridges. The analysis is based on strain measurements obtained from an in situ monitoring campaign on railway bridge KW51 in Leuven, Belgium. It is found that the quasi-static response of the structure is largely affected by the behavior of the pot bearings, which cannot be easily modeled. A methodology is proposed to avoid the incorporation of the bearings in the model adopted in the virtual sensing, where the strain data are extended by longitudinal displacement measurements at the bearings. It is shown that this methodology enables accurate strain prediction, where in this case similar results are obtained for a model based on the blueprints only and a model calibrated based on experimental modal data.

Ultimate and fatigue limit states of existing steel railway bridges – LRFD with historical steel products and connection types

AbstractThe assessment of the load‐bearing capacity and fatigue strength of existing railway bridges has been playing an increasingly important role in the infrastructure management of railway operators for several years now. Currently, many bridge structures have been in operation longer than it was foreseen during their planning and construction. In addition, the axle loads on many lines, as well as the demands on the reliability of the verification results, have steadily increased. As the material properties and construction techniques in existing structures differ to some extent from nowadays structures, e.g., riveting instead of welding, it is important to provide engineers and operators with recommendations for the assessment of existing steel bridges. This article summarises the studies conducted as part of a research project initiated by Deutsche Bahn Netz AG for possible updates to DB RiL 805, which is used for the verification of railway bridges in the Deutsche Bahn (DB) network. The studies concerned the transition of verification concepts against static and fatigue loads used in the past to limit state verifications with partial safety factors in accordance with the Eurocodes. While initially related to an upgrade of a specific operator's design recommendation, the findings in this article are of more general nature and could form the basis for similar developments of recommendations for the assessment of existing, riveted structures independently throughout Europe.

3D Displacement Measurement of Railway Bridge According to Cyclic Loads of Different Types of Railcars with Sequential Photogrammetry

In the early days of railroads in Korea, railway bridges were constructed as steel plate-girder structures, which are vulnerable to vibration and torsion. Many of these bridges have since been replaced with concrete-slab structures, which have high stability. Nevertheless, steel railway bridges still remain all over the country, and a lot of manpower and cost is being invested in the maintenance and repair of such bridges. Moreover, there have not been experimental analyses aiming to measure the cyclic loads that occur when a train enters. To ensure bridge safety, it is necessary to periodically inspect deformations. To this end, the present study proposed a sequential photogrammetric technique for measuring the deformation of a steel railway bridge for three types of railcars. Sequential stereo images of the bridge with multiple feature points are obtained using sequential photographing cameras, to determine the ground coordinates of each point as a function of time based on the space intersection from the relative orientation with coplanarity and the scale adjustment. All of these processes are performed through automated techniques using only the cameras and the targets. With this setup, the 3-dimensional dynamic motions of the bridge due to the cyclic loading of trains could be measured. In addition, the displacements by the proposed method were compared to those obtained with the 3D Laser tracker. The horizontal displacements errors did not exceed 0.5 mm and the vertical error was within 2.3 mm in root mean square error (RMSE) at camera-to-object distances of about 9 m.

Mitigation of model error effects in neural network-based structural damage detection

This paper proposes a damage detection procedure based on neural networks that is able to account for the model error in the network training. Vibration-based damage detection procedures relied on machine learning techniques hold great promises for the identification of structural damage thanks to their efficiency even in presence of noise-corrupted data. However, it is rarely possible in the context of civil engineering to have large amount of data related to the damaged condition of a structure to train a neural network. Numerical models are then necessary to simulate damaged scenarios. However, even if a finite element model is accurately calibrated, experimental results and model predictions will never exactly match and their difference represents the model error. Being the neural network tested and trained with respect to the data generated from the numerical model, the model error can significantly compromise the effectiveness of the damage detection procedure. The paper presents a procedure aimed at mitigating the effect of model errors when using models associated to the neural network. The proposed procedure is applied to two case studies, namely a numerical case represented by a steel railway bridge and a real structure. The real case study is a steel braced frame widely adopted as a benchmark structure for structural health monitoring purposes. Although in the first case the procedure is carried out considering simulated data, we have taken into account some key aspects to make results representative of real applications, namely the stochastic modelling of measurement errors and the use of two different numerical models to account for the model error. Different networks are investigated that stand out for the preprocessing of the dynamic features given as input. Results show the importance of accounting for the model error in the network calibration to efficiently identify damage.

Influence of frost and local stiffness variations on the strain mode shapes of a steel arch bridge

Natural frequencies are probably the most widely used modal characteristics in vibration-based monitoring. However, they can be highly influenced by temperature and this influence can completely mask the effect of even severe damage. This translates into a necessity for time-consuming data-normalization techniques to remove the influence of temperature and identify damage. Displacement mode shapes of homogeneous material structures are less influenced by temperature, but obtaining them in a dense grid, which is required for damage localization, is cumbersome due to the large number of sensors needed. Strain mode shapes on the other hand can be insensitive to temperature variations, while obtaining them in a dense grid is possible when fiber-optic sensors such as fiber-Bragg gratings (FBG) are used. This work presents the results of the continuous monitoring of a steel railway bridge for a period of almost two years, where modal data were collected for a wide temperature range. The bridge is instrumented with eighty FBG strain sensors, multiplexed in four fibers. The natural frequencies and strain mode shapes of ten modes have been automatically identified from operational strain time histories, on an hourly basis. A clear influence of temperature on the natural frequency of most modes is identified, especially during frost periods. On the contrary, the strain mode shapes are mostly insensitive to temperature changes and only these of some higher-order modes are slightly and uniformly influenced when frost occurs. This behavior is confirmed also by a finite element model (FE) of the bridge. Furthermore, the FE model is used to investigate the influence of local stiffness changes on the modal characteristics. A clear and local change of the modal strain amplitude is observed at the location of the reduced stiffness, especially when information from all modes is combined in a sensitive damage index.

Damage detection of in-service steel railway bridges using a fine k-nearest neighbor machine learning classifier

Minor areas of surface corrosion in steel railway bridges can grow progressively andlead to localized section losses and structural failure over time. This paper proposes a novel combined damage detection approach for the classification of various extents and degrees of cross section losses due to damages like corrosion using a k-Nearest Neighbor (kNN) machine learning classifier. A Finite Element (FE) model of an in-service railway bridge is developed and validated using vibration data from field testing and these combined FE-field data are trained and tested to classify various corrosion cases following the Australian Standard AS7636. The results show that the proposed technique is practical and highly accurate in classifying damages in steel railway bridges, even if a minor level of steel corrosion is intended to be classified. Furthermore, a comparison between the accuracies of kNN classifier and the Radial Basis Function (RBF) Gaussian kernel Support Vector Machine (SVM) is presented.

Design-related reassessment of structures integrating Bayesian updating of model safety factors

In the semi-probabilistic approach of structural design, the partial safety factors are defined by considering some degree of uncertainties to actions and resistance, associated with the parameters’ stochastic nature. However, uncertainties for individual structures can be better examined by incorporating measurement data provided by sensors from an installed health monitoring scheme. In this context, the current study proposes an approach to revise the partial safety factor for existing structures on the action side, γE by integrating Bayesian model updating. A simple numerical example of a beam-like structure with artificially generated measurement data is used such that the influence of different sensor setups and data uncertainties on revising the safety factors can be investigated. It is revealed that the health monitoring system can reassess the current capacity reserve of the structure by updating the design safety factors, resulting in a better life cycle assessment of structures. The outcome is furthermore verified by analysing a real life small railway steel bridge ensuring the applicability of the proposed method to practical applications.

Experimental validation and numerical investigation of virtual strain sensing methods for steel railway bridges

Railway bridges are subjected to intermittent excitations induced by gravity loads. The monitoring of stress time histories at fatigue prone areas (i.e., “hot spots”) for steel, heavy haul railway bridges is vital for cost-effective maintenance and, most importantly, for maintaining safety. In this regard, strain measurements are crucial for determining fatigue susceptibility and, possibly, damage to steel bridges under daily, cyclical loading conditions. While, in the limit, sensing the response of a steel bridge under operational conditions is the best way to identify hot spots to prevent damage and, possibly, collapse, it is impractical if not impossible to install strain sensors at every susceptible location. To address this issue, the current study focuses on virtual sensing of strain time histories on steel railway bridges using sparse response measurements. An augmented Kalman filter (AKF) method was adopted for input-state estimation. Since AKF estimates unknown load and response using a physical model, it is crucial to assess the effects of modeling uncertainties on estimation results. In addition to AKF, Modal Expansion (ME) was adopted for extrapolation of measured response to unmeasured locations. In contrast to AKF, which requires a full physical model, ME only relies on vibration modes. A novel application of the Singular Value Decomposition (SVD) method that facilitates data-driven strain prediction by relying on data obtained from field measurements or models was also developed and examined. The study was completed using simulated experiments and strains measured from an in-service, through girder, steel railway bridge. The three methods were compared and strengths and limitations of each were highlighted.

Damage classification of in-service steel railway bridges using a novel vibration-based convolutional neural network

Railway bridges exposed to extreme environmental conditions can gradually lose their effective cross-section at critical locations and cause catastrophic failure. This paper has proposed a practical vibration-based deep learning approach for damage classification of various extents and degrees of cross section losses due to damages like corrosion in operational railway bridges using vibration-based Convolutional Neural Networks (CNN)s. Firstly, field testing of an in-service railway bridge is conducted and the modal parameters of the bridge are obtained to validate the developed Finite Element (FE) model of the bridge. In the next phase, corrosion scenarios of the main bridge members are generated as various quantities of cross section losses of these members by the validated FE following the Australian Standard AS7636. In the deep learning part, a 1D CNN aligned with novel specific data augmentation strategies is developed to classify various acceleration responses related to each damage scenario simulated by the validated FE. Furthermore, a visualization of feature extraction and feature mapping using t-Distributed Stochastic Neighbour Embedding (t-SNE) and Gradient-weighted Class Activation Mapping (Grad-CAM) is illustrated. The case studies on the simulated and field data-validated FE model results applying background noises and variations, and the real field testing data suggest that the proposed method can reach a perfect damage classification close to an accuracy of 100%.

Optimum Stiffness of the Sleeper Pad on an Open-Deck Steel Railway Bridge using Flexible Multibody Dynamic Analysis

Optimum Stiffness of the Sleeper Pad on an Open-Deck Steel Railway Bridge using Flexible Multibody Dynamic Analysis

Pre- and Post-Rehabilitation Monitoring of a Steel Railway Bridge with Distributed and Discrete Sensors

Pre- and Post-Rehabilitation Monitoring of a Steel Railway Bridge with Distributed and Discrete Sensors

Fatigue performance of the welded details of an old, demolished steel railway bridge

This study investigates the fatigue performance of the welded details of an old disused steel railway bridge. The Saimaa canal railway bridge was demolished after 60 years of service, and as a part of the demolition process, a fillet-welded vertical web stiffener joint of an I-profile girder was taken for experimental testing. Fatigue tests were carried out on the joints tested in both as-welded and post-weld-treated conditions, using burr grinding, high-frequency mechanical impact (HFMI) treatment, and TIG dressing as improvement techniques. To examine the reduction in stress concentration due to the post-weld treatments, numerical analyses were performed as per the effective notch stress (ENS) approach and the theory of critical distance (TCD) methods, and different local approaches were used to assess the fatigue strength of the welded details. According to the obtained results, the fatigue strength of the joints in the as-welded condition was higher than that of recommended by fatigue design standards, indicating good welding quality and suggesting that the tested welded details had not experienced severe fatigue loads during service. HFMI treatment and burr grinding improved the fatigue strength by 30%, meanwhile no failures in the joint area were observed with the TIG dressing. Of the applied local approaches, the structural hot-spot stress method, 4R method, and ENS concept showed the highest accuracy in the fatigue assessments, and had lower scatter than the TCD methods.

Study of Railway Steel Bridges’ Behaviour in Order to Identify the Causes of Their Defects

Study of Railway Steel Bridges’ Behaviour in Order to Identify the Causes of Their Defects

Digital twinning of self-sensing structures using the statistical finite element method

Abstract The monitoring of infrastructure assets using sensor networks is becoming increasingly prevalent. A digital twin in the form of a finite element (FE) model, as commonly used in design and construction, can help make sense of the copious amount of collected sensor data. This paper demonstrates the application of the statistical finite element method (statFEM), which provides a principled means of synthesizing data and physics-based models, in developing a digital twin of a self-sensing structure. As a case study, an instrumented steel railway bridge of $ 27.34\hskip1.5pt \mathrm{m} $ length located along the West Coast Mainline near Staffordshire in the UK is considered. Using strain data captured from fiber Bragg grating sensors at 108 locations along the bridge superstructure, statFEM can predict the “true” system response while taking into account the uncertainties in sensor readings, applied loading, and FE model misspecification errors. Longitudinal strain distributions along the two main I-beams are both measured and modeled during the passage of a passenger train. The statFEM digital twin is able to generate reasonable strain distribution predictions at locations where no measurement data are available, including at several points along the main I-beams and on structural elements on which sensors are not even installed. The implications for long-term structural health monitoring and assessment include optimization of sensor placement and performing more reliable what-if analyses at locations and under loading scenarios for which no measurement data are available.