The subject of this master's thesis is a static and structural design of a Load-carrying steel structure of railway bridge with bottom steel deck of the span of 42 + 60 + 42 m over the Labe river in Děčín. A solid steel beam forms the main load-bearing structure of the bridge reinforced with truss-work. The truss-work is made up of an open top chord and diagonal strut without secondary vertical struts. The calculations were done in compliance with valid ČSN EN documents.

A method that uses machine learning to detect and localize damage in railway bridges under various environmental conditions is proposed and validated in this work. The developed algorithm uses vertical and lateral deck accelerations as damage- sensitive features. Indeed, an Artificial Neural Network (ANN) is trained to predict deck accelerations in undamaged condition given: previous vibration data, air temperature and characteristics of the train crossing the bridge (speed, load position and load magnitude). After an appropriate training period, the comparison between ANN-predicted and measured accelerations allows to compute prediction errors. A Gaussian Process is then used to stochastically characterize prediction errors in undamaged conditions using train speed as independent variable. Recorded vibration data leading to abnormal prediction errors are flagged as damage. The method is validated both on a simple numerical example and on data recorded on a real structure. In the latter case, an appropriate algorithm was developed with the aim of extracting vehicles characteristics from the acceleration time histories. Together with this part of the algorithm for the pre-processing of recorded accelerations, the novelty of the developed method is the addition of air temperature to the input. It allows separating between structure responses that can be flagged as damage from those only affected by environmental conditions.

A constant research of more ecological and efficient structures has enabled bridges to be more innovative through the years. Nowadays, as the need is greater than ever, a new kind of bridge is expanding in the entire world: the network arch bridges. The concept was developed by professor and engineer Per Tveit in 1955 and has been improved since then. But it is only for 10 years that many bridges of this sort have been built. The aim of the thesis is to investigate the structural behavior of these bridges and their efficiency comparing to traditional bridges. It is also proving the efficiency of the network arches used for rail traffic. To do so, a single-track railway network arch has been designed according to the Eurocodes. A 2D model has been designed to be optimal and tested under Abaqus for the loads defined in the Eurocodes. Guidelines from the literature and Per Tveit’s work have been used to determine the optimal geometry of the bridge. The steel weight needed for the 75 meters long bridge is assessed from the final design and is the main parameter to compare and evaluate the network arch structural efficiency. These results are compared with Tveit’s statements and with other structures.

With the development of high-speed railroads the dynamic behaviour of railroad bridges is increasingly important to explore. Deeper knowledge about the influence of different factors and what should be included in a model is essential if the designer shall be able to make reliable estimates of responses in existing and new structures. One factor is the soil-structure interaction (SSI), describing how the foundation of the bridge and the soil properties affect the behavior of the bridge under dynamic loading. In this thesis, the influence of including SSI in a model of a portal frame railway bridge is studied, and an analysis procedure in the frequency domain for models with frequency-dependent boundary conditions is described. A 3D finite element model of an e isting bridge has been built up, based on the theory of linear elasticity. The model has been given three different types of boundary conditions: clamped, static stiffness and frequency-dependent stiffness from SSI. Results from simulated train passages, with a train set consisting of two wagons, were compared for the different boundary conditions. The models have also been compared with measurement data from the bridge, which has given indications about which model describes reality in the best way. The results show that the model in which SSI is included by frequency dependent boundary conditions is in slightly better agreement with measurement data than the clamped model and the model with static stiffness. The model gives a slightly better damping of the free vibrations and the natural frequencies correspond better with experimental data. The difference in maximum acceleration from a train passage is very small between the different models, even if it is found that the clamped model generally has lower accelerations and hence is non-conservative. It appears that the train speed affects the maximum acceleration, the size of the free vibrations and the natural frequencies that are present in the free vibrations in the models. Further studies are suggested where it is emphasized that an analysis with longer trains, which give resonance phenomena, should be made to see how the different eigenfrequencies in the models affect the accelerations at different speeds. It is also noted that more measurements would be needed in order to draw more general conclusions about the degree of correspondence between the measurements and the models, and to calibrate the parameters of the model against measurement data.

A damage detection approach based on Artificial Neural Network (ANN), using the statistics of structural dynamic responses as the damage index, is proposed in this study for Structural Health Monitoring (SHM). Based on the sensitivity analysis, the feasibility of using the changes of variances and covariance of dynamic responses of railway bridges under moving trains as the indices for damage detection is evaluated.   A FE Model of a one-span simply supported beam bridge is built, considering both single damage case and multi-damage case. A Back-Propagation Neural Network (BPNN) is designed and trained to simulate the detection process. A series of numerical tests on the FE model with different train properties prove the validity and efficiency of the proposed approach. The results show not only that the trained ANN together with the statistics can correctly estimate the location and severity of damage in the structure, but also that the identification of the damage location is more difficult than that of the damage severity. In summary, it is concluded that the use of statistical property of structural dynamic response as damage index with the Artificial Neural Network as detection tool for damage detection is reliable and effective.

New railway lines are continuously being constructed and existing lines are upgraded. Hence, there is a need for research directed towards efficient design of the supporting structures. Increasingly advanced calculation methods can be motivated, especially in projects where huge savings can be obtained from verifying that existing structures can safely support increased axle loads and higher speeds. This thesis treats the dynamic response of bridges under freight and passenger train loads. The main focus is the idealisation of the train load and its implications for the evaluation of the vertical bridge deck acceleration. To ensure the running safety of train traffic at high speeds the European design codes set a limit on the vertical bridge deck acceleration. By considering the train–bridge interaction, that is, to model the train as rigid bodies on suspension units instead of constant moving forces, a reduction in bridge response can be obtained. The amount of reduction in bridge deck acceleration is typically between 5 and 20% for bridges with a span up to 30 m. The reduction can be higher for certain train–bridge systems and can be important also for bridge spans over 30 m. This thesis aims at clarifying for which system parameter combinations the effect of train–bridge interaction is important. To this end, a thorough literature survey has been performed on studies in train–track–bridge dynamics. The governing parameters in 2D train–bridge systems have been further studied through a parameter screening procedure. The two-level factorial methodology was applied to study the effect of parameter variations as well as the joint effect from simultaneous changes in several parameters. The effect of the choice of load model was thus set in relation to the effect of other parameter variations. The results show that resonance can arise from freight train traffic within realistic speed ranges (&lt; 150 km/h). At these resonance peaks, the reduction in bridge response from a train–bridge interaction model can be considerable. From the screening of key parameters it can furthermore be concluded that the amount of reduction obtained with a train–bridge interaction model depends on several system parameters, both for freight and passenger train loads. In line with the European design code’s guidelines for dynamic assessment of bridges under passenger trains an additional amount of damping can be introduced as a simplified way of taking into account the reduction from train–bridge interaction. The amount of additional damping is today given as function of solely the bridge span length, which is a rough simplification. The work presented in this thesis supports the need for a refined definition of the additional damping. / Nya järnvägslinjer byggs kontinuerligt och befintliga linjer uppgraderas. Det finns därför ett behov av forskning inriktad på effektiv design av de bärande konstruktionerna. Alltmer avancerade beräkningsmetoder kan vara motiverade, särskilt i projekt där stora besparingar kan erhållas från att verifiera att befintliga konstruktioner kan bära ökade axellaster och högre hastigheter. Föreliggande avhandling behandlar broars dynamiska respons under belastning av gods- och passagerartåg. Huvudfokus är att studera modelleringsalternativ för tåglasten och vilka konsekvenser de har för utvärderingen av brobanans vertikala acceleration. För att garantera trafiksäkerhet vid höga tåghastigheter definierar de europeiska normerna en maximalt tillåten vertikal acceleration i brobanan. Genom att beakta tåg-bro-interaktion, där tågkomponenterna modelleras som avfjädrade stela kroppar istället för konstanta punktlaster, kan en minskning av brons respons erhållas. Reduktionen av brobanans acceleration är typiskt mellan 5 och 20% för broar med en spännvidd på upp till 30 m. Minskningen kan vara högre för vissa tåg-brosystem och kan vara viktigt också för spännvidder över 30 m. Denna avhandling syftar till att klargöra för vilka kombinationer av tåg-broparametrar effekten av tåg-bro-interaktion är viktig. I detta syfte har en omfattande litteraturstudie genomförts inom området tåg-spår-brodynamik. De styrande parametrarna i 2D tåg-brosystem har studerats vidare i en parameterstudie. Två-nivå faktorförsök har tillämpats för att studera effekten av parametervariationer samt den ytterligare effekten av samtidiga förändringar i flera parametrar. Effekten av valet av lastmodell sattes därmed i relation till effekten av andra parametervariationer. Resultaten visar att resonans kan uppstå från godstrafik inom ett realistiskt hastighetsintervall (&lt; 150 km/h). Vid dessa resonanstoppar kan en betydande minskning av broresponsen erhållas med en tåg-bro-interaktionsmodell. Från studien av nyckelparametrar kan man vidare dra slutsatsen att reduktionen som erhålls med en tåg-bro-interaktionsmodell beror på flera systemparametrar, både för gods- och passargerartåg. Enligt de europeiska normernas rekommendationer för dynamisk kontroll av broar för passagerartrafik kan en ökad brodämpning introduceras som ett förenklat sätt att ta hänsyn till minskningen från tåg-bro-interaktion. Mängden tilläggsdämpning anges idag som en funktion av enbart brons spännvidd, vilket är en grov förenkling. Det arbete som presenteras i denna avhandling visar på behovet av en förbättrad definition av tilläggsdämpningen. / <p>QC 20140429</p>

This thesis deals with a design of a two-track railway frame bridge over a valley. Out of three proposed variants was chosen the variant number 1 - prestressed frame construction with variable beam section. According to actual EN standards the construction was designed and assessed to the serviceability limit state and the ultimate limit state.

The master thesis focuses on the design of a Load-carrying steel structure of railway bridge with bottom steel deck of the span of 28 + 45 + 23 m. The bridge is situated in the municipality of Třebestovice. Four variants of solution have been calculated. Two variants are made by plate main beams of variable profile, the next alternative was made by truss structure and the fourth variant consist of a combination of plate main beams with truss stiffener in the middle span. The most optimal solution was processed in detail. The calculations were done in compliance with valid ČSN EN norms.

The thesis deals with the variant design of the steel structure of single-track railway bridge on the line Tišnov – Nové Město na Moravě. The task is bridging the local road by a bridge structure with one field. Length of the bridge is 40 m.

The subject of master’s thesis is a structural analysis of a bearing structure of an existing historical single-track railway bridge over the Vltava river in a direction from Prague Modřany to Vrané nad Vltavou. The riveted steel bridge carries rail traffic across five fields. A part of the master’s thesis is focused on a disposal of a variant of the longest field by using a modern arch construction.