The UK has a long history in the railway industry with a large number of railway assets. Railway bridges form one of the major asset groups with more than 35,000 bridges. The majority of the bridge population are old being constructed over 100 years ago. Many of the bridges were not designed to meet the current network demand. With an expected increasing rate of deterioration due to the increasing traffic loads and intensities, the management authorities are faced with the difficult task of keeping the bridge in an acceptable condition with the constraint budget and minimum service disruptions. Modelling tools with higher complexity are required to model the degradation of assets and the effects of different maintenance strategies, in order to support the management decision making process. This research aims to address the deficiencies of the current bridge condition systems and bridge models reported in the literature and to demonstrate a complete modelling approach to bridge asset management. The degradation process of a bridge element is studied using the historical maintenance data where previous maintenance actions were triggered by a certain type of defects. Two bridge models are then developed accounting for the degradation distributions, service and inspection frequency, repair delay time and different repair strategies. The models provide a mean of predicting the asset future condition as well as investigating the effects of different maintenance strategies will have on a particular asset. The first model is a continuous-time Markov bridge model and is considered more complex than other models in the literature, the model demonstrates the advantages of the Markov modelling technique as well as highlighting its limitations. The second bridge model presented a novel Petri-Net modelling approach to bridge asset modelling. This stochastic modelling technique allows much more detail modelling of bridge components, considering: non-constant deterioration rates; protective coating modelling; limits of the number of repairs can be carried out; and the flexibility of the model allows easily extension to the model or the number of components modelled. By applying the two models on the same asset, a comparison can be made and the results further confirm the validations and improvements of the presented Petri-Net approach. Finally, optimisation technique (Genetic Algorithm) is applied to the bridge models to find the optimum maintenance strategies in which the objectives are to minimise the whole life cycle cost whist maximising the asset average condition. A hybrid optimisation that takes advantage of both bridge models, resulting in a significant time saving, is also presented.

No description available.

The buffeting response is a type of vibration caused by wind turbulence. As the bridge span and structural complexity increase, this kind of response is notable. Therefore more accurate analysis simulation methods are needed to investigate this aerodynamic phenomenon. The aim of this thesis is to review, discuss and compare the different bridge buffeting simulation approaches in the frequency domain and find the possibility of Computational Fluid Dynamic (CFD) method application in bridge buffeting prediction. In this thesis, the conventional bridge buffeting statistical analysis methods considering the influence of different parameters such as mode coupling, self-excited forces and aerodynamic admittance on the simulation results are firstly reviewed and compared. Since wind turbulence may not excite all structural vibration modes in some frequency ranges, an alternative approach based on the Proper Orthogonal Decomposition (POD) is proposed to study the effective turbulence contribution to the structural vibration. However due to the complexity of turbulence, quasi-steady theory is widely adopted and some semi-empirical functions such as aerodynamic admittance, joint acceptance are introduced to simplify the simulation. With the development of CFD method, CFD simulation of bridge aerodynamic phenomena has become possible. Since the bridge buffeting response is induced by wind turbulence, it is very important to capture the time varying characteristics of wind turbulence. In CFD technique, to close the Navier-Stokes equations and reflect unsteady characteristics, turbulence modelling is always adopted. At present Direct Numerical Simulation (DNS) is an accurate model to capture the time variation unsteady characteristic of the wind turbulence. However, the problems of civil engineering are always high Reynolds numbers, which make the simulation of aerodynamic phenomena of civil engineering impractical based on DNS method. Therefore an alternative model, known as Large Eddy Simulation (LES), becomes popular. In this thesis an unsteady inlet boundary generation technique based on an Autoregressive Moving Average (ARMA) model is proposed to simulate the unsteady inflow turbulence. 3D LES will be selected to validate the applicability of this model in the prediction of the unsteady characteristic of buffeting simulation by investigating the flow characteristic around a square cylinder under different mesh density and different LES model such as standard LES model, dynamic LES model and WALE model at Reynolds number 13,000. Before comparing the influence of inflow boundary condition with turbulence intensity (5%) on the flow around a square cylinder, an empty domain is selected to validate current inflow turbulence generation technique. Major steps of Fluid Structure Interaction (FSI), suitable for future simulation of bridge structural buffeting response, are proposed to predict the structural buffeting response induced by the inflow turbulence. To test the propose procedure of FSI, a square cylinder will be used, the across-flow oscillation of square cylinder with constant damping ratio will be considered to investigate the influence of steady inlet boundary condition and unsteady one on the response of structure. In addition different LES models are considered to compare their influence on the response of square cylinder.

This thesis reports the findings of an investigation into the feasibility of using vibration characteristics to monitor the structural health of bridges. The study is the second part of a larger project commissioned by the UK Highways Agency into the investigation of possible monitoring methods that can be used in a pass/fail/monitor inspection programme. To this end, ten one-quarter-scale 5m span reinforced concrete bridge decks were fabricated and loaded incrementally to failure in the laboratory. The dynamic properties of the decks were investigated at each of the loading increments to evaluate their sensitivity to structural cracking using both free and forced vibration. The results indicated that, for the specimens tested, natural frequencies were, in general, more sensitive to the damage introduced than mode shapes. It was found that the support conditions affected the dynamic behaviour of the decks, and indeterminate boundary conditions caused significant variation in the vibration characteristics. This presented several problems in the analysis of the modal properties and, when combined with the damage introduced through static loading, caused some modes to disappear and new modes to be measured, whilst a number of modes also displayed an increase in natural frequency. The application of finite element model updating to determine reduction in flexural stiffness in the damaged areas of the deck provided a systematic method to investigate the condition of the deck. Updating was performed based on the natural frequencies of one symmetrically and one asymmetrically loaded deck, and the cracking observed under the loading, and offered results consistent with expectations. In summary, the evidence presented in this thesis suggests that the natural frequencies of the decks are, in general, more sensitive to the damage introduced than the mode shapes and consistent trends can be observed in the natural frequency change as the damage to the deck increases. However, the application of this method to indicate the structural condition of real bridges may be limited without further investigation as the vibration characteristics were affected by a number of factors arising from the realistic nature of the specimen, such as the three-dimensional distribution of the damage and the indeterminate nature of the support conditions.

It has been two decades since Ultra High Performance Fibre Reinforced Concrete (UHPFRC) has come to the market and, so far, it has been used in only a limited number of highway bridge structures. This is due to many unanswered questions related to its structural behaviour for highway bridge structures and its high initial cost. The lack of knowledge regarding the structural behaviour is down to unavailability of appropriate test methods for this special type of concrete which behaves differently compared to normal concrete. Furthermore, the current precast production contributes to its high initial cost significantly. Therefore, this study has investigated various aspects that are restricting the potential use of UHPFRC for highway bridge applications. The investigation included extensive experimental studies and numerical modelling. In the experimental programme, various parameters ranging from material and mechanical properties, potential use of the concrete for cast in-situ applications to ductility behaviour were investigated. Furthermore, the numerical analyses were carried out to identify appropriate finite element models to predict the flexural and shear behaviour of the concrete. In the experimental work, simple and reliable test methods for material characterisation were developed. The validity of two non-destructive testing methods in studying the elastic properties of the concrete was confirmed. From this, the UHPFRC constitutive material model was obtained and used in numerical modelling. The reliability of the test methods were established by performing numerous experimental tests and similar results obtained at all times. Furthermore, the suitability of the concrete for cast in-situ applications at various temperatures by monitoring the strength development from an early age (12 hours) up to 360 days were investigated. The results showed significant strength gain of the concrete in both compression and tension within 7 days when cured at temperatures similar to site conditions (20 and 30 oC). A phenomenon related to the 90 oC curing temperature in precast production was reported and showed to have caused loss of flexural strength and toughness of the concrete. In addition, for the first time, an effective test method for studying the punching shear strength of the concrete with minimal influence of flexural stress was developed and used successfully. The results showed a reduced effective punching shear perimeter of UHPFRC slabs by half compared to normal concrete. The numerical analyses were carried out using the Abaqus finite element software. In this study, the Concrete Damaged Plasticity (CDP) and Concrete Smeared Cracking (CSC) material models with minor modifications were used to simulate the flexural and shear behaviour of UHPFRC beam and slab specimens, respectively. The results obtained here were validated against experimental studies with good agreement. The CDP model was found to replicate the linear and nonlinear structural response of the concrete with better accuracies than the CSC model. This study presents significant findings on the suitability of UHPFRC for structural applications with a lower initial cost compared to its current precast production. Furthermore, results obtained on its excellent structural behaviour in flexure and shear provides structural designers great confidence in using this concrete for highway bridge applications. The findings reported in this study contribute to the literature of UHPFRC significantly.

A number of artificially generated earthquake time-histories (AGETH) fitting to a Eurocode 8 (EC8) response spectrum are randomly generated using SIMQKE software and the average generated spectrum compares well with the EC8 one. Two Finite element (FE) smeared crack models, named Multi-crack and Craft, are well validated against experimental data of concrete and Reinforced concrete (RC) structures under monotonic and cyclic loadings. They are then used in the analysis of RC bridge piers under the AGETH. Several techniques including Fourier analysis, normalised cumulative spectrum, energy dissipation, damage index as well as probability applications are applied to quantify the structural response and damage. Based on the convergence of the representative responses under different numbers of AGETH, a minimum representative number of AGETH from 6 to 11 may be sufficient depending on the confidence band width from the mean of all damage responses. Effects of several parameters of the earthquake and structure to the dynamic response and damage of the bridge pier are investigated. Throughout these parametric studies, several of the common circumstances that structural engineers face are addressed and the proposed number of artificial earthquake time-histories required for non-linear dynamic analysis is thereby validated.

Accurate inspection data is important for efficient bridge management. Visual inspections play a key role in providing this information, but the reliability of such data has limitations. A range of techniques addressing these limitations are used in other sectors, but not to assist routine visual bridge inspection. Work has been undertaken investigating the feasibility of performing routine visual bridge inspections based on systematically collected images alone. The requirements of such a system are considered and defined. The research demonstrates that more detail can be seen in images at 1-pixel-per-mm than can be seen from 3m, and that images at this resolution can be systematically collected, processed, displayed, and inspected to complete General Inspections with results comparable to traditional routine visual inspections. No existing systems were found to be suitable for routinely providing visual inspection data; consequently a prototype was developed demonstrating the feasibility of the image-based inspection approach. The development considered hardware, image collection methodology, processing, alignment, display and interpretation. Inspectors tested and used the system to perform image-based General Inspections on several bridges. It is concluded that an image­based approach can be used to perform routine visual bridge inspections, with no loss of detail compared to traditional inspections.

Aging railroad bridge infrastructure is subject to increasingly higher demands such as heavier loads, increased speed, and increased frequency of traffic. The challenges facing railroad bridge infrastructure provide an opportunity to develop improved systems of monitoring railroad bridges. This dissertation outlines the development and implementation of a Structural Impairment Detection System (SIDS) that incorporates finite element modeling and instrumentation of a testbed structure, neural algorithm development, and the integration of data acquisition and impairment detection tools. Ultimately, data streams from the Salmon Bay Bridge are autonomously recorded and interrogated by competitive arrays of artificial neural networks for patterns indicative of specific structural impairments. Heel trunnion bascule bridges experience significant stress ranges in critical truss members. Finite element modeling of the Salmon Bay Bridge testbed provided an estimate of nominal structural behavior and indicated types and locations of possible impairments. Analytical modeling was initially performed in SAP2000 and then refined with ABAQUS. Modeling results from the Salmon Bay Bridge were used to determine measureable quantities sensitive to modeled impairments. An instrumentation scheme was designed and installed on the testbed to record these diagnostically significant data streams. Analytical results revealed that main chord members and bracing members of the counterweight truss are sensitive to modeled structural impairments. Finite element models and experimental observations indicated maximum stress ranges of approximately 22 ksi on main chord members of the counterweight truss. A competitive neural algorithm was developed to examine analytical and experimental data streams. Analytical data streams served as training vectors for training arrays of competitive neural networks. A quasi static array of neural networks was developed to provide an indication of the operating condition at specific intervals of the bridge's operation. Competitive neural algorithms correctly classified 94% of simulated data streams. Finally, a stand-alone application was integrated with the Salmon Bay Bridge data acquisition system to autonomously analyze recorded data streams and produce bridge condition reports. Based on neural algorithms trained on modeled impairments, the Salmon Bay Bridge operates in a manner most resembling one of two operating conditions: 1) unimpaired, or 2) impaired embedded member at the southeast corner of the counterweight.

Vibration induced by walking pedestrians has motivated research in the civil engineering community for many years. An area within this broad field that has received particular attention is the dynamic interaction that can occur between pedestrians and laterally flexible bridge structures. Perhaps the most notable example occurring on the opening day of London's Millennium Bridge. The enduring interest in this research problem is fuelled by two of its key features; (i) the sensitivity and adaptability of human balance to lateral motion and (ii) the spatial and temporal variation in flow characteristics exhibited by a pedestrian crowd. Both of these features are addressed herein. In this project an experimental campaign was executed with the aim of identifying the interaction mechanism by which pedestrians produce force harmonics, that resonate with the oscillating structure on which they walk. These so-called self-excited forces have been experimentally identified by others but the underlying reason for their existence has remained an open question. In an effort to address this, human balance behaviour while walking on a laterally oscillating treadmill was recorded using 3-dimensional motion capture equipment. Subsequent analysis revealed that human response to sinusoidal base motion is dominated by periodic alteration of foot placement position. This produces amplitude modulation of the lateral component of the ground reaction force and is ultimately responsible for the self-excited force harmonics. It was further revealed that human centre of mass motion while walking on an oscillating structure is predominantly passive. The passive inverted pendulum model is thus an excellent model of pedestrian frontal plane balance. The second facet of this work is concerned with developing a crowd-structure interaction model that builds upon the current state of the art. The model presented utilises the understanding of human-structure interaction identified above and employs an agent-based modelling approach. Thus, the resulting 'virtual crowd' is capable of simulating key crowd features, such as inter-subject variability and emergent velocity-density flow behaviour. Using this model, it is shown that the experimentally identified human-structure interaction mechanism can lead to large amplitude lateral deck oscillations, consistent with field observations reported in the literature. The model successfully predicts the multi-mode instability of Bristol's Clifton Suspension Bridge in the absence of step frequency tuning among the crowd. This provides supporting evidence for the model's validity. The work described above has resulted in a clearer understanding of the feedback between pedestrian balance behaviour and bridge response. Furthermore, the modelling techniques developed have potential for application in the wider study of crowd-induced vibration of dynamically susceptible structures.

This thesis focuses on the simulation of flow around bluff bodies and bridge deck sections, in which unsteady nature and vortex shedding of flow are commonly found, using computational fluid dynamics (CFD). Various turbulence models have been tested to develop understanding and proper modelling techniques for the flow around such bodies. Throughout the thesis, the turbulence models employed, mainly large eddy simulation (LES) and detached eddy simulation (DES), have been validated through comparative study with experimental work. The major part of the work discusses flow around bluff bodies ranging from a simple circular cylinder, a square cylinder to rectangular sections with various aspect ratios (1:2 to 1:8). The research section concentrates on modelling flow characteristics around bluff bodies to investigate the impact of fluid flow on them. This aids in the understanding of a more complex flow around bridge deck sections. The thesis combines investigation and discussion of the vortex shedding nature on the flow around bluff bodies, in which the simulations are done using advanced modelling techniques on high performance computing system. Work also includes a sectional wind tunnel test of the bridge deck section for the comparative study with the numerical solution. Finally, the conclusions outline the achievements and findings of the work done in this thesis and give recommendations for further research on the topic.