Search

Concrete columns are critical for the stability of structures in case of fire. In order to allow for a true Perfor-mance Based Design, the design should be based on considerations of risk and reliability. Consequently, the probability density function (PDF) which describes the load-bearing capacity of concrete columns during fire exposure has to be assessed. As second order effects can be very significant for columns, traditional probabil-istic methods to determine the PDF become very computationally expensive. More precisely, for most current numerical calculation tools (e.g. Finite Element), the computational requirements are so high that traditional Monte Carlo simulations become infeasible for any practical application. In order to tackle this, a computa-tionally very efficient method is presented and applied in this paper. The method combines the Maximum En-tropy Principle together with the Multiplicative Dimensional Reduction Method, and Gaussian Interpolation, resulting in an estimation of the full PDF requiring only a very limited number of numerical calculations. Alt-hough the result is necessarily an approximation, it gives very good assessment of the PDF and it is a signifi-cant step forward towards true risk- and reliability-based structural fire safety. Accepted version

Finite Element Reliability Analysis (FERA) has been used to evaluate the reliability of structures. Mean and variance of the structural response is often estimated with the use of approximate methods, while structural response distribution is approximated based on Monte Carlo simulation (MCS). In this paper, FERA is applied in an efficient manner with the use of a Multiplicative form of Dimensional Reduction Method (M-DRM), which can estimate accurately the statistical moments and the probability distribution of the structural response, e.g., drift of a structure. The proposed approach is combined with OpenSees FE software and illustrated through the nonlinear pushover and nonlinear dynamic analysis of a steel frame. MCS is also performed for comparison of the proposed method.

Pile-supported ports allow seawater to run below the deck, and thus may suffer structural damages during extreme coastal events such as hurricanes. These structural damages, in turn, may result to port closures that can cause significant economic losses. Risk analysis can predict the post-hazard functionality of ports though the structural damage assessment of these structures prior to coastal events. However, assumptions on the selected demand estimates may affect the estimated probability of structural damage. This research aims to shed light on the sensitivity of the wave model selection for the risk assessment of pile-supported ports when subjected to storm surge and waves. The examined structural damage is the uplift of the deck, and the risk assessment is conducted through the development of fragility curves for a typical deck-pile connection, for which fragility curves are developed for different wave models. Uncertainties are also considered in parameters affecting the demand and capacity of the examined deck-pile connection and are propagated through the Monte Carlo simulation using the Latin Hypercube Sampling. The results indicate changes to the uplift probability as a result of the selected wave model. Thus, wave model selection can alter the uplift failure probability. In addition, the study proposes parameterized fragility models to enable the uplift risk assessment across a region. The presented results aim to throw light on the proper model selection to produce more realistic risk assessment estimates towards the resilience of coastal infrastructure. Thesis Master of Applied Science (MASc)

Efficient management strategies are essential to ensure bridge safety and functionality while accommodating budget limitations. For such purpose, bridge management systems (BMSs) developed policies to predict the global bridge conditions and any performance deficiencies in the individual components. However, the current policies have some shortcomings that may limit their efficiencies. To address these shortcomings, this thesis proposes several enhancements to three major BMS policies covering inspection and maintenance planning, bearing performance assessment, and seismic screening. Inspection and maintenance are often planned using Markov Chains deterioration models derived from past inspection records. However, Markov Chains models employ impractical assumptions and neglect the subjectivity of inspections. Alternatively, parameterized fuzzy-logistic deterioration models are proposed to predict future conditions given easy-to-track parameters, such as age and last maintenance date. The proposed models can yield better maintenance predictions compared to Markov Chains and can reduce inspection costs by 30%. Bearing performance assessment policies are limited to testing bearing material or its behaviour at design displacements. In practice, bearings experience fatigue with repeated displacement cycles even if at low magnitudes, leading to bearing degradation and long-term increase in demands. Thus, a parameterized loading protocol is proposed to guide laboratory testing in assessing the fatigue life. The protocol is derived from the bearing demands attributes observed from a 3D nonlinear analysis in OpenSees for various bridge configurations and loading conditions. Seismic screening policies were developed to identify the most vulnerable bridges, giving them the highest priorities. The current policies are qualitative, relying on identifying vulnerable details, rather than quantifying the actual performance. Furthermore, the vulnerability estimates are not updated with deterioration. Thus, new risk-based screening procedures are proposed via fragility analysis of the critical components, bearings and columns, while considering their deterioration. Given the components’ fragilities, a seismic vulnerability index is computed to rank the bridge’s priority. Thesis Doctor of Philosophy (PhD) Regular inspection and maintenance of bridges are vital to the integrity of transportation networks. The planning and scheduling of such activities have become known as bridge management. Given the limitations of budget and resources, there is a high demand for optimized bridge management. The fundamental goal of this thesis is to explore potential improvements for the current practices of bridge management. Towards this goal, this thesis proposes enhanced methodologies for a more accurate prediction of future bridge conditions and maintenance needs, a better understanding of the lifespan of various bridge components, and early detection of any vulnerabilities that may be detrimental to the functionality of bridges and the safety of bridge users.

Elastomeric bridge bearings (composed of alternating rubber layers and steel laminates surrounded by a rubber cover all around) have been historically introduced between the girder and the pier or abutment to accommodate service-level movements and rotations. During a seismic event, these bearings can also experience deformation demands that are considerably larger than the service-level demands. The manufacturing process of these bearings involves compression molding during which imperfections such as vertically misaligned, rotationally misaligned, and bent steel laminates might develop. The current quality control procedures for bearings involve costly, time-consuming, destructive tests on sample bearings. Therefore, there is a need for an efficient, nondestructive methodology for quality control testing of bearings. In addition, these bearings are a key component of the bridge, and thus the effects of manufacturing imperfections on different aspects of the bearing’s performance must be understood and quantified. In the first phase of this research, an efficient vision-based assessment methodology was developed for the nondestructive identification of the internal structure of elastomeric bridge bearings and their potential imperfections. The developed methodology requires the application of three-dimensional digital image correlation (3D-DIC) and relies on the local extrema forming in displacement and strain fields on the vertical sides of compressed bearings. In the next two phases, extensive nonlinear 3D FEA studies were undertaken to evaluate the effects of imperfections on the behavior of bearings under maximum service loading and under large lateral displacements resulting from seismic actions. Based on the FEAs results, imperfections are not expected to trigger shear delamination in bearings and their increased vertical deflections due to imperfections would be also minimal. However, imperfections can appreciably increase the tensile stresses in the laminates and lead to greater extents of plasticity in them. In the third phase of this research, the examined bearings were subjected to simultaneous axial loading and five cycles of lateral displacements with increasing amplitudes, up to 200% of their total rubber thickness. It was found that imperfections do not cause a considerable difference in the overall deformed shape of bearings but can expedite the rollover of bearings in lower shear strains and increase their risks of instability. Although all the examined bearings exhibited fairly stable hysteretic response, their effective lateral stiffness and lateral force could increase up to 18% and 40%, respectively, due to imperfections. Imperfections could also increase damage measured using a plastic-energy-based damage factor by up to more than five times for bearings subjected to the five-cycle displacement protocol. However, the yielded laminates in all examined bearings under maximum service loading or cyclic lateral displacement were far from fracturing. The increase of plastic damage during a strong earthquake and due to imperfections is not expected to cause immediate failure of bearings, provided that no or limited plastic damage occurs prior to that and under the service-level condition. The proposed nondestructive methodology can be adopted by departments/ministries of transportation as an alternative to the current destructive approaches for quality control of elastomeric bearings. For bearings subjected to maximum service loading, the use of sufficiently thick laminates (e.g., 4-mm-thick) is suggested to prevent plastic damage to the laminates, and thus alleviate concerns about the service-level performance of the bearings. Considering the risks of rollover, instability, and spread of plastic damage in the laminates under cyclic loading, slender bearings with laminate thicknesses of 2 mm, or less, and slender bearings in which the majority of laminates feature relatively large rotational misalignments (in the order of 0.025 rad, or more) should be avoided. Thesis Doctor of Philosophy (PhD)

Corrugated web plate girders (CWPGs) have become an increasingly popular structural member in Canada in recent years. This is because of their economic efficiency over standard wide flange members. Although the flexural performance of such has been increasingly studied in recent years there is still advancements that can be made in their design. No research has been completed in Canada on the subject of lateral torsional buckling (LTB) strength and very minimal research has been published on sinusoidal CWPGs. In order to examine the LTB strength of a CWPG with a sinusoidally shaped web, nine specimens were loaded and failed in simply supported arrangement that favours lateral torsional buckling. Specimens were chosen to observe the difference in strength due to web thickness, web depth and variation in identical beams. All of the specimens recorded strengths that exceeded the theoretical design strengths confirming that the current design procedure is conservative. A trend of ultimate capacity increasing was observed with the increase of web thickness. The depth of the web had no significant effect on the torsional strength besides what is gained from the increased flange distance. An equivalent web thickness equation was formulated based on the results for the purpose of calculating LTB strength. To test the proposed equation a numerical analysis was run on a wider range of beams and compared with the testing results. It was determined the physical testing results can be effectively captured by the proposed equation among more than just the tested beams. Two additional analyses were prepared to lay the foundation for further investigation of the proposed equation. The first was a Monte Carlo simulation to test the risk of using the proposed equation which requires additional data. Secondly, a preliminary finite element analysis (FEA) model was developed and presented for future use to expand this research. Thesis Master of Applied Science (MASc) Corrugated web plate girders (CWPG) have grown in popularity due to their economic efficiency. No research has been presented in Canada and very minimal research has been published on the lateral torsional strength of CWPGs with sinusoidally corrugated webs. This research studied the lateral torsional buckling (LTB) strength of CWPGs through the experimental testing of physical members and a new equation for the calculation of the LTB strength is proposed. This equation and design process was then numerically tested to determine its viability as a design process.

The importance of long-term performance and safety of bridge bearings is paramount. Limitations in the inspection process as well as uncertainties in the maintenance scheduling of bridge bearings have led to a limited understanding of the lifelong performance of such components over their service life. To aid bridge owners facilitate an efficient system for repairs and management of bridge bearings, the initial step is to quantify the demands they typically experience over their service life. In this study, a framework quantifying bridge bearing demands due to thermal, traffic, and seismic loading is presented. The effect of bridge aging is accounted through corrosion prediction models. In reinforced concrete members, the reduction of the steel reinforcement and cover cracking and spalling is considered, and a uniform reduction of the steel in the case of steel members. Six bridge models are considered with variations in the number of bridge spans and two superstructure types; precast concrete girder and slab on steel girder. Monte Carlo simulation is used to account for uncertainties in corrosion model parameters and bridge loadings. The results of this study can be used for future experimental investigation of the fatigue performance of bridge bearings. Thesis Master of Applied Science (MASc)