Bridges consisting of steel plate girders and composite concrete deck slabs are common throughout North America. For a typical highway application, these composite bridges are constructed with a cast-in-place concrete deck; however, some composite bridge designs utilize precast concrete deck panels. For example, bridges built on temporary access roads which service resource industries throughout Western Canada often employ composite bridges that consist of steel plate girders and precast concrete deck panels. For spans between 18- to 36 metres, permanent bridges currently present the best economy; although, portable structures would be preferred on these temporary roads so that the bridge could be relocated after the road is decommissioned. This study proposes a shear connection between steel plate girders and precast concrete deck panels, which allows fastening, and unfastening, of these two components enabling a portable composite bridge. In total, ten connection concepts were developed during this study and a multi-criteria assessment was performed to evaluate each concept respectively. Based on the outcome of this multi-criteria assessment, and subsequent sensitivity analysis, a preferred connection was established and a finite element model was developed for the analysis of composite bridge girders. For the initial development of the finite element model, the test set up and experimental findings of a test program by other researchers was employed so that the finite element analysis results could be compared to those reported from a physical experiment. Following this initial finite element analysis, full scale composite bridge girders were modelled so that the influence of the proposed shear connection on the behaviour of a composite girder could be studied. The model was verified for its ability to capture the possible effects of flange buckling, web buckling, and lateral torsional buckling of the steel plate girder. It was then confirmed that these local responses do not influence the performance of the proposed portable composite bridge system. A parametric study was also performed in which the effect of shear connection stiffness and spacing on the behaviour of the composite girder was investigated. This parametric study allowed the sensitivity of the proposed connection to variations in these two parameters to be assessed and also allowed preliminary study of the performance of composite girders with alternative shear connection designs.

Bridges consisting of steel plate girders and composite concrete deck slabs are common throughout North America. For a typical highway application, these composite bridges are constructed with a cast-in-place concrete deck; however, some composite bridge designs utilize precast concrete deck panels. For example, bridges built on temporary access roads which service resource industries throughout Western Canada often employ composite bridges that consist of steel plate girders and precast concrete deck panels. For spans between 18- to 36 metres, permanent bridges currently present the best economy; although, portable structures would be preferred on these temporary roads so that the bridge could be relocated after the road is decommissioned. This study proposes a shear connection between steel plate girders and precast concrete deck panels, which allows fastening, and unfastening, of these two components enabling a portable composite bridge. In total, ten connection concepts were developed during this study and a multi-criteria assessment was performed to evaluate each concept respectively. Based on the outcome of this multi-criteria assessment, and subsequent sensitivity analysis, a preferred connection was established and a finite element model was developed for the analysis of composite bridge girders. For the initial development of the finite element model, the test set up and experimental findings of a test program by other researchers was employed so that the finite element analysis results could be compared to those reported from a physical experiment. Following this initial finite element analysis, full scale composite bridge girders were modelled so that the influence of the proposed shear connection on the behaviour of a composite girder could be studied. The model was verified for its ability to capture the possible effects of flange buckling, web buckling, and lateral torsional buckling of the steel plate girder. It was then confirmed that these local responses do not influence the performance of the proposed portable composite bridge system. A parametric study was also performed in which the effect of shear connection stiffness and spacing on the behaviour of the composite girder was investigated. This parametric study allowed the sensitivity of the proposed connection to variations in these two parameters to be assessed and also allowed preliminary study of the performance of composite girders with alternative shear connection designs.

The double angle connection (sometimes referred to as a cleat connection) is one of the most commonly used simple shear connections, and many investigations have been conducted on this type of connection. However, most of these investigations have focused on either the strength or the moment-rotation relationship under shear loading. Several investigations have recently been performed on the behavior of double angle connections subjected to shear plus axial tensile loads. In these investigations, analytical models and design formulas have been proposed to model the complex behavior of these connections when subjected to the combined loading. However, a complete design model has not been developed. To fulfill the need for a design procedure, double angle connections were studied for three different loading cases. The first case was used to establish the load-displacement relationship under axial tensile loads. The second case was to establish the moment-rotation relationship under shear loads. Finally, the third case was to find the effects of combined axial tensile loads and shear loads on the behavior of double angle connections. For these purposes, 3D-nonlinear finite element models were developed to simulate the connection behavior under the three loading cases. The commercial software package, ABAQUS, was used for the study. The complex phenomena of contact problems and the pretension forces in the bolts were simulated. A simplified angle model and an equivalent spring model were developed from the 3D results. / Ph. D.

Accelerated bridge construction methods are being progressively used to construct and replace bridges in North America. Unlike traditional bridge construction methods, accelerated bridge construction methods allow bridges to be built in a shortened period of time on the construction site. These methods reduce the road closure time and the traffic disruption that are associated with bridge construction. One of these methods is carried out by prefabricating the bridge elements offsite and then assembling them onsite in a time-efficient way to build the bridge. This construction method can be used to build steel-precast composite bridges, where steel plate girders are connected to full-depth precast concrete deck panels. For the expeditious construction of composite bridges, a proper shear connection detail is needed to develop composite action between the steel plate girders and the precast concrete deck panels. This research project investigated two types of shear connection that would accelerate the construction of steel-precast composite bridges. First, finite element analysis was used to study the behaviour of composite bridge girders with panel end connections. The girders were analyzed for their load-displacement behaviour, cross-sectional stress and strain profile, and connection force distributions. Secondly, experimental push tests were conducted to study the load-slip behaviour of bolted connections. The effects of steel-concrete interface condition, bolt diameter and bolt tension on the shear capacity of bolted connections were analyzed. Based on the finite element analysis results, it is concluded that the panel end connected girder exhibited strong composite action at service and ultimate load. The level of composite action decreased slightly when the panel end connection stiffness was reduced by a factor of ten. Based on the experimental results, it is concluded that the total shear capacity of the bolted connection is the sum of the friction resistance and the bolt dowel action resistance. The friction resistance of the connection depends on the interface condition and the bolt clamping force. An analytical model that can predict the ultimate shear capacity of bolted connections has been developed and recommended. The proposed model is shown to give reliable predictions of the experimental results. It should be noted that bolted connections exhibit good structural redundancy because the bolt fracture failures do not happen simultaneously.

Accelerated bridge construction methods are being progressively used to construct and replace bridges in North America. Unlike traditional bridge construction methods, accelerated bridge construction methods allow bridges to be built in a shortened period of time on the construction site. These methods reduce the road closure time and the traffic disruption that are associated with bridge construction. One of these methods is carried out by prefabricating the bridge elements offsite and then assembling them onsite in a time-efficient way to build the bridge. This construction method can be used to build steel-precast composite bridges, where steel plate girders are connected to full-depth precast concrete deck panels. For the expeditious construction of composite bridges, a proper shear connection detail is needed to develop composite action between the steel plate girders and the precast concrete deck panels. This research project investigated two types of shear connection that would accelerate the construction of steel-precast composite bridges. First, finite element analysis was used to study the behaviour of composite bridge girders with panel end connections. The girders were analyzed for their load-displacement behaviour, cross-sectional stress and strain profile, and connection force distributions. Secondly, experimental push tests were conducted to study the load-slip behaviour of bolted connections. The effects of steel-concrete interface condition, bolt diameter and bolt tension on the shear capacity of bolted connections were analyzed. Based on the finite element analysis results, it is concluded that the panel end connected girder exhibited strong composite action at service and ultimate load. The level of composite action decreased slightly when the panel end connection stiffness was reduced by a factor of ten. Based on the experimental results, it is concluded that the total shear capacity of the bolted connection is the sum of the friction resistance and the bolt dowel action resistance. The friction resistance of the connection depends on the interface condition and the bolt clamping force. An analytical model that can predict the ultimate shear capacity of bolted connections has been developed and recommended. The proposed model is shown to give reliable predictions of the experimental results. It should be noted that bolted connections exhibit good structural redundancy because the bolt fracture failures do not happen simultaneously.

京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第17884号 / 工博第3793号 / 新制||工||1580(附属図書館) / 30704 / 京都大学大学院工学研究科建築学専攻 / (主査)教授 中島 正愛, 教授 田中 仁史, 教授 金子 佳生 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

Yes / For composite cellular beams, additional deflections occur due to the loss of bending and shear stiffness at the opening positions and also due to slip in the shear connectors caused by partial shear connection. Design formulae are presented for the additional deflection of composite beams with circular openings or for cellular beams as a function of the proportionate depth of the openings. The simplified formulae are calibrated against finite element results for both cellular and solid web beams and also against measured deflections of a 15.3 m composite cellular beam test. This additional deflection is presented as a function of flexural and shear terms that are a function of the span:depth ratio. For modelling of cellular beams to determine deflections, the circular opening may be represented by an equivalent rectangular opening of length equal to 70% of the opening diameter.

Yes / This paper outlines an experiment on a 12 m long composite beam subjected to uniformly distributed loading. Although composite beams are widely used, current Eurocode design guidelines for these types of members can be over-conservative, particularly in relation to the required degree of shear connection. The tested beam comprised a concrete slab supported by profiled metal decking, connected to an asymmetric fabricated steel I-beam using welded shear studs. The specimen was assembled using unpropped construction methods and had a degree of shear connection equal to 33%, significantly lower than the minimum required amount specified in Eurocode 4. A uniformly distributed load was applied to the specimen, which was increased until the failure occurred characterized by yielding of the steel beam. The maximum bending moment of the composite beam obtained from the test was close to the plastic bending resistance according to the Eurocode 4. No concrete crushing or shear stud failure was observed and the end slips exceeded 6 mm, the limit for ductile behaviour in Eurocode 4. The test demonstrated the merits of unpropped construction, which are currently not fully exploited in Eurocode 4. The comparison and analysis suggest that the design limits governing the minimum degree of shear connection might be revised. / RFCS

no / Cellular beams are the preferred form of long span construction in multi-storey buildings. For efficient design of composite cellular beams, asymmetric sections are often manufactured in which the bottom flange is larger than the top flange. A further innovation is in the use of 80mm deep deck profiles which allows beam spacing to be increased to 4.5m, and so the effective slab width acting compositely with the long span beams is also increased. The values for shear connector (stud) resistance given in Eurocode 4 (EN 1994-1-1), when used in combination with these modern decking profiles, have led to problems in achieving the minimum degree of shear connection for composite beams in comparison to the former BS 5950-3. For secondary beams, the number of shear connectors that can be accommodated in a span is limited by the spacing of the deck ribs (typically 300mm for deep trapezoidal profiles), and it is found that even for pairs of shear connectors per deck rib, it is impossible to satisfy the shear connection rules in Eurocode 4 for long span asymmetric beams. SCI, with support from the Research Fund for Coal and Steel, is on the way to resolving this problem in design to Eurocode 4, and has completed a test on a 15.3m composite cellular beam at the University of Bradford. This is believed to be the longest composite cellular beam test ever carried out. The test was part-sponsored by ASD Westok.

Concrete-Filled Steel Tubular (CFST) columns are increasingly being used because of their many advantages, including high strength, high ductility, and higher fire resistance than conventional steel or concrete columns of the same size. In order to maximise the advantages of CFST column, composite action of the column should be ensured. In realistic structures, the load is not directly applied to the entire CFST column section and is introduced from the beam-column connection. Simple shear connections, which are usually preferred in constructions, are only connected to the external face of the steel tube and there is an issue about how this load is introduced to the concrete core, through the bond at the steel/concrete interface. There are fundamental errors in the load introduction mechanism assumed in various current design methods. Furthermore, based on this erroneous load introduction mechanism, construction methods, such as placing shear connectors inside the steel tube or using through-column plates, are recommended to ensure complete load introduction. However, these methods are either impractical or uneconomical. The aim of this project, therefore, is to develop a thorough understanding of the load introduction mechanism and to use the new insights to assess design implications, for both ambient temperature and fire safety design. The research has been conducted through physical testing, extensive numerical modelling and detailed analytical derivations. A series of new load introduction tests, in which square CFST columns are loaded through simple fin plate connections, are carried out. These tests are designed to investigate the effects of changing column lengths below and above the connection, the effectiveness of using shear connectors inside the steel tube below the connection (according to Eurocode 4) and using a cap plate on the column top for load introduction into the concrete core. The test results indicate that the connection load is introduced to the concrete core through the column length above and within the connection or the cap plate on top of the column. This is different from the currently assumed mechanism of load introduction which assumes that load introduction occurs from underneath the connection. Below the connection, there is transfer of forces from the steel tube to the concrete core, but the total force in the column remains unchanged. Consequently, using shear connectors below the connection is ineffective in increasing CFST column strength, as has been demonstrated by the tests. The physical tests are supplemented by an extensive numerical parametric study to check whether the conclusions are applicable to different design conditions and to provide data for development of a new design method. The parameters include: section geometry (square, circular, and rectangular), position of load application to CFST column, dimensions of the square column cross-section, steel tube thickness, connection length, column length above the connection, column length below the connection, and maximum bond stress at the steel-concrete interface. The numerical simulation results confirm the experimental observations. Furthermore, the numerical simulation results indicate that the entire column length and the entire perimeter of the steel-concrete interface above and within the connection are engaged in load introduction. Based on the experimental and numerical simulation results, a simple calculation method has been proposed to calculate the column cross-section resistance under compression. According to this equation, the concrete compression resistance to the composite column is the minimum of the plastic resistance or the bond strength within and above the connection. This gives rise to a “concrete strength reduction factor” to account for incomplete load introduction, being the ratio of the load introduced to the concrete core through the interface bond to the concrete plastic resistance. Based on the new load introduction calculation method and using representative values of column dimensions and concrete cylinder strength, it has been demonstrated that complete load introduction can be achieved in almost all practical arrangements of concrete-filled tubular construction. For slender CFST column design, this concrete strength reduction factor should also be used to calculate the CFST column cross-section flexural stiffness. For a CFST column under combined axial compression and bending, the concrete strength reduction factor should be used when calculating the compression force, but should be ignored when calculating the bending resistance because composite action is not necessary for bending of the CFST column. The new load introduction mechanism induces additional compression in the concrete core and possible tension in the steel tube above the connection. Therefore, the concrete core of the column above the connection in multi-storey construction should be designed to resist the additional compression force. For the steel tube, in ambient temperature design, the steel contribution ratio (steel section resistance/plastic resistance of composite cross-section) of the top floor column should be at least 0.25. For fire resistance design, the steel contribution ratio of the top floor columns, those on the floor below the top floor, and those two floors below the top floor, should not be less than 0.5, 0.33, and 0.25 respectively.