Journal of Bridge Engineering

When precast full-depth deck panels and/or decked bulb-Ts (DBTs) are used for accelerated bridge construction, transverse joints exist. In this study, continuous transverse U-bar joint details, which can provide negative moment continuity in multispan bridges, have been investigated. Four full-scale specimens connected by the developed U-bar detail together with the selected closure-pour (CP) materials, overnight cure and seven-day cure, were tested. Static and fatigue tests under tension loading were conducted. The loading demand necessary in the testing was determined based on the maximum forces in the transverse joint from an analytical study. Test results were evaluated based on tension capacity, cracking, and steel strain. Based on these test results, the developed transverse U-bar joint detail is a promising connection system.

Continuous longitudinal U-bar joint details for accelerated bridge construction are investigated. Four pairs of full-scale slabs connected by a U-bar detail with one of the selected closure-pour materials were tested under static and fatigue loadings. The loading demand necessary in the slab fatigue testing is determined on the basis of 3D finite-element (FE) parametric studies. Test results were evaluated on the basis of flexural capacity, curvature behavior, cracking, and steel strain. The fatigue loading was found to have little influence on the U-bar joint behavior. In general, crack widths were small at the service load level. No debonding between the slab and the joint was noticed. The joint moment capacity can be calculated on the basis of the assumption that U-bars were continuous reinforcement through the joint zone (i.e., not staggered as in actual specimens) reaching their yield strengths at ultimate. The actual moment capacity depends on the interaction between the closure-pour material and steel as well as the steel arrangement. Because the closure-pour material is cast-in-place and has a relative short curing time for the purpose of accelerated construction, it is very important to control its strength variation. On the basis of these tests, the U-bar joint detail is a viable connection system for the longitudinal joint in bridge decks.

Corroded cables from a cable-stayed bridge in China that had been in service for 18 years were employed to investigate the basic mechanical properties and residual fatigue life of wires and cables. First, the wires were randomly selected from the cables near the bottom anchorages and cut into segments as test specimens. The extent of corrosion of the wires was experimentally investigated. A tensile loading test was conducted on the wires to obtain the mechanical properties of the corroded single wires. The fatigue life of the corroded single wires was experimentally studied, and a dramatic degradation in fatigue life was observed. This phenomenon was interpreted using SEM images. Fatigue tests on two corroded cables were also conducted, and the test results indicated that the fatigue life of the cables had also decreased dramatically. A Monte Carlo simulation was conducted to obtain the fatigue life of cables. The simulation results indicated that the fatigue life of a cable was controlled by the small fraction of wires in the cable with the shortest fatigue lives. The fatigue life of a cable at a certain failure probability was dependent on the number of wires in the cable, but the mean fatigue life of a cable was not affected by the number of wires in the cable.

Research into the cause of the failure of the I-35W Bridge in Minneapolis, Minnesota, has indicated that some of the gusset plates were significantly overstressed. The Federal Highway Administration has recommended load rating all gusset plate connections in bridges around the country using their recommended load rating procedures, which are complex and require a large commitment of resources. Further, the procedures use approximate ultimate capacity and therefore do not address serviceability or bridge longevity, which require prediction of gusset plate yielding, not collapse. A triage evaluation procedure is proposed to enable rapid, yet conservative, identification of gusset plates that may be yielding under service loads and to eliminate gusset plates with adequate capacity from needing further or more sophisticated analysis. High-resolution finite element models of gusset plate connections from Washington State bridges and the connection identified as critical from the I-35W Bridge are used to develop the proposed procedure and evaluate the current recommendations. A complex interaction of stresses is generated in gusset plates by connecting members, and this interaction can initiate gusset plate yielding when the uniaxial stresses on Whitmore sections associated with those connecting members are well below yield. It is demonstrated that the onset of gusset plate yielding may be conservatively predicted by comparing the uniaxial stresses at Whitmore sections with $F_y/\sqrt{3}$. Additionally, a modification to the simple column analogy method recommended by FHWA for calculating the buckling capacity of gusset plates is proposed, and it is shown that gusset plate yielding occurs prior to buckling in all cases considered. The triage evaluation procedure is grounded in mechanics and conservative assumptions, enabling more rapid identification of problem gusset plates while also being reasonably conservative. The method is then applied to three bridges, from which 3 gusset plates out of 35 are identified as requiring additional investigation.

The in-plane stability of shallow parabolic arches subjected to a central concentrated load and temperature variations was investigated in this paper. The virtual work principle method was used to establish the non-linear equilibrium and buckling equations. Analytical solutions for the non-linear in-plane symmetric snap-through and antisymmetric bifurcation buckling loads were obtained. Then, the effects of temperature changes on the in-plane stability for arches with supports that stiffen under compression were studied. The results show that the influence of temperature variations on the critical loads for both buckling modes (symmetric snap-through and anti symmetric bifurcation) is significant. The critical loads for the two buckling modes are more than those only under external loads without thermal loading. Moreover, the critical loads increase with an increase of the thermal loadings. It can also be found that the effects of applying a temperature field increase when either initial stiffness coefficient $\alpha$ or the stiffening rate $\beta$ is raised. Furthermore, the effect of thermal loading on the critical load increases with the span-rise ratio $m$ for arches with any initial stiffness coefficient $\alpha$ and the stiffening rate $\beta$.

The purpose of this study is to investigate the dynamic behavior of steel deck-arch bridges with side-by-side multiple vehicle loadings moving over a rough deck. A brief description of analytical theory is provided, including vehicle model, bridge model, and numerical methods. Both vehicle and bridge are treated as three-dimensional models. The effectiveness of mode superposition and Newmark numerical methods are investigated. Most of the important factors that affect the dynamic response of bridges are considered in this study. The effects of the deck, column and arch boundary conditions are also studied, and the analytical impact factors of deck, column, and arch rib are compared with those determined on the basis of current American Association of State Highway and Transportation Officials (AASHTO) specifications. Study results are useful for both bridge engineers and bridge engineering researchers.

Research was conducted to analytically investigate the blast-response and behavior of multihazard-resistant bridge piers having circular-shaped, concrete-filled steel tube (CFST) columns. Two different analysis methods, namely a single-degree-of-freedom (SDOF) dynamic analysis and a fiber-based dynamic analysis, were used for this purpose and calibrated with the maximum residual deformations obtained from $1/4$ scale blast tests of CFST columns. It was noted that the structural response of SDOF dynamic analyses is sensitive to assumptions made in the load-mass factors needed to model structural components as an equivalent SDOF system. Fiber-based dynamic analyses showed that high-frequency modes of vibration have some influence on the structural response when subjected to blast loading. This study shows that different values of the shape factors, $\beta$ (which reduces blast pressures when applied to a circular column), must be used with different analytical methods, along with assumptions and conditions behind these different analytical methods.

Many bridges along the Gulf Coast of the United States were damaged by hurricanes, especially in the 2004 and 2005 hurricane seasons. Many more bridges are susceptible to similar damage. This research examines the structural performance of three commonly used connection details for AASHTO Type III prestressed concrete girders: headed studs, through bolts, and clip bolts. Full-scale specimens were fabricated and tested under static and dynamic cyclic load histories using expected maximum loads from previously conducted hydraulic tests of a scale model of a highway bridge spanning a coastal embayment. The load effects considered included (1) vertical uplift force, (2) lateral force, (3) combined lateral and vertical forces, and (4) dynamically applied wave force time histories. Connection performance was evaluated by comparing initial concrete cracking, strand slip, and ultimate capacity. The experimental results showed that the insert clip-bolt connections had the lowest capacity and produced significant ancillary damage to the surrounding concrete while the headed-stud connections exhibited the highest load capacity and negligible ancillary concrete damage. None of the anchorages considered would be capable of resisting newly specified vertical wave forces for large hurricane wave heights when significant air is entrapped such as those observed in Hurricane Katrina.

The earthquake engineering profession is increasingly utilizing nonlinear response history analyses (RHA) to evaluate seismic performance of existing structures and proposed designs of new structures. One of the main ingredients of nonlinear RHA is a set of ground motion records representing the expected hazard environment for the structure. When recorded motions do not exist (as is the case in the central United States) or when high-intensity records are needed (as is the case in San Francisco and Los Angeles), ground motions from other tectonically similar regions need to be selected and scaled. The modal-pushover-based scaling (MPS) procedure was recently developed to determine scale factors for a small number of records such that the scaled records provide accurate and efficient estimates of “true” median structural responses. The adjective “accurate” refers to the discrepancy between the benchmark responses and those computed from the MPS procedure. The adjective “efficient” refers to the record-to-record variability of responses. In this paper, the accuracy and efficiency of the MPS procedure are evaluated by applying it to four types of existing Ordinary Standard bridges typical of reinforced concrete bridge construction in California. These bridges are the single-bent overpass, multi-span bridge, curved bridge, and skew bridge. As compared with benchmark analyses of unscaled records using a larger catalog of ground motions, it is demonstrated that the MPS procedure provided an accurate estimate of the engineering demand parameters (EDPs) accompanied by significantly reduced record-to-record variability of the EDPs. Thus, it is a useful tool for scaling ground motions as input to nonlinear RHAs of Ordinary Standard bridges.

Results of an investigation aimed at determining bending moments and shear forces, required to design skewed concrete slab bridges using the equivalent-beam method are presented in this paper. Straight and skewed slab bridges were modeled using grillage and finite-element models to characterize their behavior under uniform and moving loads with the objective of determining the most appropriate modeling approach for design. A parametric study was carried out on 390 simply supported slabs with geometries covering one to four lane bridges of 3- to 20-m spans and with skew angles ranging from 0 to 60°. The analyses showed that nonorthogonal grillages satisfactorily predict the amplitude and the transverse distribution of longitudinal bending moments and shear forces, and can be used for the analysis of skewed slab bridges. Results of the parametric study indicated that shear forces and secondary bending moments increase with increasing skew angle while longitudinal bending moments diminish. Equations are proposed to include, as part of the equivalent-beam method for skew angles up to 60°, the increase of shear forces and the reduction of longitudinal bending moments. Equations are also given for computing secondary bending moments. A simplified approach aimed at determining the corner forces for straight and skewed bridges is proposed as an alternative to a more-refined analysis. The analyses indicated the presence of high vertical shear stresses in the vicinity of free edges that justifies suggesting to provide shear reinforcement along the slab free edges.

Seven prestressed concrete bridge girders that had been in service for 42 years, and represented two span lengths and reinforcement designs, were tested to determine their effective prestress force and ultimate shear capacity. A cracking moment test was used to determine the effective prestress force in the girders. The measured effective prestress force was compared with calculated values according to the AASHTO LRFD prestress loss equations to investigate their adequacy. The AASHTO refined method was shown to provide the most accurate results to within 10% of the measured values. An ultimate shear test was also performed on two of the girders. An external load was applied near the support and increased until the girder failed in shear. The various procedures in the AASHTO LRFD specifications were compared with the measured results. The AASHTO simplified procedure predicted only 51% and 39% of the average measured shear capacity for the short and long span girders, respectively. The strut-and-tie models were found to estimate the shear capacity more accurately. The AASHTO refined method was shown to provide the most accurate results.

This paper is a continuation of research on strengthening the load-carrying capacity of existing noncomposite bridge girders by postinstalling shear connectors. In this study, finite element models were developed to evaluate the behavior of composite beams retrofitted with postinstalled shear connectors over a wide range of variables with an objective of providing preliminary design recommendations. Variables considered in the parametric studies included beam depth, span length, and shear-connection ratio. The parametric studies showed that current simplified design approaches commonly used for partially composite beams in buildings provide good predictions of the strength and stiffness of partially composite bridge girders strengthened using postinstalled shear connectors. Use of a shear-connection ratio of less than 30% is not recommended to avoid nonductile behavior of the strengthened girder.

This study investigates vibration characteristics of a truss bridge that has been retrofitted with a fiber reinforced polymer (FRP) deck. The bridge is located on State Route 367 in Wellsburg, Chemung County, New York. It is a 140-ft-long (approximately 42.7-m-long) truss structure, originally built with a noncomposite concrete deck. The concrete deck was replaced with a much lighter fiber reinforced polymer deck to improve the bridge load capacity ratings. This study investigates the vibration characteristics of the bridge as a result of the deck replacement and impact of these changes on the structural behavior. The results indicate that the fundamental frequency for the current structure with the FRP deck is approximately 45% higher than that for the original structure with a concrete deck. The computed bridge fundamental frequency correlated well with field data. When compared to the concrete deck bridge, the FRP deck bridge resulted in lower dead load forces, higher live load forces, lower total forces, and higher live load stress ranges used in fatigue life estimates. Dynamic allowance was also determined for each of the selected members for both concrete and FRP decks. For the concrete deck bridge, the calculated dynamic allowance is higher than those used in bridge design specifications. Lower dynamic allowances were obtained for the FRP deck bridge.

In-span hinges (ISHs) are located at the bridge deck of reinforced concrete box-girder bridges and are used to transmit vertical loads between two adjacent parts of the deck. ISHs are disturbed regions because of the complex three-dimensional stress state caused by concentrated bearing loads and utility openings. The common modeling practice for ISHs is simplified two-dimensional idealization as short cantilevers following standard procedures. Such simplified analytical and design procedures lead to inefficient detailing because they do not take into account the realistic failure modes of ISHs; punching shear is one of these critical modes. In this study, the influence of reinforcement and geometrical detailing on the behavior and strength of ISHs is assessed using a computational approach. The computational model is calibrated based on previous experimental results and adopts nonlinear three-dimensional finite-element analysis (FEA), accounting for cracking of concrete and elastic–plastic behavior of reinforcement. The concrete material is modeled using the total strain rotating crack method, and the effect of compression softening is incorporated into the constitutive model. The reinforcing steel is modeled using embedded reinforcement formulation assuming a perfect bond between the concrete and the reinforcement. From the computational study, design guidelines are presented for better constructability of these disturbed regions. Findings from this study revealed that the strength of ISHs is mostly improved by increasing the amount of diagonal reinforcement in the seat and by increasing the bearing plate size.

Although the prestressed concrete (PSC) girder bridge is known to be more economical than other types of bridges in short and medium spans, a longer span has also been achieved by applying several strategies. This paper presents a systematic procedure that can be used to assess the effects of these strategies on the span. The proposed scheme adopts a graphical approach that represents a relationship between the number of prestressing tendons and the span and is derived on the basis of stress assessment equations of the girder at each stage of fabrication and in service. A quantitative evaluation for the extension of the span is performed by adopting a sample bridge. A number of advantages of the proposed scheme are apparently shown for determining why and how each strategy contributes to the span extension and for suggesting further improvement for a longer span. The results imply that increasing the strength of the girder, making the girders continuous, multistage prestressing, and the decked PSC girder are very effective.

Corrosion of the reinforcing steel can cause cover cracking and eventual spalling of reinforced concrete (RC) surfaces, resulting in costly and disruptive repairs. The present paper will compare the effect of maintenance and repair strategies on the timing, extent, and cost of remediation actions over the service life of a RC structure in a chloride environment. The paper presents a probabilistic reliability analysis, which is used to predict the likelihood and extent of corrosion-induced cracking to RC structures. A spatial time-dependent reliability model has been developed where concrete properties, concrete cover, and the surface chloride concentrations are treated as random fields. This allows for the calculation of the probability that a given extent of damage will occur for any time period. Maintenance strategies and repair efficiencies are incorporated in a Monte-Carlo event-based simulation analysis, allowing a comparison in terms of cost and number of repairs over the service life of a RC structure. Thus, the expected timing and extent of repairs can be predicted for various design parameters, inspection intervals, repair thresholds, maintenance strategies, and efficiency of repairs. Results are presented for a RC bridge deck subject to a marine environment. The life-cycle cost (LCC) analysis considers repair and user delay costs. User delay costs can be up to ten times higher than the cost of repair itself. The statistical variability of predicted LCCs can be large, with coefficients of variation exceeding one.

The paper describes the testing and finite element analysis of an experimental half-scale (approximately) wood-based, prefabricated, composite acting bridge deck that utilizes composite action between the decking and stringers. The 12-meter span (6-meter span in the model version) full-sized deck on which current developments is based uses three round log stringers with a diameter greater than 500 mm per 3.2 meter span design lane and plywood decking that can be up to 170 mm thick. The composite action enables the decks to meet strength requirements for the heaviest highway loads (M1600) specified in the Australian Standard AS 5100.2, but the model requires modification to meet the stringent deflection criteria associated with bridges. The deck modules are prefabricated at a full span length and can be as wide as can be transported within legal limits without special escort vehicles. The modules are joined on-site on a bevelled edge by using grouted rod-type technology. Other details included are aspects of the fabrication method, the shear connectors used, utilization of Strand 7 FE software, the linear and nonlinear finite element analyses used to investigate the deck performance, and the results of tests on a mixed scale model used to validate the finite element analysis.

The soil adjacent to integral bridge abutments experiences daily and annual temperature-induced cyclic loading owing to expansion and contraction of the bridge deck. This causes a particular soil response and complicated soil-structure interaction problem, with considerable uncertainties in design. This paper describes a method of calculating the effects of thermal cycling by using the results of laboratory cyclic stress-path testing within a numerical model. Samples of stiff clay and sand were tested in the triaxial apparatus under stress paths that are typical behind an integral abutment. Distinct behavior was observed for the two soils, with stiff clay showing relatively little buildup of lateral stress with cycles, whereas sand stresses continued to increase, exceeding at-rest pressure and approaching full passive pressures. To explore the implications of these findings on soil-abutment interaction and to estimate the lateral stresses acting on the abutment as a whole, a numerical i.e., (finite difference) model was developed with a soil model reproducing the sand behavior at element level. The numerical model gave good agreement with published centrifuge and field data, indicating that the stress profile specified in some current standards is conservative. The influence of abutment stiffness and wall friction is also quantified.

The dynamic interaction between a moving truck and a long span reinforced concrete culvert is studied for different truck velocities. The numerical simulation method presented in this paper is based on discretizing the culvert, surrounding soil, and roadway, using a plane strain finite element model (FEM) along with a seven-degrees-of-freedom FEM mass-spring-damper model for the truck. The time-variable coupled dynamic system is solved in a step-by-step manner using Newmark’s integration scheme. Results are presented for different parameters, including truck velocity, truck-culvert frequency ratio, damping ratio, road surface roughness, and roadway material. For the culvert under consideration, the results indicate that the road surface roughness has a significant influence on the dynamic impact factor relative to the displacement at the crown of the culvert. This influence is further magnified by variations in truck frequency, culvert damping coefficient, and truck velocity.

Permit truck models are used to consider local truck traffic in addition to national live load models in bridge design and rating practices. Weigh-in-motion (WIM) systems installed in the United States provide millions of truck records that can be used for evaluating such truck models. In this study, the standard permit vehicle in Wisconsin was evaluated by using six million WIM truck records collected in 2007. The evaluation was on the basis of statistical analyses of the maximum moments and shear in simply supported, 2-span, and 3-span continuous girders in the selected heaviest 5% of trucks in each vehicle class/group. The comparisons showed that 5-axle, short, single-unit trucks may cause larger moment/shear in bridge girders than the standard permit vehicle, and a 5-axle truck model was proposed to supplement the standard permit vehicle for possible use in bridge design and rating in Wisconsin.