Journal of Bridge Engineering

This paper provides experimentally determined estimates of the scale of fluctuation of the principal variables employed in modeling chloride-induced corrosion for reinforced concrete; i.e., the surface chloride content ($C_s$) and apparent diffusion coefficient ($D_{\mathrm{app}}$). The estimation of the scale of fluctuation, $\theta$, is based on the analysis of experimental data recorded on a bridge in South East Ireland prior to its extensive rehabilitation in 2007. In determining the scale of fluctuation the paper considers two commonly used methods; i.e., the maximum likelihood method and the autocorrelation curve-fitting method. The reliability of both methods is discussed. Introduction of the kriging statistical interpolation method is demonstrated to improve the reliability of the estimates of the scale of fluctuation. The results obtained from the analysis are compared with values in the literature proffered by other researchers.

Research has shown that exterior girders of adjacent box (AB) beam bridges act compositely with their barrier wall and curb slab assemblies, resulting in a highly asymmetric cross section. Additionally, impact damage from vehicle collision often severs or exposes strands in the exterior soffit corner of exterior girders. These effects individually (and more so in combination) have the effect of rotating the neutral axis of the girder, resulting in significant, unaccounted for biaxial flexure. Common structural rating practices analyze the member about its horizontal axis, neglecting the effects of neutral axis rotation, and thus overestimate the actual capacity of the member. A parametric study that analyzed the effects of varying levels of damage to eight prototype AB girders was conducted. In all, 106 analyses were carried out. From this study, a relationship is proposed to determine the capacity of an AB beam subject to asymmetric loading, which includes the composite behavior of the barrier wall assembly and the effect of prestressing strands lost as a result of damage to the girder. This relationship is based on easily obtained capacity predictions and is shown to result in predictions of capacity having an average absolute error of less than 6%. This study also examines, in the context of complete AB bridges, the potential for in-service loading to exceed assumed design loads.

With the magnitude of some stress cycles in a live load stress range spectrum being higher than the constant amplitude fatigue limit (CAFL) of a structural detail, the development of a fatigue crack could be expected. The current procedure in the AASHTO specifications for fatigue life estimation utilizes an equivalent constant amplitude stress range with the direct extension of the $\mathit{S}\text{-}\mathit{N}$ lines of slope −3 to below the CAFL. This approach is conservative; however, it often results in overprediction of fatigue damage and underestimation of the safe useful life of structural details. A set of bilinear $\mathit{S}\text{-}\mathit{N}$ curves with the knee at the CAFL for the AASHTO fatigue strength categories and a slope of −4 below have been produced analytically for estimating fatigue life of existing structures subjected to variable live load stresses. This paper provides information regarding the development of the new equivalent constant amplitude stress ranges for the bilinear $\mathit{S}\text{-}\mathit{N}$ curves. The results from a few examples are presented. Adoption of the procedure presented herein for fatigue life estimation is recommended.

This paper presents the results of an analytical study on inelastic behavior of a four-span continuous composite bridge girder. The objective is to evaluate the effect of pier section reinforcement on strength and ductility of the girder. The reinforcement consists of two pairs of longitudinal ribs bolted to the web of steel sections adjacent to interior piers. Nonlinear finite element analyses are carried out to determine the moment-curvature characteristic of the pier sections. The results indicate that unreinforced sections fail due to local buckling upon yielding of the compression flange whereas the reinforced sections reach their plastic moment capacity with good ductility. Idealized moment-curvature curves are developed based on the nonlinear analyses of the sections. The idealized curves are then used to evaluate the behavior of the girder. The results indicate that strength and ductility of the girder improves substantially when the pier sections are reinforced. At failure, the maximum vertical deflection of the reinforced girder is about twice the deflection of the unreinforced girder. The ultimate load carrying capacity of the reinforced girder is also significantly larger than that of the unreinforced girder.

Vortex-induced vibration of an inclined cable under wind with varying velocity profiles was investigated through computational fluid dynamics numerical simulation. As a complicated fluid-structure interaction issue, the flow field around the freely oscillating cable was simulated by Ansys CFX 11.0, and the cable oscillation as calculated using the Galerkin approach (realized by the additional subroutine, which is embedded into Ansys CFX 11.0). The shear stress transport $k\text{-}\omega$ turbulent model based on the Reynolds-averaged Navier-Stokes method was employed to simulate the behavior of turbulent flow in the CFD numerical simulation. The above computational method and turbulent model were validated through comparison of the computational results with the wind tunnel test results of a rigid circular cylinder. And then, two inclination angles, 30° and 90°, were chosen for the inclined cable; wind with a uniform velocity profile (velocity profile U) and four types of velocity profiles were used as the inlet velocity conditions. Based on the cable vibration and flow field obtained under wind with varying velocity profiles, characteristics of the cable vibrations and aerodynamic coefficients in the time domain and frequency domain as well as the wake patterns were analyzed. The results indicated that the cable vibration exhibits two types of behavior—single-mode vibration (uniform wind or the wind with small velocity changes) and multimode vibration (the wind with large velocity changes)—according to the velocity profile over the inclined cable. The single-mode vibration of the cable exhibited a standing wave response (a chessboard pattern), whereas the multimode vibration exhibited a traveling wave motion (a parallel line pattern). The lift coefficients exhibited similar features as the cable vibration under wind with varying velocity profiles. Dominant frequencies of the lift coefficients could be observed over the upper segments of the cable under wind with small velocity changes (or the entire cable length under uniform wind), but a broader frequency band appeared in the lift coefficients over the low segments of the cable under wind with large velocity changes. The wake patterns and vortex shedding lock-in regions depended on the vibration type (or the wind velocity profiles). The single-mode vibration had large lock-in regions (the entire cable length or the upper segments of the cable), and over these regions, the vortex shedding was in phase or synchronous. However, for a multimode vibration (wind with large velocity changes), the vortex shedding is irregular and complicated along the cable axis.

Field tests of two jointless bridges are presented, focusing on the magnitude and significance of earth pressure behind the abutments. The Haavistonjoki Bridge is a 56-m-long, continuous three-span bridge. Instrumentation was used to measure the horizontal displacement of an abutment, abutment rotation, abutment pile strains, earth pressures behind the abutments, superstructure displacements, frost depth, and air temperature. The measured earth pressures were compared with pressures that had been calculated on the basis of Nordic codes of practice and the Eurocodes pertaining to bridges. The bridge over the Leduån is a single-span composite bridge with a cast-in-place concrete deck on top of two steel beams. This bridge, spanning 40 m, is slender, with a 1.7-m-high superstructure. The bridge was fitted with strain and displacement gauges and short-term measurements were made using a loaded truck. The field test results for this bridge were verified with calculations based on an abutment rotation stiffness calculation model developed during the research presented in this paper.

The deck slab of a prestressed concrete (PSC) box girder bridge serves as an integral part of the structural box section, and any damage to the deck may seriously affect the load-resistance mechanism of the bridge. This paper discusses a case study of a PSC box girder that developed early damage to the deck because of the poor placement of concrete at the time of construction as established by in situ nondestructive and core tests. The possibility of total replacement of the top deck was ruled out because of the disruptive and tedious nature of the process, which could have endangered the overall stability of the box girder. Strengthening of the deck was performed by installing a series of steel joists at the underside of the deck slab and adding a thin overlay of concrete at the top to restore it to the requisite strength. This approach was best suited, for it not only improved the stress distribution in the deck, but also maintained the original load resistance mechanism provided by the prestressing arrangement in the girder. The presented scheme was easy to implement and could be used for similar damages occurring to the deck of PSC box girder bridges.

Carbon fiber-reinforced polymer (CFRP) sheets can be used to strengthen existing RC bridge girders/decks. The objective of this study is to investigate whether transient vehicle loads present during the installation of CFRP influence the bond performance between the CFRP and the concrete substrate and, therefore, the consequent strengthening effect. A total of eight RC beams were tested. Two were unstrengthened reference beams, five were strengthened with CFRP sheets while subject to transient loads, and the final beam was strengthened with CFRP sheets while subject to only a sustained static load. The test parameters included the amplitude of the transient load, the anchorage length of the CFRP sheets, and the reinforcing/strengthening ratio. The transient loads were continued for 2 days (during CFRP cure) before all five CFRP-strengthened RC beams were tested to failure in four-point flexure. The test results were compared with those of the reference beams and the one strengthened under a sustained load. It was shown that a 1-Hz sinusoidal transient load varying between 30 and 50% of the ultimate capacity of the unstrengthened beam during the installation and curing of the CFRP sheets does not affect the structural performance of CFRP-strengthened RC beams. This result demonstrates the applicability of the fiber-reinforced polymer (FRP) strengthening technique for bridge girders that are subjected to continuous vehicle loads.