Journal of Bridge Engineering

Nonintegrated multicolumn bents are the common type of substructure in existing bridges. Among all possible seismic defects in their behavior, capbeam–column joint behavior has not properly been acknowledged so far. Although for other defects some retrofit methods are available, there are only a few methods proposed for improving the shear strength of these joints. The purpose of this research was to assess the cyclic response of such bents and suggest a practical retrofit method for improving joint behavior without significant damage. To this end, a ¼-scale two-column bent, representative of existing bridges built before 1990, was tested. Using test results, a passive retrofitting method based on the increasing shear strength of joints was developed. The performance of the proposed method was examined on a new specimen similar to the first one. The results show that, unlike the reference bent, which was damaged extensively in joint regions, the joints in the retrofitted specimen did not suffer any damage, and all plastic hinges were formed at both ends of the columns, the desirable failure mode followed by seismic codes.

Although the majority of uncoated weathering steel (UWS) structures are performing well when designed and maintained in accordance with existing recommendations, exceptions to this trend exist. Additionally, the guidance on UWS use is largely qualitative. For these reasons, a data-driven method for assessing deterioration of a large sample of bridges as a function of environment was formulated, piloted, assessed, and validated. Key aspects of the methodology include the development of a database that uses graphical information systems to quantify the climate of hundreds of structures, a statistically driven process for selecting a representative sample of bridges for further evaluation, and a review of bridge performance based on owners’ evaluations and independent field evaluations. As a result of applying this methodology to UWS bridges with proximity to the Gulf Coast, environments consistently resulting in satisfactory performance were quantified; a rare combination of severe proximity to the coast, humidity, and atmospheric concentrations of chloride was also quantified and correlated with the small number of bridges with overall inferior corrosion performance in these marine environments.

Slight bracing of steel bridges to control end warping of the compression flanges near the end supports can be very effective in enhancing the load-carrying capacity of the main girders that are involved. Relatively little information is available, however, regarding the stiffness and strength requirements of end-warping bracings. The results of seven large-scale laboratory tests of different configurations of bracings, including plan bracing and corrugated metal sheets, used in a twin I-girder bridge to resist end warping at support points are presented here. Bracing forces generated in the plan bracings of the test bridge were also compared with those obtained through an approximate analysis for the preliminary analysis of such bracings. Moreover, the bracing forces present in the cross beam of the test bridge were obtained for the different end-warping brace configurations that were investigated. The bracings, the plan bracing type and corrugated metal sheets, were found to effectively enhance the load-carrying capacity of the test bridge. Only relatively small forces were generated in the bracing bars to achieve such a substantial increase (by a factor of greater than 2.0) in the load-carrying capacity of the studied bridges.

A new bridge type called a spread slab-beam bridge was developed recently by using the same concept as that used for spread box beam bridges, in which the beams are spaced apart. This paper presents an evaluation of the new bridge system in terms of design and constructability and investigates the challenges encountered during construction. Forty-four bridge geometries were designed by using standard Texas DOT (TxDOT) slab-beam types to determine the feasible design space. One of the most aggressive geometries with widely spaced slab beams was constructed at full scale. On the basis of the research findings, it was concluded that spread slab beams with a topped panelized deck provide a viable construction method for short-span bridges. Results of a shear-design check show that the standard interface shear reinforcement must be doubled for the first quarter of the span length for more aggressive designs. Larger-than-expected camber of the precast slab beams might cause construction challenges and delays. A 30% increase in standard camber estimates can be used for construction purposes. High differential temperatures together with drying shrinkage might lead to early-age longitudinal deck cracking. Care should be taken during concrete curing to avoid factors that contribute to large temperature changes in the deck.

Horizontally curved bridges have been observed to suffer severe structural damage during past earthquakes. Unseating of the deck from abutments is one of the typical modes of failure of horizontal curved bridges. This type of failure is caused primarily because of excessive in-plane rigid-body motion of decks of these bridges, mainly due to the irregular geometry of the bridge itself and seismic pounding between the deck and abutments. This paper investigates the influence of seismic pounding on rigid-body motion of horizontally curved bridges during strong earthquakes using an analytical approach. For this purpose, a 3-degree-of-freedom nonlinear model has been developed to capture main dynamic parameters affecting seismic response of horizontally curved bridges. The accuracy of this model has been verified by finite-element modeling of a typical horizontal curved bridge. An extensive parametric analysis has been performed by varying key parameters, including the size of the gap between the deck and abutments, the subtended angle of the deck, and the presence of friction. Numerical results show that in-plane response quantities of horizontally curved bridges, such as radial and azimuthal displacements of corners of the deck and rotation of the deck about the mass center, are affected noticeably because of seismic pounding. Validity and reasonableness of seismic design guideline recommendations for regular curved bridges in the presence of seismic pounding have also been investigated.

This paper proposes a generalized equivalent spectral model of pedestrian-induced forces on footbridges that can be applied in serviceability analysis, from unrestricted traffic to crowded conditions. The model is based on experimental relationships among pedestrian density, step velocity, and step frequency taken from the literature. A suitable physically based expression for the coherence function is introduced that is dependent on pedestrian density and distance among pedestrians; it reduces to perfect uncorrelation for spatially unrestricted traffic and to perfect correlation for crowded conditions. Furthermore, the harmonic content of pedestrian-induced forces is defined as a function of the pedestrian density. Based on the proposed spectral model of the loading, the classic methods of linear random dynamics can be adopted for vibration serviceability analysis of footbridges. Similar to in the wind engineering field, simple closed-form expressions are provided for the evaluation of the maximum dynamic response.

To investigate the efficiency of passive energy dissipation devices in protecting cable-stayed bridges from severe damage when subjected to high-intensity ground motions, a 1/20-scale bridge model from a typical medium-span concrete cable-stayed bridge was designed, constructed, and tested on shake tables. Yielding steel damper (YSD) and viscous fluid damper (VFD) were used as passive energy dissipation devices in the transverse and longitudinal directions, respectively. The bridge models were then excited by two ground motions with different spectral characteristics in transverse and longitudinal directions accordingly, including the Chi-Chi earthquake wave and a site-specific artificial wave. The seismic responses of the bridge models with and without the passive devices were analyzed and compared. Both numerical and test results show that the YSD applied in the transverse direction could reduce the seismic responses of the towers effectively, and the VFD applied in the longitudinal direction could reduce the displacement of the deck effectively while simultaneously limiting the seismic responses of the towers. Therefore, it is suggested that the proposed energy dissipation system, consisting of YSD transversely and VFD longitudinally along with the sliding bearing, is applied in the cable-stayed bridges for better overall seismic performance compared to the traditional isolation system that provides efficiency only longitudinally.

With the specific objective of exploring the surface pressure characteristics and further revealing the torsional vortex-induced vibration (VIV) mechanisms of a bridge deck with a particular geometry, numerous simultaneous pressure measurement campaigns were performed in a wind tunnel for aerodynamic-countermeasure–modified and unmodified sections of a section model at different angles of incidence under the conditions of smooth or turbulent flow. The mean and fluctuating pressure distributions, instantaneous pressures at typical instants, dominant pressure frequencies, pressure phase differences at the dominant frequency of individual pressure measurement taps, and the correlation coefficients among local and global torsional moments were studied, revealing the origins and mechanisms of torsional VIVs. The results demonstrate that the angle of incidence, flow conditions (smooth or turbulent), and installation of a spoiler exert significant effects on the surface pressure distributions, hence affecting the corresponding aerodynamic performance of the bridge deck. Turbulence on the top surface can potentially neutralize the vortex shedding effects and enhance immunity to torsional VIVs. The signature turbulence from the leading (fairing) edge was effectively weakened or even destroyed by sufficiently intense oncoming turbulence and/or the turbulence generated by a spoiler with an appropriate configuration and location. Therefore, potential torsional VIVs could be suppressed by the interaction of vortices generated by oncoming and signature turbulences. This knowledge is essential for a thorough evaluation of the potential for torsional VIVs for this particular bridge deck.

Shear connectors are commonly used in steel bridges to join the concrete deck and steel superstructure, providing a mechanism for shear transfer across the steel–concrete interface. The most common type of shear connector is the headed shear stud. In the current American Association of State Highway and Transportation Officials (AASHTO) LRFD Bridge Specifications on composite design, shear stud fatigue often governs over static strength, and a large number of shear connectors often result. The stud fatigue capacities presented in the AASHTO standard are largely based on a limited sample of composite fatigue tests performed in the 1960s, with limited fatigue test data at lower stress ranges leading to a somewhat arbitrary constant amplitude fatigue limit (CAFL). This often governs the composite design for bridges with moderate-to-high traffic demands. This paper presents results from an experimental and analytical study of the fatigue behavior of headed shear studs to address the lack of existing experimental data near the assumed CAFL and to better characterize the effects of fatigue uncertainty on predicted response. Results from composite push-out specimens tested at low stress ranges of between 30 and 60 MPa (4.4 and 8.7 ksi) suggest a fatigue limit of 44.8 MPa (6.5 ksi), which is near the existing limit of 7 ksi. Recommendations for modifying the existing AASHTO finite life shear stud S-N fatigue capacity curve are proposed.

This paper describes depth-calibrated live-load attenuation for the load rating of reinforced-concrete box culverts using production-simplified models. In-plane depth calibration is accomplished using a production-simplified, two-dimensional, linear-elastic, finite-element, soil-structure interaction model with results compared with those from the recommended direct-stiffness, structural-frame model. Out-of-plane live-load attenuation considers each potential critical section depth rather than the cover soil depth only. The effectiveness of depth calibration is assessed by comparing predicted live-load moments obtained from the models versus measured live-load moments obtained from full-scale culvert load tests. A load rating case study illustrates the potential for improved alignment between load rating and observed performance. Findings show that depth calibration improves current load rating practice by increasing the accuracy and precision of live-load demand predictions, particularly in culvert walls and bottom slabs. Use of the depth-calibrated soil-structure model helps close the disconnect between calculated load rating and observed structural performance by more accurately predicting both the location of the weakest critical section and the live-load magnitude. The depth-calibrated soil-structure model also moves the predicted live load toward more uniform accuracy and precision across all sections.

Fatigue cracking is crucial issue for welded steel bridges that endure repeated vehicle loads. With typical fatigue-life evaluation methods, it is difficult to consider the influence of low stress amplitude. Based on existing full-range S-N curves and their formulas for base metal, a unified fatigue-life evaluation method is recommended for welded joints. The basis for construction of full-range S-N curves is to obtain the welding residual stress and stress concentration factor. In this paper, rib-to-deck welded joints are introduced as examples to demonstrate the detailed steps of a fatigue-life evaluation procedure. From the analysis of welding residual stress, it can be concluded that residual stress is influenced by the geometric dimensions of the composite members. With an increase in the geometric dimensions, the level of residual stress is increased. In addition, static-load experiments were conducted to obtain the stress concentration factor. A hot-spot full-range S-N curve was then obtained for rib-to-deck welded joints by calibrating the formula from small quantities of fatigue experiments. An actual bridge installed with a structural health monitoring system (SHMS) was used as a case study to compare the recommended model with typical fatigue codes. The comparison results indicate that the recommended model is more conservative than typical fatigue codes.

Advances in the field of fracture mechanics over the last four decades have allowed for a greater understanding of brittle and ductile fracture, including how to account for plasticity in fracture specimens and how to statistically define data scatter in the brittle and ductile-brittle transition region. This is particularly important for structural steels used in bridge applications, which typically possess high ductility and toughness. The material toughness provisions of the fracture control plan for highway bridges in the United States were developed from a database of linear-elastic tests, many of which were considered to be invalid according to testing standards in place at the time. This, the first of two companion papers, presents a reevaluation of legacy fracture toughness data obtained from steels used in highway bridges. This reevaluation uses plasticity corrections that characterize the toughness behavior of these steels using the master curve approach. Results of the study indicate that the master curve approach can be used to accurately describe the temperature dependence and associated scatter of fracture behavior in the ductile-brittle transition region of historic bridge steels. This can be used to define cleavage fracture tolerance bounds for a given data set, which is a necessary step toward the development of a performance-based fracture design specification.

Many material toughness specifications, including a fracture control plan (FCP) for highway bridges introduced in 1978, in part rely on Charpy V-notch (CVN) requirements and associated correlations to set material toughness values at acceptable levels that will prevent brittle fracture in bridges. However, it has long been recognized that this is an indirect approach because CVN impact energy is not a direct measure of fracture toughness. Significant advances in the understanding of fracture mechanics and material behavior have taken place in the four decades since the FCP was established. These advancements are in part due to the large amount of fracture toughness data that has been generated, allowing for statistical evaluation. As a result, this provides an opportunity to evaluate CVN toughness correlations. Overall, these advances allow a more rational approach to establishing material toughness requirements for bridge steels including (1) the ability to characterize the scatter of fracture data in the ductile to brittle transition region through the master curve methodology and (2) numerous proposals for correlation procedures to relate CVN values to fracture mechanics parameters. This paper presents an analysis of 29 permutations of methods for correlating CVN to reference temperature, T_o. Results indicate that although no single method acts as a true predictor for reference temperature, the dispersion of the estimated values is quantifiable, allowing for fracture to be treated like other design limit states.

In recent years, several accidental collapses of girder bridges have occurred under the influence of eccentric heavy vehicles in China. To explore the causes and mechanisms of the collapses, a collapsed six-span, RC, continuous box-girder bridge was introduced and analyzed. By gathering onsite information about the accident, different failure forms were determined. Simplified mechanical calculations and finite-element simulations were used to determine the characteristics of the structural response caused by the load on the bridge during the accident. Finally, the sequence of the accident was reproduced, and the controlling factors leading to the structure failure were clarified. This study suggests that bridge accidents are caused not only by instability issues of integral overturning but also by strength failures of the components. Under the circumstances caused by an extreme eccentric vehicle load, the large deformation resulted in an apparent second-order effect. Noticeable lateral horizontal interaction forces between the piers and girder appeared, which led to bending failures in the piers that later caused the collapse of the superstructure.

There has been little full-scale physical experimentation to support the findings of numerous computational studies regarding the contribution of various substructure components to overall bridge-foundation system behavior. In response to this lack of experimental data, a field testing program was undertaken to investigate the in situ dynamic characteristics of a 27-m-long, three-span, precast-concrete bridge. Forced vibration testing and system identification were used to characterize the dominant modal behavior of the bridge-foundation system in each loading direction, leading to identification of the likely force-transfer mechanisms between the structure and the substructure components. It was determined that both the transverse and longitudinal responses of the bridge were dominated by the abutment stiffness, with the passive resistance of the buried settlement slab contributing significantly to the transverse response and the backfill passive resistance dominating the longitudinal response.

To overcome the limitations of traditional long-span suspension bridges with steel cables, the mechanical behavior of suspension bridges with various fiber-reinforced polymer (FRP) and hybrid FRP cables have been investigated comprehensively. The authors adopted a prototype of the proposed Messina Strait Bridge with a main span of 3,300 m to use as a suspension bridge model for analysis. The static and dynamic behaviors of the entire suspension bridge with steel and different FRP cables were analyzed by using the finite-element method. The results indicate that the application of FRP cables in a long-span suspension bridge benefits the spanning ability, improves the load-carrying efficiency, and reduces the axial force of the cable. The modal analysis results show that the long-span suspension bridge with cables of different materials have similar mode shapes, although the natural frequencies of the bridges with FRP cables are larger than that of the bridge with steel cable because of lower self-weight. It is also revealed that the adoption of FRP cables decreases the seismic responses of the suspension bridge because of their better energy-dissipation behavior, and FRP cables improve the flutter critical wind speed and contribute to the overall aerodynamic stability of large-span suspension bridges.

This paper presents an approach for the time-dependent risk assessment of bridges based on cumulative-time failure probability of bridge members. The cumulative-time failure probabilities of girders under traffic load and pier columns under scour are computed. Different scenarios of lane closure are identified and expressed in terms of failure events of bridge components. Two extreme correlation cases among the failure modes of bridge components are considered. After estimating the occurrence probabilities and consequences associated with these scenarios, the time-dependent risks are assessed. Considering the entire history of loadings and structural deteriorating processes by using the cumulative-time approach provides a more realistic basis for risk assessment of bridges than the approach based on annual failure probabilities.

To seek a proper way to calculate the preoffsetting value of saddle on a suspension bridge, some influences of structural parameters on the preoffsetting are considered, then a new method for the preoffsetting value of saddle is proposed. On the basis of the relationship of the main cable and saddle, equilibrium equations, geometric equations, and physical equations are deduced. Then, the equations are solved by the Newton-Raphson iteration method to derive the Jacobian influence matrix of preoffsetting. For convergence of the numerical calculation, constraint conditions and the range of initial values are given. Through comparison of the measured value in a 1/30-scale model test and the calculated value, the reliability of the new method is confirmed. Finally, parametric analysis of preoffsetting shows that (1) with an increase of the saddle radius, saddle preoffsetting increases in the middle-span direction; (2) with increases of the side/middle-span and rise/span ratios, saddle preoffsetting increases in the side-span direction; (3) the influence of cable self-weight on saddle preoffsetting is negligible; (4) with an increase of unstressed cable length in the side span, saddle preoffsetting increases in the middle-span direction; and (5) with an increase of unstressed cable length in the middle span, saddle preoffsetting increases in the side-span direction.