Journal of Bridge Engineering

In this study, a reliability-based calibration of live-load factors for bridge design specific to the state of Michigan was conducted. Two years of high-frequency weigh-in-motion (WIM) data from 20 representative statewide sites were analyzed, and load effects were generated for bridge spans from 6 to 122 m (20 to 400 ft), considering simple and continuous moments and shears, and single-lane and two-lane effects. Seventy-five-year statistics for maximum live load were then estimated with probabilistic projection. Bridge girders considered for the calibration included composite steel, prestressed concrete, side-by-side and spread box beams, and special long-span structural members. In some cases, it was found that Michigan load effects were greater than those previously assumed, often requiring higher load factors than those in current use. Moreover, significant variation in the required load factor was found, potentially resulting in significant inconsistencies in reliability if a single load factor is used for the design of all bridge types and load effects considered.

This paper investigates the use of kriging as a powerful tool for mode shape interpolation in flexible pedestrian bridges. Kriging can be used to interpolate a set of sparse mode shape coordinates obtained through experimental modal identification techniques onto a dense structure geometry, leading to better mode shape visualization and efficient model updating. Kriging does not require a finite-element (FE) model (analytical modes) to be available a priori, as it is purely statistical in nature. Despite this key advantage, kriging has not been adopted for mode shape interpolation due in part to the relative lack of exposure within the modal analysis community and the arbitrariness of the parameter defining the spatial correlation between discrete geometry points. The first objective of this paper is to describe kriging in detail, especially focusing on pedestrian bridges. Second, modifications to the computational procedure for kriging are presented with a new metric to automatically tune the correlation parameter and a batch process to undertake kriging for mode shape interpolation. Details of the procedure are discussed through both numerical and experimental field data obtained from flexible pedestrian bridges. It is anticipated that through the modifications described in this paper, kriging will spur greater interest from the pedestrian bridge modal analysis community.

The seismic performance of unbonded posttensioned (PT) precast segmental bridge columns (PSBCs) with circular sections was investigated. Two precast segmental columns with/without energy dissipation (ED) bars were tested under quasi-static cyclic loading. Their damage patterns and hysteretic curves were compared. Three single-degree-of-freedom hysteretic models were formulated with the key parameters identified from the tests. Cyclic loading analyses of the two tested specimens were conducted using these hysteretic models. Time history analyses were also conducted utilizing the proposed three hysteretic models to observe their applicability in modeling the dynamic responses of PSBCs. Finally, parametric analyses of PSBCs were performed under the excitation of an ensemble of 20 history ground motions to study the influences of relative yield force, elastic period, and energy dissipation coefficient on the earthquake responses of displacement ductility demand, maximum absolute acceleration, and maximum absorbed energy.

This paper describes flexural static and fatigue laboratory testing of a new connection design for spliced, prestressed precast I-girders. Intended for use as a field-spliced moment-carrying connection, the joint is prestressed at the jobsite using hydraulic jacks in the joint area. The new design was developed to achieve longer spans than currently permitted by transportation limits for precast segments. Behavior at the splice is similar to that of unbonded joints in segmental bridges. Whether the prestressed joint is considered bonded or unbonded for ultimate strength calculations, the spliced test specimens outperformed established calculated values when either loaded statically or after fatigue loading.

This paper describes shear laboratory testing of a joint design for spliced, prestressed precast I-girders. Although the connection is not intended to be placed at locations of high shear, the potential for shear-type failures was investigated. Shear testing of the spliced beam specimens is described and behavior is compared with calculated strength. The location of the unbonded strand was shown to affect the failure behavior. In the one specimen that did exhibit the characteristics of a shear failure, the vertical interface of the dry joint/closure pour was the location of failure. The ultimate shear capacity of the spliced specimens exceeded the calculated shear strength calculated per modified compression field theory and per the traditional simplified provisions. All specimens demonstrated flexural capacity exceeding calculated values, assuming an either bonded or unbonded strand. A companion paper describes flexural static and fatigue laboratory testing of the new connection.

Barriers are purposely omitted from the structural analysis for bridge design or load rating. They are not considered as primary structural members because, after a forceful collision, they may sustain some structural damage and would no longer strengthen the bridge deck. However, when completely intact, these secondary structural members do affect the distribution of applied loads, which is of interest to permitting agencies, such as the Florida Department of Transportation (FDOT). For a prestressed concrete segmental box girder bridge, both design and load rating (for oversized load permits) are determined by longitudinal and transverse analyses without considering the influence of the barriers. For the transverse analysis, the maximum moment generated from the live load is traditionally calculated from Homberg charts. These influence surfaces are based on plate behavior and idealized support conditions and are generally conservative. This moment estimation and the lack of consideration for the barriers create a conservative transverse design and load rating for the bridge. In this study, finite-element bridge models show how much the barrier affects transverse live-load moments on a prestressed concrete segmental box girder bridge. Data obtained from these models were compared with measurements obtained from a FDOT load test and also with predictions made from Homberg influence surfaces, which are often used by engineers in design. The results show that a continuous barrier reduced transverse moments, a jointed barrier behaved more similarly to no barrier, and Homberg design moments were generally conservative. In addition, the joint in the barrier played a large role in transverse live-load effects.

Finite-element simulations were used to investigate the structural demands generated by a truck colliding with a bridge pier. A reduced coupled mass-spring-damper model is proposed to simulate the collision process. The model is calibrated to the simulation data and accounts for key variables, such as truck weight, impact speed, and pier characteristics. Through comparisons to detailed finite-element analysis, the reduced model is shown to be able to reasonably represent the behavior of a set of coupled vehicle–bridge systems. On the basis of the simulation results, an argument is made that a standard design truck can be specified for impact design.

This article presents an approach for assessing time-variant risk associated with the closure of bridge lanes due to traffic loading and scour. Scenarios of lane closure are identified for both traffic loading and scour. The time-variant failure probabilities of girders under traffic loading are evaluated using a deterioration model that accounts for the decrease in resistance and a live-load model that considers the increase in maximum load effects. The annual failure probabilities of pier columns due to scour are estimated by comparing pier depth and scour depth caused by the annual peak flow. The probabilities of occurrence of the identified closure scenarios are computed on the basis of the relations between these scenarios and the failure events of girders and pier columns, and the correlations among failure modes are considered. The consequences caused by the lane-closure scenarios are evaluated by considering repair, running, and time-loss costs. Finally, the time-variant risks associated with the identified scenarios are assessed. The proposed approach is illustrated using a highway bridge as an example.

The hybrid composite beam (HCB) system is an innovative structural technology that has been recently used in bridge construction within the U.S. transportation network. This investigation focused on evaluating the in-service performance of a newly constructed HCB bridge superstructure located on Route 205 in Colonial Beach, Virginia. In the live-load-testing program, the bridge superstructure was instrumented with a series of internal and external strain gauges to evaluate the structural response. The test was conducted using tandem axle dump trucks under both quasi-static and dynamic conditions, and results obtained from the experimental investigation were used to determine three key behavior characteristics: lateral load distribution, internal load-sharing behavior, and dynamic load allowance (DLA). With respect to the live load performance, the HCB system relatively conformed to the provisions of the AASHTO specifications for beam-type bridges but did exhibit some characteristics of a flexible system when considering the dynamic response. It was also noted that the load-sharing behavior within the HCB system is noncomposite and even nonlinear as a result of local phenomena internal to the HCBs, which may be attributed to local arch bending and slippage between the tension steel and the resin matrix, suggesting that the internal components can actually behave independently.

This study evaluated the seismic performance of circular concrete-filled tube (CCFT) columns in accelerated bridge construction (ABC). In these projects, the bridge should be open to service after a couple of days. For this reason, this study evaluated the ability of CCFT columns to perform adequately under gravitational and seismic loading before the concrete reached its design strength at 28 days. A reduced seismic hazard that accounted for this temporal condition was implemented, and a performance evaluation was performed to obtain the CCFT column probability of failure. The performance of a CCFT column was compared with that of a circular reinforced concrete (RC) column of a Caltrans bridge. Probabilistic analyses using reduced seismic loads for the temporary condition indicated that CCFT columns with partial design concrete compressive strength can be used for ABC without resulting in an increased probability of failure.

Cross frames are critical structural elements in both straight and horizontally curved steel bridges. In order to properly size the brace for the strength and stiffness demands of the superstructure, an accurate model of the elements comprising the cross frame is required. Conventional details most commonly used for cross frames consist of single-angle members connected to form a truss system linking adjacent girders together. Most analyses of the bridges treat the cross frames as truss elements that primarily resist applied forces through the axial stiffness of the members. This paper documents the results of a research study that included full-scale laboratory tests to measure the stiffness and strength of cross frames utilizing both conventional and new details. The tests showed that analytical solutions, as well as computer models, that are routinely used to model the cross frames in analysis software can overestimate the in-plane stiffness of the brace by more than 100%. The primary reason for the discrepancy in the stiffness models was identified to be connection eccentricities that exist in cross frames comprised of single-angle members welded to a gusset plate through only one leg of the angle. Overestimating the stiffness of the braces during construction can lead to unsafe conditions, as well as errors in the geometry of the constructed bridge, resulting from underpredictions of deformations during concrete placement. Extensive parametric analyses were carried out using validated finite-element models to develop correction factors that can be applied to analytical and computer models to significantly improve the accuracy of the simplified models that are used for the cross-frame systems. The correction factors allow a designer to utilize a computationally efficient model for the cross frame while also including the reduction in stiffness that is caused by connection eccentricities.

Accelerated bridge construction (ABC) is gaining popularity because it offers a number of advantages over conventional methods of construction. Prefabrication of bridge components for rapid on-site assembly is a highly effective ABC approach. Bridge substructures are traditionally cast in place with columns that form plastic hinges during earthquake events. Recent studies have explored new precast column-footing connections that either emulate the seismic performance of monolithic construction (emulative) or show improved seismic performance with the use of rocking connections (nonemulative). This paper presents findings from half-scale experimental testing of one emulative and two nonemulative precast column-footing connections. The two nonemulative connections were designed and detailed to sustain limited damage that can be rapidly and cost-effectively repaired using predefined methodologies. The tested connections showed promising results for use in regions of moderate to high seismicity; however, further developments of the proposed construction and repair methodologies are required for their full potential to be realized.

In this study, push-out tests were conducted to evaluate the interface behavior between steel and concrete connected by bonding. Twelve sets of specimens with different adhesive thicknesses, modulus of elasticity, and interface roughness were investigated. Failure modes, ultimate loads, and load–slip relationships are reported and analyzed in the paper. The experimental results show that factors such as modulus of elasticity of adhesive materials, adhesive thickness, and interface roughness affect the interface behavior. An analytical model is proposed to predict the ultimate load value and the load–slip relationship. The comparison between test results and analytical predictions shows good agreement.

Many studies have been conducted to investigate the seismic fragility of regular girder bridges, whereas only a few studies have focused on irregular cable-stayed bridges. In this study, the performance of a medium-span concrete cable-stayed (MSCCS) bridge is investigated by developing its component and system fragility functions. Structural analysis software is used to build the finite-element model of the case study bridge, and the ground-motion bin is assembled to match the practical site condition. The spectral acceleration at the geometric mean of the two periods of both longitudinal and transverse directions is applied to develop the component fragility functions. Analyses show that the bearings and transverse retainers are the most fragile components, whereas the failure probabilities of the concrete towers are relatively lower. This study proposes a user-friendly and efficient framework for establishing the system fragility functions on the basis of serial system assumption. Moreover, the fragility functions are applied to investigate a reasonable earthquake-resisting system for MSCCS bridges. The results show that the fragility functions are sensitive to the stiffness of the cable restrainer, and the longitudinal constraint system performs better than the total floating system and the total rigid system, especially when the stiffness of the cable restrainers is adjusted to a certain acceptable value.

Precast deck panel bridges have a history of poor performance in Florida. In the aftermath of five localized punching shear failures, a decision was made to replace all 127 deck panel bridges in southwest Florida over a 10-year period. This paper describes the strategy used for prioritizing their replacement. Two risk indices focusing on safety and importance were developed for this purpose. The safety index was determined from available inspection records and the importance index from annual daily traffic. Forty-two bridges had already been earmarked for replacement. The 85 remaining bridges were ranked using this protocol; 39 were rated as being in good condition. The replacement order for the remaining 46 bridges was prioritized. The validity of the approach was confirmed by the subsequent localized failure of two bridges, both identified as being at high risk.

This article presents the real-world implementation of a novel monitoring system in which video images and conventional sensor network data are simultaneously analyzed to detect possible damage on a movable bridge. The monitoring system was designed to detect such problems at the onset of damage. A video stream of traffic is processed to detect and classify vehicles to determine the vehicle load and location, while strain measurements are simultaneously collected at various critical locations on the bridge for both normal and damage conditions. A series of unit influence lines can then be extracted for all of the scenarios using the image and sensor data. Because large data sets result from continuous monitoring, the system also includes a statistical outlier-detection algorithm. The proposed methodology was successfully used to detect and locate common damage scenarios on a real-world bascule bridge.

As a new high-strength material, carbon-fiber-reinforced plastic (CFRP) cable provided an opportunity for bridge structures to achieve span limits that were not possible or economical in the past. Based on existing conceptual studies on ultralong-span bridges using CFRP, a new partially earth-anchored cable bridge (PEAC bridge) system was proposed in this study. The PEAC bridge system addresses existing challenges of applying CFRP to traditional suspension and cable-stayed bridge systems, while maintaining large spanning capacity. Through a detailed comparative study on traditional and proposed long-span systems, the performance of a PEAC bridge under static and dynamic loading was evaluated. The span limit of the PEAC bridge system was calculated based on the strength of material and cost considerations.