Journal of Bridge Engineering

Horizontal cracks occur at the web of prestressed bridge girder ends during prestress release due to the spalling force, which is produced by the transfer of the prestressing force to the concrete. To avoid using a trial-and-error method to determine the maximum moment section and unify the methods to determine the arm length of the maximum moment in the conventional Gergely-Sozen (G-S) model, a modified G-S model is proposed to calculate the spalling force in a pretensioned anchorage zone. A compression dispersion model (CDM) is developed and formulated for the pretensioned anchorage zone, and the extreme edge of isostatic lines is visualized mathematically with its geometric and physical boundary conditions, which can be used to accurately predict the arm length and spalling force of the horizontal moment at any horizontal section. Based on experimental and field observations and previous finite-element method (FEM) analysis results, the junction section between the bottom flange and the web, instead of the maximum moment section, is recommended as a critical control section to calculate the spalling force in the web. Compared with the numerical and experimental results, a good agreement was obtained confirming that the modified G-S model method not only can predict the arm length of the horizontal moment, but can also evaluate the magnitude of the spalling force. A case study illustrates how to apply the proposed method to calculate the spalling force at the junction section, and the results explain why horizontal cracks often occur at the junction section.

Many research initiatives over the last two decades have had the objective of increasing the available spans of prestressed concrete bridge girders and have resulted in optimized cross-sectional shapes. To achieve these long spans, greater levels of prestress force are required. In addition, sections have been redesigned to increase the eccentricity of the prestress, which involves flattening and widening of the bottom bulb. The greater pretension forces and more slender bulbs have a number of effects at the girder end that affect both the serviceability and the ultimate behavior of the girder. This article identifies some behaviors that are potentially exacerbated when larger prestress forces are introduced and proposes a strut-and-tie modeling approach to better understand and mitigate these effects through improved girder end-region detailing. The utility of the proposed strut-and-tie model is demonstrated in two instances: (1) an investigation of the effects of strand debonding and (2) an investigation of the potential impacts of the larger prestress forces resulting from increasing girder span [specifically by adopting larger-diameter strands of 17.8 mm (0.7 in.)]. In the first instance, the strut-and-tie approach was used to establish guidance for preferred strand and strand debonding patterns. In the second instance, the impact of providing greater prestress forces through the use of larger strands was shown to have little effect on the development of transverse stresses at girder ends. The strut-and-tie approach was also demonstrated to be able to be rapidly applied over a wide range of parameters, allowing trends related to girder geometry, for instance, to be established. Thus, the utility of the strut-and-tie method in general, and the proposed model in particular, is demonstrated to provide a powerful tool for the rational analysis of the complex stress state occurring at the end regions of prestressed concrete bridge girders.

The concept of an equivalent plastic-hinge model was used to evaluate the lateral behavior of precast segmental ultra-high-performance concrete (UHPC) bridge columns. The columns consist of precast UHPC segments that were integrated by unbonded post-tensioning (PT) tendons and complemented with energy dissipation (ED) bars. The equivalent constitutive relationships of the PT tendons and the ED bars were analytically derived to meet the plane section assumption for the sectional analysis. Cyclic loading tests were conducted on three specimens to validate the proposed equivalent plastic-hinge model. The test results revealed that the proposed model showed good accuracy for predicting the lateral skeleton curve, stiffness, and opening at the base joint. Using the validated model, a parametric study was conducted to investigate the effects of nine main design parameters on the lateral behavior of the column. The parametric study showed that increasing the PT level is more economical to enhance the PT force than relying on the larger ratio of the PT tendons when the PT level is no more than 44% of the yield strength. Finally, a simplified formula for the effective stiffness was developed through regressive analysis. The effective stiffness depended on the aspect ratio, the number and unbonded length of ED bars, and axial loading ratio.

In the past, seismic-induced bearing uplift of cable-stayed bridges was seldom recognized or valued, and a conventional sliding bearing, which cannot resist uplift forces, was often used for cable-stayed bridges. In this study, identification of the bearing uplift of cable-stayed bridges was performed, and the bearing uplift effect was estimated according to a nonlinear response-history analysis of a cable-stayed bridge model subjected to longitudinal earthquake excitation. The findings revealed that deck end developed large vertical displacement during bearing uplift, and intense bearing compression force was generated through pounding of the deck against the bearings after bearing uplift. Because of bearing uplift, bearings and piers might suffer severe damage from a large bearing compression force, but the maximum response of the deck and the tower was slightly affected. In addition, bearing uplift is quite sensitive to ground motion characteristics. Particularly, bearing uplift is more prone to develop under soft soil conditions than stiff soil conditions because the predominant periods of the former are closer to the periods of significant structural modes. To avoid damage from a bearing uplift effect, uplift-restraint bearings, which introduce uplift-restraint mechanisms to sliding bearings, are recommended as substitutes for conventional sliding bearings.

This paper presents a passive control device—multidirectional cable restrainer (MDCR)—and develops the fragility curves of a curved bridge with and without MDCR, respectively. Three-dimensional finite-element models with and without MDCR were established by OpenSees software. In order to consider ground motion uncertainty, 80 ground motion records were selected from the Pacific Earthquake Engineering Research (PEER) strong motion database. Meanwhile, the parameters of MDCR, such as initial slack and stiffness of cable and the incidence angle of the motions, were treated as random variables. The probabilistic capacity model of the pier with varying gravity force was established using the response surface method (RSM). System fragility curves of the bridge with and without MDCR were developed by Monte Carlo (MC) simulation, and the correlation of the bridge components was considered in the process of sampling. The results show that of the six intensity measures (IMs) considered, spectral acceleration at the period of 1 s (Sa10) is the optimal IM for the curved bridge prototype. In addition, piers are the least vulnerable component for the system with MDCR.

This study provides the result of dynamic effects of self-propelled modular transporter (SPMT) movements during a simulated bridge transport. Accelerometers were attached to an SPMT and different transport scenarios were simulated under several loading conditions. For the vertical direction, a traverse of uneven terrain provided the peak platform acceleration (PPA), whereas a rapid start-and-stop motion controlled both longitudinal and transverse directions. The observed PPA increased as the total weight decreased. Vertical accelerations were found to be highly dependent on the SPMT speed. The horizontal accelerations were not affected by the speed of the SPMT, but they were affected by the ability of the operator to perform a braking operation. Response spectra suitable for the design for each direction were constructed from the time history graphs corresponding to these tests. The maximum pseudo accelerations (MPAs) for heavy, medium, and light load cases were 0.68, 1.07, and 2.3g in the vertical direction; they were significantly more than the horizontal direction with 0.32g for heavy, 0.47g for medium, and 0.60g for the light load case. The dynamic forces on the cargo (falsework or bridge) either increased or stayed constant as the stiffness of the system increased. Although vertical accelerations, which mostly affect the bridge superstructure components, can be reduced by limiting the SPMT speed, the horizontal accelerations can be reduced because some ductility and inelasticity within the bracing of the falsework and response modification factors may be applied to the developed spectra. Dynamic dead load impact (DLI) factors were found to be in excess of those often used for the design because of the worst-case nature of the presented results.

Due to the aging of transportation infrastructures and the ever-increasing traffic, the condition assessment of bridges has become increasingly important because it provides useful information for bridge management. A reliable condition assessment depends on the accurate prediction of extreme traffic load effects (LEs) in the remaining life of bridges. In this study, the Bayesian method is introduced for the prediction of extreme traffic LEs to improve the reliability of the prediction, and a framework for bridge condition assessment making use of the predicted LEs is proposed. To demonstrate the proposed methodology, a case study on the condition assessment of the new I-10 Twin Span Bridge (TSB) using structural health monitoring data is presented. The results show that the Bayesian method can provide more reliable predictions compared with the conventional method, because it quantifies the uncertainties inherent in the parameters and incorporates these uncertainties into the prediction. Based on the predicted traffic LEs, the condition of the bridge is assessed using the proposed framework.