Bridge Structures

Linear-elastic fracture mechanics has traditionally been used to assess the fracture initiation resistance of bridge structures in the presence of a fatigue crack. Because conventional bridge steels typically fail by brittle cleavage, this approach has worked reasonably well. The new generation of high performance steels (HPS) has significantly higher toughness compared to conventional steel. The failure mode changes to ductile rupture preceded by significant crack tip plasticity. Localized stress redistribution occurs prior to fracture, thereby minimizing the influence of local geometry on the fracture process. Under these conditions, limit load analysis provides an accurate method for fracture prediction in steel bridge members. The limit load analysis procedure greatly simplifies the computational procedures, making it a more practical tool for use by structural engineers. This paper demonstrates how limit load analysis can be applied to analyze the fracture resistance of Steel I-girders with fatigue cracks.

Estimation of fatigue cycles for railroad bridges is complicated by the variation in live-load moment as a train passes across a bridge. Railroad trains create a maximum bending moment cycle like highway vehicles but also generate other potential cycles owing to the passage of the railcars over a bridge. The magnitude of these other cycles depends on other factors besides weight. Fatigue design of railroad bridges in North America is based on the expected number of cycles for a typical unit-coal train. While the assumed train and its frequency of application of loading are adequate for design of new spans, it does not represent all loadings. This can result in error when rating for and estimating fatigue life. The current study displays the fundamental formulation for moment range, the variation in live-load moment, for railroad loadings. Train types are displayed representing different eras of weights and car types as these have changed over time. Additionally, information is provided concerning span response and impact.

The key factors to be evaluated to ascertain the remaining safe fatigue life of a railroad bridge are (a) geometric structural fatigue detail category and its stress range to cycle capacity (S – N) and (b) the applicable stress ranges. Laboratory experimentation has defined the geometric structural fatigue detail categories and their stress range to cycle capacities (S – N) at constant amplitude to a level that is generally much better defined than the applicable stress range. The applicable stress ranges depend on the load spectra and the geometric properties of the structure being evaluated. Calculated stress ranges are rarely accurate enough for realistic remaining safe fatigue life predictions. Critical stress ranges are generally quite low in relation to the ultimate stress capacity for which most analytical models have been developed, and field strain gauge testing, assuming the cost is worth the saving, is often necessary to get a more reasonable estimate of stress range to applied load. The next and often most difficult factor to evaluate is the appropriate load spectra, particularly for structures already subject to many years of loading. The present paper will discuss the issues raised above and show a number of short examples of the influence of these various factors.

The singular stress field of a transverse fillet weld toe is evaluated both analytically and numerically. A closed-form solution for the singular stress field is utilized for evaluating the stress intensity factor, K, of a crack emanating from the weld toe. The closed-form solution for the stress intensity factor is shown to be nearly identical with numerical results for two distinct crack paths. The closed-form solution provides new insight into the behaviour of the stress intensity factor for very short crack lengths, where much of the welded detail fatigue life is spent. The results have direct relevance to welded bridge details when assessing a cracked condition.

Post-weld enhancement of the fatigue resistance of common attachment details such as transverse stiffeners, cover-plates, gusset plates, bulkheads and other welded details that experience crack growth from a weld toe is essential for efficient use of modern high- performance steels. Recently fatigue performance of treated welded joints was evaluated in large-scale rolled and built-up ferrite-steel beam specimens having yield strength of 345 to 690 MPa. The welded details consisted of transverse welds at the cover-plate terminus and at the transverse stiffener to tension flange joint. The welds were treated at the toe by ultrasonic impact treatment (UIT). In as-welded condition, these details are characterized as categories C′ and E′ in the AASHTO specification. Accelerated constant amplitude fatigue tests were conducted at various treatments of tensile minimum stress and stress range, resulting in positive stress ratios up to 0.6. The test results confirmed that the post-weld impact treatments substantially improved the fatigue performance of welded details. Analyses of the data in comparison to the current AASHTO design curves for as-welded details indicated that the treatments effectively elevated the fatigue limit without changing the slope of the S – N curve in the finite life region. The extent of enhancement was found to be dependent on both minimum stress and stress range. Within the range of available test results, the increase in fatigue strength appeared to be independent of the strength of material. Fatigue design curves based on two parameters, minimum stress and stress range, was developed for the post-weld treated details as a modification to the current AASHTO specification for as-welded details. Simplified fatigue design guidelines are proposed for post-weld treated details conforming to AASHTO category C′ and E′ details. The treatment provisions should be included in the specifications for enhancing performance of both new designs and existing construction.

Many existing riveted bridges in China are often required to carry an increasing volume of traffic, usually consisting of vehicles heavier than the original design. Recently Chinese government and engineers have paid more attention to the actual fatigue life and service safety of such infrastructures, and have decided to research evaluation methods using modern technologies. Therefore, based on investigation of fatigue – damage accumulation theory and probabilistic fracture mechanics, the evaluation methods have been developed to calculate the deterministic and probabilistic remaining fatigue life of existing riveted bridges. According to the fatigue failure mechanism of existing riveted bridges, a simple fracture evaluation model and a single-angle probabilistic fracture failure model (SAPFFM) were proposed, and based on Monte-Carlo method, a fatigue reliability analysis program was developed. Finally, as a case study, the fatigue evaluation models were used to predicate the fatigue safety of Zhejiang Street Bridge. According to the evaluation results, the safe inspection intervals and maintenance strategy are determined, which can control and avoid fatigue failure accident in the bridge remaining service life.