Rib-to-deck (RTD) innovative double-sided welded joints have the potential to enhance the fatigue resistance of orthotropic steel decks (OSDs). However, welding residual stress (WRS) has been reported to be a contributing factor to fatigue cracking of steel bridges. This paper performed fatigue crack growth of the RTD innovative double-sided welded joints of OSD considering WRS. Firstly, a thermal-mechanical sequential coupling finite element model (FEM) of the RTD double-sided welded joints was established. The WRS distribution along the direction of deck thickness was analyzed. Secondly, the FEM of the OSD segment with crack was established. The fatigue crack behavior of RTD double-sided welded joints under the combined action of vehicle load and WRS was studied. Finally, the influences of initial crack on fatigue life was analyzed. Results indicate the distribution of the transverse WRS along the direction of deck thickness is tension-compression-tension stress, which has the characteristics of the sinusoidal function. The ∆ K _eff value increases significantly considering WRS than without considering WRS during crack growth. WRS has a significant effect on fatigue life, and the fatigue life of the weld toe without considering WRS is about twice that of considering WRS. The crack depth and aspect ratio ( a ₀ / c ₀ ) of the initial crack affect the fatigue life of RTD double-sided welded joints.

Weld defects such as porosity, inclusion, burn-through, and lack of penetration are difficult to detect and control effectively in an orthotropic steel deck (OSD), which will be a fatigue crack initiation site and lead to several fatigue cracking. The crack growth behavior in defective welded joints is different from that of defect-free joints. This study investigates crack–inclusion interaction for rib-to-deck welded joints in OSDs based on numerical simulation and linear elastic fracture mechanics (LEFM). A refined finite element model of a half U-rib with cracks and inclusions was established by using the FRANC3D-ABAQUS interactive technology. The full processes of the crack–inclusion interaction from approaching and penetrating were accurately simulated. Critical parameters, including the stress intensity factor (SIF), the shape factor, the growth rate, and the growth direction were analyzed. The stiff and soft inclusions amplify and shield the SIF of cracks when the crack grows to the local area of inclusions. During the entire process of crack growth, the soft and stiff inclusion accelerate and inhibit the crack growth, respectively. The stiff inclusion will lead to asymmetric growth of the crack shape, where the portion of the crack away from the inclusions has a higher growth rate. The soft and stiff inclusions will attract and repel the direction of crack growth at the proximal point, respectively.

Residual stresses are an inevitable result of the welded fabrication process of corrugated steel webs (CSWs), resulting in structures with high and unpredictable stress fields, causing unexpected failures. The residual stress field is affected by structural parameters and the welding path of CSWs. This study proposes the welding process simulation method for CSWs with element birth and death technology. The optimization design method of heat source parameters is proposed. The feasibility of the simulation method is verified by comparing the numerical results with the experimental results of relevant literature. As a part of the study, a comparison of residual stress fields upon cooling welded CSWs with bending angles of 30, 45, and 60 degrees is presented. Thereafter, the effect of two types of single-sided welding paths and double-sided welding construction processes on residual stress distribution is discussed. Generally, the study results have shown that the equivalent residual stress near the weld reaches the maximum 344 MPa, which is very close to the Q345 steel yield strength. The size of the bending angle has no major effect on the residual stress distribution pattern, but it influences the residual stress value at the bending position. The residual stress at the bending position increases with the bending angle of CSWs. Different welding paths significantly impact the residual stress of the weld toe, and selecting a reasonable welding path can effectively reduce the residual stress value by 20 to 40 MPa.

Innovative double-sided welding is expected to improve the fatigue resistance of rib-to-deck welded joints of orthotropic steel decks (OSDs). Welding crack-like defects are the crucial issue affecting the fatigue performance of rib-to-deck double-sided welded joints. This study presents a numerical simulation of three-dimensional (3D) mixed mode fatigue crack growth behavior of rib-to-deck double-sided welded joints of OSDs. Maximum tensile stress theory and equivalent stress intensity factor (SIF) were used to simulate mixed mode fatigue cracks growth. The Paris law model was employed to predict the fatigue life. Fatigue cracks of rib-to-deck double-sided welded joints were characterized by the presence of mixed mode cracks of modes I (open), mode II (shear), and mode III (tear), which was dominated by mode I. The equivalent SIF was found to be complex at the growth stage with the maximum value at the two ends of the crack front and the minimum value at the midpoint of the crack front. The crack shape became flatter in the later phase of the crack growth. The fatigue crack surface underwent deflections during crack growth, making the final crack shape exhibiting the characteristic of a spatial curved surface. The initial crack geometry showed a significant impact on the fatigue life.

Accurate bridge deflection prediction is vital for structural health monitoring. Compared with traditional methods, it can more effectively uncover complex nonlinear laws from massive monitoring data to achieve forward‐looking prediction. Although the current mainstream models have significant advantages in modeling capabilities, they still face the bottlenecks of insufficient feature representation and high computational complexity. To solve this problem, the deflection of the bridge is accurately predicted. In this study, we innovatively construct a transformer architecture by introducing a Gaussian probability density function to establish a likelihood estimation framework. Utilizing maximum likelihood estimation for parameter optimization, it achieves probabilistic interval modeling of deflection time‐series signals. For comparison, two additional encoder–decoder models are employed: The SBIGRU model refers to the Sequence‐to‐Sequence Bidirectional Gated Recurrent Unit, whereas the ASBIGRU model extends this architecture by integrating an attention mechanism to improve bidirectional sequence representation. The experimental data are based on the multisection monitoring data of a long‐span suspension bridge in China, covering ambient temperature, vehicle load effect, and deflection signal. The analysis examined correlations among datasets with multiple cross‐sections of bridges and validated the robustness of the proposed probabilistic model. The application of the transformer model produced a notably low root mean square error of 2.588 mm alongside a high coefficient of determination of 0.9693 and markedly decreased the time required for training to 28% of that demanded by established techniques. Compared to alternative modeling techniques, transformer models deliver enhanced prediction accuracy as well as greater computational efficiency. This breakthrough provides a theoretical basis for the engineering application of transformer in the intelligent monitoring of cable load‐bearing bridges.