This study integrates the fatigue test and numerical prediction to derive a comprehensive probability-stress-life (P-S-N) curve for rib-to-deck (RD) welded joints in orthotropic steel decks. Fatigue tests of RD joints are conducted to measure fatigue strength and crack growth data. Based on the test, a probabilistic fatigue crack growth (PFCG) model is established to predict the distribution of fatigue life under various stress ranges. Two machine learning tools are adopted to assist the PFCG model-based prediction, i.e., the Gaussian process regression (GPR) and dynamic Bayesian network (DBN). The GPR is used to train a surrogate model solving stress intensity factors for the PFCG prediction, using 2,000 samples generated from finite element (FE) analyses. The trained model is then validated by a new dataset of 100 FE samples. An adapted DBN model is proposed to update the PFCG model with the fatigue crack growth data measured from ten specimens. According to the result, the application of GPR can reduce the solution cost of the PFCG prediction by approximately 1,875 times. Compared with the prior PFCG model, the updated posterior model shows an improved agreement with the test data, i.e., the maximum difference in fatigue strength between model prediction and test data decreases from 12% to 3%. Based on the posterior PFCG model, the P-S-N curve of RD joints is statistically derived using sufficient numerical samples.

Rib-to-deck (RD) welded joints in orthotropic steel decks (OSDs) of bridges demonstrates two major fatigue failure models, including the toe-to-deck (TTD) cracking and root-to-deck (RTD) cracking. Generally, the sole failure model is employed in the fatigue assessment of RD joints, causing a hot dispute on the dominant failure model. In this paper, the fatigue crack growth (FCG) in RD joints has been evaluated considering uncertainties and mixed failure models. A probabilistic fatigue crack growth (PFCG) model is at first established for the RD joint, in which two crack-like initial flaws are assumed at the weld toe and root of the RD joint. After that, the gaussian process regression is used to assist and boost the PFCG simulation. Then, the PFCG model is implemented on a typical OSD with the random traffic model. Finally, the result of the PFCG model is discussed in detail, including the failure model, fatigue reliability and life prediction, and crack size evolution. It is revealed that both the TTD and RTD cracking models have a notable contribution to fatigue failure and could not be ignored. More crucial, a remarkable reduction can be observed in the fatigue reliability of RD joints when considering mixed failure models. This study not only highlights the influence of mixed failure models on the fatigue performance of welded joints, but also provide an insight into the application of novel machine learning tools in solving the traditional structural issue.

This paper has investigated the shear-slip behaviour of an innovative prefabricated composite shear stud (PCSS) connector and its application in the prefabricated steel–concrete composite bridges. A series of push-out tests are carried out on a total of 12 specimens, including 6 PCSS specimens and 6 conventional shear stud (CSS) specimens. Further comparison has been carried out between the test result and the data available from the literature. Based on the test, a high-resolution finite element (FE) analysis has been performed to reveal the load transfer mechanism of the PCSS connector at the component-level. After that, an advanced FE model has been established and validated by a full-scale test of the prefabricated composite bridge using the PCSS. With the FE model, the load-slip behaviour and slip distribution are investigated in details. The result highlights the enhanced shear capacity and ductility of the PCSS specimens compared with the CSS specimens, as well as the feasibility of PCSS connectors in composite bridges. Meanwhile, it is further revealed by the detailed investigation that the enhancement could be attributed to the lateral constraint on the concrete by the vertical steel plate in the PCSS. Besides, it is also found that the load-slip behaviour of composite bridges using the PCSS is influenced by the cracking at the seam between deck blocks. Consequently, abrupt changes can be found in the load-slip curve once the cracking occurs, which differs from the traditional composite bridges.