Abstract:Abstract: The U-rib-to-deck (RD) joint in steel bridge decks is prone to fatigue cracking under vehicle loads. It is of great practical value to predict the fatigue life of this type of joints by using appropriate theory. In order to further study the fatigue performance of RD joint, the RD joint was subjected to central loading by simulating wheel loads, and the fatigue life of RD joint under corresponding hot spot stress range was obtained by conducting fatigue tests. The 3D solid finite element model of RD joint in steel bridge decks was established by using ABAQUS, and the stress distribution of the RD joint under simulated wheel load was further analyzed. It is found that the test value is basically consistent with the simulation value. Based on the effective numerical model, the fatigue life of RD joint was compared and analyzed by using the hot spot stress method and the critical distance theory respectively. The results show that (1) the hot spot stress near the weld toe on the deck’s lower surface in RD joint is the largest when the central loading is applied, and fatigue crackings are easy to occur here; (2) Under the same finite element mesh, the hot spot stress results obtained by two-point extrapolation and three-point extrapolation are relatively close, while the characteristic stress obtained by point method using the critical distance theory is smaller than that obtained by line method. The three hot spot stress S-N curves recommended by IIW are used to evaluate the fatigue life of RD joints. The result shows that the evaluation results obtained by FAT 90 S-N curve are too conservative, while the prediction results of FAT 125 S-N curve are relatively close to the measured value. Compared with the hot spot stress method, the critical distance theory can get more accurate prediction results, in which the prediction errors obtained by the point method and the line method are 17% and 32%, respectively. In general, both the hot spot stress method and the critical distance theory can get relatively conservative fatigue life prediction results. Therefore, the two methods are feasible and safe for fatigue behavior evaluation of typical welded joints in steel bridge decks. However, the point method has higher prediction accuracy and is also more convenient in practical application.