Comparative Study and Application of Fatigue Life Prediction Methods for Rail Vehicle Spot Welding

Modeling and accurately predicting the fatigue life of a typical stainless steel spot welded rail vehicle, with thousands or even tens of thousands of welded joints, are difficult and time consuming using numerical simulation. Comparison between and optimal selection of existing methods for fatigue life prediction of spot welds are required to evaluate their fatigue life. However, evaluating the merits of the existing prediction methods based on appropriate test methods and evaluation criteria is a key issue to be addressed in spot weld fatigue resistance design research. An optimal selection method for predicting the fatigue life of spot welds was proposed. The existing methods were evaluated using two types of specimen models and test data, including tensile-shear and peel. The power function R2 was applied to quantify the dispersion of the "calculated stress-fatigue failure number" and the error factor was adopted to measure the simulation prediction accuracy. Furthermore, the mesh insensitivity of each method was tested with different mesh sizes of FEA models, and the construction time of a single weld joint was compared with the modeling efficiency. The nominal stress method, RUPP, umbrella and simplified structural stress method were compared according to the proposed method. The results show that the simplified structural stress method has the lowest dispersion, the highest prediction accuracy, and the most efficient modeling, as well as mesh insensitivity. The effectiveness and practicality of the method were verified by simulating a stainless steel body with nearly 30 000 spot welds. © 2023 Science Press. All rights reserved.