This study aims to accurately assess the long-term service performance of welded high-speed train carbodies by predicting their crack-growth life.
First, the Christopher-James-Patterson (CJP) crack tip field model was applied to analyze the crack growth rate of the welded aluminum alloy in a high-speed train carbody. Next, a high-precision surrogate model, developed by the research team, correlates multiaxial loads and stresses to transform load-time histories into dynamic structural stresses at critical points. These dynamic stresses were then corrected using the mesh-insensitive structural stress method to improve the accuracy of crack growth life predictions.
The results demonstrate that the proposed fatigue residual life prediction framework, integrating structural stress analysis with the CJP model, is mesh-insensitive and eliminates subjectivity in selecting the primary S-N curve method. This framework improves the reliability of predictions and provides valuable insights for determining overhaul cycles and implementing intelligent maintenance strategies for large welded structures.
This study presents a novel approach that combines the CJP model with structural stress analysis to predict the fatigue crack growth life of high-speed train carbodies. The proposed method overcomes limitations of traditional methods and offers more accurate and robust predictions, contributing to the field of fatigue durability assessment of welded structures.
