Fatigue life prediction methodology for welded joints considering defect effects

The paper focuses on defect detection, quantitative characterization, and fatigue crack growth life prediction considering defect effects for 1Cr11Ni2W2MoV argon arc-welded joints used in engine combustion chamber casings. First, X-CT technology was used to detect microscopic defects in different weld zones. Through three-dimensional defect reconstruction, the defect sizes and distribution patterns in various weld regions were quantitatively characterized. The maximum characteristic defect sizes were predicted using extreme value statistical analysis. Subsequently, based on fatigue crack growth test data analysis of welded joint heat-affected zones, a NASGRO equation describing the complete crack growth process in welded joints was established. Based on these findings, the model was further modified by incorporating variations in threshold values and closure parameters during short crack growth. By combining defect morphology, size characteristics, and extreme value statistical analysis results, defects were equivalently treated as semi-elliptical cracks to calculate the fatigue life of joints under different stress levels. Comparative analysis with experimental results demonstrated that the predicted lives from the modified model all fell within a factor of two scatter band, indicating its practical significance for fatigue life assessment of welded joints considering defect effects. Furthermore, the modified model was applied to extend the conventional stress-defect model (K-T diagram), establishing a stress-defect-life diagram that enables fatigue life evaluation of welded joints with various initial defect sizes under different stress conditions.