Damage tolerant design of cast components based on defects detected by 3D X-ray computed tomography

This paper presents the application of 3D X-ray computed tomography (XCT) in deterministic fatigue strength analysis of relatively large cast components and also probabilistic fatigue design of heavy-section castings. To this end, more than 100 fatigue specimens were cut from a rejected wind turbine hub casting. XCT was used to find the 3D map of defects in selected fatigue specimens. All fatigue specimens were axially tested at R = −1 at room temperature and fatigue life scatter was established. Fractography analysis of broken specimens was performed. For CT scanned specimens, the defect sizes and positions obtained by XCT, in conjunction with the (nearly homogeneous) stress distribution of the specimens were used to predict the fatigue life and also the position of the fatigue crack initiation site. The predicted fatigue life and crack initiation position for CT scanned specimens were compared with experimental results. The obtained defect distribution by XCT was used in a random defect analysis to establish the scatter of fatigue life for the analyzed specimens. The predicted fatigue life scatter was compared with experimental results. The comparison showed that random defect analysis can successfully predict the scatter of fatigue life and may be used to find the probability of failure of cast components, based on the given defect distribution.