Fatigue assessment of laser powder bed fused aluminum alloys via non-destructive examination

• A method to analyze fatigue criticality of defects via XCT is proposed. • XCT misrepresented defect shapes, leading to underestimated fatigue lives. • Corrected defect shapes improved defect sizing and critical defect identification. • Reconstructed defects gave accurate, conservative fatigue life estimates. This study evaluated the effectiveness of x-ray computed tomography (XCT) based examinations in assessing the fatigue lives of laser powder bed fused (L-PBF) aluminum alloys. L-PBF AlSi10Mg and Scalmalloy specimens with varying defect populations were examined via XCT prior to uniaxial fatigue testing. By correlating the post-fracture surface information with XCT data, the fatigue critical defects were identified and quantified; and the efficacy of XCT in accurately capturing these defects and measuring their sizes was assessed. The results indicated that lack-of-fusions (LoFs) with thin webbed features were prone to significant loss of information in XCT scans compared to bulky shaped defects, leading to frequent misidentification of critical defects and/or misrepresentation of their actual size. Accordingly, fatigue modelling relying solely on the largest detected entities in XCT scans of L-PBF aluminum alloys, which contained large critical LoFs, resulted in severely non-conservative fatigue life predictions. It was demonstrated that a distance-based criterion can help address the limitations in XCT data by allowing for the defect morphology to be reconstructed, which gave rise to improved size estimates, and in many cases, the correct identification of the critical defect. Incorporating corrected XCT data into crack growth based models enabled accurate and moderately conservative fatigue life estimates for non-destructive structural integrity assessments.