Assessing witness coupon-specimen equivalency for volumetric defects: An investigation on location dependency in additive manufacturing

Witness coupons, built alongside additively manufactured parts, can potentially be used to indicate build quality across different part locations. However, the applicability of measurements from witness coupons to actual parts/components is limited due to the difference in size and shape between them, which may result in potentially different thermal histories. Therefore, this study aims to evaluate the equivalency of volumetric defects in witness coupons and fatigue specimens using X-ray computed tomography and extreme value statistics (EVS). If such equivalency is established, coupons can be used to infer quality of the adjacent parts/components. For establishing the equivalency, the largest extreme value distributions were compared between witness coupon-specimen pairs across nine different part locations, and the results showed that the size distributions of largest defects did not vary significantly for most pairs. Subsequently, EVS estimates of 50th percentile from witness coupons at different part locations were compared to critical defect sizes obtained from fatigue specimens in the corresponding locations, and were found to be conservative.