Comparison of defect detectability between Computed Tomography inspection and CT simulation using a calibrated defect phantom

Abstract

The usage of metal Additive Manufactured (AM) components is growing in aerospace industry since the establishment of quality standards and due to the current maturity of the manufacturing systems, processing route, and inspection methodology. Moreover, possibilities of designing complex shapes by using 3D-printers enables design engineers to build lightweight structures and/or increase part functionality. However, the freedom in design is often a challenge for non-destructive testing (NDT), especially in parts with limited access, non-flat surfaces, etc. Therefore, there are few NDT methods which can be applied on such complex 3D geometries and capable of inspecting the whole part volume. Computed Tomography (CT) and Digital Xray methods are the most relevant ones offering rich information of inner defectology and the outer geometrical metrology. Simulation tools regarding manufacturing process and mechanical behavior are already considered as part of the definition phase of AM components and utilized as inputs in the design loop. However, inspectionability issues are mainly considered during the quality assurance phase. Therefore, inspection problems related with the detectability of defects above the part allowable can appear once the part have been designed, validated, and manufactured. On the other hand, tools that perform numerical simulations of X-ray imaging like SimCT could be a valuable source of additional information [1]. Firstly, for the users of XRay imaging devices in order to set up best inspection parameters and system configuration [2, 3]. Secondly, for producing CT simulations which can evaluate the inspectionability of a part, avoiding non-inspectionable regions limited by the inspection method. In this way, this information can be introduced for validating the part design for instance at PDR (Preliminary Design Review) and later at CDR (Critical Design Review) levels. In this work, an aluminum Image Quality Indicator (IQI) with calibrated defects has been analyzed with both real and simulated CT scans using different physical resolutions. In this way, a comprensive analysis for determining the limits of defect detectability by comparing both simultaion and test results has been developed. Defects of 100, 200 and 500µm diameter have been evaluated, corresponding to aerospace allowable for pores in AM hardware, depending on part criticality.