Mapping the spatial performance variability of an X-ray computed tomography inspection

Abstract Cone-beam X-ray Computed Tomography (XCT) is an advanced technique of Non-Destructive Testing (NDT) (and metrology) that is increasingly being used to assure the quality of safety-critical components, driven to a large extent by the need to inspect components of very high geometric complexity produced by Additive Manufacturing (AM) routes. It is therefore essential to understand the limitations, and hence reliability, of a given inspection. A wide range of factors are known to influence the creation of an XCT volume, but the resultant spatially varying inspection performance is not well understood, and the lack of a practical means to assess this constrains the application of XCT. This paper provides an experimental illustration of the problem using a reconfigurable artefact and describes an algorithm to map out inspection performance that yields results closely matching the experimental output. This new model-assisted inspection qualification capability should enable the more widespread usage of XCT, and hence advanced manufacturing techniques that necessitate the use of this inspection technique, for safety-critical components.