Assessing the influence of CT acquisition parameters on flaw detectability through simulation

Abstract

X-Ray Computed Tomography (XCT) is a unique tool to fully visualize and understand the nature and size of flaws in industrial parts, with a growing application in different fields such as aeronautics and more recently metal additive manufacturing. The inevitable questions underlying any XCT inspection concern the detectability limit of the measure. What size of defect will be detected with my current configuration? How can I optimize my acquisition material or parameters to improve the detectability limit? Naturally, the known characteristics of the XCT systems (detector pixel size, X-ray tube voltage and focal spot size, magnification) give a first answer to these questions, at least in terms of spatial resolution, but it is more difficult to estimate a priori the visibility of a flaw in terms of contrast. The cost of XCT inspection, the difficulty to design specimen with narrow internal defects and the influence of the geometry of the part on the XCT image quality make experimental analysis of the detectability limits difficult. The simulation brings therefore a promising alternative, provided that it gives a thorough representation of a real inspection.