Use of X-ray Tomography to Guide Data Acquisition in FA Workflows

Harris, William M

For failure analysis workflows and considering imaging techniques alone, a selection of optical and electron systems with associated analytical capabilities exist, as well as 2D or 3D X-ray imaging. Of these methods, X-ray maintains the notable strength of nondestructive operation, while being able to provide a description of interior, subsurface features.

As modern industrial X-ray tomography evolved from medical CT, initial resolution was quite low, limited primarily by X-ray source and detector technologies. Over the past 2 decades these have evolved substantially, and the capabilities of X-ray CT, microCT, and X-ray microscopy have improved accordingly. Nonetheless, for some types of materials structures (and correspondingly, failures) the X-ray resolution can be lacking – able to identify gross defects or features critical to the failure, but not a thorough or conclusive description. As a result, X-ray imaging is found typically as the first step of a workflow, providing a non-invasive overview before extracting targeted regions for higher resolution work.

This talk will cover a brief survey of the X-ray imaging tools at our disposal as well as their particular strengths (dimensional measurement, resolution requirements, sample size considerations, etc.). This will include ‘Resolution at a Distance’ technology, a unique capability achieving high resolution imaging even within ‘large’ parts. Examples will be drawn from metal alloy systems, additive manufactured parts, composite materials, energy devices, and microelectronics to demonstrate the role of X-ray to observe features including: internal porosity, cracks, inclusion/defect particles, delamination, and surface texture. Emphasis will be placed on the ability to navigate between length scales both within X-ray but also looking to higher resolution, albeit destructive, imaging techniques downstream. The talk will also briefly touch on the idea of 4D in situ imaging as a means to pre-emptively predict failure modes.