Not all scans are equal: X-ray tomography image quality evaluation. (March 2020)

• Record Type: Journal Article
• Authors: du Plessis, A.; Tshibalanganda, M.; le Roux, S.G.
• Is Part Of: Materials today communications. Volume 22(2020)
• Publication Date: 2020-03
• Subjects: X-ray tomography -- microCT -- Materials science -- Quality control -- Image quality; Materials science -- Periodicals; 620.11
• DOI: 10.1016/j.mtcomm.2019.100792
• Languages: English
• ISSNs: 2352-4928

Abstract

Graphical abstract: Highlights: Image quality in X-ray tomography can vary significantly between scans. Image quality directly influences quantification efforts. A simple set of image quality metrics is proposed. These are shown to be sensitive to most typical artifacts. Image quality evaluation allows improved confidence in the method. Abstract: X-ray microtomography is widely used in materials science and engineering applications for imaging and analysis of material structure and morphology. For this purpose, and especially in the case of routine analysis tasks for industrial materials applications, confidence in obtained measurement results are crucial. Despite great progress in this field over the last 10 years, with many high-quality commercial systems now available, the lack of a simple and widely-used image quality metric that can capture all important aspects of the quality of a microCT scan, continues to hinder wider acceptance of the technology. Various errors can occur during the microCT scan process, which can potentially mask the presence of pores, or affect the volumetric measurements of interest. In this work we demonstrate a simplified image quality metric which can easily be implemented. We show how this new image quality metric is sensitive to all typical microCT scan errors and artifacts, which makes it a valuable tool for defining a required minimum image quality for an analysis. The object used is a 10 mm cube of titanium alloy (Ti6Al4V) produced by laser powder bed fusion additive manufacturing. This type of coupon sample is useful for analysis of the additive manufacturing process, but it is critical that small pores are seen with good contrast. Identical porosity analysis workflows are applied to scans with different image qualities, which demonstrates the importance of image quality for reproducible analyses of this sample type. The results have implications in defining quality values for all forms of materials analysis using the technique. This work can further lead the way to incorporating microCT into future fully automated and standardized analysis workflows for quality control, when image quality meets a specified minimum criterion.