X-ray Computed Tomography in the Product Lifecycle: Metrological Approaches for Pre- and Post-Assembly Stages

Metrological X-ray computed tomography (CT) has emerged as a versatile technology for industrial inspection and metrology, combining non-destructive characterisation of internal and external features with dimensional measurement accuracy and traceability. This doctoral research, developed within the framework of the ITN MSCA European project xCTing, investigates the role of CT as a metrological enabler for industrial assemblies. The work is structured around two complementary dimensions of the product lifecycle: pre-assembly analysis, where CT is applied to the study of additively manufactured components before integration or assembly, and post-assembly analysis, where CT is used to recover usable information from assemblies. In the pre-assembly stream, CT was employed to study the deviations and internal defects in powder bed fusion components. The first study focused on laser-based powder bed fusion of metals and developed a CT-based methodology to quantify the effects of interrupting and resuming a building process, aiming at providing accurate device embeddings. The method defined measurands and measurement models to capture vertical shrinkage at the stop plane, layer-wise recovery to steady state, inclination effects, and porosity evolution. The results established a transferable framework to investigate interruption events aided by CT-based methods. The second study addressed laser-based powder bed fusion of polymers, applying a full-factorial design to evaluate positional and thermal effects on dimensional accuracy and porosity of individual components to be assembled. By coupling CT with in-situ thermography, it was shown that dimensional deviations -which can hinder the accuracy of assembly operations- are primarily governed by positional orientation relative to the scanning strategy, whereas porosity is strongly correlated with thermal gradients across the build platform. These findings highlight the importance of position-aware nesting and thermal management for reliable pre-assembly quality. In the post-assembly stream, CT was applied to the inspection of multi-material assemblies where metal-induced image artifacts compromise data quality and measurement accuracy. In a first study, a multi-positional approach was developed to mitigate metal artifacts in multi-material CT scans by fusing two or more CT reconstructions in different angular positions. This study proposes a method to select the best positions to be fused. In addition to this study, in order to address the intrinsic limitations of rule-based fusion, a second study introduced deep learning architectures for data fusion. The results confirmed that data-driven fusion surpasses traditional methods, providing improved recovery of fine structures and defect visibility even under challenging artifact conditions. Taken together, the research establishes CT as a multipurpose tool capable of supporting both manufacturing process optimisation and assembly inspection. The methodological contributions of the thesis include measurement frameworks with clearly defined measurands and influence quantities for additive manufacturing processes, strategies for efficient orientation planning in multi-positional scanning, and data-driven fusion methods for artifact mitigation. The dual perspective pre- and post-assembly provides a coherent framework in which CT contributes not only to inspection but also to the broader integration of metrology into advanced manufacturing workflows.