Partner event: THE 9TH CIA-CT CONFERENCE ON “APPLICATIONS OF COMPUTED TOMOGRAPHY IN INDUSTRY”

Hans Nørgaard, Head of Department of Civil and Mechanical Engineering, DTU Construct

Danilo Quagliotti, Senior Researcher, DTU Construct

Wim Dewulf, KU Leuven, Belgium

X-ray computed tomography (XCT) is increasingly being used for evaluating quality and conformance of complex products, including assemblies and additively manufactured parts. Following a brief introduction of the principles, this presentation provides insights into the current metrological performance of XCT. The error sources influencing XCT measurement results are discussed, along with related qualification, calibration and optimization procedures. Moreover, progress on performance verification testing and on the determination of task-specific measurement uncertainty is covered. Results of interlaboratory comparisons are summarized and performance in various dimensional measurement fields is illustrated.

Anne-Françoise Obaton, LNE, France

To deal with the complexity of shape that additively manufacturing (AM) enables to build, to guarantee the integrity and geometry of the AM parts, X-ray computed tomography (XCT) has no equal. What better analysis result than a three-dimensional image of the outside and inside of a part to check its conformity? However, to obtain a reliable image, the image quality is of importance and needs to be quantified. Furthermore, to help the XCT service providers to perform good quality XCT images, standards need to be developed.

The talk will address image quality enabling to perform reliable integrity inspection and dimensional measurements as well as standardization to support the development of XCT.

Marko Katić, University of Zagreb, Croatia

Use of X-ray Computed Tomography (CT) in industrial applications progressed from simple non-destructive testing of components to full-scale metrology. This progress relied on numerous improvements in design of CT scanners, as well as on improvements in methods applied to CT scans. Accurate dimensional measurement results rely on multiple assumptions when using CT as a metrology tool, some of which have already been successfully researched and implemented. SensMonCT project addresses one of the major issues still remaining in this field: accurate and objective determination of the minimum detectable detail size in a CT scan.

Valentina Aloisi, Zeiss, Germany

X-Ray Computed Tomography (CT) is increasingly being used in the industry, for performing a wide variety of tasks. Typical applications include qualitative and quantitative analysis. Examples range from classical NDT inspections (e.g. crack identification, defect analysis, porosity analysis, etc.) to dimensional quality control. One of the key aspects that may affect the result of the analysis is the CT scan data quality. This presentation will discuss new innovative solutions in the field of X-Ray Computed Tomography to enhance the scan data quality as well as AI solutions to process challenging CT scans.

Thomas Kjærgaard, Lead Project Manage, Grundfos, Denmark

Can we bring CT-scanning from the development/analysis into manufacturing? Grundfos have tested it’s use in connection with Additive manufacturing.

Sven Gondrom-Linke, Volume Graphics GmbH, Germany

Considering the current surge in artificial intelligence, machine learning and deep learning in particular, the question arises as to their impact on industrial applications.

How can the recent developments enhance the non-destructive evaluation of industrial components? And what requirements must a deep learning model meet to reliably detect, classify, and segment indications in industrial computed tomography scans? How can AI additionally help to recognize correlations between production and environmental parameters and the achieved quality of the products in order to improve them?

This presentation provides a concise overview of the success story of deep learning, extracts insights from other machine vision domains, and offers practical guidance on the appropriate utilization of deep learning for in-line inspection of manufactured components.

Erik M. Lauridsen, CEO, Xnovo Technology ApS, Denmark

X-ray imaging based on absorption contrast has evolved into being the de-facto standard when it comes to explore the internal structures of ofterwise opague components. However, X-ray imaging can do so much more than just providing information about density differences in the object under investigation. At Xnovo Technology we specializes in providing additional material insights by enabling alternative image modalities on standard X-ray microCT platforms.

Here we demonstrate examples of 3D imaging for various materials characterization needs combining absorption, diffraction and scattering contrasts on lab-scale X-ray microCT platforms.

(Illustration: Example of crystallographic imaging of minerals (left) and directional dark-field imaging of fibers (right)).

Steffen Sloth, Ph.D. Student, DTU Physics and Exruptive A/S, Denmark

The next-generation X-ray CT scanners for aviation security utilize energy-dispersive Photon Counting Detectors (PCDs) to enhance the material classification capabilities for improved threat detection. To achieve a sufficient field-of-view multiple PCDs can be stacked together into an array of detector modules. This work investigates the sensitivity of the geometry errors of the position and orientation of an array of PCD line modules. A new X-ray CT simulation tool is used to generate CT projections with known geometry misalignments. From the CT reconstructions of the simulated projections the geometry errors sensitivity factors are evaluated from the reconstructed volume.

Anders Bjorholm Dahl, DTU Compute, Denmark

To be announced

Carsten Gundlach, DTU Physics, Denmark

To be announced