X-ray Computed Tomography Data Research Articles

Image-based finite element modeling of air flow and thermal transport in Al-fiber-sintered porous materials

In this study, the pressure drop and heat transfer in a heat-transfer tube filled with a sintered porous medium comprising Al fibers were investigated using computational fluid dynamics (CFD) simulations. We reconstructed the sintered fibrous porous structure and generated a computational mesh using X-ray computed tomography data, and the simulated pressure drop and heat transfer agreed well with those reported by Enoki et al. [2021]. The differences in both the form coefficient and the permeability between two different samples can be explained from the differences in both the specific solid surface area and the porosity. Further, the CFD simulations indicated that thermal conduction of the Al solid phase enhanced the heat exchange between the air and Al fibers. To examine the effect of solid-phase thermal conduction, we generated a model of a heat-transfer tube having an ideal wire mesh structure, in which we introduced the interfacial thermal conductivity between Al fibers and the inner tube wall as additional factors. We also performed CFD simulations with four different shell-region thicknesses and found that even a very narrow gap of 5 to 10 μm heavily affected the heat-exchange performance because of the low thermal conductivity of the shell region.

The use of computed tomography and X-ray fluorescence analysis in the research of printed book from the seventeenth century: book binding, tomographic reading of the text, dendrochronological dating, pigments analysis

This paper presents the use of X-ray computed tomography and X-ray fluorescence in the analysis and expert research of the seventeenth century printed book "Eukhologīon albo Molitoslov, ili Trebnik" from Kiev. The main purpose of the survey was to confirm whether the book binding is original or whether it is a rebinding, and whether there are any fragments of the hidden older texts. Commonly used radiography is usually not able to provide sufficient information for these purposes. On the other hand, computed tomography allows a detailed and three-dimensional documentation of the bookbinding technology and the structure of the materials used, including the wooden boards. It will be presented that all elements of the weave are clearly visible, making it possible to show that there are no internal defects in the stitching and materials. It has also been convincingly shown that there are no fragments or layers of older texts in the binding, so no further invasive intervention will be necessary regarding this aspect. The paper also demonstrates the possibility of reading the text in a closed book utilising X-ray computed tomography data; this option may be advantageous for massively damaged manuscripts. It will also be shown, that thanks to detailed tomographic imaging of the wood structure of the boards, a dendrochronological survey can be successfully carried out without invasive intervention into their outer layers. From the CT data it was also found that the pigments of the letters have significantly different densities. Therefore, as part of the survey, elemental analysis of the writing was also carried out using a portable X-ray fluorescence spectrometer to confirm and clarify this finding.

A methodology for the 3D characterization of surfaces using X-ray computed tomography: Application to additively manufactured parts

Many studies highlight the significance of three-dimensional surface topography characterization in assessing its effect on the mechanical or functional properties of materials. This is especially obvious for parts made by additive manufacturing (AM), known for their complex shape and surface topographies. However, a vast majority of 3D characterizations have constraints regarding the macroscopic geometry of the parts they can probe. At the microscale, they are also unable to account for hidden surface features, e.g. notches hidden by unmelted powder particles. Even with the use of X-ray Computed Tomography (XCT) - a tool with the potential to circumvent these issues -data is often reduced to 2D or 2.5D formats for easier analysis, but this leads to a loss of information. This underscores the need for XCT data post-treatment tools to perform thorough 3D surface characterizations. Herein, we introduce a methodology for local roughness and curvature characterization of surfaces of complex shapes using XCT. This method has been designed to be user-friendly, especially for those without extensive data analysis expertise. It provides a comprehensive 3D characterization and efficiently tackles the issues caused by hidden features. After a detailed description of our methodology, we give a first illustrative example based on architected structures fabricated by Electron Powder Bed Fusion (E-PBF). By integrating roughness and curvature metrics, we also derive a parameter indicative of the stress concentrations caused by surface irregularities.

Spatial Resolution-Optimized Artifact Reduction by Combining Dual Energy X-Ray Computed Tomography Data using a Frequency Split

Spatial Resolution-Optimized Artifact Reduction by Combining Dual Energy X-Ray Computed Tomography Data using a Frequency Split

Neural network approach for shape-based euhedral pyrite identification in X-ray CT data with adversarial unsupervised domain adaptation

We explore an attenuation and shape-based identification of euhedral pyrites in high-resolution X-ray Computed Tomography (XCT) data using deep neural networks. To deal with the scarcity of annotated data we generate a complementary training set of synthetic images. To investigate and address the domain gap between the synthetic and XCT data, several deep learning models, with and without domain adaption, are trained and compared. We find that a model trained on a small set of human annotations, while displaying over-fitting, can rival the human annotators. The unsupervised domain adaptation approaches are successful in bridging the domain gap, which significantly improves their performance. A domain-adapted model, trained on a dataset that fuses synthetic and real data, is the overall best-performing model. This highlights the possibility of using synthetic datasets for the application of deep learning in mineralogy.

Prediction Models of Mechanical Properties of Jute/PLA Composite Based on X-ray Computed Tomography.

The tensile strength and modulus of elasticity of a jute/polylactic acid (PLA) composite were found to vary nonlinearly with the loading angle of the specimen through the tensile test. The variation in these properties was related to the fiber orientation distribution (FOD) and fiber length distribution (FLD). In order to study the effects of the FOD and FLD of short fibers on the mechanical properties and to better predict the mechanical properties of short-fiber composites, the true distribution of short fibers in the composite was accurately obtained using X-ray computed tomography (XCT), in which about 70% of the jute fibers were less than 300 μm in length and the fibers were mainly distributed along the direction of mold flow. The probability density functions of the FOD and FLD were obtained by further analyzing the XCT data. Strength and elastic modulus prediction models applicable to short-fiber-reinforced polymer (SFRP) composites were created by modifying the laminate theory and the rule of mixtures using the probability density functions of the FOD and FLD. The experimental measurements were in good agreement with the model predictions.

Mesoscale 3D model of irradiated concrete

The concrete biological shield in light-water reactors is exposed to neutron and gamma irradiation, which deteriorates the concrete’s mechanical properties in the long term. To assess the irradiation-induced damage, predictive mechanical models are developed and used in parallel with the characterization of irradiated concrete samples. Realistic 3D simulation domains can drastically improve a model’s prediction. In this work, we utilized x-ray computed tomography (XCT) data of a concrete specimen to reconstruct its 3D microstructure. The XCT data shows low contrast between the concrete’s aggregates and cement paste, resulting in poor image segmentation when using traditional unsupervised techniques. To address this issue, we developed and trained a 2.5D U-Net model on only 24 pre-labeled XCT layers to segment 651 layers of the XCT data. The overall F1-score of the model is approximately 96%. Then, we created a 3D finite element (FE) mesh based on the stack of segmented images. The FE model contains radiation-induced expansion, damage, and creep. The constitutive equations are adapted to each phase (aggregates and cement paste). We simulated the effects of neutron irradiation in the concrete specimen as well as the specimen’s mechanical response to uniaxial compression. Finally, model validation was performed using experimental data on similar concrete specimens in the literature.

Resonant ultrasound spectroscopy applications: Elastic moduli computation with x-ray computed tomography input for irregularly shaped objects.

Resonant ultrasound spectroscopy is a technique that uses a combination of experimentally measured resonant frequencies and physics-based computation of these frequencies to determine the entire set of single crystal elastic moduli of the material. Computation of the resonances is most often accomplished using the Rayleigh-Ritz energy minimization technique, and a basis function that requires sample with canonical geometry, such as a cylinder or a rectangular parallelepiped. Any deviation from canonical geometry can have a significant impact on the calculated resonance frequencies and the inverted elastic moduli. To overcome this limitation, this paper describes an approach that uses x-ray computed tomography data to generate accurate solid part model of components with complex geometry. The part model is then imported into an off-the-shelf finite element method (FEM) software to perform the forward problem. The FEM was combined with surrogate modeling for computation of resonance frequencies of both canonical and non-canonical samples, and ultimately, the inversion of elastic moduli.

Multi-Scale Characterization of Porosity and Cracks in Silicon Carbide Cladding after Transient Reactor Test Facility Irradiation

Silicon carbide (SiC) ceramic matrix composite (CMC) cladding is currently being pursued as one of the leading candidates for accident-tolerant fuel (ATF) cladding for light water reactor applications. The morphology of fabrication defects, including the size and shape of voids, is one of the key challenges that impacts cladding performance and guarantees reactor safety. Therefore, quantification of defects’ size, location, distribution, and leak paths is critical to determining SiC CMC in-core performance. This research aims to provide quantitative insight into the defect’s distribution under multi-scale characterization at different length scales before and after different Transient Reactor Test Facility (TREAT) irradiation tests. A non-destructive multi-scale evaluation of irradiated SiC will help to assess critical microstructural defects from production and/or experimental testing to better understand and predict overall cladding performance. X-ray computed tomography (XCT), a non-destructive, data-rich characterization technique, is combined with lower length scale electronic microscopic characterization, which provides microscale morphology and structural characterization. This paper discusses a fully automatic workflow to detect and analyze SiC-SiC defects using image processing techniques on 3D X-ray images. Following the XCT data analysis, advanced characterizations from focused ion beam (FIB) and transmission electron microscopy (TEM) were conducted to verify the findings from the XCT data, especially quantitative results from local nano-scale TEM 3D tomography data, which were utilized to complement the 3D XCT results. In this work, three SiC samples (two irradiated and one unirradiated) provided by General Atomics are investigated. The irradiated samples were irradiated in a way that was expected to induce cracking, and indeed, the automated workflow developed in this work was able to successfully identify and characterize the defects formation in the irradiated samples while detecting no observed cracking in the unirradiated sample. These results demonstrate the value of automated XCT tools to better understand the damage and damage propagation in SiC-SiC structures for nuclear applications.

(Keynote) Evidence of the Interplay of Temperature on Local and Global Battery Phenomena

Temperature is an important parameter in electrochemical processes and battery operation; for instance, open circuit voltage and resistance values of cells are temperature sensitive. Several recent studies have shown temperature within battery packs are not uniform. This non-uniformity of temperature produces localized thermal gradients within cells that can dictate cell construction, thermal management scheme, cell bussing, etc. Our research group has been working to understand how thermal gradients within battery cells effect material properties, electrochemical performance and stability. We have demonstrated through a range of electrochemical and materials characterization techniques, how dynamic external conditions, specifically thermal gradients and transients, can induce differing degradation modes of the positive and negative electrodes. The directionality and magnitude of thermal gradient dictate the degradation mode, whether loss of lithium inventory, loss of active material, etc.We have also observed strong electrochemical-mechanical coupling the result of low temperature cycling, where lithium plating on the negative electrode causes a physical deformation and collapse of the cylindrical cell jellyroll. In this talk, we will present in-operando optical microscopy and X-ray computed tomography data to show how dynamic electrical and external thermal conditions can irreversibly effect the chemistry, structure and subsequent safety of lithium-ion battery materials, components and cells. Lastly, the importance of testing materials and cells outside of isothermal laboratory conditions in favor of operating under dynamic thermal conditions, to include austere and stressful operating conditions will be discussed.

A Complete Strategy to Achieve High Precision Automatic Segmentation of Challenging Experimental X‐Ray Computed Tomography Data Using Low‐Resemblance Synthetic Training Data

It is shown that preconditioning of experimental X‐ray computed tomography (XCT) data is critical to achieve high‐precision segmentation scores. The challenging experimental XCT datasets and deep convolutional neural networks (DCNNs) are used that are trained with low‐resemblance synthetic XCT data. The material used is a 6‐phase Al–Si metal matrix composite‐reinforced with ceramic fibers and particles. To achieve generalization, in our past studies, specific data augmentation techniques were proposed for the synthetic XCT training data. In addition, two toolsets are devised: (1) special 3D DCNN architecture (3D Triple_UNet), slicing the experimental XCT data from multiple views (MultiView Forwarding), the i.S.Sy.Da.T.A. iterative segmentation algorithm, and (2) nonlocal means (NLM) conditioning (filtering) for the experimental XCT data. This results in good segmentation Dice scores across all phases compared to more standard approaches (i.e., standard UNet architecture, single view slicing, standard single training, and NLM conditioning). Herein, the NLM filter is replaced with the deep conditioning framework BAM SynthCOND introduced in a previous publication, which can be trained with synthetic XCT data. This leads to a significant segmentation precision increase for all phases. The proposed methods are potentially applicable to other materials and imaging techniques.

Semantic segmentation of defects based on DCNN and its application on fatigue lifetime prediction for SLM Ti-6Al-4V alloy.

Inevitable defects have an important impact on the fatigue lifetime of additive manufacturing materials. Therefore, it is critical to thoroughly characterize the characteristics of the defects, which requires effective semantic segmentation of the imaged defects. In this paper, a defect dataset for SLM Ti-6Al-4V alloy was obtained by synchrotron radiation computed tomography. Then a semantic segmentation method was developed based on the DeepLabV3 + network to automatically extract defects. Cropping and undersampling were introduced in the dataset pre-processing. A weighted scheme based on the ratio between the number of defect and matrix pixels was applied in the classification layer, and morphological operations were employed in image post-processing to improve the accuracy of identifying small-target defect. Finally, the above method was applied to segment the X-ray computed tomography data for two batches of SLM Ti-6Al-4V materials, and the defect segmentation results were used to predict the fatigue lifetime. The semantic segmentation method performs well with a pixel recognition accuracy of 98.2% for the test dataset, and the error in the predicted fatigue lifetime lies within the scatter band of ±2.2. This article is part of the theme issue 'Physics-informed machine learning and its structural integrity applications (Part 2)'.

Cultural Heritage Image Analysis and Support Modeling Using X-ray Computed Tomography: Focused on the Byeongsan Mask ‘Eul’ of the Hahoe Masks and Byeongsan Masks of Andong

Before conservation treatment, the Byeongsan Mask ‘Eul’ was damaged, and the face was separated on both sides and attached with stiffeners. For conservation treatment, it was necessary to remove the existing stiffeners and adhese the faces together. At this time, a support was needed to fix both faces that were separated during conservation process and to structurally stabilize the adhesion area where pressure could be concentrated even after adhesion. In order to manufacture new support, the shape analysis was first performed using the X-ray computed tomography data, and then the re-adhesion position was set. Next, modeling was conducted considering the function of the support and aesthetic conditions using the digital haptic modeling system. The completed model was printed by 3D printing using transparent material. The support produced in this study is significant in that it played a major role in the conservation process, such as being used not only to hold and store the masks but also to fix them during the conservation process. In addition, by producing support using digital methods and data, support optimized for the shape of artifact could be produced with minimal contact with artifact. This study presented a method for conserving and displaying artifacts using digital technology, and is expected to be expanded and applied to various cultural heritages.

Scalable in situ non-destructive evaluation of additively manufactured components using process monitoring, sensor fusion, and machine learning

Laser Powder Bed Fusion (L-PBF) Additive Manufacturing (AM) is among the metal 3D printing technologies most broadly adopted by the manufacturing industry. However, the current industry qualification paradigm for critical-application L-PBF parts relies heavily on expensive non-destructive inspection techniques, which significantly limits the use-cases of L-PBF. In situ monitoring of the process promises a less expensive alternative to ex situ testing, but existing sensor technologies and data analysis techniques struggle to detect sub-surface flaws (e.g., porosity and cracking) on production-scale L-PBF printers. In this work, an in situ NDE (INDE) system was engineered to detect subsurface flaws detected in X-Ray Computed Tomography (XCT) directly from process monitoring data. A multilayer, multimodal data input allowed the INDE system to detect numerous subsurface flaws in the size range of 200–1000 µm using a novel human-in-the-loop annotation procedure. Furthermore, a framework was established for generating probability-of-detection (POD) and probability-of-false-alarm (PFA) curves compliant with NDE standards by systematically comparing instances of detected subsurface flaws to post-build XCT data. We also introduce for the first time in the AM in situ sensing literature the a90/95 – the flaw size corresponding to a 90% detection rate on the lower 95% confidence interval of the POD curve. The INDE system successfully demonstrated POD capabilities commensurate with traditional NDE methods. Traditional ML performance metrics were also shown to be inadequate for assessing the ability of the INDE system’s flaw detection performance. It is the belief of the authors that future studies should adopt the POD and PFA approach outlined here to provide better insight into the utility of process monitoring for AM.

The role of defect structure and residual stress on fatigue failure mechanisms of Ti-6Al-4V manufactured via laser powder bed fusion: Effect of process parameters and geometrical factors

Undoubtedly, one of the greatest revolutions in the last few decades has been the paradigm shift in design and manufacturing of components using additive manufacturing (AM) technologies. Despite the numerous benefits of metal AM, the presence of volumetric defects and the development of residual stress during manufacturing pose significant obstacles for producing load-bearing components critical for missions. Laser powder bed fusion (LPBF) has many process parameters, each potentially affecting volumetric defects and residual stress, which may lead to varying fatigue performance. In this paper, we comprehensively investigate the criticality of volumetric defects and residual stress on fatigue properties of AMed Ti-6Al-4V. Using X-ray computed tomography (XCT) and X-ray diffraction (XRD), we characterized volumetric defects and measured residual stress of the specimens, respectively. We performed extensive statistical analysis to correlate XCT, residual stress, and fatigue data, allowing us to identify the factors impacting components' fatigue life. Our observations indicate that residual surface stress plays a crucial role in the fatigue performance of components in the absence of defects at the “critical location”. Furthermore, we found that the pore location has a more significant impact than the defect size on fatigue fracture.

Noise and blur removal from corrupted X-ray computed tomography scans: A multilevel and multiscale deep convolutional framework approach with synthetic training data (BAM SynthCOND)

Regardless of the experimental care practiced in acquiring X-ray computed tomography (XCT) data, artifacts might still exist, such as noise and blur. This is typical for fast XCT data acquisitions (e.g., in-situ investigations), or low-dose XCT. Such artifacts can complicate subsequent analysis of the data. Digital filters can moderately cure extensive artifacts. The selection of filter type, intensity, and order of application is not always straightforward. To tackle these problems, a complete sequential multilevel, multi-scale framework: BAM SynthCOND, employing newly designed deep convolutional neural networks (DCNNs), was formulated. Although data conditioning with neural networks is not uncommon, the main complication is that completely artifact-free XCT data for training do not exist. Thus, training data were acquired from an in-house developed library (BAM SynthMAT) capable of generating synthetic XCT material microstructures. Some novel DCNN architectures were introduced (2D/3D ACEnet_Denoise, 2D/3D ACEnet_Deblur) along with the concept of Assertive Contrast Enhancement (ACE) training, which boosts the performance of neural networks trained with continuous loss functions. The proposed methodology accomplished very good generalization from low resemblance synthetic training data. Indeed, denoising, sharpening (deblurring), and even ring artifact removal performance were achieved on experimental post-CT scans of challenging multiphase Al-Si Metal Matrix Composite (MMC) microstructures. The conditioning efficiencies were: 92% for combined denoising/sharpening, 99% for standalone denoising, and 95% for standalone sharpening. The results proved to be independent of the artifact intensity. We believe that the novel concepts and methodology developed in this work can be directly applied on the CT projections prior to reconstruction, or easily be extended to other imaging techniques such as: Microscopy, Neutron Tomography, Ultrasonics, etc.

Lithium Metal Battery Quality Control via Transformer-CNN Segmentation.

Lithium metal battery (LMB) has the potential to be the next-generation battery system because of its high theoretical energy density. However, defects known as dendrites are formed by heterogeneous lithium (Li) plating, which hinders the development and utilization of LMBs. Non-destructive techniques to observe the dendrite morphology often use X-ray computed tomography (XCT) to provide cross-sectional views. To retrieve three-dimensional structures inside a battery, image segmentation becomes essential to quantitatively analyze XCT images. This work proposes a new semantic segmentation approach using a transformer-based neural network called TransforCNN that is capable of segmenting out dendrites from XCT data. In addition, we compare the performance of the proposed TransforCNN with three other algorithms, U-Net, Y-Net, and E-Net, consisting of an ensemble network model for XCT analysis. Our results show the advantages of using TransforCNN when evaluating over-segmentation metrics, such as mean intersection over union (mIoU) and mean Dice similarity coefficient (mDSC), as well as through several qualitatively comparative visualizations.

A Methodology to Automatically Segment 3D Ultrasonic Data Using X-ray Computed Tomography and a Convolutional Neural Network

Ultrasonic non-destructive testing (UT) is a proficient method for detecting damage in composite materials; however, conventional manual testing procedures are time-consuming and labor-intensive. We propose a semi-automated defect segmentation methodology employing a convolutional neural network (CNN) on 3D ultrasonic data, facilitated by the fusion of X-ray computed tomography (XCT) and Phased-Array Ultrasonic Testing (PAUT) data. This approach offers the ability to develop supervised datasets for cases where UT techniques inadequately assess defects and enables the creation of models with genuine defects rather than artificially introduced ones. During the training process, we recommend processing the 3D volumes as a sequence of 2D slices derived from each technique. Our methodology was applied to segment porosity, a common defect in composite materials, for which characteristics such as void size and shape remain immeasurable via UT. Precision, recall, F1 score, and Intersection over Union (IoU) metrics were used in the evaluation. The results of the evaluation show that the following challenges have to be faced for improvement: (i) achieving accurate 3D registration, (ii) discovering suitable similar keypoints for XCT and UT data registration, (iii) differentiating ultrasonic echoes originating from porosity versus those related to noise or microstructural features (interfaces, resin pockets, fibers, etc.), and, (iv) single out defect echoes located near the edges of the component. In fact, an average F1 score of 0.66 and IoU of 0.5 were obtained.

Photothermal NDE of UD/Epoxy-Based Carbon Fibre Reinforced Laminates for Quantitative Porosity Analysis

High strength and low density make epoxy-based CFRP a highly interesting construction material for the aerospace manufacturing industry. Porosity represents an unavoidable defect and significantly weakens strength values dominated by the matrix. To evaluate the quality of safety-relevant components, non-destructive evaluation and thus the characterization of porous structures is indispensable. Pulsed thermography represents a fast, large-area and non-contact testing method that enables efficient estimation of material parameters. In this work, the authors demonstrate the quantitative estimation of porosity by pulsed thermography on a multi-axial laminate fabricated from unidirectional Prepregs for the first time. The characteristic, extensive expansion of the pores in fiber direction, is addressed by the 3D microstructure characterization of Cone beam X-ray computed tomography data. Hence, the application of effective medium theories and thus the model based porosity estimation is enabled. After the investigation of the effect of pore expansion on the effective thermal diffusivity in 3D finite element simulations, the quantitative photothermal porosity estimation on a sample with a global volume porosity of Phi =1.51% is demonstrated. The accuracy of this fast and non-contact method for porosity estimation with pulsed thermography (Delta Phi =0.63%) is comparable to the standard ultrasonic method. Consequently, an efficient estimation of porosity for large, complex shaped UD/Epoxy composite components is enabled.

A novel iterative algorithm to improve segmentations with deep convolutional neural networks trained with synthetic X-ray computed tomography data (i.S.Sy.Da.T.A)

We propose a novel iterative segmentation algorithm (i.S.Sy.Da.T.A: Iterative Segmentation Synthetic Data Training Algorithm) employing Deep Convolutional Neural Networks and synthetic training data for X-ray tomographic reconstructions of complex microstructures. In our method, we reinforce the synthetic training data with experimental XCT datasets that were automatically segmented in the previous iteration. This strategy produces better segmentations in successive iterations. We test our algorithm with experimental XCT reconstructions of a 6-phase Al-Si Matrix Composite reinforced with ceramic fibers and particles. We perform the analysis in 3D with a special network architecture that demonstrates good generalization with synthetic training data. We show that our iterative algorithm returns better segmentations compared to the standard single training approach. More specifically, phases possessing similar attenuation coefficients can be better segmented: for Al2O3 fibers, SiC particles, and Intermetallics, we see an increase of the Dice score with respect to the classic approach: from 0.49 to 0.54, from 0.66 to 0.72, and from 0.55 to 0.66 respectively. Furthermore, the overall Dice score increases from 0.77 to 0.79. The methods presented in this work are also applicable to other materials and imaging techniques.