X-ray Computed Tomography Datasets Research Articles

Detecting Porosity in Additively Manufactured Parts with Deep Learning and Artificial Computed Tomography Data

ABSTRACT Deep learning has recently gained significant attention in nondestructive testing tasks due to its ability to effectively analyze complex data and identify patterns. This paper proposes a methodology for defect detection on an X-ray computed tomography dataset utilizing the Mask R-CNN algorithm. The proposed approach generates artificial data from the original dataset and introduces a metric for classifying the intensity of the defects. The Mask R-CNN is trained for defect localization and classification, and the results show that the proposed approach performs competitively with, and in many cases better than, reported papers on the same dataset. This paper provides a detailed analysis of the proposed methodology, including experimental results and discussions, demonstrating its effectiveness and potential for future research.

High throughput automated characterization of enamel microstructure using synchrotron tomography and optical flow imaging

The remarkable damage-tolerance of enamel has been attributed to its hierarchical microstructure and the organized bands of decussated rods. A thorough characterization of the microscale rod evolution within the enamel is needed to elucidate this complex structure. While prior efforts in this area have made use of single particle tracking to track a single rod evolution to various degrees of success, such a process can be both computationally and labor intensive, limited to the evolution path of a single rod, and is therefore prone to error from potentially tracking outliers. Particle image velocimetry (PIV) is a well-established algorithm to derive field information from image sequences for processes that are time-dependent, such as fluid flows and structural deformation. In this work, we demonstrate the use of PIV in extracting the full-field microstructural distribution of rods within the enamel. Enamel samples from a wild African lion were analyzed using high-energy synchrotron X-ray micro-tomography. Results from the PIV analysis provide sufficient full-field information to reconstruct the growth of individual rods that can potentially enable rapid analysis of complex microstructures from high resolution synchrotron datasets. Such information can serve as a template for designing damage-tolerant bioinspired structures for advanced manufacturing. Statement of significanceThorough characterization and analysis of biological microstructures (viz. dental enamel) allows us to understand the basis of their excellent mechanical properties. Prior efforts have successfully replicated these microstructures via single particle tracking, but the process is computationally and labor intensive. In this work, optical flow imaging algorithms were used to extract full-field microstructural distribution of enamel rods from synchrotron X-ray computed tomography datasets, and a field method was used to reconstruct the growth of individual rods. Such high throughput information allows for the rapid production/prototyping and advanced manufacturing of damage-tolerant bioinspired structures for specific engineering applications. Furthermore, the algorithms used herein are freely available and open source to broaden the availability of the proposed workflow to the general scientific community.

Simulation-Assisted Augmentation of Missing Wedge and Region-of-Interest Computed Tomography Data.

This study reports a strategy to use sophisticated, realistic X-ray Computed Tomography (CT) simulations to reduce Missing Wedge (MW) and Region-of-Interest (RoI) artifacts in FBP (Filtered Back-Projection) reconstructions. A 3D model of the object is used to simulate the projections that include the missing information inside the MW and outside the RoI. Such information augments the experimental projections, thereby drastically improving the reconstruction results. An X-ray CT dataset of a selected object is modified to mimic various degrees of RoI and MW problems. The results are evaluated in comparison to a standard FBP reconstruction of the complete dataset. In all cases, the reconstruction quality is significantly improved. Small inclusions present in the scanned object are better localized and quantified. The proposed method has the potential to improve the results of any CT reconstruction algorithm.

On the hydrodynamics of macroporous structures: Experimental, CFD and artificial neural network analysis

Porous metallic structures play a critical role in mass and heat transfer processes due to their high surface areas, fixed porosity, and high stiffness – so understanding their fluid flow behaviour is crucial in designing materials that perform efficiently in mass and heat transfer. In view of this, a multi-disciplinary approach is employed to study the hydrodynamics of aluminium foams produced by a liquid melt infiltration technique using experimental, computational fluid dynamics (CFD) modelling and simulation, as well as artificial neural network (ANN) machine learning backpropagation. X-ray computed tomography datasets were used to characterize pore-structure-related properties of replicated materials, followed by three-dimensional advanced imaging of workable representative volume elements. Hydraulic flow information was acquired for the porous matrices using the constant-head permeameter technique. Experiments showed the permeability and Forchheimer coefficient dependence on pore-structure-related properties for fluid-flowing within the pre-Forchheimer and fully developed Forchheimer regimes. Flow permeability of 8.479 × 10−09m2 was highest in the material with the widest mean pore openings (0.212 mm) and lowest (1.291 × 10−09m2) in the material with the narrowest mean pore openings (0.106 mm). Conversely, Forchheimer coefficients were higher for materials with lower porosities and lower for materials with higher porosities. CFD calculations accurately predicted the fluid properties of metallic foams, as well as the influence of intrinsic foam properties on permeability and the Forchheimer coefficient. The ANN model framework was also able to provide valuable information about the hydrodynamics of these materials. Convolution and non-linearity of the ANN model were improved by adding supplementary neurons to the hidden layers allowing deviations within 0.3 and 9.0 percent to be attained.

NDTFlix: Collaborative Remote Analysis of X-ray Computed Tomography Datasets

When analysing data collected from non-destructive testing, several areas of expertise are required, such as on imaging modalities, data analysis and visualisation, or the domain background in terms of the inspected material. Typically, this requires a team of experts, who are not necessarily situated at the same working place. We therefore propose a system supporting collaborative remote analysis facilitating easy access to a common workspace. The system supports a collaborative analysis of datasets by connecting interactions in a lightweight web-based viewer with those in an advanced analysis software package on a desktop PC. Our system supports both co-located and distributed collaboration on large complex datasets and allows for a simple spontaneous cross-device access. This allows a wide range of devices and immersive systems to be supported regardless of their capabilities. Viewpoints and annotations can be shared as visual cues for the analysis of collaborators. The proposed system will be presented using a case study of analysing an adaptive manufactured aeronautics component. The preliminary findings showed positive feedback of the experts regarding the creation of a flexible and fast collaborative analysis set-up.

IntelliScan: Improving the quality of x-ray computed tomography surface data through intelligent selection of projection angles.

X-ray computed tomography (XCT) enables the dimensional measurement and inspection of highly geometrically complex engineering components that are unmeasurable using optical and tactile instruments. Conventional XCT scans use a circular scan trajectory where X-ray projections are acquired with a uniform angular spacing; this approach treats all projections as being of equal importance, in practice, some projections contain more object information than others. In this work we capitalize on this concept by intelligently selecting projections with a view to improve the quality of surface models extracted from an XCT data-set. Our approach relies on using a priori object information to select X-ray projections in which the surfaces of the object are aligned with a ray-path, thus ensuring the surface of the object is fully sampled. Results are presented showing that the proposed method is able to reduce CAD comparison errors by 16%, reduce surface form error by 3%, and improve edge contrast by 14% for a machined aluminium component.

Investigation of Silicon-Based Anodes for Li-Ion Batteries Using X-Rays and Neutron 3D/4D Imaging Techniques

Silicon-based anode materials are one of the most promising approaches to further increase the energy density of lithium-ion batteries. However, current materials are limited by poor cycling stability and rapid capacity fading, mainly caused by the massive volume expansion of Si during lithiation and subsequent strain on the material composite. [1] Furthermore, this electrode swelling also results in continuous solid electrolyte interface (SEI) growth, which hinders the migration of Lithium and leads to permanent capacity loss. [2] [3]To optimize these materials, analytical techniques able to probe the local 3D morphology and Li content are necessary. By applying methods such as neutron computed tomography (NCT), X-Ray computed tomography (XCT) and synchrotron scattering micro-tomography (SMT) structural ageing and changes in the distribution of lithium in the various components can be characterized and quantified for different charging states and recharge cycle numbers.This study investigates an anode material based on a dual phase alloy system of amorphous Silicon (a-Si) with crystalline iron silicide (c-FeSi2) and graphite. [4] [5] Li-ion battery coin cells containing this silicon-graphite composite anode material were industrially produced and aged by performing 1, 300, and 700 charge-discharge cycles, corresponding to remaining capacities of 100, 70 and 50 %, respectively. NCT and XCT scans were acquired at the NeXT instrument of the ILL in Grenoble (France) [6], and SMT measurements performed at ID31 of the neighboring ESRF. These datasets were reconstructed and evaluated using specially developed data processing pipelines. By performing multi-modal registration, the NCT and XCT scans were aligned and segmented to combine the complementary datasets and create a 4D model of the Li-ion battery coin cells.Figure 1 shows a sample of the combined NCT (cyan) and XCT datasets (red). Cropped horizontal slices of Li-Ion battery coin cells are depicted side by side with the main components labelled. Comparing one of the highly cycled cells with 50% remaining capacity (top) with the reference cell without electrolyte (bottom), trapped Lithium can be observed in the anode layers, visible as cyan colored blotchy areas in the otherwise dark component. The pixel size is approximately 5 µm.The change of lithium distribution in the components between different cycle numbers and charge states in the attenuation-based tomography techniques of NCT / XCT is quantified by modelling expected attenuation parameters to the observed values. The diffraction-based tomography method of SMT was modeled using expected lattice parameters. From this information, trapped Lithium can be identified and changes in the distribution are analyzed to point out possible degradation and failure mechanisms.We acknowledge financial support from the European Union’s Horizon 2020 research and Innovation program No. 875514 (ECO2LIB) as well as No. 847439 (InnovaXN) under the Marie Skłodowska-Curie grant agreement.[1] S. Tardif et al, https://doi.org/10.1021/acsnano.7b05796 [2] T. Vorauer et al, https://doi.org/10.1038/s42004-020-00386-x [3] P. Kumar et al, https://doi.org/10.1002/smll.201906812 [4] C. Berhaut et al, https://doi.org/10.1021/acsnano.9b05055 [5] C. Berhaut et al, https://doi.org/10.1016/j.ensm.2020.04.008 [6] A. Tengattini et al, https://doi.org/10.1016/j.nima.2020.163939 Figure 1

Probing the Role of Multi-scale Heterogeneity in Graphite Electrodes for Extreme Fast Charging.

Electrode-scale heterogeneity can combine with complex electrochemical interactions to impede lithium-ion battery performance, particularly during fast charging. This study investigates the influence of electrode heterogeneity at different scales on the lithium-ion battery electrochemical performance under operational extremes. We employ image-based mesoscale simulation in conjunction with a three-dimensional electrochemical model to predict performance variability in 14 graphite electrode X-ray computed tomography data sets. Our analysis reveals that the tortuous anisotropy stemming from the variable particle morphology has a dominating influence on the overall cell performance. Cells with platelet morphology achieve lower capacity, higher heat generation rates, and severe plating under extreme fast charge conditions. On the contrary, the heterogeneity due to the active material clustering alone has minimal impact. Our work suggests that manufacturing electrodes with more homogeneous and isotropic particle morphology will improve electrochemical performance and improve safety, enabling electromobility.

FDM sintering shrinkage rate measurements by X-ray computed tomography

In this work, we report on several possible approaches to the measurements of fused deposition modelling (FDM) sintering shrinkage rates (SSR) from X-ray computed tomography (XCT) data. Metallic parts were produced using FDM technology which relies on an isotropic scaling factor applied to the STL model of the part to be produced to compensate for the shrinkage happening during the sintering of the printed part. We report on whether the shrinkage is isotropic and discuss 4 different approaches to perform the shrinkage measurements based on XCT datasets.

In situ x-ray computed tomography of zinc–air primary cells during discharge: correlating discharge rate to anode morphology

Zinc–air batteries have gained significant attention as safe battery alternatives, with high theoretical energy densities and a high abundance of their constituent materials. However, barriers to their widespread adoption include the need to improve their cycling lifetime, as well as stability and avoiding degradation mechanisms such as zinc dendrite growth and hydrogen-producing side reactions. X-ray computed tomography (CT) is a widely used technique for the study of batteries. In situ / operando x-ray CT has been increasingly used to study the zinc anode of zinc–air batteries to evaluate the interesting morphological changes occurring during the reaction from zinc (Zn) to zinc oxide (ZnO) during discharge (vice versa during charge). However, several studies have been carried out using synchrotron x-ray sources, which have limited availability for users. In this work, we present a comprehensive study of the discharge of commercial, primary zinc–air batteries using a laboratory-based x-ray source for in situ x-ray CT measurements. Four different discharge rates are investigated (C/30, C/60, C/90 and C/150), with tomograms collected at various stages throughout each discharge. Results confirm that with decreasing C-rate (i.e. decreasing discharge current) a greater volume of zinc is reacted, with average mass utilisations of 17%, 76%, 81% and 87% for C/30, C/60, C/90 and C/150, respectively. Furthermore, quantification using x-ray CT datasets showed that there is a direct correlation between the volume of zinc remaining in the cell and the state-of-charge of the cell, which deviated from linearity for the longer C-rates. Finally, a potential new mechanism for shape change is discussed, where a Zn particle is replaced with a pore of a similar volume. As well as improvements in statistical relevance gained from multiple repeats for each C-rate, the results presented here could be used in both modelling of battery performance, as well as consideration for future anode design concepts.

Influence of pore structural properties on gas hydrate saturation and permeability: A coupled pore-scale modelling and X-ray computed tomography method

Gas hydrate, as an alternate hydrocarbon source, has attracted significant attention in past decades. A precise estimation of permeability of the gas hydrate-bearing formation is essential for predicting the flow behaviors and the associated gas production performance. In this research, the influence of gas hydrate saturation on pore structural properties and then affect permeability was investigated using a three-dimensional micro X-ray computed tomography dataset that records an experiment of xenon hydrate formation in a sand pack at selected times during the experiment. Unlike the previous work, the goal of this work is to characterize pore space evolution, during gas hydrate formation, using a set of key microscopic pore characteristics, i.e. pore and throat radii, pore throat ratio, coordination number and tortuosity, by applying pore network analysis, on larger and therefore more representative sub-volumes. The same segmented volume of that dataset was used in this work to estimate gas hydrate saturation and permeability, recalculated by the lattice Boltzmann method, on those larger sub-volumes at selected snapshots. Besides, the analysis provides further insights into the links of gas hydrate localities and local pore characteristics and therefore their controls on the permeability. New evidence of semi-quantitative nature emerges that grain-coating gas hydrate formed at low gas hydrate saturations play a crucial role on reducing permeability, while pore-filling and/or cementing gas hydrate become dominating at high gas hydrate saturations, and these can be explained by gas hydrate formation mechanisms.

Segmenting Water Observed with X-Ray Computed Tomography in Gas Diffusion Layers of Polymer Electrolyte Fuel Cells Using Convolutional Neural Networks

Gas diffusion layers (GDLs) are responsible for oxygen delivery to the catalyst layer and water management in polymer electrolyte fuel cells (PEFCs). With x-ray computed tomography (CT) – a non-destructive technique, one can observe water transport through the GDL, and also assess local wettability. With ability of X-ray CT to capture dynamics of water transport the quantity of data and the time required to process the data increases. There is a need to further automate data processing and analysis through machine learning; in particular, convolutional neural networks (CNNs). CNNs can be applied to a multitude of image processing tasks such as finding centers of rotation for CT data [1], removal of reconstruction artifacts [1], and denoising of data sets [2]. However, CNNs are probably most known for their ability to perform segmentation/labeling.Although CNNs have proven to be widely applicable to image processing and analysis, it has also become apparent that individual pairings of architecture and application must be validated [1]. As such, we focus on the use of CNNs to segment synchrotron X-ray CT data sets for water within various GDLs. The water is tracked for an in-situ evaporation/condensation experiment. Knowledge of the water's location and shape as a function of time is necessary to understand local GDL’s wettability and its effect on PEFC performance. The challenge these samples present to more traditional segmentation methods is the low contrast between water and carbon fibers. Depending on the composition of a GDL, binder, polytetrafluoroethylene (PTFE), and the microporous layer (MPL) may also be present in the same narrow window of gray-scale values. Separating these phases is essential in evaluating the material's water transport properties. We will present results from both our efforts with CNNs and the specific PEFC experiment that produced the data sets.

Improved surface extraction of multi-material components for single-source industrial X-ray computed tomography

The paper demonstrates a novel methodology for surface extraction of multi-material components (MMCs) on industrial X-ray computed tomography (CT) datasets. The methodology is based on a combination of fuzzy C-means clustering (FCM) and region growing (RG) methods. FCM, used as a pre-processing step, allows proper classification and improvement of different objects boundaries present on industrial X-ray CT datasets. Afterwards, application of RG method enables accurate segmentation of classified and improved X-ray CT datasets. The performance of presented approach has been tested on two CT datasets acquired on an industrial X-ray CT system NIKON XT H 225. It was also compared against two commercial industrial software VGStudio Max v3.1 and GOM Inspect v2018. Obtained results from application of the proposed approach show significant improvement in surface extraction of MMCs in CT datasets, especially in cases of low-density materials such as polymers. Verification has been conducted by obtaining reference measurements using contact coordinate measuring machine (CMM) Contura G2 by CARL ZEISS.

Coupled ptychography and tomography algorithm improves reconstruction of experimental data

Three-dimensional (3D) x-ray microscopy by ptychographic tomography requires elaborate numerical reconstructions. We describe a coupled ptychography-tomography reconstruction algorithm and apply it to an experimental ptychographic x-ray computed tomography data set of a catalyst particle. Compared to the traditional sequential algorithm, in which ptychographic projections are reconstructed to serve as input for subsequent tomographic reconstruction, the coupled ptychography-tomography algorithm reconstructs the 3D volume with higher spatial resolution over a larger field of view. Coupling the data from different projections improves the overall reconstruction, and the ptychographic sampling in individual projections can be coarsened beyond the point of overlap between neighboring scan points, still leading to stable reconstructions.

Ring artifact suppression in X-ray computed tomography using a simple, pixel-wise response correction.

We present a pixel-specific, measurement-driven correction that effectively reduces errors in detector response that give rise to the ring artifacts commonly seen in X-ray computed tomography (CT) scans. This correction is easy to implement, suppresses CT artifacts significantly, and is effective enough for use with both absorption and phase contrast imaging. It can be used as a standalone correction or in conjunction with existing ring artifact removal algorithms to further improve image quality. We validate this method using two X-ray CT data sets acquired using monochromatic sources, showing post-correction signal-to-noise increases of up to 55%, and we define an image quality metric to use specifically for the assessment of ring artifact suppression.

Erratum to: A micro X-ray computed tomography dataset of South African hermit crabs (Crustacea: Decapoda: Anomura: Paguroidea) containing scans of two rare specimens and three recently described species.

CITATION: Landschoff, J., Du Plessis, A. & Griffiths, C. L. 2018. A micro X-ray computed tomography dataset of South African hermit crabs (Crustacea: Decapoda: Anomura: Paguroidea) containing scans of two rare specimens and three recently described species. GigaScience, 7(4):giy022, doi:10.1093/gigascience/giy022.

Reactive Transport Simulation of Fracture Channelization and Transmissivity Evolution.

Underground fractures serve as flow conduits, and they may produce unwanted migration of water and other fluids in the subsurface. An example is the migration and leakage of greenhouse gases in the context of geologic carbon sequestration. This study has generated new understanding about how acids erode carbonate fracture surfaces and the positive feedback between reaction and flow. A two-dimensional reactive transport model was developed and used to investigate the extent to which geochemical factors influence fracture permeability and transmissivity evolution in carbonate rocks. The only mineral modeled as reactive is calcite, a fast-reacting mineral that is abundant in subsurface formations. The X-ray computed tomography dataset from a previous experimental study of fractured cores exposed to carbonic acid served as a testbed to benchmark the model simulation results. The model was able to capture not only erosion of fracture surfaces but also the specific phenomenon of channelization, which produces accelerating transmissivity increase. Results corroborated experimental findings that higher reactivity of the influent solution leads to strong channelization without substantial mineral dissolution. Simulations using mineral maps of calcite in a specimen of Amherstburg limestone demonstrated that mineral heterogeneity can either facilitate or suppress the development of flow channels depending on the spatial patterns of reactive mineral. In these cases, fracture transmissivity may increase rapidly, increase slowly, or stay constant, and for all these possibilities, the calcite mineral continues to dissolve. Collectively, these results illustrate that fluid chemistry and mineral spatial patterns need to be considered in predictions of reaction-induced fracture alteration and risks of fluid migration.

A micro X-ray computed tomography dataset of fossil echinoderms in an ancient obrution bed: a robust method for taphonomic and palaeoecologic analyses

BackgroundTaphonomic and palaeoecologic studies of obrution beds often employ conventional methods of investigation such as physical removal and extraction of fossils from their host rock (matrix) by mechanical preparation. This often-destructive method is not suitable for studying mold fossils, which are voids left in host rocks due to dissolution of the original organism in post-depositional processes.FindingsMicrocomputed tomography (µCT) scan data of 24 fossiliferous rock samples revealed thousands of Paleozoic echinoderms. Digitally "stitching" together individually µCT scanned rock samples within three-dimensional (3D) space allows for quantifiable taphonomic data on a fossil echinoderm-rich obrution deposit from the Devonian (Emsian) of South Africa. Here, we provide a brief step-by-step guide on creating, segmenting, and ultimately combining sections of richly fossiliferous beds to create virtual models suited for the quantitative and qualitative taphonomic analyses of fossil invertebrate assemblages.ConclusionsVisualizing the internal features of fossiliferous beds in 3D is an invaluable taphonomic tool for analyzing delicate fossils, accounting for all specimens irrespective of their preservation stages and with minimal damage. This technique is particularly useful for analyzing fossiliferous deposits with mold fossils that prove to be difficult to study with traditional methods, because the method relies on the large density contrast between the mold and host rock.

Visual analysis of void and reinforcement characteristics in X-ray computed tomography dataset series of fiber-reinforced polymers

Fiber-reinforced polymers (FRPs) are of great importance in various industries because of their superior properties as compared to conventional materials, their versatile processing, and their wide application possibilities. To fulfil the high-quality standards in its respective applications, industrial 3D X-ray computed tomography (XCT) is increasingly used. It enables an accurate, non-destructive characterization of material features such as inclusions, voids, fibers, or other reinforcements, which is of core importance for material and component design. In this work we present FeatureAnalyzer, a generalization of the previously introduced PorosityAnalyzer tool, which allows to analyze dataset series as generated for exploring the parameter space of image processing workflows (including pre-filtering, segmentation, postprocessing or quantification) applied to XCT datasets of fiber-reinforced polymers. With a scatter plot matrix (SPLOM), the characteristics of the features of interest may be examined in more detail regarding the used input and output parameters. Individual results may be selected in the SPLOM and analyzed using 2D slice views and 3D renderings. For this work, three different samples (sample #1 - #3) were scanned by means of XCT and were evaluated by using FeatureAnalyzer. The samples #1 and #2 have a porosity value of approximately 1.7 vol. %. By using the FeatureAnalyzer in combination with SPLOM, the threshold parameters could be analyzed before the over-segmentation of voids occurs. Additional evaluations by parallel coordinates clearly show, that sample #2 has a higher number of spherical voids in the center of the specimen compared to sample #1. By evaluating the resin content of sample #3, the individual layer thickness could be measured. The source code of the tool is available on Github: https://github.com/3dct/open_iA/

A micro X-ray computed tomography dataset of South African hermit crabs (Crustacea: Decapoda: Anomura: Paguroidea) containing scans of two rare specimens and three recently described species.

BackgroundAlong with the conventional deposition of physical types at natural history museums, the deposition of 3-dimensional (3D) image data has been proposed for rare and valuable museum specimens, such as irreplaceable type material.FindingsMicro computed tomography (μCT) scan data of 5 hermit crab species from South Africa, including rare specimens and type material, depicted main identification characteristics of calcified body parts. However, low-image contrasts, especially in larger (>50 mm total length) specimens, did not allow sufficient 3D reconstructions of weakly calcified and fine characteristics, such as soft tissue of the pleon, mouthparts, gills, and setation. Reconstructions of soft tissue were sometimes possible, depending on individual sample and scanning characteristics. The raw data of seven scans are publicly available for download from the GigaDB repository.ConclusionsCalcified body parts visualized from μCT data can aid taxonomic validation and provide additional, virtual deposition of rare specimens. The use of a nondestructive, nonstaining μCT approach for taxonomy, reconstructions of soft tissue structures, microscopic spines, and setae depend on species characteristics. Constrained to these limitations, the presented dataset can be used for future morphological studies. However, our virtual specimens will be most valuable to taxonomists who can download a digital avatar for 3D examination. Simultaneously, in the event of physical damage to or loss of the original physical specimen, this dataset serves as a vital insurance policy.