Edge Of Sample Research Articles

Differences between stenotic bicuspid and tricuspid aortic valves: a histological comparison

Abstract Background Aortic valve stenosis is a narrowing of the aortic valve opening due to leaflet calcification and thickening. Thickening of the leaflets involves the original leaflet but also the presence of a superimposed tissue (SIT), defined as an extra layer of tissue on top of the original leaflet which has been shown to form mainly in response to mechanical stress. Patients with bicuspid aortic valve (BAV) are known to be at higher risk of developing aortic stenosis and at an early age as compared to patients with tricuspid aortic valves (TAV). However to date, no systematic histo-morphometric comparison has been made between BAV and TAV in order to unravel potential difference in the underlying mechanisms of valve degeneration. Purpose Aim of this study was to compare the extent and distribution of calcifications and SIT between stenotic BAV and TAV. Methods BAV (n=14) and TAV (n=12) samples were collected from patients undergoing aortic valve replacement due to aortic stenosis. Of each patient, at least one leaflet was used for analyses, for a total of 27 BAV leaflets and 17 TAV leaflets. Samples were fixed in 4% paraformaldehyde solution, and sectioned in the midline of the leaflet. In case of a leaflet with raphe, the leaflet was sectioned in the midline between the raphe and the commissure side of the leaflet. Sections were stained for elastin fibers using Weighert’s Resorcin Fuchsin. Thickness of the leaflets was calculated by dividing the cross sectional area by the length of the leaflet. Calcifications were visualized using Alizarin Red staining and quantified as a percentage of the total cross sectional area of one section. SIT was defined as an extra layer of tissue outside the elastin lamina (Figure). Results BAV samples showed in average thicker leaflets as compared to TAV (1.92 vs 1.31 mm, p <0.001). Although the thickness of the original leaflet was the same between the two groups, SIT formation was significantly thicker in the BAV as compared to TAV (0.67 vs 0.30 mm, p <0.001). This difference was mostly due to a significant thicker SIT in the free edge of BAV samples as compared to TAV (0.78 vs 0.39 mm, p=0.003), while the SIT at the level of the leaflet body was comparable between the two groups. The percentage of leaflet calcification was similar between BAV and TAV (23% vs 20% respectively, p=0.669). Interestingly, while calcification in TAV samples was observed only on the aortic side, in BAV samples calcification was also observed on the ventricular side (n=5; 36% Figure). Conclusions BAV showed thicker leaflets as compared to TAV due to a larger contribution of SIT formation particularly at the free edge. Extent of calcification was similar between BAV and TAV but calcification at the ventricular side of the leaflets was found only in BAV samples. These observations suggest significant differences in the mechanisms underlying BAV and TAV pathology possibly in relation to distinct mechanical stress.

Sample Thickness and Edge Proximity Influence Spatial Behavior of Filaments and Treatment Uniformity of RF Cold Atmospheric Pressure Plasma Jet

AbstractThe ability of atmospheric pressure plasma jets to treat complex non-planar surfaces is often cited as their advantage over other atmospheric plasmas. However, the effect of complex surfaces on plasma parameters and treatment efficiency has seldom been studied. Herein, we investigate the interaction of the atmospheric pressure plasma slit jet (PSJ) with block polypropylene samples of different thicknesses (5 and 30 mm) moving at two different speeds. Even though the distance between the slit outlet and the sample surface was kept constant, the treatment efficiency of PSJ ignited in the Ar and $$\hbox {Ar/O}_2$$ Ar/O 2 gas feeds varied with the sample thickness due to the plasma parameters such as filament count and speed being affected by the different distances of the ground (the closer the ground is, the higher the discharge electric field). On the other hand, the $$\hbox {Ar/N}_2$$ Ar/N 2 PSJ diffuse plasma plumes were less affected by the changes in the electric field, and the treatment efficiency was the same for both sample thicknesses. Additionally, we observed a difference in the efficiency and uniformity of the PSJ treatment of the edges and the central areas in some working conditions. The treatment efficiency near the edges depended on the duration of the filament contact, i. e., how long the local electric field trapped the filaments. Conversely, the treatment uniformity near the edges and in the central areas was different if the number of filaments changed rapidly as the discharge moved on and off the sample (the 5 mm samples treated by easily sustained Ar PSJ).

Extraction of agricultural plastic greenhouses based on a U-Net Convolutional Neural Network Coupled with edge expansion and loss function improvement

ABSTRACT Agricultural plastic greenhouses (APGs) are crucial for sustainable agricultural planting, and accurate spatial distribution information acquisition is crucial. Deep learning network models can extract target features from remote sensing images more effectively than traditional interpretation methods, which face challenges like high workloads and poor repeatability. In this study, we aim to enhance the inventorying of Agricultural Plastic Greenhouses (APGs) by improving the extraction accuracy of their locations and numbers through remote sensing techniques. Utilizing GF-7 satellite imagery, we propose an enhanced U-Net convolutional neural network (CNN) model that incorporates edge information expansion and a joint loss function to optimize performance. The primary objective is to provide a rapid and accurate method for mapping APGs, which is crucial for effective agricultural management and environmental monitoring. The U-Net network’s accuracy was enhanced by 1.1% after expanding 3 × 3sample edge information, and further by 1.9% by combining edge extension and loss function constraints. Our results demonstrate that the modified U-Net model significantly improves extraction accuracy compared to traditional methods, thereby facilitating better inventory management and planning for agricultural cash crops. This advancement not only supports farmers in optimizing resources but also contributes to sustainable agricultural practices by enabling precise monitoring of APG distribution. Implication Statement Compared to traditional interpretation methods, which suffer problems such as heavy workloads, small adaptation ranges and poor repeatability, deep learning network models can better extract target features from remote sensing images. In this study, we used GF-7 image data to improve the traditional U-Net convolutional neural network (CNN) model. The Canny operator and Gaussian kernel (GK) function were used for sample edge expansion, and the binary cross-entropy and GK functions were used to jointly constrain the loss. Finally, APGs were accurately extracted and compared to those obtained with the original model. The results indicated that the APG extraction accuracy of the U-Net network was improved through the expansion of sample edge information and adoption of joint loss function constraints.

Topological protection revealed by real-time longitudinal and transverse transport measurements

Topology is essential for achieving unchanged (or protected) quantum properties in the presence of perturbations. A challenge facing the application is the variable protection levels displayed in real systems associated with the reconstructive behaviors of the dissipationless modes. Despite various insights on potential causes of backscattering, the edge-state-based approach is incomplete because the bulk states also contribute indispensably. This study investigates sample-scale reconstruction where dissipationless modes are global objects instead of being restricted to the sample edge. An integer quantum Hall effect hosted in a Corbino geometry is adopted and brought to the verge of a breakdown. Two independent and simultaneous detections are performed to capture transport responses in both longitudinal and transverse directions. The real-time correspondence between orthogonal results confirms two facts. 1. Dissipationless modes undergo frequent reconstruction in response to electrochemical potential changes, causing dissipationless current paths to expand transversely into the bulk while preserving chirality. A breakdown only occurs when a backscattering emerges between reconstructed dissipationless current paths bridging opposite edge contacts. 2. Topological protection is subject to an interplay of disorder, electron-electron interaction, and topology. The proposed reconstruction mechanism qualitatively explains the robustness variations, beneficial for protection optimization.

Ce-doped ZnO nanonails synthesized by a simple thermal evaporation method for photocatalytic degradation

One-dimensional (1D) nanomaterials have garnered significant scientific and technological attention due to their potential applications in electronics devices, gas sensing, energy conversion, and photocatalysis. However, the development of a simple and low-cost technique for 1D nanostructure synthesis remains a challenge. The doping of ZnO nanostructures has proven to be an effective way to improve their internal properties. In this work, one-dimensional ZnO and Ce-doped ZnO nanorods and nanonails were synthesized by a simple thermal evaporation method using different weight ratios of ZnS and CeO2 precursor powders in a catalyst-free process. The effects of cerium doping concentration on the structural, morphological, and optical properties of the as-prepared samples were investigated. XRD analysis confirmed that the crystal structure was converted from the cubic ZnS phase to the hexagonal ZnO phase with increasing annealing temperatures. The UV–Vis spectra showed that the optical absorption edge of Ce-doped ZnO samples was slightly red-shifted. Additionally, the band gap energy decreases with increasing cerium concentration. SEM images showed that the morphology of the structures obtained changed from nanorods to nanonails as the Ce content increased. The incorporation of Ce ions into the ZnO matrix has been successfully confirmed by HRTEM, Raman, and EDX analysis. Photocatalytic activity studies showed that Ce-doped ZnO samples exhibited significantly enhanced photocatalytic performance compared to pure ZnO for the degradation of rhodamine B under UV radiation. These results may suggest the use of heterogeneous photocatalysis as an efficient, cheap, and environment-friendly alternative for the removal of pollutants from water and the use of Ce-doped ZnO as a promising candidate.

Meandering conduction channels and the tunable nature of quantized charge transport

The discovery of the quantum Hall effect has established the foundation of the field of topological condensed matter physics. An amazingly accurate quantization of the Hall conductance, now enshrined in quantum metrology, is stable against any reasonable perturbation due to its topological protection. Conversely, the latter implies a form of censorship by concealing any local information from the observer. The spatial distribution of the current in a quantum Hall system is such a piece of information, which, thanks to spectacular recent advances, has now become accessible to experimental probes. It is an old question whether the original and intuitively compelling theoretical picture of the current, flowing in a narrow channel along the sample edge, is the physically correct one. Motivated by recent experiments locally imaging quantized current in a Chern insulator (Bi, Sb)[Formula: see text]Te[Formula: see text] heterostructure [Rosen et al., Phys. Rev. Lett. 129, 246602 (2022); Ferguson et al., Nat. Mater. 22, 1100-1105 (2023)], we theoretically demonstrate the possibility of a broad "edge state" generically meandering away from the sample boundary deep into the bulk. Further, we show that by varying experimental parameters one can continuously tune between the regimes with narrow edge states and meandering channels, all the way to the charge transport occurring primarily within the bulk. This accounts for various features observed in, and differing between, experiments. Overall, our findings underscore the robustness of topological condensed matter physics, but also unveil the phenomenological richness, hidden until recently by the topological censorship-most of which, we believe, remains to be discovered.

Identification of gamma irradiated polycarbonate-polybutylene terephthalate stiff polymer blend products using HS-GC/MS, FTIR and UV/Vis spectroscopy

In this study, the fragmentation products of Makro-blend stiff polymer due to gamma exposure were identified using Head space (HS), Gas Chromatography, and Mass Spectrometry (HS-GC/MS) techniques and Fourier Transform Infra-Red (FTIR) spectroscopy. As well the electronic transitions due to gamma exposure were examined using UV/Vis spectroscopy. For the un-irradiated and 10 kGy irradiated samples, one product was detected by (HS-GC/MS) method with a clear peak of Dimethoxyethane (MW = 90), which has a peak area percentage of 98.58% and 97.13%, respectively. With increasing gamma doses the number of identified components was increased. At 1200 kGy, ten components were detected: Biethylene (MW = 54), Dimethoxyethane (MW = 90), 1,1-Dimethoxy-2-propanone (MW = 118), 1,1-Dimethoxy-2-propanol (MW = 120), 1,1-Dimethoxycyclopentane (MW = 130), 1,1-Dimethoxyheptane (MW = 160), 1,1-Dimethoxyoctane (MW = 174), Dimethyl terephthalate (MW = 194), 4-(Diethoxymethyl) benzaldehyde (MW = 208), and Terephthalate dihexyl (MW = 334), which have peaks area percentages of 4.22, 2.98, 25.84, 2.2, 10.65, 38.86, 1.57, 5.91, 1.22, and 6.48, respectively. The recorded mass spectrum of each component confirms the molecular ion peak. The FTIR measurements demonstrate the decrease in the detected band intensities following gamma radiation. Accordingly, a chain scission occurs, which causes the carbonate bond to scission, resulting in the creation of a hydroxyl group and the (–H) abstraction from the backbone of the irradiation blend samples. The absorbance edge of the gamma-exposed samples shifted towards a higher wavelength. This pattern shows that the band gap energy is declining. The use of FTIR and HS-GC/MS in mass spectrometry analysis will be useful tools for examining radiation-induced polymeric fragments and advancing our understanding of the process underlying the creation of these fragments in polymers.

Competition between microstructural factors affecting growth of abnormally large grains in thin Cu foils

Grain boundary types and local boundary curvatures are generally considered to be important microstructural factors controlling grain boundary migration during grain growth. In this work, grain growth in thin copper foils is studied during annealing at a temperature of 1040 °C near the melting point by ex-situ experiments and Monte Carlo simulations. Few grains, stimulated by slight deformation at the sample edge, are observed to grow abnormally into the cube-oriented recrystallized microstructure with columnar grains spanning the foil thickness. The grain boundaries of these abnormally growing grains and the grain sizes in the adjacent polycrystalline recrystallized regions are analyzed. The experimental results suggest that spatial heterogeneities in the distribution of small recrystallized grains have a significant effect on the migrating boundaries. Potts model simulations confirm that grain boundary segments with small grains in front are more likely to migrate than segments facing coarser grains. The simulations also demonstrate the importance of grain morphology. Altogether, this work highlights the effects of a heterogeneous recrystallized microstructure on abnormal grain growth in thin foil samples.

A graph residual generation network for node classification based on multi-information aggregation

The key to improving the performance of graph convolutional networks (GCN) is to fully explore the correlation between neighboring and distant information. Aiming at the over-smoothing problem of GCN, in order to make full use of the relationship among features, graphs and labels, a graph residual generation network based on multi-information aggregation (MIA-GRGN) is proposed. Firstly, aiming at the defects of GCN, we design a deep initial residual graph convolution network (DIRGCN), which connects the initial input through residuals, so that each layer node retains part of the information of the initial features, ensuring the localization of the graph structure and effectively alleviating the problem of over-smoothing. Secondly, we propose a random graph generation method (RGGM) by utilizing graph edge sampling and negative edge sampling, and optimize the supervision loss function of DIRGCN in the form of generation framework. Finally, applying RGGM and DIRGCN as inference modules for modeling hypotheses and obtaining approximate posterior distributions of unknown labels, an optimized loss function is obtained, we construct a multi-information aggregation MIA-GRGN that combines graph structure, node characteristics and label joint distribution. Experiments on benchmark graph classification datasets show that MIA-GRGN achieves better classification results compared with the benchmark models and mainstream models, especially for datasets with less dense edge relationships between nodes.

Defect characterizations of N-rich GaNAs ternary alloys

This paper presents a comprehensive analysis of defects in As-diluted GaN alloys. GaNAs crystal layers with an arsenic content of 1.8, 2.9, 4.1, and 5.5 % are grown on GaN buffer layers using the metalorganic vapour-phase epitaxy (MOVPE) method. Since these alloys have potential in electronics due to the modification of the electronic structure caused by As, it is extremely important to determine their quality. Complementary techniques for the characterization of the alloys are used. Densities of screw and edge dislocations are obtained using X-ray diffraction measurements performed, respectively, from the surface and edge of samples. In turn, a population of vacancies is determined by Doppler broadening variable energy positron annihilation spectroscopy (DB-VEPAS) and variable energy positron lifetime positron annihilation spectroscopy (VEPALS), which provide the distribution of vacancies as a function of depth. An increase in dislocation density is observed accompanied by a rise in the concentration of vacancy agglomerations and reduction in Ga-vacancy-hydrogen associates in the function of As-content.

A smart surveillance system utilizing modified federated machine learning: Gossip‐verifiable and quantum‐safe approach

SummaryEdge computing has the capability to process data closer to its point of origin, leading to the development of critical autonomous infrastructures with frequently communicating peers. The proposed work aims to evaluate the effectiveness of security and privacy mechanisms tailored for distributed systems, particularly focusing on scenarios where the nodes are closed‐circuit television (CCTV) systems. Ensuring public safety, object tracking in surveillance systems is a vital responsibility. The workflow has been specifically crafted and simulated for the purpose of weapon detection within public CCTV systems, utilizing sample edge devices. The system's primary objective is to detect any unauthorized use of weapons in public spaces while concurrently ensuring the integrity of video footage for use in criminal investigations. The outcomes of prior research on distributed machine learning (DML) techniques are compared with modified federated machine learning (FML) techniques, specifically designed for being Gossip verifiable and Quantum Safe. The conventional federated averaging algorithm is modified by incorporating the secret sharing principle, coupled with code‐based McEliece cryptosystem. This adaptation is designed to fortify the system against quantum threats. The Gossip data dissemination protocol, executed via custom blockchain atop the distributed network, serves to authenticate and validate the learning model propagated among the peers in the network. It provides additional layer of integrity to the system. Potential threats to the proposed model are analyzed and the efficiency of the work is assessed using formal proofs. The outcomes of the proposed work demonstrate that the trustworthiness and consistency are meticulously preserved for both the model and data within the DML framework on the Edge computing platform.

Radiofrequency assisted hot air drying of corn kernels: Drying characteristics, uniformity, quality, and energy consumption

In order to clarify the applicability of radio frequency assisted hot air drying (RFHAD) for harvested cereals, the drying characteristics, quality, and energy consumption of RFHAD with various air temperatures (40 °C, 50 °C, 60 °C, 70 °C) and electrode gaps (90 mm, 100 mm, 110 mm, 120 mm) for corn kernels was investigated by comparing with these of hot air drying (HAD). Results showed that the temperature of corn kernels was higher than that of hot air during RFHAD, hot air mainly played a convective cooling role, and the cooling intensity enhanced with the decrease of air temperature or electrode gap. However, RFHAD inherited the non-uniform characteristic of RF heating, showing a higher temperature at the edges of corn samples, but it could be improved at higher or lower air temperatures than that at 40 °C and larger electrode gaps. The drying efficiency, color difference, and crack rate of corn kernels increased with the increase of air temperature or decrease of electrode gap. Especially, for corn kernels with similar temperature by RFHAD (50 °C air temperature, 100 mm electrode gap) and HAD (60 °C air temperature), the drying time of RFHAD was 25% shorter than that of HAD, meanwhile, the color difference and crack rate were at a smaller level. The energy consumption of RFHAD was reduced at small electrode gaps and high air temperatures since the drying time was significantly shorter than that of HAD. The results provide useful assistance for developing effective strategies for RFHAD of corn kernels.

Correcting directional dark field x-ray imaging artefacts using position dependent image deblurring and attenuation removal

In recent years, a novel x-ray imaging modality has emerged that reveals unresolved sample microstructure via a “dark-field image”, which provides complementary information to conventional “bright-field” images, such as attenuation and phase-contrast modalities. This x-ray dark-field signal is produced by unresolved microstructures scattering the x-ray beam resulting in localised image blur. Dark-field retrieval techniques extract this blur to reconstruct a dark-field image. Unfortunately, the presence of non-dark-field blur such as source-size blur or the detector point-spread-function can affect the dark-field retrieval as they also blur the experimental image. In addition, dark-field images can be degraded by the artefacts induced by large intensity gradients from attenuation and propagation-based phase contrast, particularly around sample edges. By measuring any non-dark-field blurring across the image plane and removing it from experimental images, as well as removing attenuation and propagation-based phase contrast, we show that a directional dark-field image can be retrieved with fewer artefacts and more consistent quantitative measures. We present the details of these corrections and provide “before and after” directional dark-field images of samples imaged at a synchrotron source. This paper utilises single-grid directional dark-field imaging, but these corrections have the potential to be broadly applied to other x-ray imaging techniques.

ShaleSeg: Deep-learning dataset and models for practical fracture segmentation of large-scale shale CT images

Shale gas plays a crucial role as an exploration resource for human life, and its contribution to energy production has been steadily increasing. The success of hydraulic fracturing operations is highly dependent on the understanding of shale fracture propagation at large scales. Currently, the segmentation of large-scale shale fractures is primarily carried out through time-consuming manual or semi-automated methods. The identification of shale fractures is similar to image segmentation, in which deep-learning holds the potential to achieve accurate performance close to human experts. However, its application encounters challenges when dealing with large-scale shale images due to factors such as a lack of datasets, severe artifacts, and limited sample sizes. In this study, we compiled the first publicly available large-scale shale CT dataset, which consists of a total of 4313 scanned images, with over 700 images coming from scanning experiments on 300 cm3 cubic shale samples. We also proposed a new data augmentation method for shale CT images to overcome challenges posed by severe artifacts and small-sample sizes. This method involves simulating image artifacts and reducing high edge gradients along the sample edges, which has proven to be effective in facilitating the efficient training of deep-learning models. We've also conducted several ablation studies to determine the best model design, namely ShaleSeg. ShaleSeg was able to accurately identify multi-scale fractures, achieving a precision of 0.84 and an F-score of 0.81, which is close to the manual results. In the case of minor fractures, precision increased by an increase significantly from 0.191 to 0.513, compared to the traditional algorithm. ShaleSeg significantly improves the accuracy and efficiency of fracture segmentation for large-scale shale images. This advancement will be crucial for a thorough comprehension of the fracture mechanism in shale.

Detecting Near-Surface Sub-Millimeter Voids in Additively Manufactured Ti-5V-5Al-5Mo-3Cr Alloy Using a Transmit-Receive Eddy Current Probe.

Additive Manufacturing (AM) Direct Laser Fabrication (DLF) of Ti-5Al-5V-5Mo-3Cr (Ti5553) is being developed as a method for producing aircraft components. The additive manufacturing process can produce flaws near the surface, such as porosity and material voids, which act as stress raisers, leading to potential component failure. Eddy current testing was investigated to detect flaws on or near the surface of DLF Ti5553 bar samples. For this application, the objective was to develop an eddy current probe capable of detecting flaws 500 µm in diameter, located 1 mm below the component's surface. Two initial sets of coil parameters were chosen: The first, based on successful experiments that demonstrated detection of a near surface flaw in Ti5553 using a transmit-receive array probe, and the second, derived from simulation by Finite Element Method (FEM). An optimized transmit receive coil design, based on the FEM simulations, was constructed. The probe was evaluated on Ti5553 samples containing sub-surface voids of the target size, as well as samples with side-drilled holes and samples with holes drilled from the opposing inspection surface. The probe was able to effectively detect 80% of the sub-surface voids. Limitations included the probe's inability to detect sub-surface voids near sample edges and a sensitivity to surface roughness, which produces local changes in lift-off. Multifrequency mixing improved signal-to-noise ratio when surface roughness was present on average by 22%. A probe based on that described in this paper could benefit quality assurance of additively manufactured aircraft components.

A finite-strain plate model for the magneto-mechanical behaviors of hard-magnetic soft material plates

In this paper, we propose a plate model, within the framework of finite-strain elasticity, to study the magneto-mechanical behaviors of hard-magnetic soft material plates. The 2D vector plate equation is derived from the 3D governing equations through a series expansion and truncation approach. The displacement and traction boundary conditions on the edges of plate samples are also proposed. Compared with the other plate or shell theories, the current plate model naturally incorporates the elastic incompressibility and the magneto-mechanical coupling effect of the material, which doesn't involve any a priori hypotheses on the geometry and deformation of the plates. Thus, it is applicable for modeling the large deformations of hard-magnetic soft material plates. To show the efficiency of the plate model, we study a benchmark problem, i.e., the plane-strain bending deformations of plate samples induced by magnetic and mechanical loads. Some analytical solutions of the plate equation system are derived, which provide accurate predictions on the magneto-mechanical response of the plate samples with different thickness-length ratios and magnetization directions. To promote applications of the plate model, we further derive a simplified plate equation system and write out its weak form, which is then implemented into the finite element software COMSOL Multiphysics. The efficiency of the simplified plate equation system is also verified through some typical examples. It is found that the numerical results obtained from the 2D plate model show good consistency with those of the 3D volume model, while the plate model exhibits obvious advantages in the aspects of numerical efficiency.

Spontaneous time-reversal symmetry breaking by disorder in superconductors

A growing number of superconducting materials display evidence for spontaneous time-reversal symmetry breaking (TRSB) below their critical transition temperatures. Precisely what this implies for the nature of the superconducting ground state of such materials, however, is often not straightforward to infer. We review the experimental status and survey different theoretical mechanisms for the generation of TRSB in superconductors. In cases where a TRSB complex combination of two superconducting order parameter components is realized, defects, dislocations and sample edges may generate superflow patterns that can be picked up by magnetic probes. However, even single-component condensates that do not break time-reversal symmetry in their pure bulk phases can also support signatures of magnetism inside the superconducting state. This includes, for example, the generation of localized orbital current patterns or spin-polarization near atomic-scale impurities, twin boundaries and other defects. Signals of TRSB may also arise from a superconductivity-enhanced Ruderman-Kittel-Kasuya-Yosida exchange coupling between magnetic impurity moments present in the normal state. We discuss the relevance of these different mechanisms for TRSB in light of recent experiments on superconducting materials of current interest.

Expediting structure–property analyses using variational autoencoders with regression

We present a machine learning approach that expedites structure–property analysis in materials, bypassing traditional feature extraction and exploratory data analysis techniques. This objective is accomplished by employing a variational autoencoder (VAE) structure that is modified to include a regressor network for property prediction (VAE-Regression). This modification allows for direct linkage of imaged features and quantitative part properties within the VAE latent space. We first demonstrate our approach using 2D optical micrographs and corresponding four-point bend fatigue life data from laser beam powder bed fusion additively manufactured Ti-6Al-4V coupons. The VAE-Regression model extracts spatial features, predicts fatigue life, and identifies features of porosity defect governing fatigue behavior such as pore clusters, pores near sample edges, and jagged pore morphologies. These features corroborate fatigue literature on physics-based modeling and experimentation. We then demonstrate the versatility of our methodology using binder jet additively manufactured WC-Co coupons, where porosity and microstructural discontinuities are known to lower the three-point bend transverse rupture strength, but the interaction between the WC and Co are yet to be completely understood. We attempted to understand these interactions using our VAE-Regression architecture. Within our dataset, we show that coarser WC grains surrounded by larger Co pools indicate lower strength, while finer WC grains with smaller Co pools indicate higher strength. This machine learning approach using image-based data will likely prove to be critical in understanding and identifying structure–property relationships in new materials and manufacturing processes.

Pre-yielding and necking process of double network hydrogels revealed by sample geometry effects

Understanding the yielding and necking mechanisms of double network (DN) materials is crucial for establishing structure-property correlations. While previous studies have primarily focused on how macro-yielding behavior changes with network structure and swelling, in this work, we study the local internal damage in the pre-yielding process to unveil the yielding and necking mechanisms for typical DN gels. Through birefringence retardation imaging during tensile deformation on DN gels of various sample geometries, our findings reveal the following key points: 1) Initiation and Growth of Damage Zones: Prior to macroscopic yielding, damage zones initiate from the sample edges and gradually grow towards the sample center with elongation. 2) Rapid Propagation and Yielding: Beyond a certain stress threshold, damage zones rapidly propagate under constant loading. Two damage zones eventually merge at the sample center, resulting in yielding. 3) Intrinsic Stress Determination: The stress at this point is intrinsic and determined by the structure of the two networks. 4) Pseudo Size-Dependency: Samples with insufficient width exhibit a pseudo size-dependency of yielding stress, as yielding occurs before reaching the critical stress. To explain the origin of intrinsic yielding stress, we introduce an intrinsic effective crack length (cI) as a measure of stress concentration around the internal crack tip of the damaged zone in the first network. Beyond this length, the influence on internal stress concentration around the damage zone is effectively screened by the load-bearing effect of the stretchable second network. The estimated cI, dependent on the microstructure of the two networks, was approximately 10 times the mesh size of the first network for typical DN gels.

Signal-distribution-based crack detection for divertor monoblock inspection using eddy current testing

This study evaluated the applicability of eddy current testing to the surface inspection of a divertor plasma-facing unit in a Tokamak fusion reactor. Artificial slits were introduced on the surfaces of some samples. The eddy current testing was performed to detect these slits. A quantitative detection evaluation method based on evaluating the distribution differences is proposed. The results reveal that the signals of slits near the edge are easily buried by the edge signals and difficult to detect directly from the signal amplitude image. With the aid of the proposed method, not only the slit in the middle of the sample surface but also the slits near the sample edges were successfully detected by eddy current testing.