Search Results (47,454)

Cold recycled mixtures with bitumen emulsion (CRME) serving in seasonally frozen regions are susceptible to mechanical deterioration under repeated freeze-thaw (F-T) cycles, which is primarily manifested as interfacial damage and crack propagation. However, the micro-mechanical processes associated with the transmission and dissipation of frost-heaving stresses induced by water-ice phase transition within the interfacial transition zone (ITZ) between reclaimed asphalt pavement (RAP) and asphalt mortar remain to be further characterized. In this study, a numerical simulation approach coupling frost heave effects with the phase transition of water-ice particles was developed based on X-ray computed tomography (CT) and the discrete element method (DEM), and the micro-mechanical parameters of the RAP-asphalt mortar ITZ were determined through laboratory experiments. Combined with acoustic emission (AE) monitoring, the damage evolution characteristics of cold recycled mixtures and the associated interfacial damage mechanisms under freeze-thaw action were systematically investigated. The results indicate that the optimal micro-parameters of the RAP-asphalt mortar ITZ can be taken as approximately 85% of those of virgin asphalt mortar. After 20 freeze-thaw cycles, the number of shear cracks and tensile cracks in ITZ on RAP surface reached 493 and 92, respectively, which were much higher than 11 and five on the surface of new aggregate. ITZ was the main control weak area of freeze-thaw damage. Compared with the unfrozen specimens, the minimum effective contact number of mortar decreased by 1.63%, 4.52% and 8.52% respectively after 5, 10 and 20 freeze-thaw cycles, and the total effective contact number decreased from 75,842 to 69,383. Freeze-thaw cycles significantly reduce the strain energy storage capacity of CRME: the maximum energy storage capacity of the adhesive spring decreased from 2.15 J in the non-freeze-thaw state to 1.28 J in 10 cycles (a decrease of 40.47%) and 1.16 J in 20 cycles (a decrease of 46.05%), and the damage mode changed from brittle fracture to interface-controlled energy dissipation. The proposed water-ice phase transition-based DEM framework provides a reliable numerical tool for investigating freeze-thaw damage mechanisms and supporting durability-oriented design of cold recycled pavement materials. Full article

This study presents a full-scale steady CFD methodology for evaluating flow quality in a large-scale low-speed wind tunnel (8 m × 6 m test section, 130 m/s). The tunnel geometry is resolved at 1:1 scale, the damping screens and honeycomb are modeled as porous media, and results are validated against wind tunnel test data and the GJB 1179A-2012 acceptance criteria. The baseline configuration reproduces the axial static pressure gradient within the acceptance criterion but substantially overpredicts turbulence intensity, dynamic pressure coefficient, and both velocity direction deviation angles. Damping screens, modeled as porous jumps, provide the dominant correction, bringing all metrics within the acceptance limits. Adding honeycomb yields incremental improvement: porous zone modeling preserves or improves all metrics, whereas a porous jump representation pushes velocity direction deviations beyond the limit. Between RNG k-ε and SST k-ω, only turbulence intensity is closure-dependent, with RNG k-ε closer to experiment. At Ma ≈ 0.38, compressibility does not alter the flow quality assessment, confirming that incompressible assumption is sufficient. By replacing costly physical trials with a validated CFD workflow, these findings provide a practical, resource-efficient reference for the CFD-based evaluation, design, and retrofit of wind tunnel infrastructure that underpins renewable-energy and energy-efficiency research. Full article

Manual judgment in coal-slurry flotation suffers from dynamic hysteresis and sensitivity to illumination and dust. In contrast, froth images characterize macroscopic flotation appearances, while tailings ash content reflects the internal pulp quality, and the two sources provide complementary information. This paper proposes a flotation condition recognition method based on the fusion of froth images and tailings ash content data. Visualization via violin plots shows the high correlation between ash content and flotation states. In this framework, froth images are designated as primary inputs and ash content as auxiliary signals, with an optimal fusion-weight ratio determined through ablation experiments. Multimodal models based on ResNet18, MobileNetV2, and MobileNetV3 are compared, and image-only and tailings-ash-only models are used to assess the value of multimodal fusion. Experiments demonstrate that Multimodal MobileNetV3 achieves an accuracy of 94.65%, outperforming Multimodal MobileNetV2 at 87.98%, Multimodal ResNet18 at 86.13%, the image-only model at 86.43%, and the tailings-ash-only model at 81.63%. t-SNE visualization indicates that multimodal fusion enhances feature discrimination and clustering. This method effectively improves classification accuracy and provides a new approach for intelligent working condition recognition in coal preparation plants. Full article

Federated learning (FL) has emerged as an important distributed machine learning paradigm allowing many users to train a model together without sharing raw data. However, FL’s decentralized design makes it very susceptible to both Byzantine and coordinated model poisoning attacks, where a few malicious rounds of gradients are strategically inserted to reduce the overall integrity of the model. Even the existing Byzantine-resilient aggregation methods such as geometric median, Krum, and trimmed mean are vulnerable to sophisticated, coordinated poisoning attacks that leverage statistical gaps in outlier detection. In this article, we present FedSentinel, a novel Byzantine-resilient federated learning framework that combines cryptographic gradient attestation with adaptive trust-weighted aggregation to protect against coordinated model-poisoning attacks, which are among the most serious challenges. Three key innovations are introduced in FedSentinel Cryptographic Gradient Attestation Protocol (CGAP) that utilizes commitment schemes and zero-knowledge range proofs to ensure the integrity of the gradients and verify that submitted updates satisfy predefined norm and direction constraints; in the current protocol configuration, CGAP provides gradient integrity verification rather than full gradient confidentiality from the server Dynamic Trust-Weighted Robust Aggregation (DT-RoA): the computation and updating of per-client trust scores based on historical gradient consistency and cross-validation signals; and Coordinated Attack Detection Engine (CADE): based on spectral analysis of the gradient covariance matrices, it detects and isolates colluding Byzantine clients. By participating in up to 100 clients on CIFAR-10, CIFAR-100, FEMNIST, and Sentiment140 datasets, extensive experiments conducted under up to 30% Byzantine adversaries reveal that FedSentinel achieves 91.36% average global accuracy, whereas state-of-the-art defenses such as FLTrust, FLAME, RoFL, ShieldFL, and DnC achieve 83.83–86.94%. FedSentinel outperforms these defenses by 4.42–7.53% in terms of accuracy, while decreasing attack success rates by 53.2% under coordinated Byzantine backdoor attacks. The proposed framework offers a promising approach for federated learning that is verifiable and trustworthy in adversarial environments. Full article

Background/Objectives: Veterinary vaccine manufacturing is essential for controlling transboundary animal diseases, strengthening animal health systems, and supporting livestock development, particularly in Southeast Asia where livestock production continues to expand. However, modernization toward Pharmaceutical Inspection Co-operation Scheme (PIC/S) Good Manufacturing Practice (GMP) compliance remains challenging because it requires coordinated development across regulatory, technical, organizational, financial, and governance domains. This study aimed to develop an evidence-informed strategic framework for veterinary vaccine manufacturing modernization through a structured and systematically documented framework development methodology integrating multidisciplinary evidence and practice-informed implementation experience. Methods: An evidence-informed framework development study was conducted through the multidisciplinary integration of multiple evidence sources, including international regulatory guidance, peer-reviewed scientific literature, multidisciplinary feasibility assessment, and practice-informed implementation experience derived from a government-commissioned consultancy project. Framework development followed a structured methodology comprising evidence collection, evidence categorization, multidisciplinary evidence synthesis, framework development, iterative refinement, and expert-informed validation. Only generalized methodological principles and potentially transferable implementation knowledge were incorporated to ensure confidentiality of project-specific information. Results: Multidisciplinary evidence synthesis identified ten evidence domains encompassing regulatory, technical, organizational, economic, legal, environmental, governance, workforce, infrastructure, and implementation considerations. These domains were translated through a structured process into an evidence-informed strategic framework organized around four sequential modernization phases—Design, Construction, Validation, and Continuous Improvement—supported by cross-cutting strategic enablers. Iterative refinement through expert review, technical consultation, and implementation verification strengthened the framework’s scientific consistency, regulatory alignment, and practical applicability. Conclusions: Veterinary vaccine manufacturing modernization should be regarded as a multidisciplinary organizational transformation rather than solely as an infrastructure or engineering project. Beyond proposing a strategic framework, this study provides a transparent, systematically documented and traceable methodology for translating multidisciplinary evidence into strategic decision-support tools. The proposed framework may assist governments, regulatory authorities, vaccine manufacturers, and researchers in developing evidence-informed modernization strategies for veterinary biological manufacturing systems, particularly in low- and middle-income countries, subject to adaptation to specific regulatory, technical, institutional, and resource contexts. Full article

Accurate detection of sub-millimeter defects in reactor core-plate cotter-pin holes is essential for nuclear safety. However, underwater inspection images often suffer from low signal-to-noise ratios, weak boundary responses, and pseudo-edge interference, resulting in unstable localization of defects. Existing deformable and attention-based detectors remain vulnerable to sampling drift and semantic–boundary inconsistency under such conditions. To address these challenges, an Edge-Geometry-Guided Deformable Detection Network (EGD-Net) is proposed for underwater defect detection. EGD-Net introduces an edge-geometry-constrained deformable sampling mechanism that embeds edge-confidence priors into deformable convolution to improve boundary-aware feature sampling. A cross-level semantic–geometric alignment strategy is designed to enhance the interaction between defect semantics and geometric boundary cues, while a top-down feedback recalibration mechanism improves multi-scale response consistency for weak defects. Experiments on the Core-Plate Pin-Hole Defect (CPHD) dataset demonstrate that EGD-Net achieves the highest AP@[0.5:0.95] on both datasets while maintaining competitive or superior Precision, Recall, and F1-score while reducing engineering center error under a fixed operating point. Performance across the two complementary domains suggests its robustness to variations between coupon images and practical underwater inspection scenes. These results indicate that EGD-Net provides a reliable solution for boundary-sensitive localization of underwater sub-millimeter defects in nuclear inspection. Full article

Shale gas is a major unconventional energy resource in China. Its low porosity and permeability require large-scale volumetric fracturing to create conductive fracture networks. However, most induced fractures are propped discontinuously because shale reservoirs are geometrically complex. Fracture conductivity and proppant placement efficiency therefore directly control stimulation performance. Following SY/T 6302-2009, this study used linear flow-through experiments and a large-scale visual fracture simulation system to investigate the effects of proppant particle-size distribution, injection sequence, flow rate, and closure pressure on fracture conductivity and placement. The results show that the 20/40:40/70 mesh dual-particle-size combination at a 3:2 ratio provides the best overall performance. A fine-particle content of no more than 16.7% limits conductivity loss and improves the match between particle size and fracture aperture. Multilayer placement at fracture corners distributes high-stress loading and maintains conductivity. Injecting 70–140 mesh fine proppant before 40–70 mesh coarse proppant at 3.6 m³/h improves transport distance, coverage, and placement uniformity. The optimized scheme maintains stable conductivity at closure stresses of 10–80 MPa and achieves at least 95% propped-area coverage. These findings provide experimentally supported parameters for discontinuous propping and can inform shale gas fracturing design. Full article

Reliable inspection of insulator defects from unmanned aerial vehicle (UAV) remote sensing imagery is essential for the safe operation of power transmission systems. However, the task remains challenging due to fine-grained defect patterns, thin structures, large-scale variations, and complex backgrounds in aerial scenes. To address these issues, we propose Insulator-DETR, an end-to-end detection transformer specifically designed for UAV-based insulator defect inspection. The proposed framework preserves the set-prediction paradigm of DETR while introducing task-oriented modifications. In the encoder, a Parallel Attention MLP integrates multi-scale spatial perception and phase-aware token interaction to enhance the representation of fine textures and structural details. In the decoder, a Masked Linear Attention mechanism combines efficient global modeling with local contextual aggregation, enabling accurate localization of irregular defects. Furthermore, a convolutional feed-forward design is adopted to strengthen spatial interactions. Extensive experiments on two newly annotated UAV insulator-defect datasets and a public benchmarks demonstrate that Insulator-DETR consistently outperforms state-of-the-art detectors in both detection accuracy and recall. Full article

As a key pixel-level analysis technology, semantic segmentation is widely deployed in autonomous driving and medical imaging. Fully supervised segmentation relies on labour-intensive pixel-wise annotations, so weakly supervised semantic segmentation (WSSS) with only image-level labels has attracted wide attention. Existing Vision Transformer (ViT)-based WSSS methods suffer from two critical limitations: ViT’s global self-attention mechanism leads to insensitivity to local small target features and incomplete foreground activation; its class-agnostic attention maps frequently misactivate background regions as foreground objects, introducing heavy noise. To tackle these two issues, this paper proposes a single-stage weakly supervised segmentation algorithm based on local–global class labelling comparison. First, we design a local–global class labelling comparison (LTG) module. By feeding both original images and randomly cropped local patches into ViT, we adopt InfoNCE contrastive loss to align local class tokens with global class tokens, enhancing the feature integrity of local target regions and suppressing background false activation. Second, a class-aware stimulus module (CSM) is embedded into ViT’s multi-head attention branch. It injects category semantic constraints into self-attention to generate class-aware attention maps, guiding the model to focus on real foreground targets and reduce background interference. Finally, we construct a feature fusion class-aware activation map (FFCAM) by fusing ViT global output features and CSM class-aware attention features to generate high-quality pseudo-labels for segmentation training. Extensive experiments are conducted on PASCAL VOC 2012 and MS COCO 2014 datasets. Our method achieves 78.5% mIoU on the validation set of PASCAL VOC 2012 and 50.9% mIoU on MS COCO 2014, showing competitive numerical performance among the compared single-stage ViT-based WSSS approaches. Ablation experiments verify the independent and joint effectiveness of LTG, CSM and FFCAM. The proposed method effectively improves the completeness of target activation regions and suppresses background noise, and it maintains strong generalization for slender, small and texture-sparse objects. In future work, we will further lightweight the ViT backbone to reduce computational overhead for embedded deployment. Full article

This study aims to systematically review the current applications and development trends of virtual reality (VR) technology in blue space research. It clarifies the main research methods, technical pathways, application scenarios, and existing knowledge gaps, and integrates the advantages of VR technology with the characteristics of blue spaces to provide references for future urban renewal and urban management research. Following the PRISMA-ScR guidelines, this study searched literature published up to January 2026 in the Web of Science and Scopus databases. A total of 41 original studies were included. A scoping review approach was adopted, employing descriptive statistics, cross-analysis, and classification methods. Existing studies mainly focus on blue space types such as rivers and lakes, with immersive head-mounted displays as the primary method. Most experiments adopt single session, short term exposure designs, suggesting potential positive effects of virtual blue spaces in reducing stress, improving mood, and restoring attention. Other studies demonstrate the application of VR technology in environmental education, design assessment, and urban management within blue spaces. As an indirect exposure and supplementary intervention tool, VR shows significant potential in blue space research, particularly in supporting vulnerable populations, large-scale urban planning and design assessment, and risk management. However, current research still faces several limitations, including insufficient multisensory integration, limited capability in simulating dynamic water environments, lack of methodological standardization, uneven sample distribution, absence of longitudinal studies, and weak interactive and social dimensions. The findings highlight the need for further research on these specific aspects. Full article

This paper presents a speech emotion recognition approach and its implementation on a field-programmable gate array (FPGA) with efficient neural network inference. The proposed approach employs a hybrid neural network architecture that integrates convolutional neural networks (CNNs), long short-term memory (LSTM) networks, and fully connected neural networks (FCNNs) to effectively capture both local and temporal features from speech signals. This architecture is well suited for modeling time-series data and is applied to speech signals in this paper. For FPGA implementation, one-dimensional CNNs (1D CNNs) are employed to extract features from Mel-frequency cepstral coefficients (MFCCs), thereby reducing computational complexity and improving the efficiency of CNN-based processing. The experiments are conducted using 10-fold cross-validation, achieving average recognition rates of 96.12%, 97.79%, 96.91%, 94.01%, and 98.09% on the RAVDESS, EMO-DB, IEMOCAP, BAUM-1s, and eNTERFACE’05 databases, respectively. According to the recognition accuracy, the proposed method performs better than existing related studies. The emotion recognition model is then deployed on an FPGA using the proposed inference algorithms, and the accuracies obtained on the FPGA are identical to those achieved on the PC. These results verify the successful deployment of the designed model on the FPGA platform. Full article

Background/Objectives: Health literacy is a critical determinant of health, closely linked to behaviours, outcomes, and broader social determinants. Given that young people navigate distinct social contexts and health-related challenges compared to adults, understanding how to support their health literacy development during this important period is essential for lifelong wellbeing. This study aimed to explore the perceptions of academics and community-based practitioners regarding the concept of health literacy in young people and the importance of supporting its development. Methods: Semi-structured interviews were conducted with 30 participants (academics n = 17, health and wellbeing practitioners n = 9, policy/research n = 4) based across Europe and Australasia, with experience in youth health literacy. Data were analysed using reflexive thematic analysis. Results: Three key themes were identified: (i) Building lifelong health agency: From early foundations to transitional autonomy; (ii) Navigating novel health changes; and (iii) Structural inequities and intergenerational barriers in health literacy. Conclusions: Participants emphasised that youth health literacy interventions should move beyond didactic, individual-level information delivery toward fostering critical digital competencies and addressing structural barriers through participatory co-design with young people. Full article

Additive manufacturing (AM) technologies are increasingly being integrated into work environments, making the coexistence of production and office activities within hybrid workplaces more common and raising concerns regarding indoor air quality (IAQ) in shared or adjacent workplace environments. Although emissions associated with AM processes have been examined, limited attention has been given to IAQ conditions in hybrid workplaces where production and office functions share the same indoor environment. This study investigates the IAQ conditions associated with selected powder bed fusion and vat photopolymerization operations in a hybrid workplace through real-time environmental monitoring and employee perception analysis. Field experiments were conducted during pre-printing, printing, post-printing, and post-processing phases. Two custom-designed instruments, the Indoor Network-Based Air-Sensing Ecosystem (INBASE) and the Indoor Air Quality Polling Instrument (AIRPOLL), were developed to monitor particulate matter, volatile organic compounds, formaldehyde, benzene, ethanol, nitrogen dioxide, carbon dioxide, and oxygen and collect employee feedback. Findings showed that pollutant concentrations varied by process type, operational phase, and monitored space. Higher concentrations, particularly of volatile organic compounds, were observed in production spaces, while most other parameters remained within reference values. Employee feedback indicated positive perceptions in office spaces and greater concerns in production spaces. The findings support the importance of ventilation, spatial separation, and continuous monitoring in AM-integrated hybrid workplaces. Full article

This article traces how nontraditional preservice teachers experience the mediation of academic research writing, and how the transition from mediation to independence unfolds in their accounts when the supporting structure is withdrawn. The case centers on the first research seminar of a teacher education program in which 13 experienced paraprofessionals, most of them the first in their families to enter higher education and carrying long histories of academic difficulty, design and write a structured self-regulation intervention study with a pupil. From a qualitative analysis of three retrospective focus groups, five themes were constructed: (1) arriving with academic apprehension; (2) holding the standard as recognition; (3) the graduated scaffold of the first seminar; (4) concentric circles of mediation and support; and (5) the abrupt fading of the scaffold in the move toward independence. The findings highlight that participants valued scaffolding not only for the support it provided but for the way responsibility for the work was gradually handed over. Despite this gradual structure, the handover toward independence was experienced by some students as an abrupt, sudden transition, which provoked ambivalent feelings. The discussion proposes two participant-grounded sensitizing concepts, the effort ladder and the concentric circles of mediation, as interpretive lenses for scaffolding theory. We discuss the unresolved tensions that participants hold between the importance of high academic standards and the need for accommodations, and the implications for teacher education. Full article

Ultra-fine zirconia-based powder used in thermal barrier coatings is a core raw material for high-performance coatings, but its small particle size causes clogging during discharge. The discrete element method (DEM) can simulate powder flow and is an effective numerical tool to study clogging, yet discrete element parameters of this powder are lacking. In this work, irregular particles were simplified as soft spheres and scaled from D50 = 1.2 µm to D50 = 240 µm using an improved scaling scheme derived from classical particle scaling and similarity principles. The Hertz-Mindlin with Johnson-Kendall-Roberts Version 2 (JKR V2) contact model in EDEM was adopted, with the angle of repose as the calibration response. The two-level fractional factorial design showed that the particle-particle coefficient of static friction had the largest absolute main effect, followed by JKR surface energy and the particle-stainless steel coefficient of static friction. Steepest ascent tests determined optimal ranges: 0.7–0.9 for particle-particle static friction and 0.17–0.27 for JKR surface energy; the particle-stainless steel friction was fixed at its median of 0.45. An orthogonal test combined with linear interpolation yielded the calibrated parameter combination: 0.8 and 0.19, respectively. The simulated angle of repose was 53.87°, which lies within the experimentally measured range of 50–56° and differs by 0.37% from the experimental mean of 53.67°, showing good agreement between the simulation and experiment under the present calibration condition. The calibrated parameters may provide reference values for DEM simulations of similar static or quasi-static conditions; however, their applicability to other granular flow conditions, including discharge, requires further independent validation. Full article