Search Results (393)

An altermagnet, characterized by its distinctive magnetic properties, may hold potential applications in diverse fields such as magnetic materials, spintronics, data storage, and quantum computing. As a prototypical altermagnet, RuO₂ exhibits spin polarization and demonstrates the advantageous characteristics of high electrical conductivity and low thermal conductivity. These exceptional properties endow it with considerable promise in the emerging field of thermal spintronics. We studied the electronic structure and thermoelectric properties of RuO₂; the constructed RuO₂/TiO₂/RuO₂ all-antiferromagnetic tunnel junction (AFMTJ) exhibited thermally induced magnetoresistance (TIMR), reaching a maximum TIMR of 1756% at a temperature gradient of 5 K. Compared with prior studies on RuO₂-based antiferromagnetic tunnel junctions, the novelty of this work lies in the thermally induced magnetoresistance based on its superior thermoelectric properties. In parallel structures, the spin-down current dominates the transmission spectrum, whereas in antiparallel structures, the spin-up current governs the transmission spectrum, underscoring the spin-polarized thermal transport. In addition, thermoelectric efficiency emphasizes the potential of RuO₂ to link antiferromagnetic robustness with ferromagnetic spin functionality. These findings promote the development of efficient spintronic devices and spin-based storage technology for waste heat recovery and emphasize the role of spin splitting in zero-magnetization systems. Full article

Resin Transfer Molding (RTM) is a widely used composite manufacturing process where liquid resin is injected into a closed mold filled with a fibrous preform. By applying this process, large pieces with complex shapes can be produced on an industrial scale, presenting excellent properties and quality. A true physical phenomenon occurring in the RTM process, especially when using vegetable fibers, is related to the absorption of resin by the fiber during the infiltration process. The real effect is related to the slowdown in the advance of the fluid flow front, increasing the mold filling time. This phenomenon is little explored in the literature, especially for non-isothermal conditions. In this sense, this paper does a numerical study of the liquid injection process in a closed and heated mold. The proposed mathematical modeling considers the radial, three-dimensional, and transient flow, variable injection pressure, and fluid viscosity, including the effect of liquid fluid absorption by the reinforcement (fiber). Simulations were carried out using Computational Fluid Dynamic tools. The numerical results of the filling time were compared with experimental results, and a good approximation was obtained. Further, the pressure, temperature, velocity, and volumetric fraction fields, as well as the transient history of the fluid front position and injection fluid volumetric flow rate, are presented and analyzed. Full article

This study employs Particle Image Velocimetry (PIV) technology to investigate the statistical properties and flow structures of the turbulent boundary layer over smooth walls and riblet walls with yaw angles of 0, ±30° in both clear water and liquid–solid two-phase flow fields. The results indicate that, compared to the smooth wall, streamwise riblet walls and 30° divergent riblet walls can reduce the boundary layer thickness, wall friction force, comprehensive turbulence intensity, and Reynolds stress, with the divergent riblet wall being more effective. In contrast, convergent riblet walls have the opposite effect. The addition of particles leads to an increase in boundary layer thickness and a reduction in wall friction resistance, primarily by reducing turbulence fluctuations and Reynolds stress in the logarithmic region of the turbulent boundary layer. Moreover, the two types of drag-reduction riblet walls can decrease the energy content ratio of near-wall streak structures and suppress their motion in the spanwise direction. Their impact on burst events is mainly characterized by a reduction in the number of ejection events and their contribution to Reynolds shear stress. In comparison, convergent riblet walls have the complete opposite effect and also enhance the intensity of burst events. The addition of particles can fragment streak structures and suppress the intensity and number of burst events, acting similarly on drag-reduction riblet walls and further strengthening their drag reduction characteristics. Full article

Fairness is a fundamental virtue in machine learning systems, alongside with four other critical virtues: Accountability, Transparency, Ethics, and Performance (FATE + Performance). Ensuring fairness has been a central research focus, leading to the development of various mitigation strategies in the literature. These approaches can generally be categorized into three main techniques: pre-processing (modifying data before training), in-processing (incorporating fairness constraints during training), and post-processing (adjusting outputs after model training). Beyond these, an increasingly explored avenue is the direct modification of existing algorithms, aiming to embed fairness constraints into their design while preserving or even enhancing predictive performance. This paper presents a comprehensive survey of classical machine learning models that have been modified or enhanced to improve fairness concerning sensitive attributes (e.g., gender, race). We analyze these adaptations in terms of their methodological adjustments, impact on algorithmic bias and ability to maintain predictive performance comparable to the original models. Full article

Process control and quality assurance of wire + arc additive manufacturing (WAAM) and automated welding rely heavily on in-process monitoring videos to quantify variables such as melt pool geometry, location and size of droplet transfer, arc characteristics, etc. To enable feedback control based upon this information, an automatic and robust segmentation method for monitoring of videos and images is required. However, video segmentation in WAAM and welding is challenging due to constantly fluctuating arc brightness, which varies with deposition and welding configurations. Additionally, conventional computer vision algorithms based on greyscale value and gradient lack flexibility and robustness in this scenario. Deep learning offers a promising approach to WAAM video segmentation; however, the prohibitive time and cost associated with creating a well-labelled, suitably sized dataset have hindered its widespread adoption. The emergence of large computer vision models, however, has provided new solutions. In this study a semi-automatic annotation tool for WAAM videos was developed based upon the computer vision foundation model SAM and the video object tracking model XMem. The tool can enable annotation of the video frames hundreds of times faster than traditional manual annotation methods, thus making it possible to achieve rapid quantitative analysis of WAAM and welding videos with minimal user intervention. To demonstrate the effectiveness of the tool, three cases are demonstrated: online wire position closed-loop control, droplet transfer behaviour analysis, and assembling a dataset for dedicated deep learning segmentation models. This work provides a broader perspective on how to exploit large models in WAAM and weld deposits. Full article

The present study introduces an AI (Artificial Intelligence) framework for surface roughness assessment in milling operations through sound signal processing. As industrial demands escalate for in-process quality control solutions, the proposed system leverages audio data to estimate surface finish states without interrupting production. In order to address this, a novel classification approach was developed that maps audio waveform data into predictive indicators of surface quality. In particular, an experimental dataset was employed consisting of sound signals that were captured during milling procedures applying various machining conditions, where each signal was labeled with a corresponding roughness quality obtained via offline metrology. The formulated classification pipeline commences with audio acquisition, resampling, and normalization to ensure consistency across the dataset. These signals are then transformed into Mel-Frequency Cepstral Coefficients (MFCCs), which yield a compact time–frequency representation optimized for human auditory perception. Next, several AI algorithms were trained in order to classify these MFCCs into predefined surface roughness categories. Finally, the results of the work demonstrate that sound signals could contain sufficient discriminatory information enabling a reliable classification of surface finish quality. This approach not only facilitates in-process monitoring but also provides a foundation for intelligent manufacturing systems capable of real-time quality assurance. Full article

Given the exceptional capabilities of femtosecond laser processing in achieving high-precision ablation for air-film cooling hole fabrication on turbine blades, it is imperative to develop an advanced monitoring methodology that enables real-time feedback control to automatically terminate the laser upon complete penetration detection, thereby effectively preventing backside damage. To tackle this issue, a spectrum-domain coherent imaging technique has been developed. This innovative approach adapts the fundamental principle of fiber-based Michelson interferometry by integrating the air-film hole into a sample arm configuration. A broadband super-luminescent diode with a 830 nm central wavelength and a 26 nm spectral bandwidth serves as the coherence-optimized illumination source. An optimal normalized reflectivity of 0.2 is established to maintain stable interference fringe visibility throughout the drilling process. The system achieves a depth resolution of 11.7 μm through Fourier transform analysis of dynamic interference patterns. With customized optical path design specifically engineered for through-hole-drilling applications, the technique demonstrates exceptional sensitivity, maintaining detection capability even under ultralow reflectivity conditions (0.001%) at the hole bottom. Plasma generation during laser processing is investigated, with plasma density measurements providing optical thickness data for real-time compensation of depth measurement deviations. The demonstrated system represents an advancement in non-destructive in-process monitoring for high-precision laser machining applications. Full article

Optimized earth-rock allocation is key in the construction of large-scale navigation channel projects. This paper analyzes the characteristics of a large-scale navigation channel project and establishes an earth-rock allocation system in phases and categories without a transit field. Based on the physical characteristics of the earthwork and stonework used to design a differentiated transport strategy, a synergistic optimization model is built with economic and ecological benefits. As a solution, this paper proposes a sustainable earth-rock allocation optimization method that integrates the improved ant colony algorithm and firefly algorithm, and establishes a two-stage hybrid optimization framework. The application of the Pinglu Canal Project shows that ant colony optimization via division and cooperation combined with the firefly algorithm reduces the transportation cost by 0.128% compared with traditional ant colony optimization; improves the stability by 57.46% (standard deviation) and 59.09% (coefficient of variation) compared with ant colony optimization through division and cooperation; and effectively solves the problems of precocious convergence and local optimization of large-scale earth-rock allocation. It is used to successfully construct an earth-rock allocation model that takes into account the efficiency of the project and the protection of the ecological system in a dynamic environment. Full article

To address the unclear coupling mechanism between bubble detachment behavior and acoustic characteristics in gas–liquid two-phase flow, this paper systematically studied bubble behavior and acoustic characteristics under different conditions by building a high-precision synchronous measurement system, combining acoustic signal analysis and bubble dynamics observation. The influence mechanism of liquid surface tension, dynamic viscosity, and orifice diameter on the critical gas flow velocity of bubble flow transition was analyzed, and a flow pattern classification criterion system was established. The experimental results showed that the bubble flow state could be divided into three states according to the characteristics of the acoustic signals: discrete bubble flow, single-chain bubble flow, and dual-stage chain bubble flow. The liquid surface tension and dynamic viscosity had no significant effect on the critical gas flow velocity of the transition from discrete bubble flow to single-chain bubble flow, but significantly increased the critical gas flow velocity of the transition from single-chain bubble flow to dual-stage chain bubble flow. The increase in the orifice diameter reduced the critical gas flow velocity of the two types of flow transition. In addition, the Weber number (We) and Galileo number (Ga) were introduced to construct a quantitative classification system of flow pattern, which provided theoretical support for the optimization of industrial gas–liquid two-phase flow. Full article

Enhancing energy efficiency is essential for proton exchange membrane fuel cells (PEMFCs) operating in a dead-ended anode (DEA) mode. This study proposes the integration of a buffer tank, positioned between the mass flow meter and the fuel cell, to reduce hydrogen loss during purge events. The buffer tank stores hydrogen when the purge valve is closed and releases it when the valve opens, thereby stabilizing anode pressure, minimizing hydrogen waste, and improving overall system efficiency. The effectiveness of the buffer tank is experimentally evaluated under varying load currents, hydrogen supply pressures, purge intervals, and purge durations. The objective is to determine the optimal purge duration that maximizes energy efficiency, both with and without the buffer tank. The results show that the buffer tank consistently improves energy efficiency. Under optimal conditions (0.1 bar, 8 A, 0.1 s purge duration, and 20 s purge interval), efficiency increases by 3.3%. Under non-optimal conditions (0.1 bar, 1 A, 0.1 s purge duration, and 20 s interval), the improvement reaches 71.9%, demonstrating the buffer tank’s effectiveness in stabilizing performance across a wide range of operating conditions. Full article

Nanomaterials with large specific surface area (SSA) have emerged as pivotal platforms for energy storage and environmental remediation, primarily due to their enhanced active site exposure, improved mass transport capabilities, and superior interfacial reactivity. Among them, polymeric carbon nitride (g-C₃N₄) has garnered significant attention in energy and environmental applications owing to its visible-light-responsive bandgap (~2.7 eV), exceptional thermal/chemical stability, and earth-abundant composition. However, the practical performance of g-C₃N₄ is fundamentally constrained by intrinsic limitations, including its inherently low SSA (<20 m²/g via conventional thermal polymerization), rapid recombination of photogenerated carriers, and inefficient charge transfer kinetics. Notably, the theoretical SSA of g-C₃N₄ reaches 2500 m²/g, yet achieving this value remains challenging due to strong interlayer van der Waals interactions and structural collapse during synthesis. Recent advances demonstrate that state-of-the-art strategies can elevate its SSA to 50–200 m²/g. To break this surface area barrier, advanced strategies achieve SSA enhancement through three primary pathways: pre-treatment (molecular and supramolecular precursor design), in process (templating and controlled polycondensation), and post-processing (chemical exfoliation and defect engineering). This review systematically examines controllable synthesis methodologies for high-SSA g-C₃N₄, analyzing how SSA amplification intrinsically modulates band structures, extends carrier lifetimes, and boosts catalytic efficiencies. Future research should prioritize synergistic multi-stage engineering to approach the theoretical SSA limit (2500 m²/g) while preserving robust optoelectronic properties. Full article

Video stabilization is a critical technology for enhancing video quality by eliminating or reducing image instability caused by camera shake, thereby improving the visual viewing experience. It has deeply integrated into diverse applications—including handheld recording, UAV aerial photography, and vehicle-mounted surveillance. Propelled by advances in deep learning, data-driven stabilization methods have emerged as prominent solutions, demonstrating superior efficacy in handling jitter while achieving enhanced processing efficiency. This review systematically examines the field of video stabilization. First, this paper delineates the paradigm shift from classical to deep learning-based approaches. Subsequently, it elucidates conventional digital stabilization frameworks and their deep learning counterparts along with establishing standardized assessment metrics and benchmark datasets for comparative analysis. Finally, this review addresses critical challenges such as robustness limitations in complex motion scenarios and latency constraints in real-time processing. By integrating interdisciplinary perspectives, this work provides scholars with academically rigorous and practically relevant insights to advance video stabilization research. Full article

Background: Pharmaceutical compounding remains a predominantly manual process with limited innovation, particularly in non-sterile applications. This study explores the implementation of an automated compounding platform based on 3D printing to enhance precision, efficiency, and adaptability in pediatric corticosteroid formulations. Methods: Personalized hydrocortisone dosage forms were prepared in a hospital pharmacy setting using a proprietary excipient base and standardized procedures, including automated dosing and syringe heating when required. Three dosage forms—3.2 mg gel tablets, 2.8 mg water-free troches, and 1.2 mg orodispersible films (ODFs)—were selected to demonstrate the platform’s versatility and to address pediatric needs for varying strengths and dosage types. All products were prepared using a reproducible semi-solid extrusion (SSE)-based workflow with the consistent API-excipient blending and automated deposition. Results: Analytical testing confirmed that all formulations met pharmacopeial criteria for mass and content uniformity. The ODF and troche forms achieved rapid drug release, exceeding 75% within 5 min, while the gel tablet showed a slower release profile, reaching 86% by 60 min. Additionally, in-process homogeneity testing across syringe printing cycles confirmed the consistent API distribution. Conclusions: The results support the feasibility of integrating automated compounding technologies into pharmacy workflows. Such systems can improve accuracy, minimize variability, and streamline the production of customized pediatric medications, particularly for drugs with poor palatability or narrow therapeutic windows. Overall, this study highlights the potential of automation to modernize non-sterile compounding, and to better support individualized therapy. Full article

This research presents an essential oil (EO) distillation method with improved efficiency, called recurrent water and steam distillation (RWASD), as well as the testing of a 500 L prototype essential oil distillation machine (500 L PDM). The raw material used was 100 kg of lime fruit. At each distillation time point, the test result was compared with that obtained via water and steam distillation (WASD), and different raw material grid configurations were taken into consideration. It was found that distillation using the RWASD method increased the amount of EO obtained from limes by 53.69 ± 2.68% (or 43.21 ± 2.16 mL) compared with WASD. The results of gas chromatography mass spectrometry (GC-MS) analysis of bioactive compounds from the distilled EO revealed that important compounds were present in amounts close to the standards reported in many studies; namely, β-myrcene (2.72%), limonene (20.72%), α-phellandrene (1.27%), and terpinen-4-ol (3.04%). In addition, it was found that the temperature, state of saturated steam, and heat distribution during distillation were relatively constant. The results showed the design, construction, and heat loss error values of the 500 L PDM were 5.90 ± 0.29% and 7.83 ± 0.39%, respectively, leading to the use and percentage of useful heat energy to stabilize at 29,880 ± 1,494 kJ/s and 22.47 ± 1.12%, respectively. Additionally, the shape of the grid containing the raw material affects the temperature distribution and the amount of EO distilled, with values 10.14 ± 0.51% and 8.07 ± 0.40% higher for the normal grid (NS), respectively, as well as an exergy efficiency of 49.97 ± 2.49%. The highest values found for exergy in, exergy out, and exergy loss were 294.29 ± 14.71 kJ/s, 144.76 ± 7.23 kJ/s, and 150.22 ± 7.51 kJ/s, respectively. The obtained results can be further developed and expanded to promote the application of this method in SMEs, serving as basic information for the development of the EO distillation industry. Full article

Additive manufacturing (AM), or 3D printing, enables efficient fabrication of complex and customized components. Despite its growth across industries, users frequently encounter print failures due to design errors, process limitations, and inadequate monitoring. While existing research has explored various aspects of these failures, much of it remains fragmented, with limited consolidated overviews that map common problems, troubleshooting strategies, and guidelines across the AM workflow. This study conducted a systematic literature review using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology to identify and categorize common 3D printing problems and their solutions. Relevant studies published between 2000 and 2024 were extracted from major databases. A total of 126 peer-reviewed articles were selected and analyzed. Three major categories of recurring challenges were identified: (1) design and pre-processing errors; (2) geometric errors and dimensional deviations; (3) failures in in-process error detection and response. A variety of mitigation strategies have been proposed across the literature, including STL and slicing optimization, thermal management, machine calibration, and sensor-based real-time monitoring. These approaches reflect the multifactorial nature of 3D printing failures, which often arise from the complex interplay of design, material, and process parameters. This review provides a structured summary of failure types and mitigation strategies across the AM workflow. Full article