Engineering reports : open access最新文献

High-strength galvanized steel wire (HSGSW), as a critical load-bearing component in bridge cable systems, is highly susceptible to environmental corrosion during long-term service, posing significant threats to the safety and durability of bridge structures. To address this issue, this study systematically investigates the corrosion behavior and mechanical performance degradation of HSGSW in the presence of N,N′-dimethylethanolamine (N,N′-DMEA), an organic corrosion inhibitor. Electrochemical accelerated corrosion tests combined with weight loss measurements were conducted to quantitatively evaluate the inhibition efficiency of N,N′-DMEA under varying concentrations and exposure durations. SEM and EDS were employed to characterize the microstructural evolution and surface chemical composition of corroded specimens. The effects of corrosion inhibition treatment on the mechanical degradation of HSGSW were further analyzed based on load–displacement curves obtained from tensile tests. The results indicate that N,N′-DMEA forms a protective adsorption film on the steel surface, significantly enhancing corrosion resistance, with an optimal inhibitor concentration of 0.08 mol·L⁻¹. As corrosion progresses, the corrosion products evolve into a dark, porous structure primarily composed of Fe, leading to the formation of localized pits and inducing stress concentration, which alters the fracture mode from a typical cup-and-cone morphology to a mixed splitting–milling fracture. Inhibitor concentrations not exceeding 0.08 mol·L⁻¹ show a positive correlation with inhibition efficiency, while increased current density results in reduced efficiency. Notably, under equivalent corrosion conditions, specimens treated with the inhibitor exhibited significantly higher ultimate tensile strength than untreated ones, with an estimated service life extension of approximately 150%. This study provides a novel technical approach for the corrosion protection of HSGSW used in bridge cables and offers valuable engineering guidance for ensuring the long-term safe operation of cable-supported bridges.

Brain tumors remain a major neurological challenge, where timely and accurate diagnosis is critical for improving patient outcomes. Although several reviews have examined machine learning (ML) and deep learning (DL) techniques for brain tumor analysis, most existing surveys either focus on a single methodological family or lack a comparative perspective across emerging computational paradigms. This review addresses that gap by providing an integrated analysis of ML, Convolutional Neural Networks (CNNs), Transformer-based models, Generative Adversarial Networks (GANs), and hybrid ensemble frameworks for tumor detection, classification, and segmentation using magnetic resonance imaging (MRI). Unlike prior reviews, we systematically evaluate the clinical applicability, dataset limitations, and reproducibility concerns of these models while identifying unresolved issues such as interpretability, data scarcity, and domain generalization. Furthermore, we synthesize trends in multimodal learning, federated frameworks, and explainable AI, offering actionable insights for translating research advances into clinical practice. This critical perspective highlights not only the state of the art but also the pathways required for developing robust, transparent, and clinically viable artificial intelligence (AI)-driven diagnostic systems.

Wire arc additive manufacturing (WAAM) has emerged as a cost-effective and scalable approach for producing large and complex metallic components. However, its industrial deployment faces persistent challenges in process stability, real-time quality assurance, and data transparency. This review provides a comprehensive analysis of the individual applications of artificial intelligence (AI) and Blockchain technologies in WAAM, emphasizing their distinct contributions and future potential for convergence. AI techniques such as artificial neural networks (ANN), support vector machines (SVM), deep learning (DL), adaptive neuro-fuzzy inference systems (ANFIS), and reinforcement learning (RL) are critically examined for their roles in process modeling, defect prediction, adaptive control, and toolpath optimization. Concurrently, Blockchain's decentralized and tamper-proof framework is analyzed for its capacity to enhance data integrity, certification, traceability, and supply chain transparency within WAAM ecosystems. A patent landscape analysis identifies AI-related and blockchain-related filings, reflecting the rapid global expansion of intelligent and secure additive manufacturing research. Despite these advancements, current studies predominantly address these technologies independently, with limited integration between intelligent decision-making and secure data management. The review highlights key research gaps, methodological constraints, and offers actionable directions toward developing hybrid AI–Blockchain frameworks tailored for autonomous, traceable, and industry-ready WAAM systems.

Accurate and timely health status assessment of power converter systems is crucial for ensuring the reliability and safety of power equipment. Conventional health assessment methods for power converters often rely on static models or fixed-weight Health Index (HI), which lack adaptability to evolving degradation patterns and fail to prioritize recent operational data, limiting prediction accuracy and timeliness. In this study, a rolling prediction framework is proposed for health status assessment of key components in power converter systems, which is built upon an adaptively weighted HI and rolling Support Vector Regression (SVR). First, the HI is constructed from multiple degradation-related features, where an inverse standard deviation weighting scheme is applied to dynamically capture the relative contribution of each feature, yielding an adaptive and interpretable HI. Then, a rolling prediction mechanism is introduced using an SVR model to characterize the nonlinear relationship between raw features and the HI. In this framework, the training set is continuously updated through a sliding time window, while exponentially decaying weights are applied to emphasize more recent data. Finally, two experiments on circuit breakers and Insulated-Gate Bipolar Transistors (IGBT) are conducted to demonstrate the effectiveness of the proposed method.

This research formulates a two-phase mathematical model to investigate the dynamics of a Maxwell dusty fluid across a linearly stretching surface embedded within a Darcy–Forchheimer porous medium, influenced by a magnetic field and varying thermal conductivity. Dusty fluid flows are significant in industries such as oil transportation, gas cleaning, and car exhaust control. The governing partial differential equations are reduced to a system of ordinary differential equations using similarity transformations and solved numerically via the bvp4c solver in MATLAB. The model's reliability is verified by comparing its results with previously published results. Parametric analysis reveals that increasing the magnetic field strength, Maxwell fluid parameter, and Forchheimer number decreases the velocities of both the fluid and dust phases, while increasing the temperature. The dusty-phase temperature is more sensitive to thermal conductivity and fluid–particle interactions. The local Nusselt number increases with thermal conductivity but drops with magnetic and Maxwell parameters, implying a lower heat transfer rate. These findings provide a deeper scientific understanding of how viscoelastic particulate flows transmit heat and momentum.

This study presents a novel approach for sandy soil stabilization through the alkali activation of recycled construction glass powder, aimed at mitigating wind erosion. The investigation commenced with a comprehensive evaluation of the physical and chemical properties of the activated glass waste, followed by laboratory tests, including wind tunnel experiments, particle size analysis, compaction, unconfined compressive strength, x-ray spectroscopy, FTIR, SEM, permeability, and vane shear tests, on samples prepared with varying sodium hydroxide (NaOH) solution as an alkaline activator, glass powder contents, and spraying rates. Results indicated that a molar concentration of 3 M containing 25 g/L of glass powder applied at 2 L/m² produced a protective layer of 7.5–8 mm, reducing wind erosion to nearly undetectable levels. Thermal assessments confirmed the stability of the geopolymerization process at temperatures up to 50°C, while enhanced mechanical performance was evidenced by increased surface shear strength and a characteristic brittle failure mode under unconfined compressive loading. These findings validate the efficacy of alkali-activated recycled glass powder as a sustainable solution for environmental management and infrastructure protection.

Unmanned aerial vehicle (UAV) roles and applications are rapidly growing, extending their variety and functionality. However, the battery dependency considerably limits the UAV's flight duration and coverage. To overcome associated challenges, a wireless power transfer (WPT) system has emerged as a viable solution, eliminating human assistance in battery depletion. Nevertheless, lateral misalignment in such systems can significantly degrade performance. In this regard, multiple-input single-output (MISO) systems have shown potential in addressing this challenge. This paper, therefore, proposes the development and modeling of a MISO WPT system with high robustness that also addresses the lateral displacement issues encountered in UAV powering. Initially, a defected ground structure-based resonator is designed with a 50-by-50 mm² area. Subsequently, two coupled resonators at 55 mm formed the WPT system. Performance validation under perfect alignment and lateral misalignment revealed the system's efficiency to reach 98% and decrease to 33% under ±25 mm shift, accordingly. The obtained results leave room to realize a MISO WPT system with two resonators composing a transmitter and a single receiver. Furthermore, the 1.25 mm isolating substrate was embedded between adjacent resonators on the transmitter to mitigate interference. The developed MISO WPT system demonstrated stable efficiency exceeding 50% under lateral misalignment.

Securing slope stability in the Nepalese Himalayas today prevents against future disasters. This research integrates numerical and experimental approaches for the analysis of the slope stability of the Ramche landslide using a combination of Back Analysis (BA) and Spectral Element Method (SEM) implemented in the open source Specfem3D_Geotech software. This study focuses on the determination of the factor of safety (FoS) under fully saturated and fully dry conditions, excluding vegetation and earthquake loading. Laboratory tests on collected soil samples provided cohesion and friction angles. The Ramche landslide is a post-failure slope geometry and soil profile data was used from such an area as a case study to develop a SEM model using AutoCAD and SW_DTM, meshing in CUBIT software and Specfem3D_Geotech tools is used for displacement analysis at different Strength Reduction Factors (SRF). A BA method was used to determine representative soil parameters under dry conditions, assuming zero cohesion and adjusting the friction angle until the SRF reached unity, yielding insitu friction angles between 33° and 36° for different sections. Displacement of the slope and SRF was visualized in Paraview, and corresponding graphs were plotted using Tecplot. Additionally, Phase2 was utilized for the validation of numerical simulation results, yielding a correlation coefficient of 99.02% for SRF and 87.04% for maximum displacement with low statistical errors (RMSE = 0.22266, MSE = 0.04957, MAE = 0.2 for SRF and RMSE = 2.73959, MSE = 7.50539, MAE = 2.16685, for displacement). This study concludes that due to the steep slope and water table fluctuation, the landslide remains highly vulnerable, particularly during the monsoon season. The results indicate that the FoS remains below one, with stability achieved only under fully dry conditions. Mitigating measures such as groundwater draw down and slope modification were found effective in improving slope stability.

Aiming at the problem of insufficient individualization and dynamic demand in higher vocational English learning, this paper proposes a dynamic learning path recommendation method based on real-time knowledge graph updates. By constructing a domain knowledge graph covering vocabulary, grammar, listening, speaking, reading, and writing skills in higher vocational English, and integrating multi-source data such as learners' answering behavior and changes in mastery, an incremental graph real-time update mechanism is designed to achieve dynamic adjustment of knowledge point relationships. Furthermore, based on the basic learning experience of higher vocational students, this system combines learner profiles with a graph neural network (GNN) model to generate personalized, explainable optimal learning paths. The system supports real-time status tracking and personalized optimization. In a pilot study with 150 students, the system achieves an update latency of only 3.44 s under a load of 1000 times per hour, achieving an 86.7% recommendation accuracy rate, a 52.0% improvement in learning efficiency after 6 weeks, and a cumulative user satisfaction score of 75%. These results demonstrate that, compared to traditional static recommendation approaches, our system offers significant improvements in real-time responsiveness, recommendation precision, and learning effectiveness, thereby providing a feasible technical solution with real-time response capabilities for intelligent English learning in higher vocational colleges.