Hasil untuk "Mechanical engineering and machinery"

Quan Qian, Yi Qin, Yi Wang et al.

Abstract Deep learning has gained a great achievement in the intelligent fault diagnosis of rotating machineries. However, the labeled data is scarce in actual engineering and the marginal distribution of data is discrepant under different conditions. Transfer learning provides a feasible way to overcome these difficulties. Considering the effect of noise on the transfer fault diagnosis, this work puts forward a new deep transfer learning network based on convolutional auto-encoder(CAE-DTLN) to implement the mechanical fault diagnosis in target domain without labeled data. In the proposed framework, CAE is used as the feature extractor as it has the ability of noise removal. Moreover, both CORrelation ALignment (CORAL) loss and domain classification loss are integrated to enhance the effect of domain confusion. The proposed model is applied to the fault transfer diagnosis of planetary gearboxes under different working loads and noise levels, and it is compared with other typical fault transfer diagnosis models. The experimental results show that CAE-DTLN has higher diagnosis accuracy and stronger generalization ability. The average diagnostic accuracy of CAE-DTLN is over 99%. Moreover, the proposed transfer learning model has better anti-noise performance.

Dong Wang, Lubo Zhou, Liyun Xie et al.

IntroductionHigh-voltage circuit breakers are core control and protection equipment in power systems, and their operational status directly affects device stability and power grid security. Improving the accuracy of their fault detection is a key demand for the operation and maintenance of power equipment.MethodsThis study proposes a fault detection method for high-voltage circuit breakers based on multi-source information and motion analysis. First, a 1-dimensional recurrent neural network (1DRNN) is used to analyze voiceprint and current signals to extract feature data related to the mechanical state of the operating mechanism. Second, a physics-informed transfer learning network model consisting of a Common Feature Learning Network (CFLN) and a Mechanical Feature Learning Network (MFLN) is constructed to explore shared features between multi-source signals and mechanical parameters and extract specific features of individual mechanical parameters in a targeted manner. Meanwhile, a multi-head attention mechanism is integrated to enhance the model’s ability to capture key features, and a physics-based loss function is designed to improve the physical consistency of the model during mechanical parameter identification.ResultsExperimental verification shows that the proposed method achieves a fault diagnosis accuracy of over 93% for high-voltage circuit breakers, and the model can still maintain high diagnostic stability and detection accuracy under noise interference conditions.DiscussionThrough the design of deep fusion of multi-source signals and embedding of physical information, this method makes up for the information defects of single-signal diagnosis, solves the problem of lack of physical consistency in data-driven models, and improves the environmental adaptability of fault diagnosis models, providing a practical technical solution for the intelligent fault diagnosis of high-voltage circuit breakers.

Ömer Elri, Serkan Savaş

Manual notes and scattered messaging applications used in managing business processes compromise data integrity and abstract project tracking. In this study, an integrated system that works simultaneously on web and mobile platforms has been developed to enable individual users and teams to manage their workflows with concrete data. The system architecture integrates MongoDB, which stores data in JSON format, Node.js Express.js on the server side, React.js on the web interface, and React Native technologies on the mobile side. The system interface is designed around visual dashboards that track the status of tasks (To Do-In Progress-Done). The urgency of tasks is distinguished by color-coded labels, and dynamic graphics (Dashboard) have been created for managers to monitor team performance. The usability of the system was tested with a heterogeneous group of 10 people consisting of engineers, engineering students, public employees, branch managers, and healthcare personnel. In analyses conducted using a 5-point Likert scale, the organizational efficiency provided by the system compared to traditional methods was rated 4.90, while the visual dashboards achieved a perfect score of 5.00 with zero variance. Additionally, the ease of interface use was rated 4.65, and overall user satisfaction was calculated as 4.60. The findings show that the developed system simplifies complex work processes and provides a traceable digital working environment for Small and Medium-sized Enterprises and project teams.

Manambedu Vijayakumar Raja, Himanshu Thaker, Subba rao Katragadda et al.

The shift towards Industry 5.0 from Industry 4.0 represents the paradigm shift in industry, not only highlighting automation and efficiency but also human-centered innovation, resilience, and sustainability. Central to this transformation is the synergy between Artificial Intelligence (AI) and Data Science with mechanical automation to produce intelligent, adaptive, and collaborative industrial environments. This review identifies the new frontier of human-centered AI in Industry 5.0 as the intersection of data-driven intelligence, mechanical engineering, and human-robot collaboration (HRC). It methodically examines how models of AI/Machine Learning (ML), such as explainable AI (XAI), prediction analytics, and systems of human-in-the-loop are redefining mechanical automation into cognitive, user-oriented settings. A systematic methodology based on major scientific databases was employed in order to choose more than 150 high-impact articles published between the years 2015 and 2025. Fundamental enabling technologies like collaborative robots (cobots), digital twins, cyber-physical systems, and edge AI are discussed in detail, with special emphasis on how they facilitate ergonomic, transparent, and secure interaction between humans and machines. In addition, the review discusses how data science frameworks are implemented to maximize the performance, trust, and well-being of humans in automated machinery systems. The paper also identifies some key missing gaps, such as the absence of scalable explainability of industrial AI, poor integration of ergonomic models with robotics, and difficulty in implementing real-time feedback systems from humans. In overcoming these challenges, this review provides a research and development pathway towards ethically oriented, resilient, and inclusive production. The research is expected to be used as a basis of reference by academics, engineers, and policy-makers who are leading the humanity-oriented shift of smart production systems.

Peter Trusov, Pavel Gladkikh

The most important element of mathematical models of thermomechanical processing of metals and alloys is the constitutive model. In recent decades, multilevel physically-oriented constitutive models (CMs) have found widespread application. The first two-level model was the rigid-plastic theory of J. Taylor,a rigorous mathematical justification of which was developed by J. Bishop and R. Hill (TBH type models). The main disadvantage of this model is the uncertainty of the choice of active slip systems when more than 5 systems are activated. Despite this, the TBH models have become widespread, and its basic provisions have been preserved in many later developments. It seems that limiting the number of active slip systems to 5 has no physical justification and is determined only by the numerical procedure for implementing the model. Since the 1970s, elastic-viscoplastic models have emerged; it has been shown that as the velocity sensitivity parameter tends to zero, the macroparameters determined in the modeling converge to a solution using an elastic-plastic model. However, the system of equations becomes rigid, requiring the use of implicit schemes and extremely small time steps, which significantly reduces the computational efficiency. The paper proposes a modification of elastic-plastic model of the TBH type, in which a procedure for overcoming the above-mentioned drawback is proposed. To compare the computational efficiency of the elastic-plastic and elastic-viscoplastic models, a series of numerical experiments was carried out.

Mengyu Rong, Xianfang Tan, Na Gao et al.

Aqueous zinc-based energy storage systems offer high theoretical specific capacity, low cost, intrinsic safety, and environmental compatibility, positioning them as promising candidates for next-generation energy storage and conversion technologies. However, issues such as zinc dendrite growth, hydrogen evolution reaction (HER), and surface passivation hinder their practical deployment. To address these challenges, a hollow nanotubular magnesium silicate (denoted MgSi) interfacial layer was constructed on the zinc metal anode (Zn@MgSi). The unique layer structure and negatively charged surface of MgSi facilitate the desolvation of [Zn(H2O)6]2+ by stripping water molecules, while temporarily immobilizing Zn2+ to suppress random diffusion. The combined effects of the electric field-guided Zn2+ distribution and rapid ion transport through the layer structure co-regulate Zn2+ flux, leading to uniform, dendrite-free zinc deposition. Consequently, the Zn@MgSi symmetric cell demonstrates a high Zn2+ transference number (0.64), extended cycling life exceeding 1600 h at 1 mA cm−2, and stable operation for 200 h at 5 mA cm−2. Furthermore, zinc-ion hybrid capacitors employing Zn@MgSi electrodes exhibit excellent cycling stability over 5000 cycles. This work highlights the efficacy of artificial interfacial layers in stabilizing zinc metal anodes and provides valuable insights into the development of advanced aqueous zinc-ion energy storage systems.

A. Kirby, T. Nishino, L. Lanzilao et al.

<p>Turbine–wake and farm–atmosphere interactions can reduce wind farm power production. To model farm performance, it is important to understand the impact of different flow effects on the farm efficiency (i.e. farm power normalised by the power of the same number of isolated turbines). In this study we analyse the results of 43 large-eddy simulations (LESs) of wind farms in a range of conventionally neutral boundary layers (CNBLs). First, we show that the farm efficiency <span class="inline-formula"><i>η</i>_f</span> is not well correlated with the wake efficiency <span class="inline-formula"><i>η</i>_w</span> (i.e. farm power normalised by the power of front-row turbines). This suggests that existing metrics, classifying the loss of farm power into wake loss and farm blockage loss, are not best suited for understanding large wind farm performance. We then evaluate the assumption of scale separation in the two-scale momentum theory <span class="cit" id="xref_paren.1">(<a href="#bib1.bibx26">Nishino and Dunstan</a>, <a href="#bib1.bibx26">2020</a>)</span> using the LES results. Building upon this theory, we propose two new metrics for wind farm performance: turbine-scale efficiency <span class="inline-formula"><i>η</i>_TS</span>, reflecting the losses due to turbine–wake interactions, and farm-scale efficiency <span class="inline-formula"><i>η</i>_FS</span>, indicating the losses due to farm–atmosphere interactions. The LES results show that <span class="inline-formula"><i>η</i>_TS</span> is insensitive to the atmospheric condition, whereas <span class="inline-formula"><i>η</i>_FS</span> is insensitive to the turbine layout. Finally, we show that a recently developed analytical wind farm model predicts <span class="inline-formula"><i>η</i>_FS</span> with an average error of 5.7 % from the LES results.</p>

Nayaab Azim, Sadath Ullah Khan Mohammed, Evan Phaup et al.

In recent years, the field of software engineering has experienced a considerable increase in demand for competent experts, resulting in an increased demand for platforms that connect software engineers and facilitate collaboration. In response to this necessity, in this paper we present a project to solve the lack of a proper one-stop connection platform for software engineers and promoting collaborative learning and upskilling. The idea of the project is to develop a web-based application (NEXAS) that would facilitate connecting and collaborating between software engineers. The application would perform algorithmic matching to suggest user connections based on their technical profiles and interests. The users can filter profiles, discover open projects, and form collaboration groups. Using this application will enable users to connect with peers having similar interests, thereby creating a community network tailored exclusively for software engineers.

Agrawal Naman, Ridwan Shariffdeen, Guanlin Wang et al.

Large Language Models (LLMs) are becoming increasingly competent across various domains, educators are showing a growing interest in integrating these LLMs into the learning process. Especially in software engineering, LLMs have demonstrated qualitatively better capabilities in code summarization, code generation, and debugging. Despite various research on LLMs for software engineering tasks in practice, limited research captures the benefits of LLMs for pedagogical advancements and their impact on the student learning process. To this extent, we analyze 126 undergraduate students' interaction with an AI assistant during a 13-week semester to understand the benefits of AI for software engineering learning. We analyze the conversations, code generated, code utilized, and the human intervention levels to integrate the code into the code base. Our findings suggest that students prefer ChatGPT over CoPilot. Our analysis also finds that ChatGPT generates responses with lower computational complexity compared to CoPilot. Furthermore, conversational-based interaction helps improve the quality of the code generated compared to auto-generated code. Early adoption of LLMs in software engineering is crucial to remain competitive in the rapidly developing landscape. Hence, the next generation of software engineers must acquire the necessary skills to interact with AI to improve productivity.

Sten Grüner, Nafise Eskandani

The integration of Industry 4.0 technologies into engineering workflows is an essential step toward automating and optimizing plant and process engineering processes. The Asset Administration Shell (AAS) serves as a key enabler for creating interoperable Digital Twins that facilitate engineering data exchange and automation. This paper explores the use of AAS within engineering workflows, particularly in combination with Business Process Model and Notation (BPMN) to define structured and automated processes. We propose a distributed AAS copy-on-write infrastructure that enhances security and scalability while enabling seamless cross organizational collaboration. We also introduce a workflow management prototype automating AAS operations and engineering workflows, improving efficiency and traceability.

Lennart Werner, Pol Eyschen, Sean Costello et al.

Accurate real-time estimation of end effector interaction forces in hydraulic excavators is a key enabler for advanced automation in heavy machinery. Accurate knowledge of these forces allows improved, precise grading and digging maneuvers. To address these challenges, we introduce a high-accuracy, retrofittable 2D force- and payload estimation algorithm that does not impose additional requirements on the operator regarding trajectory, acceleration or the use of the slew joint. The approach is designed for retrofittability, requires minimal calibration and no prior knowledge of machine-specific dynamic characteristics. Specifically, we propose a method for identifying a dynamic model, necessary to estimate both end effector interaction forces and bucket payload during normal operation. Our optimization-based payload estimation achieves a full-scale payload accuracy of 1%. On a standard 25 t excavator, the online force measurement from pressure and inertial measurements achieves a direction accuracy of 13 degree and a magnitude accuracy of 383 N. The method's accuracy and generalization capability are validated on two excavator platforms of different type and weight classes. We benchmark our payload estimation against a classical quasistatic method and a commercially available system. Our system outperforms both in accuracy and precision.

S. Afolalu, T. Ogedengbe, S. Monye et al.

Tribology, the science, technology, engineering, and friction and wear, is reviewed in terms paper by providing a comprehensive review of the principles of tribology and their critical applications in engineering. The road starts with delving into the historical and conceptual basis of tribology and placing it in the context of a wide range of industrial applications, ranging from automotive to aerospace engineering, and manufacturing to biomedical engineering. The study highlights the critical contribution of tribology to reduce friction and wear of machines, and thus improve the machine reliability, energy efficiency and sustainability through material and surface engineering. It concludes with an emphasis on the interdisciplinary nature of tribology and its central role in technological innovation, machinery performance improvement and in tackling the sustainability challenges in engineering. Finally, future research directions in nanoscale tribological studies and advanced materials development are proposed to further optimize the design and operation of mechanical systems.

Bingcheng Li, Qingfeng Wang, Xiangshuo Xi et al.

The volume throttle speed control system, known for its speed regulation accuracy and energy efficiency, is widely used in construction machinery. When combined with a double-variable hydraulic source that allows simultaneous control of motor speed and pump swashplate angle(SWA), the system gains greater flexibility but also introduces additional nonlinearities. To address this issue and enhance power matching performance, a nonlinear flow model is first developed, considering motor speed, SWA, and pump supply pressure. Based on this model, a control structure is designed that integrates a continuous sliding mode controller to improve accuracy and robustness under dynamic conditions. Second, a dynamic variable pressure margin (VPM) strategy is proposed to achieve real-time power matching. It analytically computes the target pressure margin (PM) based on a characterization factor λ , allowing users to flexibly balance energy consumption and machine responsiveness. Using load data from a 2-ton electric excavator’s digging cycle as a benchmark, the double-variable hydraulic source with the proposed control structure showed an improvement of approximately 24.9% in control performance compared to traditional hydraulic sources. The application of the VPM strategy resulted in power matching optimizations of 23.0% and 15.3% for the double-variable and traditional hydraulic sources, respectively. These results validate the superiority of the proposed control structure and the effectiveness of the VPM strategy.

F. Veiga, A. Gil-Del-Val, E. Iriondo et al.

This paper presents the experimental work developed to measure the learning process through concept map analysis. The development of a concept map is requested by the students for each chapter or theme of the subject. As a result, maps from engineering courses have been analyzed. The measurements carried out consider several parameters, such as individual and team map building, student progressive knowledge level, and map complexity. Concerning the complexity analysis, the focus is qualitative, and it is based on the data extracted from the concept maps elaborated by the students. The study, conducted during the 2018–2019 academic year, included students from various academic levels and institutions, such as the Public University of Navarra UPNA and the University of the Basque Country UPV-EHU, covering first-degree students of Bachelor's Degree in Mechanical Engineering and first-degree students of Master's Degree in Industrial Engineering at UPNA, third-degree students of Bachelor's Degree in Mechanical Engineering at UPV-EHU. The data collected from 37 individual maps in Industrial Drawing, 31 group maps in Industrial Drawing, 12 individual maps in Design of Machinery, and 12 group maps in Design of Machinery, along with a control group of 79 students who did not participate in any activity, provided valuable insights into the effectiveness of concept maps for evaluating understanding levels and learning outcomes across various engineering subjects and academic levels. The learning outcome of the students is treated to obtain the level of understanding of complex systems shown by the students through the concept maps previously drawn and the questionnaire answered by each student about the achievement of learning results through the use of concept maps. This work shows the research methodology established and the learning results achieved qualitatively: measuring the maps by means of a rubric, self-assessment based on a survey, and through the questionnaires. Also, the results obtained in the final exams have been compared. From the observed results, this methodology is presented as a suitable alternative for evaluating the correct acquisition of concepts in online teaching situations.

Chunhua Zhao, Baoping Tang, Lei Deng

Compressed sensing (CS) can significantly improve the transmission efficiency of large amounts of vibration data in wireless sensor networks (WSNs) for mechanical vibration monitoring. To address the issue of irrecoverable measurements loss due to unstable communication links in WSN, this article proposes a missing-measurements-tolerant CS (MMTCS) in WSN for mechanical vibration monitoring. First, the embedded compressed sampling (ECS) is designed to compressed sampling the original signals in the acquisition nodes, thereby enhancing transmission efficiency. Moreover, the article analyzes the missing measurements perturbation error caused by compressed sampling and measurements loss in wireless transmission. An objective optimization function is derived for missing measurements. Combined residual adaptive sparse reconstruction (CRASR) is proposed for accurate data reconstruction. The experimental results demonstrate that the proposed method achieves a better trade-off between reconstruction accuracy and reconstruction time in comparison with other popular methods. More importantly, the proposed method can achieve satisfactory fault detection accuracy for rotating machinery under some degree of compressed sampling and missing measurements. This is of great value to practical engineering applications and provides a practical and effective solution.

Yu-qian Shi, jiuhe yang

Dynamics of a periodically excited vibro-impact system with soft impacts is investigated. Essential features of period-one multi-impact motion group and correlated transition characteristics in low-frequency range are discussed in detail by the way of two-parameter bifurcation space providing qualitative domains for different periodic motions. The main focus is given to the effect of sensitive parameters including constraint stiffness k 0, clearance threshold b, and damping parameter ζ on the system response. The low-frequency characteristics in the finite-dimensional parameter space are particularly explored. It is found that the increase of k 0 induces multi-type bifurcation of period-one double-impact symmetrical motion, which induces a rich variety of periodic motions, and period-one multi-impact motion group orbit primarily exist in the small-clearance b and low-frequency ω zone. Based on the evolution irreversibility of adjacent period-one multi-impact orbit, the mechanism of singularies appearing in pairs and two different transition zones (hysteresis and liguliform zones) is studied, the result of which provides a theoretical reference value for the common low-frequency vibration instability phenomenon in the field of mechanical engineering. For small-damping coefficient ζ, period-one multi-impact motion has a large quantity, and the main bridge for the transition of adjacent period-one multi-impact motion is liguliform zone, which embraces period-one multi-impact asymmetrical motion and period-n multi-impact subharmonic motion and a certain chaotic zone. For large-damping coefficient ζ, the amount of period-one multi-impact motion group is reduced, and the main bridge for the transition of adjacent period-one multi-impact motion is hysteresis zone, where adjacent period-one multi-impact orbits can coexist according to initial conditions. As designing and renovating impact mechanical equipment, the reasonable matching law of dynamic parameters can be determined through two-parameter bifurcation space, which is conducive to making the system work in stable periodic motion and obtaining larger instantaneous impact velocity.

Huisheng Ou, Jiabin Liu, Zhenhua Tan et al.

Jiahao Lv

This study introduces an innovative deep learning architecture, amalgamating Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks, termed the CNN-LSTM model. Its efficacy in both identifying and anticipating mechanical failures is explored through an examination of vibration datasets sourced from actual industrial machinery. The assessment delves into the model's capabilities across various fault categories and severities. Findings indicate that the CNN-LSTM model exhibits remarkable precision in fault identification, with forecasted outcomes largely aligning with actual fault occurrences, thus corroborating its diagnostic efficacy. A comparative analysis against traditional diagnostic techniques further elucidates the superior performance of the proposed model, as evidenced by its enhanced accuracy, recall, and F1 score metrics. Such results underscore the deep learning model's advanced precision and dependability when addressing intricate fault prediction tasks. This study proves the superior performance of CNN-LSTM model in the task of mechanical fault diagnosis and prediction. This discovery provides strong evidence for the application of deep learning in industrial field, and provides new tools and methods for solving practical engineering problems.

Redeil N. Arreza, Alec Nowell A. Ranara, Trisha Kerstin C. Tan et al.

Significant volumes of wastewater, particularly from the textile industry, pose environmental concerns due to the presence of hazardous substances such as ortho-toluidine (OT). The photo-Fenton process can be used to break down and remove this hazardous organic compound. Previous studies on the photo-Fenton process have focused on local optimization of operating variables without considering cost factors. The photo-Fenton process is studied in this paper with UVA irradiation, Fe2+ dosage, and H2O2 concentration considered as variables. The study uses fuzzy optimization in a multi-objective framework for making decisions to determine the optimal values of OT degradation with its corresponding cumulative uncertainty error (YA), and the total operating cost (CT), both of which are essential for assessing the techno-economic feasibility of the process. The Pareto front was generated from the objective functions to establish the boundary limits for YA and CT. The results show an overall satisfaction level of 71.81% for the objective functions, indicating a partially satisficing solution for maximizing OT degradation while minimizing operating cost. The optimum conditions of the variables require 85.70 W m−3 UVA irradiation, 0.5177 mM for Fe2+ dosage, and 7.85 mM for the H2O2 concentration. These conditions yielded an OT degradation value of 83.22% and a total operating cost of 768.61 USD·m−3. Comparison with previous literature showed an OT degradation efficiency that was 16.78% lower. However, this tradeoff in the process efficiency is offset by a total operating cost that is 2.28 times cheaper, emphasizing the cost-effectiveness of the fuzzy optimized solution.

Lucia Mazzapioda, Francesco Piccolo, Alessandra Del Giudice et al.

Abstract Single lithium-ion conducting polymer electrolytes are promising candidates for next generation safer lithium batteries. In this work, Li+-conducting Nafion membranes have been synthesized by using a novel single-step procedure. The Li-Nafion membranes were characterized by means of small-wide angle X-ray scattering, infrared spectroscopy and thermal analysis, for validating the proposed lithiation method. The obtained membranes were swollen in different organic aprotic solvent mixtures and characterized in terms of ionic conductivity, electrochemical stability window, lithium stripping-deposition ability and their interface properties versus lithium metal. The membrane swollen in ethylene carbonate:propylene carbonate (EC:PC, 1:1 w/w) displays good temperature-activated ionic conductivities (σ ≈ 5.5 × 10–4 S cm−1 at 60 °C) and a more stable Li-electrolyte interface with respect to the other samples. This Li-Nafion membrane was tested in a lithium-metal cell adopting LiFePO4 as cathode material. A specific capacity of 140 mAhg−1, after 50 cycles, was achieved at 30 °C, demonstrating the feasibility of the proposed Li-Nafion membrane.