Gear hobbing is a highly efficient and typical tooth manufacturing method in which hob accuracy directly affects the quality of machined gears. Therefore, geometric error inspection and precision grade evaluation of gear hobs, characterized by discontinuous helical tooth surfaces and periodic distribution of chip grooves, are essential prerequisites for enhancing the machining accuracy of gears. A novel probing-based measurement method for identifying the cutting edge of gear hobs was proposed. By applying the principles of differential geometry, the basic worm helical surface equation and rake surface equation of Archimedes gear hobs were derived, and a mathematical model of the hob cutting edge profile was established. The measurement path for the tooth profile error of the hob cutting edge is planned based on the complex spatial morphology of the hob. The hob cutting-edge profile error measurement experiment was conducted at the GMC 400 Gear Measurement Center. The measurement data were processed in accordance with the calculation equation for the hob tooth profile error in GB/T6084-2016, and the precision grade of the hob was evaluated in terms of the modulus range and tolerance value. The minimum error was 6.3 mu m(grade 2 A) and the standard deviation of the repeatability experiment was less than 0.15 mu m. The experiment verified the correctness of the proposed method and provided a reliable technical scheme for high-accuracy measurement of gear hobs.

Residual stresses, induced during manufacturing due to solidification shrinkage and phase transformations in large-scale structural components of precision machine tools, may significantly compromise the components’ accuracy and service life. Thermal aging, widely employed for stress relief, requires rigorous optimization of its process parameters to maximize effectiveness. This study investigates the evolution of residual stresses in HT300(ASTM Class 30/EN-GJL-300) cast iron crossbeams during thermal aging using a heating rate of 50–55 °C/h, a holding temperature of 550 °C, and a cooling rate of 15–20 °C/h. A thermomechanical-coupled finite element analysis, validated by non-destructive magnetic measurements (with a mean relative error <15

Rotating machinery fault diagnosis remains challenging in practical vibration measurement scenarios, where variable operating regimes, noise interference, and non-stationary degradation hinder reliable interpretation of measured vibration signals. To address these challenges, this paper proposes a prior-guided dynamic fusion network (PGDFNet), termed PGDFNet, for robust vibration signal-based fault diagnosis through collaborative modeling of temporal dynamics, time-frequency structures, and statistical priors. Specifically, a gated temporal Mamba branch is designed to enhance fault-sensitive impulsive responses and capture long-range temporal dependencies from raw vibration signals. Meanwhile, a spatial-frequency Mamba branch models local spectral textures, multi-scale time-frequency structures, and global spatial-frequency dependencies from continuous wavelet transform representations. In addition, a prior-guided dynamic fusion module incorporates statistical priors into the dual-branch fusion process to adaptively coordinate temporal and time-frequency features under noisy measurement conditions. By integrating these components, PGDFNet improves the robustness and discriminability of vibration-based fault representations. Experiments on the CWRU, PU, and JNU datasets demonstrate that the proposed method outperforms representative baselines in classification accuracy, Gaussian and complex-noise robustness, class-imbalance robustness, and cross-domain transfer capability. Ablation, interpretability, and efficiency-complexity analyses further verify the effectiveness of the proposed components and the favorable trade-off between diagnostic performance and computational overhead.

Moisture-curable one-component polyurethane (1K-PU) adhesives are widely used in bonding and sealing applications due to their excellent flexibility and sealing properties. Under the condition that the adhesive is not fully cured before entering the service phase, insufficient interface adhesion and reduced load-bearing capacity may lead to joint loosening, posing safety risks. Although the mechanical properties of polyurethane adhesives have been widely studied, the quantitative characterization of mechanical-property evolution during the curing process remains limited. This study proposes a multiscale-coupling method for mapping the curing time to mechanical properties. In situ Fourier-transform infrared (FTIR) monitoring tracks the real-time conversion of isocyanate groups, enabling the identification of parameters for the curing reaction kinetics model, which predicts the degree of cure (DOC) over time as a constraint introduced in molecular dynamics (MD) simulations. Based on the principles of moisture curing, MD simulations model gradual crosslinking and generate crosslinked network structures at different curing degrees. The mechanical parameters of adhesives are computed under boundary conditions, establishing a quantitative, cross-scale mapping relationship between curing time and mechanical performance. The proposed method addresses the challenge of obtaining mechanical properties of the adhesive during the curing process and elucidates the modulus growth mechanism driven by crosslinking densification and reduction in free volume. It provides a practical basis for evaluating the mechanical properties of low-elastic-modulus adhesives throughout the curing phase and for optimizing assembly processes.

The motorized spindle is the core component of a CNC grinding machine. During the machining process, the working conditions are complex and variable. The uneven temperature distribution caused by internal and external heat sources in the motorized spindle leads to nonlinear and time-varying thermal deformation, particularly, axial thermal elongation can exceed several tens of micrometers, significantly compromising the grinder's machining accuracy. To effectively implement thermal error compensation with high accuracy and cost-efficiency, a CNN-BiLSTM-AM based thermal error prediction model for the motorized spindle is proposed, which combines CNN for spatial feature extraction, BiLSTM for capturing long temporal dependencies, and an attention mechanism to enhance feature relevance. The model's prediction performance and robustness were validated through thermal characteristic experiments under various working conditions. In the predictions for Tests 2 and 3, the maximum residuals between the predicted and measured axial thermal deformation were 1.326 mu m and 1.495 mu m, with R2 values as high as 0.991 and 0.992, MAEs as low as 0.524 mu m and 0.589 mu m, and RMSEs of 0.596 mu m and 0.691 mu m, respectively. All metrics outperformed those of comparative methods. Furthermore, the thermal compensation experiments were performed to verify the model's practical applicability.

Thermally induced positioning error (TIPE) in ball screw feed drives critically affects the machining accuracy of high-precision machine tools. Existing thermal error models are challenged by their reliance on extensive labeled datasets, which increases data acquisition costs and prolongs production cycles. In addition, these models utilize sparse temperature measurements as inputs, which are insufficient for reconstructing the full temperature field and therefore limit the characterization of non-uniform TIPE distributions under complex feed motions. To overcome these limitations, this study proposes a physics-informed digital twin (PIDT) framework for TIPE prediction that does not require labeled TIPE data for supervised training or parameter fitting, while explicitly incorporating the time-varying nut trajectory. First, an independent reduced thermal domain of the screw shaft is established by assimilating measured boundary temperatures, while a convolution-based physics-informed solver reconstructs the transient temperature field induced by the moving nut and stationary bearings. Subsequently, the TIPE profile along the feed axis is derived by integrating the axial temperature distribution and is dynamically constrained by a novel in-process end-displacement (IPED) measurement, thereby enabling online correction without interrupting machining. Across four operating conditions with different feed rates and reciprocating ranges, the proposed model achieves accurate and robust prediction under parameter uncertainty, reducing the maximum TIPE from 42.3, 56.7, 58.3, and 53.7 mu m to 2.9, 4.6, 4.1, and 3.6 mu m, respectively. This framework provides an efficient, physically interpretable, and cost-effective solution for TIPE compensation under complex and variable machining conditions.

Tool wear prediction is a critical technology for ensuring the quality and efficiency of precision machining. Existing deep learning methods based on entropy features suffer from limitations such as insufficient feature sensitivity and poor adaptability to multi-operating conditions. This study proposes a tool wear prediction method that integrates Adaptive Multiscale Weighted Slope Entropy (AMWSLE) with a 1DCNN-BiLSTM-Attention model. The implementation process is as follows: First, Variational Mode Decomposition (VMD) is used to decompose vibration signals into multi-scale Intrinsic Mode Functions (IMFs), addressing the non-linearity and non-stationarity of the signals. Second, the AMWSLE algorithm is developed, which enhances the ability to characterize the gradual tool wear process through dynamic threshold adjustment, multi-scale weight assignment, and 5-level symbolic coding optimization. Finally, a 1DCNN-BiLSTM-Attention model is constructed to realize the extraction of temporal features and regression prediction of wear amount. Experimental validation shows that under 9 sets of cutting conditions, the proposed method achieves an average RMSE of 13.01 and an average R² of 97.21%. Notably, it reduces the Mean Absolute Error (MAE) by 17.79%-38.60% compared with benchmarks such as SVM, GRU, and conventional slope entropy, underscoring the proposed method’s significant performance gain.

Discrete manufacturing environments face increasing challenges in managing work-in-process (WIP) inventory due to growing product customization and demand volatility. While Value Stream Mapping (VSM) has been widely used for process improvement, traditional approaches lack the ability to dynamically control WIP levels while optimizing multiple performance dimensions simultaneously. This research addresses this gap by developing an integrated framework that synergizes Multi-Dimensional Value Stream Mapping (MD-VSM) with multi-objective optimization, functioning as a specialized digital twin for dynamic WIP control. The framework employs a four-layer architecture that connects real-time data collection, multi-dimensional modeling, dynamic WIP monitoring, and execution control through closed-loop feedback mechanisms. A mixed-integer optimization model is used to balance time, cost, and quality objectives. Validation using a high-fidelity simulation, parameterized with real-world industrial data, demonstrates that the proposed approach yielded up to a 31% reduction in inventory costs while maintaining production throughput and showed a 42% faster recovery from equipment failures compared to traditional methods. Furthermore, a comprehensive sensitivity analysis confirms the framework’s robustness. The system demonstrated stable performance even when key operational parameters, such as WIP upper limits and buffer capacity coefficients, were varied by up to ±30%, underscoring its reliability for real-world deployment. These findings provide manufacturers with a validated methodology for enhancing operational efficiency and production flexibility, advancing the integration of lean principles with data-driven, digital twin-based control systems.

Data scarcity is one of the key bottlenecks in the application of machine learning in the field of materials discovery. In this challenge, transfer learning can leverage existing consistent large-scale data to assist in property prediction on small datasets, thereby opening up more possibilities for materials development. With the discovery of many-dimensional materials and the challenges posed by the miniaturization of transistors, the range of electrode materials available for transistors is extremely broad, but it is difficult to explore them through traditional experimental methods. Therefore, in the face of scarce data on electrode contact characteristics, this study proposes a cross-scale hybrid transfer learning framework that integrates first-principles calculations with a 2D materials database. By utilizing large-scale potential height data obtained through PBE functional calculations, the framework achieves high-precision predictions of DFT-1/2 method and HSE06 functional calculation results, with an MSE controlled within 0.04 eV. The research results indicate that this learning framework accelerates the process of screening electrode materials for MoS2, providing important theoretical guidance and technical support for the design and optimization of new electronic devices.

Titanium alloy is widely utilized in the manufacturing of crucial components for aerospace and other industries due to its excellent physical and mechanical properties. Flexible abrasive disc grinding provides significant advantages in titanium alloy machining, yet the interaction between the abrasive disc and the titanium alloy surface is extremely complex. This study aims to reveal the mechanism of the flexible abrasive disc grinding process through the study of the interaction of multi-grains and the material removal process. Specifically, the motion trajectory of a single grain and the contact area formula on the workpiece surface were derived. Additionally, a simulation model for multi-grains grinding of TC17 titanium alloy was established. The study further explores the contribution of individual grains to the total grinding force within the multi-grains model, as well as the material removal mechanism in flexible abrasive disc grinding. The simulation results indicated that the interaction among the abrasive grains and the surface morphology of the contact area significantly impacts the grinding force. The actual count of effective abrasive grains participating in the grinding process was fewer than that predicted by geometric calculation. The grinding tracks display a crescent shape, and as the distance from the grinding center increases, their depth and width decrease. When compared to the experimental data, the prediction error of the simulation model for the normal grinding force was 6.3

In the digital measurement of aircraft assembly,the accuracy of large-size measurement field construc-tion is highly dependent on the stability of the reference points laid on the tooling.The position of the reference points of large-sized tooling is very susceptible to thermal drift due to changes in ambient temperature,leading to a reduction in the accuracy of the measurement field or even failure.Therefore,this paper takes a combined large-scale tooling as an example to construct a numerical model for predicting the thermal drift of the reference points of large-scale tooling under a non-uniform temperature field;constructs a proxy model for the thermal drift of the tool-ing based on a large amount of thermal drift data obtained from the simulation of the aforementioned model using BP neural network;and formulates a program for improving the accuracy of the measurement field based on the aforementioned proxy model.The temperature and coordinate measurement data collected in the field at the refer-ence points of the tooling are used to verify the validity and correctness of the proxy model,and the temperature-co-ordinate drift data at the reference points obtained from the model are compared and analyzed.The results show that the average relative errors of the simulation results are below 18%,and the average relative errors of the BP neural network results are below 26%,which can effectively improve the measurement field construction accuracy.

Due to their inherently low structural stiffness, thin-walled parts are prone to elastic deformation under cutting forces during machining, which is one of the key factors limiting the improvement of machining accuracy. To address the challenge of deformation prediction in flank milling, this paper proposes the force-induced deformation model for flank milling of thin-walled parts based on CWE-driven meshing strategy and two-stage static condensation. First, the milling force prediction model is developed by integrating the cutter-workpiece engagement (CWE) and instantaneous undeformed chip thickness (IUCT), providing high-accuracy load input for deformation computation. Then, by integrating the geometric evolution features of the engagement zones, the CWE-driven meshing strategy is designed. Focusing on the machining zones and CWE sub-domains, the two-stage static condensation method is proposed to reduce the dimensionality of the stiffness matrix. By coupling the material removal process with the stiffness matrix reduction method, efficient stiffness updates are achieved during deformation analysis. Finally, the force-induced deformation iterative prediction model for flank milling is developed based on the equivalent mapping of cutting forces, the mapping from deformation to radial depth variation, and calibration of machining parameters. The accuracy and practicality of the proposed model are validated through comparisons between experimental results and model predicted data in terms of cutting forces, machining errors, and dynamic displacements. Its computational efficiency is further demonstrated by comparing computation times with global stiffness matrix. This study provides both theoretical support and practical pathway for precision machining of thin-walled parts.

The evolution of the automotive industry demands gearboxes that offer improved specifications in terms of power transfer and shifting performance. The high-quality standards required for transmission component manufacturing processes render it necessary to measure shaft parts. To realize online diameter inspection of shaft parts in the gearbox production line, a machine vision-based non-contact metrology method with Canny-Steger subpixel edge detection is proposed. The interference background was separated from the measured target using ROI technology. Morphological closure operations and bilateral filtering preprocessing were performed on the collected images to achieve filtering and denoising, respectively. With the extraction of pixel-level edges using the Canny operator as a basis, the Steger algorithm was used to detect the subpixel edges. The edges and normal directions are derived from the Hessian matrix. Subsequently, Taylor polynomials are applied to the pixel grayscale in the normal direction to obtain the grayscale distribution function, the extreme values of which are solved to obtain the subpixel edge point positions. The fitting data are employed to evaluate the machining errors of the parts, significantly reducing redundant calculations in the Steger algorithm process and improving the detection speed. The proposed method was validated through synthetic and real experiments. Experimental results demonstrate that the measurement method proposed in this paper can rapidly measure shaft parts. Its diameter measurement accuracy can reach 3 mu m, and the repetitive measurement accuracy can reach 2 mu m. Compared with pixel-level edge detection, the accuracy has been improved by 60%.

Residual stresses in large-scale structural components are recognized as critical compromisers of dimensional stability and operational longevity in precision machine tools. In order to clarify the evolution of residual stress across casting to thermal aging processing chains, the stress formation mechanism was elucidated through an integrated multiphysics framework addressing geometric nonlinearities and heterogeneous thermal dynamics. A ProCAST-ABAQUS computational workflow was established to quantitatively track stress evolution across thermal manufacturing stages: casting and thermal aging. A 38 % reduction in residual stress was experimentally validated at the optimized pouring temperature of 1430(degrees)C relative to the industrial standard of 1370(degrees)C. Concurrently, 79.3 % peak stress attenuation was achieved through the implementation of gradient cooling protocols during thermal aging. This methodology provides quantitative process controls with <20 % prediction-measurement variance, significantly enhancing dimensional stability in precision castings. It simultaneously establishes theoretical foundations and delivers actionable strategies for distortion mitigation in heavy-section manufacturing and machine tool performance optimization.

This paper focuses on analyzing the use of the Support Vector Machine (SVM) classifier in forecasting the career progression of college students. In this case, the research seeks to evaluate the performance of SVM in the prediction of students’ job outcomes regarding factors like GPA, extra curriculum activities, and internship. This dataset was taken with these attributes and after completing the exploration a feature selection by the Recursive Feature Elimination (RFE) was used. The model compiled the data with 80% of data for training, with the 20% of data that were used for testing, the model’s overall accuracy in prediction stood at 87%. Evaluation metrics such as precision, recall, and F1-score were used to validate the model’s performance across five distinct career paths: Academia, industry, entrepreneurship, government, and freelancing. In general, high accuracy in identifying academic and government careers was reported while freelancing and entrepreneurship were less successfully predicted possibly because of their unbound lifestyle. As stated, the study shows that SVM can indeed be used for career counseling in the educational sectors since students can be given an objective model to follow. Future enhancement includes the addition of personality variables and career choice to improve prediction for the less defined occupation types such as freelance work and self-employment.

Continuous high glucose levels in blood could cause heart disease, liver dysfunction, kidney failure, and blindness to diabetics. To detect glucose levels non-invasively and precisely, it is urgent to construct flexible and sensitive sensors to monitor glucose levels in time. Herein, a flexible non-enzymatic electrochemical glucose sensor was fabricated with gold nanoparticles/ZnO nanorods (AuNPs/ZnONRs) heterojunctions modified on stainless steel wire sieve (SSWS), and further it was stimulated under UV irradiation for enhanced electrocatalytic capacity. The AuNPs not only as excellent electron carrier facilitate electron immigration at the solid/liquid interface, but also as favorable catalyst oxidize glucose. More importantly, under the synergistic effect of the AuNPs/ZnONRs heterojunctions and the UV irradiation, the AuNPs efficiently separate UV- generated electron/hole pairs within the ZnONRs and more holes react with H2O to produce sufficient hydroxyl, which as oxidants oxidize glucose. Consequently, the sensitivity of the AuNPs/ZnONRs/SSWS glucose sensor significantly raises from 36.4 mu A center dot mM-1 center dot cm- 2 to 344.3 mu A center dot mM-1 center dot cm- 2 over a wide linear range of 14.0 mM, and the low detection limit diminishes from 2.1 mu M to 0.21 mu M. Additionally, the sensor displays desirable selectivity, bending capability, and stability, and it could also reliably analyze glucose contents in sweat and fruit juices. This work demonstrates a novel and effective method for improved electro-catalytic capacity of non- enzymatic electrochemical glucose sensors using the synergistic effect of the AuNPs/ZnONRs heterojunctions and UV irradiation, and it could be used to enhance the sensing performance of other electrochemical biosensors.

Spindle thermal error is a major cause of machining inaccuracies. Conventional compensation approaches typically require large datasets across multiple operating conditions, resulting in increased costs and machine downtime. This study proposes a physics-informed digital twin (PIDT) framework for thermal error prediction with minimal data, comprising two sequential core components. First, a dynamic temperature field realtime simulation (DTFRTS) reconstructs the transient temperature fields based on real-time inputs such as spindle speed, motor power, and collected temperatures. It dynamically updates thermal boundary conditions and employs convolution-based finite-difference scheme for computational efficiency. Second, a transient temperature field-based spatiotemporal integrator (TTF-STI) converts simulated temperatures into axial thermal elongation error (ATEE) by physically integrating along the spindle's structural path, requiring only six sparse ATEE samples for calibration. Experimental results validate the high-fidelity temperature field by the DTFRTS model (mean error: 0.43 degrees C) and demonstrate that TTF-STI delivers high-accuracy ATEE prediction (average deviation: 1.88 mu m), outperforming benchmark data-driven models. The two models are embedded in an oriented thermal error compensation digital twin system (OTMDTS) deployed on an edge computing platform, achieving real-time prediction and reducing ATEE by over 98% in compensation tests. The proposed PIDT framework is a robust, efficient, and scalable solution for real-time spindle thermal error prediction and compensation in industrial applications.

Hydrate-induced pipeline blockages pose critical risks to energy infrastructure reliability and operational safety. This study proposes a novel acoustic-based structural health monitoring system integrating linear frequency modulated (LFM) signals with adaptive matched filtering, specifically designed for early blockage detection in complex pipeline networks, including those in built environments. Validated via controlled experiments simulating industrial-scale conditions, the system achieved millimeter-level positioning accuracy (error <1 %) across varying blockage intensities (20-100 %), representing a 40 % precision improvement over conventional methods. The system's capability to resolve multi-stage blockage boundaries enhances predictive maintenance strategies for ensuring energy-efficient pipeline operations and infrastructure integrity management. By enabling real-time identification of incipient hydrate formation phases through acoustic signature analysis, this methodology supports sustainable infrastructure management through timely remedial interventions. Its applicability extends to intelligent building service pipelines, regional energy distribution networks, and deep-sea energy infrastructure, demonstrating significant potential for improving failure prediction and enhancing the resilience of built environment systems.

ABSTRACT Conductive hydrogels hold significant promise in the fields of flexible electronics and smart sensing applications owing to their outstanding stretchability and electrical conductivity. In this study, a dual‐network hydrogel based on polyacrylamide (PAAM) and sodium alginate (SA) networks was constructed to fabricate highly sensitive strain sensors. Carbon nanotubes (CNTs) were incorporated into the hydrogel system, and polypyrrole (PPy) nanoparticles were synthesized using SA as a soft template. During preparing process, ultrasonic oscillation was employed to facilitate the uniform distribution of PPy in the hydrogel matrix, and then a CNTs–PPy conductive pathway was formed, thereby enhancing the conductivity and sensitivity of the hydrogel. The prepared PAAM–SA–CNTs–PPy hydrogel exhibits high electrical conductivity (4.84 S/m), superior mechanical properties (including a tensile strength of 683.5 kPa at 812%), high toughness, good self‐healing, and adhesion properties. The assembled strain sensors demonstrate high sensing capability, a wide measurement range (the gauge factor of 2.67 for 0%–400% strain and the gauge factor of 5.25 for 400%–800% strain), rapid response times (126 ms), and good stability. The developed hydrogel‐based conductive strain sensor has the potential to be excellent candidates as wearable sensing devices for monitoring human movements.

Long-term high blood glucose levels brings extremely detrimental effect on diabetic patients, such as blindness, renal failure, and cardiovascular diseases. Therefore, there is an urgent need to develop highly flexible and sensitive sensors for precisely non-invasive and continuous monitoring glucose levels. Herein, we present a highly flexible and sensitive wearable sensor for non-enzymatic electrochemical glucose analysis with vertically aligned mushroom-like gold nanowires (v-AuNWs) chemically grown on stainless steel wire sieve (SSWS) as integrated electrode. Owing to the unique nanostructures and excellent catalysis of the v-AuNWs, the asfabricated glucose sensors exhibit superior flexibility and excellent electro-catalytic capability. In detail, these sensors display rapid response towards glucose within 5 s, and the sensor constructed with v-AuNWs for growth time of 15 min shows the highest sensitivity of 180.1 mu A mM-1 cm-2 within a wide linear range of 6.5 x 10-4 mM-12.0 mM and the lowest detection limit of 0.65 mu M (S/N = 3). It is noteworthy that due to the good ductility of the v-AuNWs and their strong contact with the SSWS substrate, these glucose sensors exhibit no obvious response variation after repeated bending for 100 times at bending angle of 180 degrees. Additionally, the glucose sensors display superior anti-interfering capability as well as desirable repeatability. More importantly, these glucose sensors can be attached on human skin to determine sweat glucose reliably and analyze glucose concentration in human serum in vitro.