Journal of Engineering and Thermal Sciences
Not a member yet
1200 research outputs found
Sort by
An ensemble model with convolutional neural network by DS evidence fusion for bearing fault diagnosis
Get PDFBearing fault diagnosis is crucial for ensuring the safety and reliability of rotating machinery. In recent years, artificial intelligence technology based on machine learning has made substantial progress in the field of bearing fault diagnosis. Most existing models for bearing fault diagnosis are built using big data and deep learning algorithms and can achieve high diagnostic accuracy with sufficient fault data. However, there still exist two open issues, 1) in practical engineering, acquiring fault sample data is challenging, and it is difficult to obtain a sufficient number of samples to train the hyperparameters of deep learning models. 2) Fault diagnosis models based on individual classifiers rely heavily on prior knowledge for signal feature extraction and the selection of network structures and parameters, making it difficult to guarantee the model’s effectiveness. This paper proposes an integrated diagnostic model called DS-ELM that employs multiple extreme learning machine modules with different parameters as subclassifiers. The outputs of these modules are then fused via DS evidence fusion theory to obtain the final diagnostic result. This ensemble model has better flexibility and robustness which significantly improves the accuracy and stability of the diagnostic model. Overall, the proposed DS-ELM provides a new solution for bearing fault diagnosis. In addition, the superiority of the reported technique is confirmed via experimental bearing fault data from Case Western Reserve University
Reduced-order modeling of digital twins for a high-voltage circuit breaker
Get PDFThe dynamic characteristics of operating mechanisms are of great significance to the reliability of high-voltage circuit breakers (HVCBs). To achieve dynamic analysis and optimization design of HVCBs, this paper presents a reduced-order digital twin model for a 252 kV HVCB under the OpenModelica simulation platform. To validate the effectiveness of the model, simulated closing and opening stroke curves are compared with experimental results, demonstrating the high efficiency and accuracy of the reduced-order digital twin model. Furthermore, a simulation analysis is performed to examine the mechanical dynamics of the HVCB under potential fault conditions, such as abnormal spring driving force. The analysis reveals that a reduction in the preload of the closing spring slows the operating mechanism’s movement, increases closing time, and may even cause closing failure. Similarly, a decrease in the preload of the opening spring reduces the moving contact’s speed, prolongs the opening time, and may result in opening failure. The proposed digital twin modeling method offers designers a systematic method for quantifying the dynamic responses of HVCBs
Energy analysis of living stumps slope based on Hilbert-Huang Transform and marginal spectrum
Get PDFA large-scale shaking table model test on a slope with living stumps was designed and conducted. Under various types of seismic waves and excitation intensities, acceleration data from monitoring points on both sides of the living stumps were collected. Hilbert-Huang Transform (HHT) was innovatively applied to study the dynamic response of slopes with living stumps under seismic loading, overcoming the limitations of traditional Fourier Transform and Wavelet Transform. The variation patterns of Hilbert energy and marginal spectral characteristics under different seismic excitations were analyzed, providing new insights from both time-frequency domain and energy perspectives. The research conclusion showed that: (1) Under different seismic waves, the horizontal peak acceleration inside the living stumps slope shows the elevation amplification effect, and increases with the intensity of excitation. Additionally, the existence of living stumps causes a difference in horizontal acceleration on both sides, and the absolute value of the difference is positively correlated with elevation and excitation intensity. (2) Under different seismic waves, Peak of Hilbert energy spectrum (PSHEA) is positively correlated with excitation intensity and elevation. With the increase of elevation, the increase of PSHEA increases gradually when the excitation intensity increases. PMSA is positively correlated with excitation intensity, but at low frequencies (1-3 Hz), Peak of marginal spectrum (PMSA) is negatively correlated with elevation; while at high frequencies (7-11 Hz), PMSA is positively correlated with elevation. (3) With increasing elevation and excitation intensity, the total seismic Hilbert energy continues to accumulate and reaches the maximum at the top of the slope. During the propagation of seismic waves, the living stumps and the rock-soil composite play the characteristics of filtering the low-frequency components and amplifying the high-frequency components, causing the total seismic Hilbert energy in the low-frequency (1-3 Hz) component to gradually decrease and transfer to the high-frequency (7-11 Hz) component, resulting in a significant increase in seismic Hilbert energy in the high-frequency component. (4) The superposition of incident wave and reflected wave near the living stumps, and the absorption of seismic Hilbert energy by the living stumps make the PSHEA, PMSA, and total seismic Hilbert energy on the outside of the living stumps always smaller than the inside, resulting in different dynamic responses on either side of the living stumps. The living stumps show attenuation effect on seismic Hilbert energy, and the attenuation degree increases with the increase of excitation intensity and elevation. The study provides a theoretical basis for the seismic design of living stumps slopes
Effect of Si addition on phase structure and wear resistance of CoCrFeMoNi alloy coatings
Get PDFCoCrFeMoNi high entropy alloy coating was prepared on Q235 substrate by plasma cladding method. The phase structure, morphology characteristics, element distribution, microhardness, and wear resistance for this alloy without and with Si doping were investigated by XRD, OM, SEM, EDS, microhardness tester, and friction-wear tester, respectively. The results show that CoCrFeMoNi alloy is composed of a single FCC phase, while Si-containing alloy is composed of FCC main phase and HCP phase. Both alloys have a typical dendritic structure. There is a layer of isotropic fine-grained region near the fusion line, and a columnar crystal region away from the fusion line. After adding Si element, the enrichment of Mo element in the interdendrite region and Co element in the dendrite region significantly decreased, which is related to the Si-containing alloy can provide a liquid environment with longer duration, lower viscosity, and greater fluidity. The change of Cr element enrichment from interdendrite region to dendrite region is the result of comprehensive competition of mixing enthalpy, atomic radius difference, electronegativity, density, and melt flowability between alloying elements. The friction coefficients of the two alloys show a rapid increase first and then gradually stabilize with the increase of time. After adding Si element, the hardness and wear resistance of the alloy are greatly improved, which is mainly related to the increase of the lattice distortion of FCC phase, the formation of high-strength HCP phase and the reduction of internal defects
Structure design and sensitivity analysis of flexible ultrasonic transducer array
Get PDFTo investigate the influence of element parameters on the performance and acoustic field of flexible ultrasonic transducer arrays, this study employs finite element multiphysics simulation software to analyze various parameters of flexible ultrasonic transducers within a multiphysics coupled field. The analysis begins with simulating the width and thickness of piezoelectric materials in a single-element ultrasonic transducer structure. Simulation results indicate that the electromechanical coupling coefficient of the ultrasonic transducer exhibits a quasi-sinusoidal relationship with width. When the piezoelectric material width is 1.8 mm, the electromechanical coupling coefficient reaches its maximum at a thickness of 0.4 mm. Subsequently, simulations were conducted on various parameters of the flexible ultrasonic transducer array. Key investigations included the effects of piezoelectric unit count, inter-unit spacing, and frequency on the ultrasonic focusing performance of linear phased array transducers. Findings indicate that the focusing capability of flexible ultrasonic transducer arrays improves with reduced spacing and increased unit count. However, due to varying practical application requirements and manufacturing precision constraints, array parameters should be selected by comprehensively considering real-world factors. Overall, this study employs multiphysics coupling simulation to visually demonstrate how array element parameters influence the performance of flexible ultrasonic transducers. It provides valuable reference for advancing flexible ultrasonic technology from laboratory research toward commercial application
Improvement of the system for reporting the state of the electrified railway contact line
Get PDFEnsuring the stable and reliable operation of electrified railways requires continuous monitoring of the overhead contact line (CL), whose mechanical displacement under wind loads can lead to interruptions in power transmission, pantograph detachment, and safety hazards. Traditional inspection and monitoring systems are limited in responsiveness and cannot provide real-time information about the dynamic state of the CL. This study presents an improved contact line deviation reporting system based on distributed Signal Processing and Transmission Modules (SPTM) and Signal Reception Modules (SRM) connected through a GSM wireless communication network. Each vibration sensor installed on the catenary wire continuously measures the displacement amplitude, converts the analog signal into digital form, and transmits it to the dispatcher or driver in real time. The developed modules were implemented using microcontrollers with embedded wireless interfaces, allowing autonomous operation powered by solar-assisted batteries and ensuring electromagnetic protection under high-voltage (25 kV) conditions. Field experiments were carried out on an electrified railway test section near the Tashkent depot to evaluate the system’s performance in real environmental conditions – including wind speeds of 5-18 m/s, ambient temperatures from –10 °C to +38 °C, and during snow and rain. The results confirmed stable data transmission up to 1 km distance with signal delay below 0.8 s and detection accuracy above 95 %. The proposed system thus enables real-time monitoring, automatic warning, and high reliability of communication even under harsh weather conditions, significantly improving the safety and efficiency of train operation. The novelty of this work lies in the practical validation of a GSM-based monitoring network for contact line deviation detection that integrates autonomous power supply, environmental robustness, and real-field reliability testing – aspects that are rarely demonstrated in previous studies
Rocking responses of free-standing rigid blocks on flexible foundation with a slope under earthquakes: structure-foundation interaction effects
Get PDFTo simulate the dynamic rocking responses of free-standing rigid blocks on a flexible foundation with a slope, a modified concentrated spring model (CSM) is proposed. The shaking table model test of a rigid steel frame structure was conducted, and the test results indicated that the rocking response of the structure can be effectively predicted by the proposed model. Utilizing the modified CSM, the interaction effects between the structure and foundation on the rocking of the free-standing rigid blocks on a flexible foundation with a slope were investigated. The influences of spring stiffness, viscous damping, and the frictional coefficient on the rocking of the rigid blocks under pulse-type ground motions were analyzed numerically. The results indicated that decreasing the spring stiffness and increasing the viscous damping coefficient of the modified CSM can reduce the number of impacts and rocking duration, while the friction coefficient has a significantly non-linear effect on the rocking response
Muti-objective optimization of tuned liquid column damper design parameters for vibration control under wind load
Get PDFTuned Liquid Column Dampers (TLCDs) are widely used as passive devices for vibration control in structures dominated by wind loads, utilizing the oscillation of liquid in a U-shaped container to dissipate energy. The effectiveness of TLCDs is significantly influenced by key design parameters, like mass ratio, tuning frequency ratio, and head loss coefficient. This study developed governing equations of TLCDs and investigated the influence of external load excitation spectrum on the vibration mitigation performance. A multi-objective optimization approach based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II) was proposed to identify optimal design parameters of TLCDs under various loading conditions. The results revealed that the external excitation spectrum played a crucial role in determining the optimal parameters, and the damper performance was distinctly different when the excitation frequencies changed. This optimization method was validated through a 200-meter high tower, where the optimized TLCD significantly enhanced vibration control performance at a wide range of wind directions. These findings offered valuable insights for the application of TLCDs in complex environments with varying external load characteristics
Study of the aerodynamics of a full-scale tractor-trailer
Get PDFThe wind tunnel campaign evaluated 10 configurations of a tractor-trailer combination using a Volvo VNL860 tractor and the 30 ft NRC trailer. The test results include balance measurements of drag force, side force, and yawing moment. The baseline tractor-trailer combination had a blockage-corrected drag coefficient of 0.457 at 0° yaw and a wind-averaged drag coefficient of 0.527. The combined removal of the side extenders, their extensions, and the roof air deflector (in Configurations 6 and 7) resulted in the largest increase in drag (6 % relative to baseline) of all of the aerodynamic packages that were evaluated. The smallest change in drag coefficient resulted from the removal of the tractor skirt extensions (aerodynamic package 4), but it should be noted that the tractor skirt extensions were only evaluated at 0 yaw and would have more of an effect at higher yaw angles. The purpose of this paper is to study the drag coefficient of the entire Tractor through wind tunnel experiments. At present, there are few tests on the resistance coefficient of Tractors, and there are also few published papers on this topic. By removing Aerodynamic packages, this helps to understand the impact of different Aerodynamic packages on the overall drag coefficient of the vehicle
Small targets detection in low-resolution remote sensing images based on super-resolution joint optimization
Get PDFWhile convolutional neural networks have driven remarkable progress in remote sensing object detection, persistent challenges remain in detecting small targets within low-resolution imagery due to their limited pixel representation and feature degradation during hierarchical downsampling. To address this, this study proposed the joint super-resolution and detection network (JSRDN), which synergistically optimizes SR reconstruction through task-specific detection feedback, significantly enhancing small target recognition in LR remote sensing imagery. Firstly, generator in generative adversarial network incorporates improved residual blocks, enabling enhanced perception of complex deep-level features in the SR reconstruction process. Then, a perceptual loss function is introduced into the adversarial training process, which captures perceptual discrepancies in high-level features between reconstructed images and original HR references. After that, an edge-enhancement network is designed to dynamically detect edges in intermediate features restored by the generator, prioritizing edge influence across network layers to generate discriminative features for target recognition. Furthermore, the JSRDN implements detection-driven feedback by backpropagating object recognition loss through the generator, enforcing the super-resolution process to prioritize detection-salient feature recovery. Evaluated on 64×64 low-resolution COWC datasets, JSRDN achieves 0.1819 dB peak signal-to-noise ratio (PSNR) and 7.18 % average precision (AP) improvements over the deep residual dual-attention network (DRDAN), with ablation studies and visualizations confirming its balanced optimization of reconstruction fidelity and detection-oriented feature learning. This technology can provides valuable support for small target measurement and opens new opportunities in the field