Search Results (12)

Rail surface defects pose significant risks to the operational efficiency and safety of industrial equipment. Traditional visual defect detection methods typically rely on high-quality RGB images; however, they struggle in low-light conditions due to small, low-contrast defects that blend into complex backgrounds. Therefore, this paper proposes a novel defect segmentation method leveraging a dual-stream feature fusion network that combines polarization images with DeepLabV3+. The approach utilizes the pruned MobileNetV3 as the backbone network, incorporating a coordinate attention mechanism for feature extraction. This reduces the number of model parameters and enhances computational efficiency. The dual-stream module implements cascade and addition strategies to effectively merge shallow and deep features from both the original and polarization images. This enhances the detection of low-contrast defects in complex backgrounds. Furthermore, the CBAM is integrated into the decoding area to refine feature fusion and mitigate the issue of missing small-target defects. Experimental results demonstrate that the enhanced DeepLabV3+ model outperforms existing models such as U-Net, PSPNet, and the original DeepLabV3+ in terms of MIoU and MPA metrics, achieving 73.00% and 80.59%, respectively. The comprehensive detection accuracy reaches 97.82%, meeting the demanding requirements for effective rail surface defect detection. Full article

The rail transit switch machine ensures the safe turning and operation of trains on the track by switching switch positions, locking switch rails, and reflecting switch status in real time. However, in the detection of complex rail transit switch machine parts such as augmented reality and automatic inspection, existing algorithms have problems such as insufficient feature extraction, large computational complexity, and high demand for hardware resources. This article proposes a complex scene rail transit switch machine parts detection network YOLO-SMPDNet (YOLO-based Switch Machine Parts Detecting Network). The YOLOv8s backbone network is improved, and the number of network parameters are reduced by introducing MobileNetV3. Then a parameter-free attention-enhanced ResAM module is designed, which forms a lightweight detection network with the improved network, improving detection efficiency. Finally, Focal IoU Loss is introduced to more accurately define the scale information of the prediction box, alleviate the problem of imbalanced positive and negative samples, and improve the relative ambiguity of CIoU Loss in YOLOv8s on the definition of aspect ratio. By validating the performance of YOLO-SMPDNet on a self-made dataset of rail transit switch machines, the results show that YOLO-SMPDNet can significantly improve detection accuracy and real-time performance and has robust comprehensive detection capabilities for rail transit switch machine parts and good practical application performance. Full article

To lighten the workload of train drivers and enhance railway transportation safety, a novel and intelligent method for railway turnout identification is investigated based on semantic segmentation. More specifically, a railway turnout scene perception (RTSP) dataset is constructed and annotated manually in this paper, wherein the innovative concept of side rails is introduced as part of the labeling process. After that, based on the work of Deeplabv3+, combined with a lightweight design and an attention mechanism, a railway turnout identification network (RTINet) is proposed. Firstly, in consideration of the need for rapid response in the deployment of the identification model on high-speed trains, this paper selects the MobileNetV2 network, renowned for its suitability for lightweight deployment, as the backbone of the RTINet model. Secondly, to reduce the computational load of the model while ensuring accuracy, depth-separable convolutions are employed to replace the standard convolutions within the network architecture. Thirdly, the bottleneck attention module (BAM) is integrated into the model to enhance position and feature information perception, bolster the robustness and quality of the segmentation masks generated, and ensure that the outcomes are characterized by precision and reliability. Finally, to address the issue of foreground and background imbalance in turnout recognition, the Dice loss function is incorporated into the network training procedure. Both the quantitative and qualitative experimental results demonstrate that the proposed method is feasible for railway turnout identification, and it outperformed the compared baseline models. In particular, the RTINet was able to achieve a remarkable mIoU of 85.94%, coupled with an inference speed of 78 fps on the customized dataset. Furthermore, the effectiveness of each optimized component of the proposed RTINet is verified by an additional ablation study. Full article

U75V heavy rail steel production uses an aluminum-free deoxidation process; however, large particles of MgO–Al₂O₃ inclusions form in the steel, which has a great impact on product quality. In this paper, we try to explain how spinel inclusions, which affect the metallurgical quality of heavy rail steel, are produced by thermodynamic and experimental methods, and then determined measures for avoiding such inclusions. The formation mechanism of spinel inclusions in U75V heavy rail steel was determined through the analysis of nozzle clogging in the pouring process and typical inclusions in steel. The results show that there are two types of spinel inclusions in heavy rail steel: one is pure Mg–Al spinel inclusions and the other is Mg–Al spinel inclusions coated with calcium aluminate. The small, pure Mg–Al spinel inclusions were precipitated during the solidification of the molten steel, and the precipitation temperature was related to the composition of the molten steel. The large spinel inclusions were derived from clogging of the submersed nozzle. Mg–Al spinel inclusions coated with calcium aluminate were transformed from CaO–SiO₂–Al₂O₃–MgO complex inclusions in the steel during cooling, and the formation temperature was related to the content of Al₂O₃ and MgO in the inclusions. The content of Al₂O₃ and MgO in the inclusions was the key to the formation of the Mg–Al spinel inclusions. Therefore, in order to control the production of spinel inclusions in steel, it is necessary to strictly control the content of impurity elements such as magnesium and aluminum in the alloy auxiliary materials, to reduce the secondary oxidation of liquid steel and to reduce the erosion of refractory materials. Full article

The damage mode of U75V rail steel in application is determined by its rolling wear behavior. In this paper, the pearlite microstructure of U75V steel is characterized to investigate the relationship between wear and fatigue behavior. The results show that, with decreasing of pearlite lamellar spacing, the wear resistance of the steel increases and the contact fatigue resistance decreases. The spacing decreasing causes the change of the wear mechanism from abrasive wear to adhesive wear, as well as the damage mode from wear damage to fatigue damage. The smaller the pearlite lamellar spacing is, the stronger the deformation of the cementite lamellar is. The thin cementite lamellar is hardly broken in the rolling friction to pile up a large number of dislocations in the ferrite matrix and the work hardening degree was improved. So, the plastic deformation layer is difficult to remove, and fatigue cracks are easy to initiate and extend to the interior of the material. Full article

U75V rail steel is widely used in railways in China, including train tunnels in mountain and subsea projects, where it suffers from selective corrosion near welded joints. To ensure adequate railway service life, this study examines the effect of the normalisation process on the electrochemical behaviour of U75V rail-welded joints (URWJs) manufactured by flash butt welding (FBW) using potentiodynamic polarisation and electrochemical impedance spectroscopy (EIS). Corrosion morphology and elemental distribution analyses were performed to investigate the corrosion behaviour. The results show that the grains within the joints became finer and more homogeneous after normalisation, with a lower corrosion rate and higher corrosion resistance. It is demonstrated that fewer corrosion products were formed on the surface of the normalised joints after electrochemical test, and the corrosion resistance of the URWJs improved, owing to the formation of denser passivation films caused by normalisation. These mechanisms of corrosion response help explain corrosion failure in railway lines, as well as also help optimise the welding process and normalising processes to obtain a corrosion-resistant microstructure and ensure the quality of welded joints. Full article

To study the effect of Ce on the morphology of manganese sulfide, we added different contents of Ce into U75V heavy rail steel. The composition and morphology of sulfide in steel were analyzed. The inclusions’ number, size, and aspect ratio were analyzed by automatic scanning electron microscope ASPEX. The results show that the inclusions in heavy rail steel without Ce are elongated MnS and irregular Al-Si-Ca-O inclusions. With the increase of Ce from 52 ppm to 340 ppm, the composition of main inclusions changes along the route of Ce₂O₂S-MnS → Ce₂O₂S-MnS-Ce₂S₃ → Ce₂O₂S-Ce₃S₄-Ce₂S₃ → Ce₂O₂S-Ce₃S₄-CeS. Ce has a noticeable spheroidization effect on MnS, which can make inclusions finely dispersed. When Ce content is 139 ppm, the average size of inclusions is the smallest. The mechanism of Ce-modified MnS was discussed by combining experimental results with thermodynamic calculations. Finally, the effect of Ce treatment on inhibiting MnS deformation was verified by simulated rolling. Full article

Surface temperature is a critical operating parameter that influences the phase transition time and rolling quality of U75V rail steel plates in the rolling process. There is still no extensive online detection system for the surface temperature of rail steel plates due to the hazardous environment, incorrect surface emissivity, and complex backgrounds. In this paper, online surface temperature detection equipment based on multi-spectral photography was built for high-temperature rail steel plates in the rolling processes. Then, the emissivity model for a high-temperature environment, colorimetric thermometry, and noise filtering methods were investigated to improve the accuracy of the temperature detection results of rail steel plates. Finally, the surface temperature of the U75V rail steel plate during three rolling passes could be calculated online point by point, and the greatest error was approximately 0.82% due to the blackbody calibration experiments. The results not only have a positive effect on understanding the temperature declination process of low-alloy rail steel plates during the rolling process, but could also benefit the control of the cooling rate and optimize the rolling model during rolling passes. Full article

Laser cladding is considered to be a highly complex process to set up and control because it involves several parameters, such as laser power, laser scanning speed, powder flow rate, powder size, etc. It has been widely studied for metal-part coating and repair due to its advantage in controllable deposited materials on a small target substrate with low heat-affected distortion. In this experiment, laser cladding of U75V and U20Mn rail steels with Inconel 625 powder was captured by an infrared camera with image analysis software to monitor the laser cladding process in order to determine the quality of the cladded substrates. The cladding temperature, thermal gradient, spot profile, and cooling rate were determined from infrared imaging of the molten pool. The results showed that cladding temperature and molten pool’s spot closely related to the laser cladding process condition. Infrared imaging provided the cooling rate from a temperature gradient which was used to correctly predict the microhardness and microstructure of the HAZ region. This approach was able to effectively detect disturbance and identify geometry and microstructure of the cladded substrate. Full article

In the present paper, the continuous-drive friction welding (CDFW) technology has been successfully applied to join the U75V rail steel. The base metal (BM) of U75V rail steel is lamellar pearlite, and the weld zone could be clearly divided into three subzones (i.e., heat affected zone, thermo-mechanical affected zone (TMAZ), and central weld zone (CWZ)). Electron back-scattered diffraction examinations revealed the martensitic evolution in TMAZ and CWZ, suggesting that the experienced high temperature, severe plastic deformation, and fast cooling rate induce the microstructure transition during the CDFW process. The hard and brittle martensite structure explains the raised microhardness profiles and the reduced impact absorption energy of the as-welded joints. The CDFW process parameters govern the joint properties via influencing the welding heat input and plastic deformation by spindle speed and friction pressure at the friction stage, and the plastic deformation layer (flash) extrusion by upsetting pressure at the upsetting stage. More favorable results could be obtained at small set values of spindle speed (1800 rpm) and friction pressure (75 MPa) with less heat input and plastic deformation, and a large set value of upsetting pressure (175 MPa) with more flash extrusion, whose tensile strength reached 94.3% of that of the BM. Full article

This paper presents methods for continuous condition monitoring of railway switches and crossings (S&C, turnout) via sleeper-mounted accelerometers at the crossing transition. The methods are developed from concurrently measured sleeper accelerations and scanned crossing geometries from six in situ crossing panels. These measurements combined with a multi-body simulation (MBS) model with a structural track model and implemented scanned crossing geometries are used to derive the link between the crossing geometry condition and the resulting track excitation. From this analysis, a crossing condition indicator $C_{{\lambda }_1-{\lambda }_2,\gamma }$ is proposed. The indicator is defined as the root mean square (RMS) of a track response signal γ that has been band-passed between frequencies corresponding to track deformation wavelength bounds of ${\lambda }_1$ and ${\lambda }_2$ for the vehicle passing speed (f = v/$\lambda$). In this way, the indicator ignores the quasi-static track response with wavelengths predominantly above ${\lambda }_1$ and targets the dynamic track response caused by the kinematic wheel-crossing interaction governed by the crossing geometry. For the studied crossing panels, the indicator $C_{1-0.2 m,\gamma }$ (${\lambda }_1=1$ and ${\lambda }_2=0.2$) was evaluated for γ = u, v, or a as in displacements, velocities, and accelerations, respectively. It is shown that this condition indicator has a strong correlation with vertical wheel–rail contact forces that is sustained for various track conditions. Further, model calibrations were performed to measured sleeper displacements for the six investigated crossing panels. The calibrated models show (1) a good agreement between measured and simulated sleeper displacements for the lower frequency quasi-static track response and (2) improved agreement for the dynamic track response at higher frequencies. The calibration also improved the agreement between measurements and simulation for the crossing condition indicator demonstrating the value of model calibration for condition monitoring purposes. Full article

In this paper is given a three-dimensional numerical simulation of the eddy current welding of rails where the longitudinal two directions are not ignored. In fact, usually it is considered a model where, in the two-dimensional numerical simulation of rail heat treatment, the longitudinal directions are ignored for the magnetic induction strength and temperature, and only the axial calculation is performed. Therefore, we propose the electromagnetic-thermal coupled three-dimensional model of eddy current welding. The induced eddy current heat is obtained by adding the z-axis spatial angle to the two-dimensional electromagnetic-thermal, thus obtaining some new results by coupling the numerical simulation and computations of the electric field and magnetic induction intensity of the three-dimensional model. Moreover, we have considered the objective function into a weak formulation. The three-dimensional model is then meshed by the finite element method. The electromagnetic-thermal coupling has been numerically computed, and the parametric dependence to the eddy current heating process has been fully studied. Through the numerical simulation with different current densities, frequencies, and distances, the most suitable heat treatment process of U75V rail is obtained. Full article