Search

To solve the problems of poor mixing effect and poor sphericity for the silicon nitride particles in the dry granulation chamber, the influence of the inclination parameters for the open turbine typed axial‒radial combined structure on the mixing effect of silicon nitride particles prepared by dry method was studied. The mathematical model of Eulerian‒Eulerian gas‒solid two phase flow was established, and the physical model of the axial‒diameter combined structure granulation chamber was simplified. The computational fluid dynamics (CFD) method was used to analyze the volume distribution and velocity field of silicon nitride particles in granulation chamber in the axial structure under the different inclined angles. The sphericity of silicon nitride particles was tested by experiments, and the accuracy of numerical simulation results was verified. The numerical analysis results show that, when the inclination angle is 45°, the volume fraction of silicon nitride particles is 0.3~0.7 (accounting for 28%), showing the better circulating fluidity and the best mixing effect, compared with the inclination angle of 30° and 60°. The experimental results show that, the best average sphericity of the silicon nitride particles is 0.72 when the inclination angle of the axis‒diameter combined structure is 45°, which can effectively improve the mixing effect of silicon nitride particles.

The paper partially covered Active Constrained Layer Damping (ACLD) cantilever beams' dynamic modeling, active vibration control, and parameter optimization techniques as the main topic of this research. The dynamic model of the viscoelastic sandwich beam is created by merging the finite element approach with the Golla Hughes McTavish (GHM) model. The governing equation is constructed based on Hamilton's principle. After the joint reduction of physical space and state space, the model is modified to comply with the demands of active control. The control parameters are optimized based on the Kalman filter and genetic algorithm. The effect of various ACLD coverage architectures and excitation signals on the system's vibration is investigated. According to the research, the genetic algorithm's optimization iteration can quickly find the best solution while achieving accurate model tracking, increasing the effectiveness and precision of active control. The Kalman filter can effectively suppress the impact of vibration and noise exposure to random excitation on the system. [ABSTRACT FROM AUTHOR]

The semantic segmentation network based on U-Net (U-shaped network structure) in Si3N4 ceramic bearing cylindrical roller surface defect segmentation has the problems that low accuracy of defect feature extraction and large number of model parameters. It cannot meet the segmentation accuracy requirements of different size defects. The semantic segmentation network (D-A-IU-Net-Semantic segmentation network model with embedded multi-scale defective features) with embedded multi-scale defective feature fusion module is proposed. D-A-IU-Net inherits the encoding–decoding features of U-Net semantic segmentation network in structure and has the ability to sense multi-layer defective features. To reduce the number of model parameters and the network model training time, the D-A-IU-Net network model is downsampled only three times. To address the loss of detailed semantic information during downsampling, the network adopts a depthwise separable convolution layer with stride = 2 instead of the maximum pooling layer in the coding module. The segmentation accuracy is further improved. For overcoming the low segmentation accuracy of tiny defects on Si3N4 ceramic bearing cylindrical roller surface, the Atrous Spatial Pyramid Pooling Module (ASSP module-Hollow space convolution pool pyramid) is added to the bottom of the D-A-IU-Net network model to replace the model structure at the bottom of the U-net network model. To avoid the performance degradation caused by the increasing depth of the network model, two identity blocks are added at the bottom. A dataset of Si3N4 ceramic bearing cylindrical roller surface defects is prepared for validation. The experimental results show that the mIoU (mean Intersection over Union) of the D-A-IU-Net network model reaches 90.70%, the mPA (mean Pixel Accuracy) is improved to 95.13%, and the single image prediction time is 46.53 ms. The region segmentation efficiency of Si3N4 ceramic bearing cylindrical roller surface defects is improved. [ABSTRACT FROM AUTHOR]

To solve the problems of noise coverage defect and low contrast between the defect and the background of ZrO2 ceramic bearing balls, a surface defect extraction algorithm based on shearlet transform image enhancement for ZrO2 ceramic bearing balls is proposed. According to the shape characteristics of ceramic bearing balls, the surface defect image acquisition platform is built to collect and analyze surface defect images. Gaussian filtering weakens the scatter-particle noise in the image, and the threshold corrects the coefficient generated by the shearlet transform. After shearlet transform, the relatively low-frequency and high-frequency parts appear. The low-frequency part reflects the edge information of defects, and the high-frequency part reflects the edge and texture information of defects. Thus, the integrity of the defect is ensured, and an enhanced surface defect image is obtained. The gray histogram of the enhanced image is observed. The optimal threshold is selected by the histogram threshold segmentation method, and the process of defects being completely extracted from the background is realized. Experimental results showed that the extraction rates of pits, scratches, and cracks in ZrO2 ceramic bearing balls' surface images are 95.00%, 92.50%, and 92.50%, respectively. [ABSTRACT FROM AUTHOR]

This investigation centers on exploring the plastic deformation removal mechanism and dislocation dynamics in the nanogrinding of single‐crystal silicon carbide featuring a randomly rough surface. To simulate this, an approach combining the Weierstrass–Mandelbrot fractal surface function with molecular dynamics is employed. Randomly rough surface contours are generated using the Weierstrass–Mandelbrot fractal surface function. By adjusting the fractal dimension, an ideal representation of single‐crystal silicon carbide with randomly rough surfaces is obtained. By combining plastic deformation detection methods, the process of workpiece sliding, plowing, and chip removal is studied, and the plastic deformation removal mechanism of random rough surface grinding is explored. Combining postprocessing methods such as sheer strain and dislocation trajectory, the morphological changes and transformation mechanisms of dislocations are analyzed. Notably, at grinding depth of 33.18 nm, the activation of slip systems induces dislocation formation, enabling plastic deformation. Furthermore, at depths exceeding 144.00‐nm, the emergence of stacking faults on the random rough surface is observed. Throughout the grinding procedure, plastic deformation of debris occurs, leading to the formation of plastic bulges on both sides of the tool, and the debris generated during processing does not impact the defects encountered during secondary grinding of the rough surface. [ABSTRACT FROM AUTHOR]

To extract the fuzzy contour features, tiny depth features of surface microcracks in the Si3N4 ceramic bearings roller. An adaptive nano feature extraction and multiscale deep fusion coupling method is proposed, to sufficiently reconstruct the three‐dimensional morphology characteristics of surface microcracks. Construct an adaptive nano feature extraction method, form the surface microcrack image scale space and the Gaussian difference pyramid function equation, realize the detection and matching of global feature points. The sparse point cloud is obtained. Through polar‐line correction, depth estimation, and fusion of feature points on the surface microcracks image, a multiscale depth fusion matching cost pixel function is established to realize a dense point cloud reconstruction of surface microcracks. The reconstruction results show that the highest value of the local convex surface reconstructed by the dense point cloud reaches 1183 nm, and the lowest local concave surface is accurate to 296 nm. Compared with the measurement results of the confocal platform, the relative error of the reconstruction result is 24.6%. The overall feature‐matching rate of the reconstruction reaches 93.3%. It provides a theoretical basis for the study of surface microcrack propagation mechanism and the prediction of bearing life. [ABSTRACT FROM AUTHOR]