In order to improve the detection efficiency and image quality of Si 3 N 4 ceramic bearing balls surface defects, digital image processing technology is used to analyse the information characteristics of Si 3 N 4 ceramic bearing balls surface. A multi-scale decomposition enhancement algorithm for surface defect images of Si 3 N 4 ceramic bearing balls based on the stationary wavelet transform is proposed. By building the surface defects detection system of Si 3 N 4 ceramic bearing balls, the image enhancement program based on stationary wavelet transform with index low-pass filtering and nonlinear transform enhancement is designed. Finally, the effectiveness of the algorithm is verified by experiments. The experimental results show that the algorithm is applied to the surface defects image of Si 3 N 4 ceramic bearing balls can effectively weaken the background noise and surface grinding texture, and enhance the contrast between defects and background clearly. In addition, the binary image is obtained by an adaptive threshold binary algorithm. After removing the tiny points by morphological opening operation, the defects are accurately and completely segmented, and then the Canny operator is used for edge detection to extract the edge contour of defects. When the decomposition level is set to 3, the average calculation time is 0.88 s, which are relatively short and have sufficient precision, and the algorithm can be extended to other kinds of ceramic ball surface damage detection.
Due to the influence of mechanical vibration, high temperature creep and other factors, Si3N4 turbine blades are prone to surface defects. Besides, traditional algorithms are incapable to detect and classify surface defects simultaneously. Aiming at solving these problems, an algorithm for defect detection and classification of Si3N4 turbine blades based on convolutional neural network is proposed. The detection and classification network of this algorithm is optimized based on YOLOv5 network, the PAN structure and FPN structure of YOLOv5 are replaced by BiFPN structure. We establish the dataset of Si3N4 turbine blades, which is expanded by data enhancement. For the purpose of achieving a higher level of feature fusion, the PAN and FPN structures of the Neck part are replaced by BiFPN structure. As a result, the accuracy of detecting and classifying the surface defects by this algorithm is as high as 97.4%, and the detection speed is as low as 16ms. This optimized algorithm is able to solve the problems of traditional detection methods such as heavy workload, long time consuming and low accuracy. The algorithm provides a feasible approach for the quality detection of Si3N4 turbine blades and has certain engineering application value.
The sandstone from the dry to saturated state shows obvious deterioration characteristics. Taking the sandstone of a slope in the Three Gorges Reservoir area as the research object, uniaxial/triaxial compression of sandstone samples with different water-bearing states (dry, natural, and saturated) is carried out to study the changes in macromechanical properties of sandstone under different water-bearing states. Combined with NMR and SEM, the characteristics of microstructure of sandstone under different moisture conditions were studied. The results show that, with the increase of water content, the macromechanical parameters of sandstone gradually decrease, and the fine and microstructure characteristics are characterized by the gradual increase in the number of pores and the gradual increase in pore size. Based on the PFC2D software, considering the weakening effect of water on the partial cementation from a mesoscopic point of view, it is proposed to use soft and hard contacts to simulate the changes in the degree of cementation between particles under different water-bearing conditions and to study the impact of sandstone micromechanical parameters with changes in water content. Related research results can provide theoretical guidance for the stability evaluation of wading rock mass engineering.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.