No abstract

In this research, SiC f /SiC-B 4 C composites were prepared through slurry impregnation and reaction melting infiltration (RMI). To explore the damage evolution of this composites after oxidation in wet atmosphere, the changes of damage behavior of composites before and after oxidation were studied using acoustic emission (AE) pattern recognition (PR) technology. Results show that since the damage behavior of the composites changes after oxidation, the optimal number of clusters transforms from 3 to 5. In addition, as the oxidation time increases, the area of concentration of damage events is gradually broadened towards lower stress direction due to the decrease of damage threshold of composites. Meanwhile, because the BN interface is gradually oxidized, the density of C3-cracks (axial yarns cracks) near the fracture and the proportion of Cluster-4 (the signals of transverse yarns cracks) will increase, for which the rising of interfacial shear stress should be responsible. Through this research, not only can the damage evolution of SiC f /SiC-B 4 C composites after oxidation be understood, but also the method employed in this research can be as guide for studying other similar materials.

In this research, for studying the influence of size and heat treatment temperature of initial Al(OH) 3 on the physical properties of porous mullite ceramics, porous mullite ceramics were prepared by in situ reaction sintering of amorphous silica and treated Al(OH) 3 . The transition phases χ-Al 2 O 3 , к-Al 2 O 3 , and stable phase α-Al 2 O 3 can be obtained in turn when the treatment temperature of raw Al(OH) 3 is 500, 1000, and 1500 • C, respectively. The coarser the raw Al(OH) 3 , the higher the strength of porous mullite ceramics. When the sintering temperature is 1500 • C, the bending strengths of PS500-C, PS1000-C, and PS1500-C (PSx-C represents that the specimen was prepared by the coarse grade Al(OH) 3, which was previously treated at x • C) are 40.3 ± 2.1, 54.9 ± 5.2, and 64.8 ± 4.8 MPa, respectively. In addition, although the activated Al 2 O 3 can decrease the formation temperature (∼100 • C) of porous mullite ceramics, the strength and density of porous mullite ceramics prepared by activated Al 2 O 3 will decrease at the same sintering temperature. It is believed that the increase of defects and pores during the phase transformation should be responsible for this phenomenon.

Herein, aluminum titanate (Al2TiO5, AT) flexible ceramics with different cooling rates (program‐controlling cooling (10 and 20 °C min−1), air cooling (≈400 °C min−1), and water cooling (>1000 °C min−1) are fabricated by reaction sintering of Al2O3 and TiO2. The effect of cooling rates on microstructures and the mechanical properties of AT ceramics are systematically investigated. The microcrack width and density of AT ceramics with different cooling rates are statistically calculated. In addition, the effect of a microcracks structure on the mechanical properties of AT ceramics is deeply studied. Results show that the increase in cooling rate will increase the width and density of microcrack, leading to a suitable increase in flexibility and a slight decrease in strength. When AT is cooled in air, the best bending strain of 1.10% can be obtained, with a median microcrack width of 0.533 μm and a density of 6831.88 ± 374.87 N mm−2. It is hopeful that through this research, the effect of cooling rate on the mechanical properties of AT flexible ceramics can be deeply understood, and the strategy used in this research can be used in other exploration of similar material.

Purpose The current popular image processing technologies based on convolutional neural network have the characteristics of large computation, high storage cost and low accuracy for tiny defect detection, which is contrary to the high real-time and accuracy, limited computing resources and storage required by industrial applications. Therefore, an improved YOLOv4 named as YOLOv4-Defect is proposed aim to solve the above problems. Design/methodology/approach On the one hand, this study performs multi-dimensional compression processing on the feature extraction network of YOLOv4 to simplify the model and improve the feature extraction ability of the model through knowledge distillation. On the other hand, a prediction scale with more detailed receptive field is added to optimize the model structure, which can improve the detection performance for tiny defects. Findings The effectiveness of the method is verified by public data sets NEU-CLS and DAGM 2007, and the steel ingot data set collected in the actual industrial field. The experimental results demonstrated that the proposed YOLOv4-Defect method can greatly improve the recognition efficiency and accuracy and reduce the size and computation consumption of the model. Originality/value This paper proposed an improved YOLOv4 named as YOLOv4-Defect for the detection of surface defect, which is conducive to application in various industrial scenarios with limited storage and computing resources, and meets the requirements of high real-time and precision.