The oxide film of AFA-OC6 in sCO 2 was mainly composed of thin and continuous Al 2O 3 and (Cr, Mn) 3O 4 at low temperatures or after short exposure time, while the oxide film showed a complex multilayer structure as the temperature and exposure time increased 31. 30 compared the CO 2 pressure compatibility of several commercial Fe- and Ni-based structural alloys, and found that the mass gain of Al-containing materials was the lowest.
Alumina (Al 2O 3) owns a lattice of corundum type, which is the same as Cr 2O 3, while the thermodynamic stability of Al 2O 3 is higher 22 and is expected to offer better protection to the materials 27, 28, 29 exposed to high temperature and corrosive environment. Previous research showed that the mass gain of AFA steels in 800 ☌ air 25 and supercritical water 26 is quite low for the reason that a continuous Al 2O 3 layer was formed. Thus, the alumina-forming austenitic (AFA) stainless steels that initially developed to improve the creep resistance 19, 20, 21, 22, 23, 24 have attracted more and more attention. To solve this problem, a material that not only owns high oxidation resistance in sCO 2, but also keeps the advantages of easy processing and low cost, is needed. Large area spallation of oxide film and many porosities were observed on the surface of 310 and 316 stainless steels exposed to sCO 2 for only 500 h 10, which cannot meet the requirements for the applications in sCO 2-cooled nuclear reactor, especially the cladding materials. But the stability of these Cr oxide films in high-temperature sCO 2 is still insufficient 10, 16, 17, 18. For the austenitic stainless steels and F/M steels exposed to low-temperature environment (such as subcritical water), Cr 2O 3 and Cr-containing oxide layers are formed on the surface, which plays the most important protective role 15. The thickness of oxide film on T22 steel was beyond 32 μm after 200 h exposure in 550 ☌ sCO 2 9. High-temperature corrosion resistance of F/M steels is poor 14. Among them, Ni-based alloys have high radioactive residue, but their economic cost is too high to be applied on a large-scale 13. However, the failure of materials under the operating environment has gradually become one of the key issues that limit the development of the sCO 2 system 8.Ĭurrently, the conventional structural and cladding materials that may be used in sCO 2 cooled nuclear reactor mainly include ferritic/martensitic (F/M) steel 9, austenitic stainless steel 10, 11, and nickel-based alloy 12. The sCO 2 cooled nuclear reactor has become one of the most promising Generation IV nuclear reactors 3, 4, 5, 6, 7.
#Texture transformer 2 crack crack
We also built a comprehensive pavement crack dataset containing 156 high-resolution manually annotated CCD images and made it publicly available on Zenodo.With the advantages of high compact ability, good compressibility, and high heat transfer efficiency 1, 2, supercritical carbon dioxide (sCO 2) has been considered as a potential fluid for different energy systems, such as nuclear reactors.
Specifically, the average precision, recall, ODS, IoU, and frame per second (FPS) of the LETNet on three testing datasets are approximately 93.04%, 92.85%, 92.94%, 94.07%, and 30.80FPS, respectively. The quantitative comparison demonstrates that the proposed LETNet outperformed four advanced deep learning-based models with respect to both efficiency and effectiveness. In addition, a defect rectification module is further developed to reinforce the network for hard sample recognition. To take advantage of these rich features, a skip connection strategy and an efficient upsampling module is built to restore detailed information.
By designing a convolution stem and a local enhancement module, both low-level and high-level local features can be compensated. In the LETNet, Transformer is employed to model long-range dependencies. Therefore, we propose a locally enhanced Transformer network (LETNet) to completely and efficiently detect pavement cracks.
Even though convolutional neural networks (CNNs) have achieved impressive performance in this task, the stacked convolutional layers fail to extract long-range contextual features and impose high computational costs. Precisely identifying pavement cracks from charge-coupled devices (CCDs) captured high-resolution images faces many challenges.
0 kommentar(er)
