Search

As a detection tool to identify metal or alloy, metallographic quantitative analysis has received increasing attention for its ability to evaluate quality control and reveal mechanical properties. The detection procedure is mainly operated manually to locate and characterize the constitution in metallographic images. The automatic detection is still a challenge even with the emergence of several excellent models. Benefiting from the development of deep learning, with regard to two different metallurgical structural steel image datasets, we propose two attention-aware deep neural networks, Modified Attention U-Net (MAUNet) and Self-adaptive Attention-aware Soft Anchor-Point Detector (SASAPD), to identify structures and evaluate their performance. Specifically, in the case of analyzing single-phase metallographic image, MAUNet investigates the difference between low-frequency and high-frequency and prevents duplication of low-resolution information in skip connection used in an U-Net like structure, and incorporates spatial-channel attention module with the decoder to enhance interpretability of features. In the case of analyzing multi-phase metallographic image, SASAPD explores and ranks the importance of anchor points, forming soft-weighted samples in subsequent loss design, and self-adaptively evaluates the contributions of attention-aware pyramid features to assist in detecting elements in different sizes. Extensive experiments on the above two datasets demonstrate the superiority and effectiveness of our two deep neural networks compared to state-of-the-art models on different metrics.

Cu2+ exhibited a good activation effect on specularite. However, its microscopic activation mechanism needs further study. Additionally, Cu2+ was mainly present in the flotation solution as Cu2+, Cu(OH)+, and Cu(OH)2 at pH = 7. Therefore, density functional theory (DFT) calculations were used to investigate the effect of Cu species such as Cu2+, Cu(OH)+, and Cu(OH)2 adsorbed on the crystal structure and properties of the specularite (0 0 1) surface. The adsorption mechanism of different Cu components on the surface was also further clarified by the analyses of the adsorption model, adsorption energy, partial density of states (PDOS), charge transfer, and bond properties. In addition, the obtained results are discussed. Based on the obtained results, it can be concluded that the geometric structure and electronic properties on the surface changed after adsorbing Cu components and that the O3–Fe1–O1 structure was more susceptible to the adsorbates. The adsorption engines results show that Cu components could be spontaneously adsorbed onto the specularite (0 0 1) surface with adsorption energies of −0.76, −0.85, and −1.78 eV, corresponding to Cu2+, CuOH+, and Cu(OH)2, respectively. Therefore, the adsorption stability of the Cu species on the specularite surface increased in the order of Cu2+ &lt, Cu(OH)+ &lt, Cu(OH)2. Additionally, the adsorption sites for Cu species on the surface were different. Cu2+ interacted mainly with O atoms on the surface, forming Cu–O complexes, while Cu(OH)+ and Cu(OH)2 acted mainly through the O atom of –OH, interacting with Fe atoms to form Cu–O–Fe complexes. The formation of Cu–O and Cu–O–Fe complexes increased the adsorption sites for sodium oleate, with more hydrophobic species being generated to improve the floatability of specularite.