Analysis of steel corrosion‐induced surface damage evolution of magnesium oxychloride cement concrete through gray‐level co‐occurrence matrices. (29th June 2020)

• Record Type: Journal Article
• Authors: Wang, Penghui; Qiao, Hongxia; Li, Yuanke; Feng, Qiong; Chen, Kefan
• Is Part Of: Structural concrete. Volume 21:Number 5(2020)
• Page Start: 1905
• Page End: 1918
• Publication Date: 2020-06-29
• Subjects: color heat maps -- gray‐level co‐occurrence matrices -- MOCC -- region of interest -- surface cracks; Reinforced concrete -- Periodicals; 624.1834
• DOI: 10.1002/suco.202000182
• Languages: English
• ISSNs: 1464-4177

Abstract

Abstract: In order to get the characteristics of crack formation on a concrete surface, a coupled method of industrial cameras and gray‐level co‐occurrence matrices (GLCM) is used to analyze the formation of magnesium oxychloride cement concrete surface cracks. This method requires fewer samples than the previous edge detection method, typing analysis, and artificial neural network method and can provide more damage parameter information without being limited to the geometric information of local features. According to the GLCM analysis of the surface crack formation of concrete under the effect of a rust expansion force, it is concluded that the larger the concrete protective layer, the longer the energization time required. When the concrete surface crack reaches 0.2 mm, the test pieces with protective layers of 14, 24, 34, and 44 mm need to be energized for 270, 465, 673, and 792 hr, respectively. With the increase of cracks on the concrete surface, the contrast keeps increasing, while the correlation, energy, and homogeneity gradually decrease. In color heat maps, the number of matrix peaks gradually increases with the formation of cracks, and the bandwidth of the horizontal projection of matrix peaks gradually decreases, and there is a tendency to extend along the main diagonal of the matrix. The selection of the size of the region of interest region will have a certain impact on the size of the four eigenvalues in GLCM, but it will not change its development law.