Evaluation and prediction on abrasion resistance of hydraulic concrete after exposure to different freeze-thaw cycles

• Published in: Construction & building materials 2022-01, Vol.316, p.126055, Article 126055
• Main Authors: Zhu, Xiangyi, Bai, Yin, Chen, Xudong, Tian, Zhenghong, Ning, Yingjie
• Format: Article
• Language: English
• Subjects: Fractal dimension; Freeze-thaw damage; Hydraulic abrasion; Projection pursuit regression; Surface morphology indexes
• ISSN: 0950-0618; 1879-0526
• DOI: 10.1016/j.conbuildmat.2021.126055

•FT cycle and hydraulic abrasion tests were carried out on hydraulic concrete.•Effect of different FT cycles on abrasion resistance of the concrete was studied.•A prediction model for the abrasion resistance was established by PPR theory. Considering that the hydraulic concrete structures in cold regions will suffer from the combined action of freeze-thaw (FT) damage and hydraulic abrasion, the morphology indexes such as the abrasion depth, abrasion volume and fractal dimension of hydraulic concrete subjected to different FT cycles (0, 100, 200, 300 and 400 FT cycles) are determined by the underwater method. On this basis, a prediction model for the abrasion resistance of concrete subjected to different FT damage is established by using the projection pursuit regression (PPR) theory. The results show that with the aggravation of FT damage, the number of isolated matrices and abrasion depth on the concrete surface continue to increase, and the surface roughness parameters increase exponentially, which significantly affects the evolution of abrasion depth, abrasion volume and fractal dimension of hydraulic concrete with abrasion time, showing a continuous decline in abrasion resistance. And a power function is used to establish the prediction models of the morphology indexes of hydraulic concrete subject to different FT cycles, which have high prediction accuracy. In addition, the relationship functions between abrasion depth, abrasion volume and fractal dimension are proposed, and both of them have a good linear correlation. Eventually, the PPR model based on morphology indexes is proved to be a promising method to predict the abrasion performance of hydraulic concrete structures in cold regions.