Residual strength and stiffness estimation for RC columns damaged in earthquake through surface crack texture analysis

A novel non-destructive non-contact methodology based on surface crack image processing is developed in this paper for residual strength and stiffness estimation of seismically damaged reinforced concrete (RC) columns. The texture geometry measures, such as succolarity, lacunarity, and fractal dimensions, provide quantitative insights into the irregularity and intricacy of the crack patterns. To the authors' knowledge, analyzing the succolarity and lacunarity dimensions of the crack patterns has not been employed for strength and stiffness loss evaluation in RC elements yet, although they have been used in other scientific fields. The extensive database of 425 crack textures collected from cyclic tests on 112 non-circular RC column specimens with diverse structural and geometric features serves as a resource for the development and verification of the suggested method. Based on available structural information and fractal geometry measures extracted from the seismically damaged RC column images, 6 regression-based closed-form equations were developed and the results present the efficiency of these measures for quantifying the complexities of the image as well as estimating the strength and stiffness loss in RC column with various ultimate failure modes. Finally, case studies and applications of the damage assessment of RC columns by the proposed model are presented. This innovative approach holds great promise in replacing traditional assessment methods that rely on subjective visual inspection and human judgment, by implementing an analytical procedure centered around image processing, ultimately leading to a fully automated system.