Rapid post-earthquake damage evaluation of RC shear walls in resilient community

The post-earthquake damage evaluation is usually conducted by visual inspection, which is unsafe and subject to human error. In this paper, a rapid non-destructive methodology is developed that uses crack images to assess damage in rectangular reinforced concrete shear walls. This computer vision-oriented method is based on a databank of surface crack maps collected by the authors from quasi-static experiments. Succolarity, Lacunarity, and fractal dimensions of the specimens are used to measure the complexity of crack patterns. Fifteen closed-form equations are proposed that use the structural characteristics and crack map indices obtained from damaged wall images to estimate the peak drift ratio that the shear walls experience during an earthquake. Results reveal that the predictive equations which are the nonlinear combination of the fractal geometry have the highest accuracy among the equations. The accuracy of the equation is slightly increased when concrete compressive strength is added to the input variables.