Crack is one of the common diseases in hydraulic concrete structures. Its traditional safety monitoring models only consider the loading factors such as hydraulic head, temperature and time without taking into account the parameters of the crack itself. The crack depth and initial opening are considered here besides the usual factors based on further analysis of the impacting factors of crack opening displacement (COD) according to Fracture Mechanics. In this way, the improved safety monitoring model can reflect both the mutation and the gradient of crack. Due to the fact that in the usual cases the crack depth and initial opening are not easily detected and measured, a method is designed to inverse them based on the finite element method (FEM) and the artificial neural network (ANN), and then a criterion of abnormality diagnosis for crack is established. Finally, based on the improved safety monitoring model and the criterion of abnormality diagnosis, a case of concrete gravity arc dam is accomplished.
展开▼
