Reinforced concrete structural damage diagnosis by using artificial neural network

Typical defects exist in reinforced concrete structures include honeycomb, crack, and scaling and strength deduction for concrete, and corrosion and decreased section area for steel members. These defects are a result of many factors such as unproper construction and maintenance, overloading and environmental impact. The existing of defects certainly weakens the structures and reduces the expected life time of structures. Diagnosis and repair in time would be the responsibility of civil engineers. The purpose of this study is to develop a diagnosing model for reinforced concrete structures by using of backpropagation neural network technique to assess the severity and location of defects. Theoretical analysis of a simply-supported reinforced concrete beam in specified size (i.e., rectangular cross section and 4 meter span) by finite element program is performed to generate training and testing samples for neural network assessing task. The efficiency of the developed neural network model for the damage assessment is verified