INVESTIGATION OF NATURAL FREQUENCIES REDUCTION, AN APPROACH TO DAMAGE DIAGNOSIS IN STRUCTURES

摘要

This paper present an approach based on reduction of natural frequencies of structures in order to identification of damage. In the proposed method, the crack identification procedure consists of three steps. Firstly, using finite element method, three natural frequencies of a variable cross-section beam for different cracks depth and locations are obtained. In the second step, two types of networks are created and trained. The neural networks inputs are first three natural frequencies and the outputs of first and second networks are corresponding locations and cracks depth, respectively. In third step, some of natural frequencies of variable cross-section beam with distinct crack conditions are applied as inputs for trained neural networks. Finally obtained results of two types of neural networks are compared with each other.