Amir Banimahd - Ardakan University, Faculty of Engineering, Khatami Blvd
Navid Soltani - Ardakan University, Faculty of Engineering, Khatami Blvd.

An analytical model is utilized for identification of cracks in pipes with the circular hollow section. The model simulates a circumference transverse crack as a massless spring at the crack location, which is developed based on the fracture mechanics [1]. Therefore, the pipe can be divided into two parts in the left and right hand of the crack, see Fig. 1&2. The mass matrix of the pipe is also represented according to the consistent mass matrix definition and assumed unchanged after the crack occurrence. The dynamic approach is used in order to identify the crack location and size in a pipe. Moreover natural frequencies of the analytical model is used instead of the experimental data [2]. Therefore, the dynamic properties of a cracked pipe is determined. The main aim of the study is finding the crack identification using the dynamic behavior of the cracked pipe [3], which can be defined as a reverse problem. Solving such a problem with the gradient method may be an expensive and time-consuming procedure. The problem can be solved by an iterative numerical optimization method. Among numerical optimization techniques, the particle swarm optimization is an efficient and fast convergence method which is applyed to find the unknown parameters as location and depth of cracks in the study.

Crack identification, PSO Algorithm, Hollow section.