Crack defects identification in composite elastic bodies of complex geometry

Abstract

Crack defects identification in composite elastic bodies of complex geometry

Borisova E.V., Vasiliev P.V.

Incoming article date: 15.12.2014

This paper proposes a method of constructing a simulation model of the pipe, reinforced by a thin annular plate, and further analysis and definition of internal defects based on the ultrasonic non-destructive testing methods and the artificial neural networks. The influence of various physical and geometrical parameters of training neural networks was performed. Optimal structure of neural network and the type of training vectors was established.

Keywords: Crack, coating layer, functional coatings, finite-element method, artificial neural networks