Composite Skin/Stringer Panel Damage Detection Based on Modal Strain Energy and Neural Network Technique

摘要

This paper proposed a new method to detect the damage of composite skin/stringer panel structure using modal strain energy combined with neural network. The change ratio of element modal strain energy is chosen as damage indicator because of it's highly sensitivity to the location and severity of structure damage. Neural network here play the role of a tool to identity the damage according to the change ratio of modal strain energy. To achieve this, a three layers neural network model is built and the BP arithmetic is used. The proposed method is validated using a numerical simulation of a composite skin/stringer panel with damages in some elements of its FEM mode, which are simulated by reducing elements' material stiffness properties. The result shows that, this method is robust, accurate and highly efficient with the maximal error limited in 10%.