Crack Monitoring On Aerospace Structure Based On Lamb Waves And Binary Tree Support Vector Machines

In recent years, with the rapid development of China’s aviation industry, the security issue isurgent attention. Aerospace structures are often subjected to a variety of loads which may lead tocrack and the stress concentration portion especially is easy to produce micro crack. If the crackthat is small or located inside the structures is not detected in time, aircraft catastrophic accidents willoccur. It is important and crucial for impeding the structural degradation, increasing the flight safetyand reducing the maintenance cost by applying structural heath monitoring system to obtaininformation about the length and distribution of the crack.This dissertation studies the damage identification and signal processing methods in structuralhealth monitoring of aircraft structures. An active method is applied to monitor the structurecontinuously and online to ensure the structural safety. This technique can detect the crack lengthin/near real time. First, To overcome the difficulty in identifying the crack in key parts of aerospacestructure, one crack length monitoring method based on matching pursuits (MP) method and binarytree support vector machines (BT-SVM) classification algorithm was developed. In this method,Lamb wave signals were decomposed into a linear expansion of several chirplet atoms by MP method,the matching parameters were computed and the feature vectors were extracted, and then the featurevectors were trained and tested by BT-SVM classification algorithm. Second, a lug joint model wascreated with a certain ratio and the effect of crack extension on Lamb wave signals propagation wasstudied, and the result showed that the amplitude of the received Lamb wave signals had someattenuation with the increasing of the crack length, which laid the foundation for the subsequentexperiments At last, fatigue loading experiments were carried out in lug joints and tail reinforcedframes of aircraft. The results showed that this method can monitor the crack change on line, whichpresents a new approach for monitoring the crack.