Joint damage identification using chaotic excitation and attractor analysis

In their operational lifetime, engineering structures suffer joint damage inevitably under environmental loads, it may affect integrity and functionality of structures, even causes catastrophic events. Here, a method using chaotic excitation and attractor geometric property analysis to identify joint damage was presented. A chaotic excitation signal was used to excite a structure, and the measured structural acceleration response was reconstructed in state space. A new feature based on attractor local variance was extracted and used to identify joint damage. The parameters influencing feature extraction were also analyzed. A test with a cantilevered beam for joint damage identification was designed and the joint damage conditions were quantified by varying the fastener preload. Five different joint damage conditions were identified. The results showed that the proposed approach works well in detecting joint damage, the feature varying is monotonic with increase in damage, and the location for measuring response and calculation parameters have influence on damage identification.