Multi-bolt looseness state monitoring using the recursive analytic based active sensing technique

Abstract

Monitoring the looseness degree of multi-bolt connections is significant for ensuring the safety of steel structures. Among existing structural health monitoring techniques, the active sensing is preferred due to its sensitivity to structural damages. However, the energy based indicator, such as wavelet packet energy index (WPEI), is prone to saturation and fluctuation when monitoring multi-bolt looseness. In order to resolve this problem, the recurrence plots (RP), phase trajectory and recurrence quantification analysis (RQA) were employed to quantify the multi-bolt looseness by extracting nonlinear features from stress wave signals. Repeated experiments were conducted on a steel beam assembly with three bolts to verify the effectiveness of the proposed method. The experiments showed that the recursive entropy index could effectively eliminate the energy saturation and fluctuation, which enhances the performance to monitor the multi-bolt looseness state. In addition, the phase trajectory and RP could intuitively detect the location of the loosened bolt.