Bolt looseness monitoring using dynamic mode decomposition with piezoelectric active sensing

Abstract

Bolt joints are widely used in multiple industries due to its ease of operation and low-cost. Under the influence of multiple external factors, bolt joint will loosen, which poses risks to structural safety. In this paper, a bolt loosening monitoring approach is proposed based on piezoelectric active sensing and dynamic mode decomposition (DMD). To the best of the author’s knowledge, this research represents the first attempt to detect bolt looseness using DMD method. A high-frequency dynamic response signal of the bolt joint is obtained based on piezoelectric active sensing. DMD is used to decompose the high-frequency dynamic response signal into a series of dynamic mode components. The bolt looseness mainly affects the localized high-frequency and high-order dynamic characteristics of the bolted structure. This paper innovatively proposes that the high frequency component signal constructed by the higher order dynamic component more directly reflect the localized dynamic characteristics of the bolted joint. Then, the signal energy of the constructed high frequency component signal is developed into a health index (HI) to characterize the bolt loosening degree. The experiments were conducted to verify the effectiveness of the proposed method. The experiment results shows that the proposed HI monotonically increases as the tightening of the bolt over the whole range of bolt-preload. Furthermore, the research incorporates different sensor mounting positions and loosening of bolts with various diameters, demonstrating the applicability of the proposed method. The proposed method is significant in detecting loosening of bolt joints, especially in bolt early loosening, and enhances the performance of the piezoelectric active sensing method.