A comprehensive review of loosening detection methods for threaded fasteners. (1st April 2022)

• Record Type: Journal Article
• Authors: Huang, Jiayu; Liu, Jianhua; Gong, Hao; Deng, Xinjian
• Is Part Of: Mechanical systems and signal processing. Volume 168(2022)
• Publication Date: 2022-04-01
• Subjects: Thread fasteners -- Loosening detection -- Contact characteristics -- Structural dynamic characteristics -- Visual inspection; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2021.108652
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: Outlining various loosening detection methods systematically and thoroughly. Loosening detection methods can be categorised into sensor-based, vision-based and percussion-based methods. The advantages and disadvantages of different methods are compared. The development trend of bolt looseness detection technology is prospected. Abstract: Loosening of threaded fasteners can cause preload decline, induce bolt fatigue fracture, and severely compromise the reliability of mechanical products. Loosening detection is an effective method for early prevention of severe loosening behaviour. This study classifies various detection methods into sensor-based, vision-based and percussion-based methods and systematically summarises their research progresses. The sensor-based method implants or sticks sensors on the mechanical structure with bolted joints, and achieves loosening detection by exploiting the variation on measurement parameters of sensors. It can be divided into explicit detection and implicit detection. The former requires accurate experimental calibration whereas the latter requires to extract sensitive loosening features. The percussion-based method applies a hammer to knock the mechanical structure and receives the audio signal. Like implicit sensor-based methods, loosening severity is evaluated by extracting sensitive features from the received audio signal. The vision-based method captures the images of threaded fasteners and calculates the rotational angle or the length of exposed bolt for loosening detection. The implicit sensor-based, percussion-based, and vision-based methods can only detect several discrete loosening states and be applied mainly to a single bolted joint. It is considered essential and significant to develop a self-powered sensor capable of signal wireless transmission and to conduct precise preload detection by establishing the quantitative relationship between loosening features and preloads using deep learning algorithms.