Percussion-based Health Monitoring of Bolted Lap Joints

Abstract

Bolted joints are critical to guaranteeing structural safety and performance, yet traditional monitoring methods tend to be expensive and intrusive. In practice, pretensioned bolts inevitably suffer from preload loss, making the reliable detection of loosening essential. This study introduces a Structural Health Monitoring (SHM) approach based on vibration and percussion, utilising audio-emission signals generated under controlled impact. A single bolted lap joint was tested, with the resulting audio and vibration responses recorded and analysed in both the time and frequency domains to assess bolt tightness. The impact range is kept between 300-325 N for tightness levels of 2 Nm, 10 Nm and 20 Nm, considered as loose, semi-tight and tight, respectively. The features such as signal energy, peak to RMS ratio and kurtosis from loose to tight conditions vary by 60%, 34% and 67%, respectively for vibration signals and 62%, 62%, 38% and 52% for sound signals. Utilising the Fast Fourier Transform (FFT) for analysis, a correlation was identified between bolt tightening levels and frequency amplitudes. Both signal types exhibited a slight increase in frequency corresponding to an increase in bolt torque. Sensor fusion is found to be an effective method for validating results and can enhance SHM of bolted connections in automotive, aerospace and aircraft structures.