Estimating bolt tension from vibrations: Transient features, nonlinearity, and signal processing

Marie Brøns*, Jon Juel Thomsen, Si Mohamed Sah, Dmitri Tcherniak, Alexander Fidlin

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

Monitoring and control of tension in bolted joints is a difficult task that has received long-time attention. A newly proposed technique is to hammer-impact the bolts and estimate the tension based on the vibration response. This present work conducts a thorough experimental investigation of two different bolted structures to identify the potentials of the technique, particularly examining damping ratios and nonlinearity by appropriate signal processing. The applicability of the method is checked by conducting tests with a real multi-bolt structure. For larger tension, the squared first bending natural frequency of a bolt increases approximately linearly with bolt tension. This study investigates the sensitivity of that feature with respect to impact force, i.e. the nonlinearity in the bolt's frequency response. The damping ratio is estimated and observed to decrease with tension, and is overall reproducible for the tested impact forces, though with significant variation for small bolt tension. A time–frequency analysis provides insights into the variations observed in the measured linear damping ratios and natural frequencies. The time-dependent damping ratios generally depend on acceleration amplitude, especially for small amplitudes and small bolt tension. In contrast, the instantaneous natural frequencies are found to be practically independent of amplitude. The absence of significant nonlinearity is encouraging for the potential of a hammer impact-based technique for estimating bolt tension.

Original languageEnglish
Article number107224
JournalMechanical Systems and Signal Processing
Volume150
Number of pages17
ISSN0888-3270
DOIs
Publication statusPublished - 2021

Keywords

  • Bolted joints
  • Damping ratio
  • Nonlinearity
  • Tension estimation
  • Time–frequency analysis

Cite this