Early detection of gear-tooth bending-fatigue damage by the Average-Log-Ratio ALR algorithm

Research output: Contribution to journalArticlepeer-review

3 Scopus citations


This work illustrates and explains an application of the Average-Log-Ratio, ALR, gear-damage detection algorithm for a case of gear-tooth bending fatigue. Gear-vibration rotational-harmonic spans between adjacent tooth-meshing harmonics are combined in ALR, thereby greatly reducing the masking effects resulting from the meshing action of the mating gear with the damaged gear. The first observation of unambiguous tooth fracture is identified. Moreover, 100,000 revolutions of the failing gear before this first observation of fracture, gear damage initiation, likely plastic deformation, is identified by ALR, thereby illustrating the potential for early warning of impending gear failure. Computation of kurtosis, FM4, using the same data, is provided for comparison of the detections provided by ALR and FM4, indicating earlier detection of damage initiation provided by ALR in comparison to that provided by FM4.

Original languageEnglish (US)
Article number108922
JournalMechanical Systems and Signal Processing
StatePublished - May 15 2022

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Signal Processing
  • Civil and Structural Engineering
  • Aerospace Engineering
  • Mechanical Engineering
  • Computer Science Applications


Dive into the research topics of 'Early detection of gear-tooth bending-fatigue damage by the Average-Log-Ratio ALR algorithm'. Together they form a unique fingerprint.

Cite this