In-depth study of direct and derived acoustic emission parameters and their information content for classifying lubricated tribological sliding contacts

Abstract

Acoustic signals (AE) have become a valuable tool for monitoring various tribological systems. This study investigates three key tribological behaviors: the initial running-in process, severe wear, and the lubrication state of journal bearings using four AE postprocessing methods. These methods are friction power-scaled quadratic mean (SRMS), power spectrum (PSD), normalized power spectrum (nPSD), and a statistical hit-rate (HR) analysis of the raw AE signal. The research draws on data from four tests conducted in three set-ups: Ring-on-Disc (RoD), Ring-on-Liner (RoL), and Journal-bearing-adapter (JBA) test configurations. AE measurements were taken at both 1 Hz and 1 kHz for the quadratic mean value (RMS), as well as at 900 kHz for short time intervals during each test to capture the raw signal data. In conjunction with other measured parameters like friction coefficient, temperature, contact potential, and wear, we establish correlations between the acoustic parameters and tribological behaviors. The findings indicate that SRMS and nPSD are valuable for identifying the end of the initial running-in phase. All four methods are effective in recognizing high wear states, with nPSD and HR even distinguishing between different wear phenomena. The lubrication state of journal bearings is best assessed using nPSD, allowing users to monitor certain geometric changes in the bearing.