Fiber optic strain measurement for machine monitoring

Abstract

Monitoring machines during operation is an important issue in measurement engineering. The usual approach to monitoring specific machine components is using strain gauges. Strain gauges, however, may sometimes not be used if conditions are harsh or installation space is limited. Fiber optic sensors seem to be an alternative here, but dynamic health monitoring has been dificult so far. The focus of this field study is to measure vibration characteristics of machine parts during operation using fiber optic sensors with the objective of early damage detection. If that was possible, downtime and maintenance costs could be minimized. Therefore a field test for dynamic fiber optic strain measurement on a roller bearing was carried out. The test setup consisted of the bearing built into a gear test stand and equipped with an array of fiber Bragg grating sensors. Fifteen fiber sensors were interrogated with a sample rate of 1 kHz and the vibration pattern was extracted. The radial load distribution was measured with high spatial resolution and a high degree of compliance with simulation data was found. The findings suggest that fiber optic health monitoring for machine components is feasible and reasonable. Especially with the help of distributed sensing on various components extensive health monitoring on complex technical systems is possible.