Visibility of gearbox backlash oscillation and baselining for future assessment of deterioration
As part of routine testing of a number of rotating machines on this asset, three large gas compressors were tested (sizes ranging from 1 – 9 MW). Each compressor has a single-stage speed increasing gearbox, with known numbers of teeth on the gears. The spectra in each case show very clear signals at frequencies corresponding to the “hunting tooth” frequency and multiples of it (the hunting tooth frequency is a measure of when the same combination of teeth on the two gears mesh with one another and is given by the Lowest Common Multiple of the tooth numbers on the two gears). Vibration levels at the gearboxes are low. Whilst a signal simply at this one frequency would indicate damage to one of the teeth on one gearwheel, and one of the spaces between the teeth on the other gearwheel, the presence of peaks at a whole series of multiples of this frequency combined with the absence of significant vibration signals points to the phenomenon of backlash oscillation, where slight eccentricities and slight variations in tolerances on the gear teeth lead to dynamic behaviour of the two gears effectively “rattling” against one another. Even though the motor is continuously driving, and the compressor is continuously being driven, these dynamic behaviours mean that at the level of individual teeth, the load can be changing and even reversing. Within reasonable limits, this probably has little impact on the longevity of the equipment, though it probably results in higher noise levels. This situation has now been baselined and future deterioration can be compared against this baseline.
What to do next
If you are interested in applying Faraday Predictive condition monitoring systems to your plant, give us a call on 0333 772 0748, or email us on info@faradaypredictive.com.
If you want to develop your business case for investment in Faraday Predictive condition monitoring systems, take a look at this.
And if you want to know more about our condition monitoring systems themselves, click here.