Monitoring gear wear with transmission error

Abstract

The use of transmission error (TE) in gear diagnostics has recently become more feasible. TE is a parameter that describes the meshing error between two gears, i.e., the deviation in angular position from the theoretically perfect, rigid involute case, and can be measured very accurately using shaft encoders installed on the unloaded part of the input and output shafts. It is hence one of the measurements closest to the source – the internal excitation from the gearmesh. Wear induces geometric deviations on gear tooth profiles, which will manifest as a geometric TE (GTE). To perform effective gear wear monitoring, one requires reliable sensors or measurements that are sensitive to anomalies due to wear, especially when wear is at an early stage and hard to detect. Vibration, which is usually measured on the gearbox casing, tends to be affected by the complex transfer path from the source to the measurement point and thus poses challenges in monitoring wear. Conventional wear analysis techniques, including image acquisition and characterisation of worn gear surfaces, often involve disruption of the operation and offline analysis, causing some delay. TE, as a measurement much closer to the source, has considerable benefits compared to vibration as well as traditional wear analysis.This paper demonstrates a comparative study on TE measured from dry and lubricated wear experiments, offering new insights on the effect of different wear mechanisms on TE signals. The results show that TE is sensitive to the evolution of gear wear and could be used to perform trending effectively. Such a powerful measurement with good trending ability could revolutionise gear prognostics. With the embedded sensor trend associated with Industry 4.0 and Internet of Things, the use of inbuilt encoders will likely become more prevalent in industry, allowing the use of TE measurements in practical gear monitoring.