[115] Gear prognostics: an ISO-based predictive algorithm for lifetime estimation of operating gearboxes

L Bergamini, A Zippo, G Iarriccio, M Molaie and F Pellicano
Università degli Studi di Modena e Reggio Emilia, Italy 

For the realisation of optimised, effective gearboxes, both the set of design rules and the study of operation conditions interconnected to each other have to be taken into account. Condition monitoring is an essential tool where the study of gearbox operation is tackled and many techniques have been developed to detect the onset of failures due to fatigue or wear. However, these techniques fall short where the prediction of gearbox conditions is concerned, as in the case of tracking the useful lifetime during operation, ie when long-term predictions are needed, prognostics approaches must be developed. Within this framework, a prognostic approach/software is developed.

The technique is able to estimate the useful lifetime of a gearbox on the basis of design features (safety factors and parameters), standards and codes (ISO) and realistic/measured load spectra that is representative of the operation condition of the gearbox. In order to have an effective prediction, the digital twin of the gearbox has to be set and the reference safety factors as well as parameters for failure condition (for example bending failure, pitting, etc) must be properly selected and set. Once the history of the applied load spectrum is recorded and suitably filtered by the software, the lifetime of the gearbox is estimated, along with the remaining lifetime, remotely and in real-time.

This tool opens up the possibility of combining condition monitoring and prediction for a wider, comprehensive study of gearbox operation. Moreover, it potentially allows the detection of machinery misuse and planning accurate usage-dependent maintenance in an industrial environment.