This paper falls within the context of diagnosis of rotating machines in speed variableregime. Based on simulation signals, this work has the purpose to find relevant indicatorsfor the diagnosis of gear transmissions in a variable regime. Two indicators are proposed;the first indicator is the RMS value applied to the vibration signal divided by itscorresponding instantaneous frequency. The second is the normalised gear frequency byaveraging speed. The gear frequency and averaging speed are estimated from thespectrogram. To test the proposed indicators simulate signals have been used. Thesesignals are the results of a dynamic modelling of the gears transmission and arecalculated by using the Newmark integration diagram. This dynamic modelling takes intoaccount the eccentricity defect.

This paper discusses the characteristics of time-frequency estimators to be used in theidentification of systems with localized non-linearities. The common idea underlying thisresearch is that, for certain classes of structural response signals, the availability ofa limited number of experimental data can be partially obviated by taking into account the“localisation” in time of the frequency components of the signals. Time-frequencytechniques for structural identification are reported that extend the definition ofinstantaneous time-frequency estimators and Gabor instantaneous estimators were extractedfrom non-stationary vibration signals. In order to foresee their validity on the basis ofmeasured data, methods were applied to seismic responses obtained from numerical testsconducted on steel frames. The results obtained made it possible to evaluate thecharacteristics of time-frequency identification techniques as well as their efficiencywhen applied to non-linear structures.