For α = 0, (Figs 9.a, 9.c, 9.e and 9.g), the power spectrum level for the faulty bearing is higher than that of the good bearing for both simulated and measured signals. The increase in the dB difference is observed across the whole frequency range. As the comparison is made at α = 0, no conclusion could be drawn as to the source of excitation, as this increase in the dB difference could be the result of an increase in the stationary noise, which is still present at α = 0. Note that in the current simulations, the dB difference in the low frequency regions has a better correspondence with the experimental one. For α = Ω (Fig 9.b, 9.d, 9.f and 9.h), it is noticed that the highest increase in the dB difference appears mainly in the low frequency region (fault modulates the gearmesh frequencies). The increase in the dB difference at α = Ω implies the existence of an extended fault in the inner race (as no gear component or stationary noise are present in this comparison). In the measured signal the increase is noticed mainly in the low frequency region, as the fault is quite smooth (the presence of the oil film helps in providing such smoothness) in particular with a smooth exit so that the higher frequency region was not excited. In the simulated signals (both LPM and condensed model) there is an increase in the dB difference in the high frequency region, very likely due to the fact that the exit from the fault is still rather abrupt. The inclusion of the casing model through the Craig-Bampton reduction method, and then through the convolution of the extracted forces by the impulse responses increases the similarities with the experimental results and shows a much better correspondence both in the low and high frequency regions when compared to the initial LPM model and the direct LPM- Craig Brampton reduced model. As this is still an ongoing work, more refinements and adjustments are expected to take place to improve the model. More experimental cases for extended inner and outer race faults are planned and should be compared to simulated results at different speeds and loads. 8. Conclusions The simulation of the dynamic behavior of a complex structure such as a gearbox is carried out using finite element model reduction technique. The dynamic reduction was based on the Craig-Bampton method and the reduced mass and stiffness matrices of the casing were incorporated into the lumped parameter model of the gearbox. The simulation was carried out for an extended inner race fault. In order to extend the validity of the combined /reduced model, the forces were extracted from combined/reduced model and convolved with the impulse responses corresponding to the FRFs of the whole gearbox (casing and internals). The current approach has the advantage of maintaining a dynamic interaction in the low/mid frequency regions and improves the validity at higher frequency where it is the modal density rather than the individual modes that are of interest. The inclusion of the casing model through the Craig Bampton reduction method and then through the convolution of the extracted forces with the impulse responses, increases the similarities with the experimental results and shows a much better correspondence both in the low and high frequency regions when compared to the initial LPM model and the direct LPM- reduced Craig-Brampton model. As this is still an ongoing work more refinements and adjustments are expected to take place to improve the model. More experimental cases for extended inner and outer race faults are planned and should be compared to simulated results at different speeds and loads. Acknowledgements This research was supported by the Defence Science and Technology Organisation (DSTO) through the Centre of Expertise in Helicopter Structures and Diagnostics at UNSW. 408
RkJQdWJsaXNoZXIy MTMzNzEzMQ==