166 S.K. Cheah Fig. 17.8 Comparison of strain in gage #1 in frequency & time domain 0 50 100 150 200 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Frequency, Hz Normalized Strain Amplitude, μ Analytical Measured*2 0 0.05 0.1 0.15 0.2 0.25 0.3 −3 −2 −1 0 1 2 3 Time, s Normalized Strain, μ Analytical 0 0.05 0.1 0.15 0.2 0.25 0.3 −3 −2 −1 0 1 2 3 Time, s Normalized Strain, μ Measured*2 Fig. 17.9 Comparison of strain in gage #2 in frequency & time domain 0 50 100 150 200 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Frequency, Hz Normalized Strain Amplitude, μ Analytical Measured*2 0 0.05 0.1 0.15 0.2 0.25 0.3 −3 −2 −1 0 1 2 3 Time, s Normalized Strain, μ Analytical 0 0.05 0.1 0.15 0.2 0.25 0.3 −3 −2 −1 0 1 2 3 Time, s Normalized Strain, μ Measured*2 17.6 Concluding Remarks The application of measured ODS onto the generator chassis FEA model using large mass method via harmonic analysis yields results that shows some correlation. The bolt alternating load matches well but the strain data overestimated the response at 1st and 1.5th order indicating some modes were overrepresented. Unfortunately there was no opportunity to perform a modal test on the chassis in a free-free state to allow for model correlation and updating. Any model discrepancy (e.g. incorrect dynamic spring stiffness) is expected to contribute to the errors. Another large deficiency of this approach is that rotational degrees of freedom are not captured in the ODS. Depending on the selected accelerometer measurement locations, this may be important. While the above case study shows reasonable results, confidence in this approach needs to be substantiated for other systems. Acknowledgements The author would like to thank Steve Seidlitz, Maria Wood and James Hogben for the exceptional vibration test data. Gunjan Maheshwari, Alastair Clifford contributed to the excellent strain measurement. Special thanks to Hari Prasad Konka for creating the initial FEA models. This work was supported by Cummins Power Generation.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==