Abstract:The blade tip timing measurement of rotating mistuned bladed-disk vibration is modeled and analyzed by a newly proposed method. The motion of the blades is simplified by a 2-degree-of-freedom lumped mass model responding to traveling wave excitations. A deconvolution algorithm is established to recognize the amplitudes and natural frequencies of system modes based on the undersampled tip displacement data collected by several circumferentially distributed probes. Theoretical analysis shows that the algorithm could obtain accurate amplitudes given a sufficient candidate vector that includes the orders of dominant vibrational components. The robustness is demonstrated by a numerical study which reveals that the uncertainty of the magnitude caused by normally distributed random noise with 0.1 standard deviation is within +/- 5% as long as the leakage effect is insignificant. The proposed method is applied to analyze a 2F/4EO vibration of a fan in a real aero-engine test. The response curves generated by the present method successfully reconstruct the feature of multiple peaks observed in the measurements. More importantly, the maximum amplitude obtained by the present method agrees with that measured by the strain gauge, while the traditional SDOF method underestimates the value by 38.2% on average.

Vibrational Response Analysis of Rotating Mistuned Bladed-Disk Using a Deconvolution Method Based on Blade Tip Timing