Hilbert transform-based frequency response function estimation from run-up response of rotors

Transient operating conditions such as run-up and run-down are often seen in automotive and other machines with variable speed rotating elements. To estimate the frequency response function from swept sine responses (obtained during run-up), a new approach is proposed which utilizes Hilbert transform and stationary wavelet transform. This method is an alternate tool for conventional swept sine frequency response function estimators like tracking vectors and co-quad analysers. The output from this algorithm is validated using frequency response function obtained using exact frequency response function expression and the conventional Fourier transform-based frequency response function. Both MATLAB simulations and experiments have been carried out in this study to estimate the frequency response function. Modal parameters are extracted from the frequency response function obtained using Hilbert transform.