A method for the harmonic removal in operational modal analysis of rotating blades

The operational modal analysis, OMA, allows estimating the dynamic properties of a structure, natural frequencies, damping ratios, and mode shapes, without measuring the input forces. According to the main hypothesis concerning the input excitation, i.e., stochastic with frequency independent spectra (at least in the frequency band of interest), it is not theoretically possible to apply the OMA procedures in structures characterized by the presence of harmonic components in the excitation loading. In this paper, an approach capable to identify the presence of harmonic excitations, acting together with a broad band stochastic loading, and then to remove their effects in the modal parameter estimate is presented. The approach is based on the joint use of the statistical parameter called “entropy” and the already developed output-only procedure based on the application of the Hilbert transform properties to the output response signals. The capability to improve the OMA procedures is investigated numerically and through whirl tower experimental tests of a rotating blade in which both stochastic and harmonic contributions to the dynamic excitations have been provided by the perturbations arising from the operative conditions. A sensitivity analysis has been also performed to evaluate the effects of the filtered responses, in the time domain, on the statistical characterization, required to distinguish the operational frequencies from the natural ones.