Uncertainties in an application of operational modal analysis

Operational modal analysis is widely used to estimate the mechanical properties of vibrating systems. The several available mathematical techniques, based upon different mathematical formulations, have the common feature that the unmeasured excitation is modeled as a random process specified by some probabilistic model. In practical applications, the length of the measurement is necessarily limited and the probabilistic model adopted to represent the excitation does not necessarily apply. This, together with measurement errors, leads to uncertainties of different nature that affect the estimation of the modal parameters. We discuss the effect of these uncertainty through the assistance of a wide numerical simulation in which a realistic identification process is reproduced isolating different sources of uncertainty. Besides, the identification activity of the experimentalist (e.g. rejection of clearly unphysical results) is mimed by an automated software. The conclusions derived from the simulations are verified by applying the same procedures to a real case study for which a large measurement database is available