Time domain operational modal analysis in support of aeroelastic flight testing

Operational Modal Analysis (OMA) for modal identification is currently expanding from the civil to the aerospace engineering field. This is due to the capability to extract modal parameters from operational conditions with a non-deterministic input, even though assumed random in space and time. Thus, no assumptions shall be made on the boundary conditions for the structure under test, since those are actual in service. Furthermore, the need for a random excitation source could be easily satisfied in flight, setting the aircraft in a straight and level, unaccelerated condition. These peculiarities are fostering the interest in OMA for flutter testing applications. Nevertheless, when compared to classical Flight Vibration Testing techniques, OMA is more sensitive to measurement chain noise and sensors quantity. The latter is quite often critical when dealing with the narrow spaces inside high performance aircraft, requiring a thorough optimization in terms of sensors number and positioning. In this paper, frequency and time domain OMA techniques are applied, and their effectiveness evaluated on simulated random response data generated from the Finite Element Model of a typical high-performance aircraft wing, the AGARD 445.6. Optimal sensors positions are identified for an increasing number of measurement points. Natural frequencies and mode shapes are identified with different OMA techniques and compared to the true data source, even introducing output noise components, in an effort to identify the best condition with a minimum number of sensors for an acceptable modal parameters estimation accuracy.