The spin test is a standard industrial practice employed for the qualification of rotor blades and disks. The expected results are the modal properties of blades and assemblages at different rotation velocities. If a significant dynamic coupling among the blades exists, global vibration modes appear, reflecting into a set of closely spaced natural frequencies for each mode family. In case of perfectly-tuned bladed disks, the circumferential structure of the mode shapes is known and can be exploited during the identification process so that traditional single-dof models may be applied. On the contrary, the mode irregularities produced by mistuning prevents the use of single-dof models requiring the development of more sophisticated approaches. In this work, we propose a multi-dof identification technique organized as follow: 1) the FRF of the bladed disk in the neighborhood of a resonance crossing is identified by the wavelet transform of the measured response; 2) the modal parameters of the system are estimated using a mixed stochastic-deterministic subspace algorithm formulated in the frequency domain. The procedure is validated using a realistic numerical simulation.
Turbine blades are critical components in thermal power plants and their design process usually includes experimental tests in order to tune or confirm numerical analyses. These tests are generally carried out on full-scale rotors having some blades instrumented with strain gauges and usually involve a run-up and/or a run-down phase. The quantification of damping in these conditions is rather complicated, since the finite sweep velocity produces a distortion of the vibration amplitude in contrast to the Frequency-Response Function that would be expected for an infinitely slow crossing of the resonance. In this work, we show through a numerical simulation that the usual identification procedures lead to a systematic overestimation of damping due both to the finite sweep velocity, as well as to the variation of the blade natural frequency with the rotation speed. An identification procedure based on the time-frequency analysis is proposed and validated through numerical simulations.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.