The vibration monitoring of rotor blades by Blade Tip Timing (BTT) has extraordinary potentials as the experimental setup is much simpler and robust than the strain-gauge / telemetry system and the whole set of blades of the disk is observed. The main drawback is that the blade vibration is heavily undersampled, at least in the common Nyquist sense. Besides, the sampling times are irregularly distributed due to possible variations of the rotor speed and the non-uniform spacing of the sensors along the circumference of the casing.

When non-synchronous vibrations are concerned, BTT data are processed following two alternative approaches. The traditional method, usually referred to as all-blade spectrum, constructs a uniformly-sampled signal by enqueuing the deflection of all the blades measured by a single sensor. The sampling frequency of this signal *f~s~*=*N~b~ f~rev~* (*N~b~* being the number of the blades and *f~rev~* the rotor speed) is often high enough to represent correctly the blade vibration, at least as far as the low-order modes are concerned. The drawback of this method is that the vibration of the disk is observed in the fixed reference frame. Its translation into the rotating reference frame introduces ambiguities related to the circumferential shape of the vibration, which is estimated afterwards by comparing spectra obtained from different sensors. This method is simple and efficient, but produces results that may be hard to interpret, unless the blade vibration is very well correlated along the disk like in the case of flutter.

Other methods follow an approach that is somehow orthogonal to the previous one. They collect the vibration of a single blade as it is measured by all the available sensors encountered during the revolution. This leads to a non-uniformly sampled signal representing the vibration observed in the rotating frame. The circumferential shape of the vibration is obtained afterwards comparing spectra obtained for different blades.

This paper tries to bridge the gap between the two mentioned approaches introducing a technique for the spectral analysis of non-uniformly sampled, multi-channel signals. The building blocks of this technique are the Orthogonal Matching Pursuit (OMP), already applied for processing BTT data, and the Proper Orthogonal Decomposition (POD). The method is applied to two realistic bladed disks whose characteristics are opposite in terms of interblade mechanical coupling. Both experimental and simulated data are used to support the discussion.


Multi-Channel Spectral Analysis of Non-Synchronous Vibrations of Bladed Disks Measured by Blade Tip Timing

Division / Sponsor:
International Gas Turbine Institute (IGTI)

Paper Proceedings: https://asmedigitalcollection.asme.org/GT/GT2020/volume/84058

Turbo Expo - Virtual Conference

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.