New Possibilities for Analyzing Complex Asynchronous Blade Vibrations From Tip-Timing Data Using a Sparse Spectral Analysis Method

Abstract

Measurements in engine operation provide key information for understanding the blades vibrating behaviors, especially while observing complex asynchronous vibrations, such as flutter. The studied vibration is performed by the spectral analysis of the measured vibrating mechanical responses. In this context, tip-timing measurement technology enables to get a large amount of information as it monitors all the blades. However, tip-timing technology generates sub-sampled and non-uniform sampled signals. Thus, conventional spectral estimation methods lead to aliased frequency components on the spectrum. Therefore, specific estimation methods have been developed to overcome this issue. This paper presents the tip-timing measurements analysis of a non synchronous vibration test case. It uses a new sparse method with block-OMP algorithm in order to estimate time-frequency diagrams of each blade separately. The results are compared to the results of the well-known asynchronous tip-timing method, called All Blade Spectrum. This real test case highlights the limitations of the latter method, as it involves too much uncertainty, while the sparse method with block-OMP enables to characterize these complex asynchronous blade vibrations.