Using Tiptiming and Strain Gauge Data for the Estimation of Consumed Life in a Compressor Blisk Subjected to Stall-Induced Loading

Using tip timing technology to record blade vibratory behavior has grown to become an industry standard over the past decade. Typically, the technology gets used during engine prototype testing to verify safe operation of the blades and thus the engine through synchronous and non-synchronous excitation events. Another common application is blade health monitoring, where the technique is used to detect deviations in natural frequencies and/or amplitudes compared to the virgin state. In both cases, acquired response data are used to establish that blade stresses remain below the high cycle fatigue limit. More rarely is tip timing data used as basis for remaining life estimation. As an example of how tip timing technology can be used beyond traditional resonance clearance for new blade designs, this paper presents an assessment of the fatigue damage incurred to a transonic compressor rotor subjected to stall-induced dynamic loading. The compressor rotor in question is equipped with tip timing, as well as strain gauges for a limited set of airfoils. The dynamic loads at stall are non-synchronous and highly erratic in nature, leading to quasi-static response of multiple modes. To facilitate a conceptually straightforward time domain finite life fatigue analysis, different strategies are employed to reconstruct the stress-time signal from tip timing data. This in turn allows for quantification of accumulated damage cycles, which is here done through simplified and traditional rainflow counting techniques. Additionally, a non-standard way of processing of tip timing data was employed to overcome one of tip timing method drawbacks — frequency aliasing. As an approach the nonuniform Fourier transform was applied to the same data sets. The results obtained are thoroughly evaluated and compared with strain gauges results highlighting the benefits and limitations of the respective approaches for highly complex stress-time histories such as stall events.