Blade Tip Timing (BTT) methods have been increasingly implemented for blade health monitoring. However, there are two drawbacks of current signal analysis methods preventing them from applying to on-line monitoring: firstly, current on-line monitoring requires the manual judgment of the resonance region which is time-consuming. Secondly, existing BTT resonance signal analysis methods are not suitable for on-line monitoring. Spectral-analysis based method present spectral aliasing while the computational complexity of the sparse based method is usually high. In this paper, we propose an adaptive on-line blade health monitoring method that includes two steps: automatic resonance region recognition and parameters identification in resonance area. For the former step, we demonstrate different methods for synchronous and asynchronous resonance, we use the cross-correlation to judge the occurrence of synchronous vibration and use linear estimation to determine the appearance of asynchronous vibration. For the latter step, an iterative adaptive least-squares periodogram is adopted for its trade-off between spectral aliasing and computational complexity. The effectiveness of the above steps is firstly verified using different simulation data separately. Then, the laboratory data is used to test the effectiveness of the whole method. Finally, the online monitoring function of the proposed method is verified by the engineering data with both synchronous and asynchronous resonances.
Complex blade responses such as a rotating stall or simultaneous resonances are common in modern engines and their observation can be a challenge even for state-of-the-art tip-timing systems and trained operators. This paper analyses forced vibrations of axial compressor blades, measured during the bench tests of the SO-3 turbojet. In relation to earlier studies conducted in Poland with a small number of sensors, a multichannel tip-timing system let us observe simultaneous responses or higher-order modes. To find possible symptoms of a failure, blade responses in a healthy and unhealthy engine configuration with an inlet blocker were studied. The used analysis methods covered all-blade spectrum and the circumferential fitting of blade deflections to the harmonic oscillator model. The Pearson coefficient of correlation between the measured and predicted tip deflection is calculated to evaluate fitting results. It helps to avoid common operator mistakes and misinterpreting the results. The proposed modal solver can track the vibration frequency and adjust the engine order on the fly. That way, synchronous and asynchronous vibrations are observed and analysed together with an extended variant of least squares. This approach saves a lot of work related to configuring the conventional tip-timing solver.
18 November 2019 in Journal of Vibration and Acoustics
Blade tip timing (BTT) includes a number of uncertainties that discourage its use. One of the main ones is the shift in the equilibrium position of the blade tip due to steady (non-oscillatory) bending and/or twisting of the blade, and axial movement of the bladed disk (blisk)-shaft system. This results in a shift in the effective measurement position of the probe relative to the blade chord, resulting in errors in the tip vibration measurement which can translate to a huge error in the corresponding stress estimate, which relies on calibration against finite element (FE) models. Previous experimentally validated research by the authors introduced a method for quantifying steady movement of a single type (axial, lean, or untwist), using BTT data from not more than two probes. In this paper, a development of the previous method is presented that provides a solution for the case of simultaneous types of blade steady movements. Additional probes are used for determining the direction, but these can be placed at any angular positions. The developed method is validated using a BTT simulator of a blisk, and accurate results obtained. The simultaneous axial and lean movements can be accurately determined when the untwist is negligible, and an uncertainty level can be specified when the untwist is not negligible. The untwist itself can be calculated accurately in all cases of simultaneous movements. Guidelines for the use of the method in different scenarios are provided.
29 May 2025 in IEEE Transactions on Instrumentation and Measurement
Ensuring the structural integrity and operational reliability of aero-engine compressors requires precise vibration measurement of rotating blades. Blade Tip Timing (BTT) is a widely used non-contact measurement technique, but its accuracy is inherently influenced by the blade’s twisted geometry and probe positioning. This study introduces a calibration framework that improves BTT precision by integrating Finite Element (FE) modeling to account for blade geometry and vibration characteristics. By extracting vibration mode shapes from the FE model and converting them into an equivalent measurement plane, this method enhances the reliability of BTT measurements. The calibration is validated using Laser Doppler Vibrometry (LDV), which serves as a high-precision reference in controlled laboratory conditions. Results show that this approach significantly reduces measurement discrepancies from 10% to a range of 2–3%, strengthening BTT’s robustness in industrial applications where LDV is not deployable. Additionally, the study establishes a crucial restriction on sensor positioning near the blade tip, identifying areas where BTT probes may fail to capture meaningful data depending on the mode shape. This work lays the foundation for standardized calibration methodologies in turbomachinery health monitoring, ultimately improving predictive maintenance strategies and diagnostics.
17 June 2021 in Journal of Engineering for Gas Turbines and Power
This paper illustrates a fundamentally different approach applied to blade tip-timing (BTT) analysis in the background of works performed through the long history of this method. New innovative approach is based on the precise separation of measured data that was published in previous works. In this respect, the construction of the signal profile is intended for fast and accurate definition of the regulation function served to expression of the complete speed fluctuation and its instabilities. Following, the presented signal profile is demonstrated on four variants of data processing in two significantly different operations, where it allows the application of a one-step approach. Finally, these results are also compared with a two-step method used in previous cases without the signal profile. The benefit of the suggested procedure of the signal profile calculation is evident in the possibility of processing more complex functions in deterministic methods without limiting the characteristic advantages of these methods.
1 September 2021 in Transactions of the Canadian Society for Mechanical Engineering
The blade vibration measurement is crucial for gas turbine engines to ensure safe operations. One of the techniques is blade tip-timing (BTT), which is under the assumption that rotor speed is constant and depends on a once-per-revolution (OPR) timing reference to calculate the blade tip displacement and to identify the blade sequence. However, this assumption is incorrect for transient conditions, and the installation of an OPR sensor sometimes is not allowable and reliable. These reasons greatly limit the application of the BTT technique. This paper proposes a self-correcting (SC)-BTT method to realize the blade vibration measurement under different operating conditions without the use of an OPR sensor. The proposed method is based on polynomial fitting, and a reference probe is used to correct high-order fitting coefficients. Numerical results show that the SC-BTT method can greatly reduce the fitting error caused by blade pitch and vibrational parameters. Experimental results demonstrate that the proposed method is capable of removing the limitation of the lack of an OPR sensor and overcoming the drawbacks of the OPR system, such as the failure of the OPR sensor or low-speed resolution. For three investigated cases, the relative errors of derived rotor speed are below 0.12%. The relative error of blade peak-to-peak amplitude (PPA) and initial phase angle are around 3% at the resonance region with engine order (EO) 2.