An improved blade tip timing dual-probe method of synchro-resonance frequency identification for blade damage detection

• Published in: Mechanical systems and signal processing Vol. 203; p. 110731
• Main Authors: Dong, Jiannan, Li, Hongkun, Cao, Hongwei, Fan, Zhenfang, Chen, Yugang
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.11.2023
• Subjects: Blade damage detection; Blade tip timing; Dual-probe method; Frequency identification; Synchro-resonance; Frequency identification; Blade damage detection; Blade tip timing; Dual-probe method; Synchro-resonance
• ISSN: 0888-3270; 1096-1216
• DOI: 10.1016/j.ymssp.2023.110731

•Synchro-resonance frequency of BTT signals is recovered based on a sparse model of frequency domain transformation.•The minimum number of probes is changed from four to two by the improved measurement method of blade departure time from the probe.•The parameterized matrix formed by the probe layout is analyzed subject to the limitations of RIP and Spark constant.•A method of synchro-resonance frequency identification without prior knowledge is proposed to extend frequency recovery to frequency identification.•The effectiveness of the proposed method was verified by blade vibration tests and damage detection of the blades. Vibrational state monitoring of rotating blades is essential for the safe operation of turbomachinery. Hence, it is particularly important to detect damage to rotating blades during operation. Blade tip timing (BTT) is a promising non-contact monitoring technology for rotating blades. Blade damage will lead to changes in its natural frequency, which can be used as a diagnostic indicator to determine whether there is damage. However, the BTT signal is undersampled to make it not truly reflect the vibration frequency of the blade. Additionally, the number and layout of the BTT probes also limit the application of this technology. In this work, an improved identification method of synchro-resonance frequency without prior knowledge based on the BTT dual-probe is proposed to detect the blade damage. Firstly, the four-channel vibration displacement sequence is obtained by the random layout of dual-probe, which overcomes the limitation of quantity and layout respectively. Then, based on the sparse recovery model in the frequency domain, the multi-scale adjustment fused atomic rationality judgment method is adopted for synchro-resonance frequency identification. The effectiveness and the feasibility of the proposed method are well verified by the experimental data. The monitoring of blade damage status can be realized by comparing the synchro-resonance frequency of the healthy and damaged blades. Thus, the proposed method can provide a practical engineering method for damage monitoring of rotating blades.